https://invisible-island.net/autoconf/
2024-09-09
* AcSplit/CF_WITH_MAN2HTML: s/http/https/
2024-08-10
* AcSplit/CF_INSTALL_PREFIX: ci-aclocal generated header
* AcSplit/CF_GETOPT_HEADER:
prefer getting unistd.h via AC_HEADER_STDC, which is normally prerequisite
* AcSplit/CF_UTF8_LIB:
s/AC_HAVE_HEADERS/AC_CHECK_HEADERS/, for consistency
2024-08-03
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
Eli Zaretskii has built a native Windows port of pkg-config which uses
DOS/Windows pathnames. Provide for that in the search-list.
* AcSplit/CF_LIB_RULES, AcSplit/CF_PATH_SYNTAX, AcSplit/CF_PROG_INSTALL:
use CF_GLOB_FULLPATH
* AcSplit/CF_GLOB_FULLPATH:
split the two-part glob into separate variables so that shell variables can
be used for the parts in a case-statement.
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR, AcSplit/CF_WITH_VERSIONED_SYMS:
use CF_GLOB_FULLPATH
2024-07-28
* config.guess: 2024-07-27
From: Paul Eggert <eggert@cs.ucla.edu>
Port config.guess to strict POSIX.1-2024 c17
* config.guess (set_cc_for_build):
Try the c17 compiler, standardized in POSIX.1-2024. See:
https://pubs.opengroup.org/onlinepubs/9799919799/utilities/c17.html
While we're at it, prefer c99 to c89.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2024-07-20
* AcSplit/CF_GLOB_FULLPATH: ci-aclocal generated header
2024-07-16
* AcSplit/CF_SRAND:
add --enable-arc4random option so that mutt can have the feature normally
disabled, while lynx can have it normally enabled
2024-07-06
* AcSplit/CF_NCURSES_PTHREADS:
package names are not necessarily the same as library names
2024-07-05
* AcSplit/CF_WITH_IMAKE_CFLAGS:
change default, so that imake is only checked for if requested.
2024-06-22
* AcSplit/CF_BUILD_CC:
revert - doesn't entirely fix lynx with pdcurses and breaks ncurses
2024-06-13
* AcSplit/CF__DEFINE_SHLIB_VARS:
eliminate trailing backslashes, to help with Debian #1071581 (needs update
to autoconf-252)
* COPYING.asc: sign
* COPYING: bump
2024-06-12
* AcSplit/AM_WITH_NLS:
don't define IGNOTE_MSGFMT_HACK on the command-line because that goes into
the ".pc" and "*-config" files. Replace this by a check which can define
HAVE__NL_MSG_CAT_CNTR, which is more informative -- should only occur with
GNU gettext.
* AcSplit/CF__INTL_BODY: replace ifdef for _nc_msg_cat_cntr
* AcSplit/CF_CHECK_LIBTOOL_VERSION, AcSplit/CF_LIBTOOL_VERSION:
provide for LT_PACKAGE_VERSION set via libtoolize (Debian #1071581)
* AcSplit/CF_VERSION_INFO:
make this work with autoheader (Debian #1071581)
2024-06-11
* AcSplit/CF_CURSES_FUNCS:
modify to work with autoheader (Debian #1071581)
2024-06-04
* AcSplit/CF_NETLIBS: improve symbol used for detecting inet library
2024-06-03
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: add missing CPUs, vendors, and OSes.
Add the following entries to the tables of valid CPUs, vendors, and
operating systems:
CPUs: v810
Vendors: acorn, atari, be, ccur, commodore, digital, intergraph, rebel,
rom68k, rombug, semi, siemens, sni, stratus, tektronix
OSes: atheos, chorusos, psos, syllable, wince
Evidence elsewhere in this script, and/or in config.guess, indicates
that these are all supposed to be valid.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: improve consistency of shorthands and X-Y partial names.
It seems to me that if ‘fooco’ expands to ‘cpu-fooco-foonlyos’,
where ‘cpu’ is whatever CPU was hip at the time Fooco was making
computers, then ‘cpu-fooco’ should also expand to ‘cpu-fooco-foonlyos’,
and ‘cpu-foonlyos’ should as well, except when ‘foonlyos’ is something
generic like ‘sysv’ or ‘bsd’. Implement a bunch of cases where we
didn’t do this quite right. Probably there are more.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: Treat i386-pc-rtmk-nova as a 4-part system name.
Handle ‘i386-nova’ as shorthand for a 4-part system name,
kernel=rtmk os=nova, instead of a canonicalized OS name with
a dash in the middle, os=rtmk-nova.
RTMK <https://rtmk.sourceforge.net/> is a microkernel.
Nova <https://www.nongnu.org/nova> is an OS written on top of that
microkernel. The short system name ‘i386-nova’ was therefore defined
to canonicalize to ‘i386-pc-rtmk-nova’, I think prior to the general
introduction of 4-part system names. ‘i386-pc-rtmk-nova’ still makes
sense under the rules for interpreting 4-part names, but nowadays the
logic for validating canonicalized name components rejects OS names
with a dash in the middle. Feeding kernel=rtmk os=nova to that logic,
instead, does the Right Thing.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: remove two unreachable case clauses.
These attempt to perform normalizations that were already done by
code above this point, so they are unnecessary. There are already
test cases for these normalizations.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: normalize nextstep/openstep consistently.
Normalize the names for NeXT’s operating systems consistently.
The version number bump from 2 to 3 went along with a marketing
name change from “nextstep†to “openstepâ€, which was being honored,
or not, depending on what shorthand you put in and what CPU you
selected (!)
Consistent with the treatment of “Solaris 1†as an alias for sunos4
and “SunOS 5†as an alias for solaris2, and also with what config.guess
appears to do on these machines (which I do not have), config.sub will
now emit CPU-next-nextstep, -nextstep2, -openstep3, -openstep4 no
matter how it was spelled on input.
These OSes historically were used with several other CPUs besides m68k
and sparc, but I think one CPU for which they have a single-component
shorthand, and one for which they don’t, is enough testing.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: don’t override OS for blackfin, m68knommu, parisc.
A similar issue to the previous patch: make the OS default for the CPU
aliases “blackfin,†“m68knommu,†and “parisc†actually be a default
and not an override. There have been many OSes for PA/RISC. Both
blackfin and m68knommu are definitely used in bare-metal configuration
as well as with an OS. And in general, CPU aliases should behave as
much like the canonical CPU name as possible.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: don’t override OS for c90-cray.
Fix a typo in the logic for handling the OS default for c90-cray,
causing c90-cray-[anything] to get overridden to c90-cray-unicos.
I’m not aware of any non-Unicos operating systems for the C90, but
let’s not be a pointless obstacle for anyone who wants to write one.
(For testing purposes I used Linux, because config.sub’s generic logic
accepts it in combination with any CPU whatsoever.)
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: handle cNN-convex and convex-cNN shorthands correctly.
convex-cNN and cNN-convex (where NN in 1, 2, 32, 34, 38) are supposed
to be shorthands for cNN-convex-bsd, but the code to handle them
was in the wrong switch statement, making it unreachable. Fix.
https://en.wikipedia.org/wiki/Convex_Computer says there was also a C4
machine in this product line, at least on paper, but it’s not clear to
me whether they ever shipped one.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: accept knetbsd-gnu, netbsd-gnu, uclinux-gnu.
There is evidence elsewhere in config.sub, and in config.guess,
indicating that knetbsd-gnu, netbsd-gnu, and uclinux-gnu were
supposed to be acceptable combinations of $kernel and $os, but
they are currently being rejected. Add them to the list of
acceptable kernel-os combinations, and add test cases for them.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: consistently use sun as default vendor for solaris*.
config.sub had a rule that os=sunos* implies vendor=sun unless the
vendor has already been set for some other reason, but no such rule
for os=solaris*, and that meant `./config.sub sparc-solaris2` was
inconsistent with what you’d get from config.guess on that system.
Add the missing rule.
Unlike the previous patch, this changes config.sub output for some
systems still actually in use today, but hardly anything looks at
the vendor field unless the OS field is something super generic like
“bsd†or “sysvâ€, so I think we can get away with it.
This patch also adds tests for the mapping of ‘sunos5’ to ‘solaris2’
and ‘solaris1’ to ‘sunos4’, which is logically unrelated *except*
that some of those test cases will fail in the absence of the above
change.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-05-27
From: Zack Weinberg <zack@owlfolio.org>
config.sub: correct Bull DPX/2-[123]00 and DPX/20 canonical names.
config.sub was mapping all single-component shorthands matching the
shell glob `dpx2*` to ‘m68k-bull-sysv3’, which is incorrect. Of the
four different machines once manufactured by Groupe Bull whose model
number can be abbreviated ‘dpx2something’, only two were m68k-based:
the DPX/2-200 and DPX/2-300. The DPX/2-100, on the other hand, had an
i386 CPU. And the DPX/20 was a completely different beast, released
years later and based on the POWER architecture.
I’ve chosen to keep making ‘dpx2’ map to ‘m68k-bull-sysv3’ as the
DPX/2-200 and /2-300 seem to have been the most widely used. config.sub
now understands ‘dpx2100’, ‘dpx2200’, and ‘dpx2300’ as shorthands for
‘i386-bull-sysv3’, ‘m68k-bull-sysv3’, and ‘m68k-bull-sysv3’ respectively.
It also accepts ‘dpx21xx’, ‘dpx22xx’, and ‘dpx22xx’ as equivalent to
these, but no other variations; in particular, ‘dpx2xxx’, which was in
the test suite, no longer works, as it is ambiguous.
(As far as I can tell, there weren’t actually any submodels of any of
these systems, so I could be persuaded to remove the …xx aliases.
I kept them mainly because of the test suite formerly testing dpx2xxx.)
Finally ‘dpx20’ now maps to ‘rs6000-bull-bosx’. Both the ‘rs6000’ and
‘bosx’ components of this name are debatable, but I *think* it is what
config.guess would do on this machine.
Sources for Bull DPX/2whatever history:
- https://oldskool.silicium.org/stations/bull_dpx20.htm
- https://www.feb-patrimoine.com/english/bull_dpx2.htm
- https://www.feb-patrimoine.com/english/unix_and_bull.htm
Note in particular “Que cette machine soit estampillée BULL ou IBM,
elle faisait tourner AIX (l'Unix d'IBM)†on the first of these pages,
which is why I say “bosx†is debatable…
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-03-12
From: Zack Weinberg <zack@owlfolio.org>
config.sub: flatten and sort long case-alternative lists
Several of these lists are going to be modified substantially in the
next few patches, this will make it much easier to see what is going on.
No functional change.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2024-03-12
From: Zack Weinberg <zack@owlfolio.org>
config.sub: disable shellcheck warning SC2162 globally.
With the version of shellcheck I have (0.9.0), the existing attempts
to disable SC2162, above each use of read without -r, do not work
because they are not on the line *immediately* previous to the read
command — there’s always a “saved_IFS=$IFS†line in between. Since
this script must not use read -r at all, we should just be disabling
SC2162 globally.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2024-06-02
* AcSplit/CF_SOCKS5:
modify to work with Dante version of socks5, tested with Debian and Fedora
2024-06-01
* AcSplit/CF_NCURSES_ADDON:
change the configure option from --enable-widec to --with-screen=XXX, to
make it easy to link with ncurses test-packages.
2024-05-18
* AcSplit/CF_WITH_ADA: ci-aclocal generated header
2024-05-13
* config.guess: 2024-04-03
From: Petr Vorel <petr.vorel@gmail.com>
config.guess: Avoid implicit int in definitions of main.
Implicit ‘int’ (e.g. ‘extern foo();’ meaning the same thing as
‘extern int foo();’) was dropped from the C standard in its 1999
edition. Twenty-five years later, free C compilers are finally
starting to make this an error by default, so let’s not use it
anymore in config.guess probe programs.
(Note: As of this writing, GCC 14 and Clang 16 are both more lenient
for ‘main() { … }’ specifically than for other uses of implicit int.
Still, the writing is clearly on the wall.)
We continue to use ‘int main() { … }’, instead of ‘int main(void) { … }’,
because these programs may be compiled by truly ancient compilers that
do not recognize the keyword ‘void’. This leaves open the possibility
of a compiler that errors by default on an empty argument list in a
function definition, which, prior to the 2024 C standard, is technically
still an “old-style†function definition; but we can worry about that
if and when it comes up.
Signed-off-by: Petr Vorel <petr.vorel@gmail.com>
Co-authored-by: Zack Weinberg <zack@owlfolio.org>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2024-05-11
* AcSplit/CF_FUNC_GETTIME:
check for both monotonic and realtime clocks, and add a check for clock_gettime
in the "rt" library for Solaris
2024-05-01
* AcSplit/CF_ANSI_CC_CHECK:
reset CFLAGS/CPPFLAGS on each loop, to keep the options from accumulating
2024-04-24
* AcSplit/CF_FIND_TDLIB:
ensure that $cf_cv_tdlib_package is set when --without-pkg-config
* AcSplit/CF_FIND_TDLIB:
correct check for side-by-side library, and add a compile/link check for it
2024-04-21
* AcSplit/CF_TD_SIG_ARGS:
correct test-program, which was returning a value from a void function
2024-04-09
* AcSplit/CF_ADD_OPTIONAL_PATH: add --verbose messages to show progress
* AcSplit/AM_WITH_NLS: corrected option-name
* AcSplit/CF_WITH_BROTLI, AcSplit/CF_WITH_BZLIB, AcSplit/CF_WITH_IDNA,
AcSplit/CF_WITH_ZLIB:
improve --verbose mode of CF_ADD_OPTIONAL_PATH
2024-04-08
* AcSplit/AM_WITH_NLS: only define ENABLE_NLS if we found gettext()
* AcSplit/AM_WITH_NLS: drop $POSUB symbol
* AcSplit/CF_ADD_OPTIONAL_PATH:
add --verbose messages to show the result from this macro
* AcSplit/AM_WITH_NLS:
add "--with-libintl-prefix" option (patch by Urs Janßen)
* AcSplit/CF_OUR_MESSAGES: correct usage-message (patch by Urs Janßen)
* AcSplit/CF_PIC_HACK:
make "readelf" tool configurable (report by Sam James)
2024-04-02
* AcSplit/AM_WITH_NLS: restore a chunk needed for updating $LIBS
2024-04-01
* AcSplit/CF_DEFAULT_SHELL: egrep vs grep -E
* AcSplit/CF_INSTALL_MAN: use "-n" option with gzip'd manpages
2024-03-31
* AcSplit/CF_FUNC_WORDEXP: ci-aclocal generated header
2024-03-30
* AcSplit/CF_MANPAGE_FORMAT:
add symbol $cf_manpage_coptions, to provide for passing gzip "-n"
* AcSplit/CF_MAN_PAGES: use $cf_manpage_coptions if set
2024-03-29
* AcSplit/CF_SHARED_OPTS: build-fix for clang on Solaris
* AcSplit/CF_CURSES_SETBEGYX:
rather than defining ERR on failure, simply define the macro only if successful
2024-03-17
* AcSplit/CF_NGROUPS: check for headers needed in compile-checks
* AcSplit/AM_GNU_GETTEXT, AcSplit/AM_WITH_NLS, AcSplit/CF_BUNDLED_INTL:
remove obsolete scripting which supported bundled intl-directory
* AcSplit/CF_BUILD_CC:
allow $BUILD_CPPFLAGS and $BUILD_LIBS to be overridden when not cross compiling,
fixes build of Lynx with PDCurses
2024-03-12
* AcSplit/CF_PIC_HACK: provide for -fPIC being found in $CC_SHARED_OPTS
allow for readelf on MinGW (discard error message)
* AcSplit/CF_SHARED_OPTS:
trim a space after "-R" options on Solaris which interferes with builds
using clang. The space might be needed for some older versions of Sun cc.
* config.sub: 2024-03-12
From: Jan Klötzke <jan.kloetzke@kernkonzept.com>
config.sub: recognize L4Re OS
* config.sub ($os == l4re*): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (aarch64-unknown-l4re, x86_64-pc-l4re): New
entries.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2024-03-02
* AcSplit/CF_MB_LEN_MAX: ci-aclocal generated header
2024-02-24
* config.sub: 2024-02-24
From: Jonathan Schwender <jonathan.schwender@huawei.com>
config.sub: Recognize OpenHarmony OS
Add support for [OpenHarmony] targets.
The `ohos` targets are available [in `LLVM`] and
documented in the [Rust target documentation].
Known targets in Rust:
- aarch64-unknown-linux-ohos
- armv7-unknown-linux-ohos
- x86_64-unknown-linux-ohos
Known targets in Clang:
- arm-linux-ohos
- aarch64-linux-ohos
- x86_64-linux-ohos
There are also some additional targets available in clang,
e.g. `liteos-ohos`, but I don't know much about those
targets, so I'm leaving `liteos-ohos` out of scope for this
patch.
[OpenHarmony]: https://gitee.com/openharmony/docs/
[in `LLVM`]: https://github.com/llvm/llvm-project/blob/main/clang/lib/Driver/ToolChains/OHOS.cpp
[Rust target documentation]: https://doc.rust-lang.org/rustc/platform-support/openharmony.html
* config.sub ($os == ohos*): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (aarch64-linux-ohos): New entry.
Signed-off-by: Jonathan Schwender <jonathan.schwender@huawei.com>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2024-01-25
* config.sub: 2024-01-26
From: Mike Hommey <mh@glandium.org>
config.sub: recognise ARM64E machine type
ARM64E is a custom ABI for AArch64 8.3+ introducing pointer
authentication codes. While technically just an ABI, it is treated
as its own machine type, with triples in the format arm64e-*.
* config.sub (arm64e): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (arm64e-apple-darwin20.0.0,
arm64e-apple-ios): New entries.
Signed-off-by: Mike Hommey <mh@glandium.org>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2024-01-22
* AcSplit/CF_SRAND: s/arc2random/arc4random/ (fix typo)
* AcSplit/CF_SRAND: drop MirOS-specific "arc4random_push()"
2024-01-21
* AcSplit/CF_ADD_LDFLAGS: ci-aclocal generated header
2024-01-07
* AcSplit/CF_FUNC_GETPWUID: ci-aclocal generated header
* AcSplit/CF_COLOR_CURSES, AcSplit/CF_NCURSES_BROKEN:
fix unused-variable warnings
* AcSplit/CF_SLANG_UNIX_DEFS, AcSplit/CF_XOPEN_CURSES:
prefer "#error" with standard C
* AcSplit/CF_SRAND: reduce conversion warnings
* AcSplit/CF_ANSI_CC_CHECK: fix a nested-extern warning
* AcSplit/CF_BOOL_DEFS, AcSplit/CF_NGROUPS:
fix unused-variable warnings
* AcSplit/CF_CURSES_TERM_H, AcSplit/CF_NCURSES_BROKEN,
AcSplit/CF_NCURSES_VERSION:
prefer "#error" with standard C
2024-01-02
* AcSplit/CF_FUNC_SETGROUPS: ci-aclocal generated header
2024-01-01
* AcSplit/CF_NL_LANGINFO_1STDAY: fix compiler warning
2023-12-27
* config.sub: 2023-12-02
From: Aleksandar Rikalo <aleksandar.rikalo@syrmia.com>
config.sub: recognize nanoMIPS CPUs
* config.sub: Recognize nanoMIPS CPUs.
* testsuite/config-sub.data (nanomips): New entry.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-12-17
* AcSplit/CF_CHECK_1_DECL, AcSplit/CF_CHECK_NESTED_PARAMS,
AcSplit/CF_COMPTYPE, AcSplit/CF_CPP_CONCATS, AcSplit/CF_CPP_EXPANDS,
AcSplit/CF_FUNC_SYSTEM, AcSplit/CF_LOCKFILE, AcSplit/CF_PW_GECOS,
AcSplit/CF_SYS_SELECT_TIMEVAL:
reduce compiler warnings
2023-12-13
* AcSplit/CF_CC_INIT_UNIONS, AcSplit/CF_STDIO_UNLOCKED:
fix unused-variable warning
* AcSplit/CF_XOPEN_CURSES: fix writable-strings warning
2023-12-10
* AcSplit/CF_SRAND:
add check for -lbsd needed with Debian to link arc4random
add check for limits.h needed to define INT_MAX
add default includes (and limits.h) to improve configure checks
* AcSplit/CF_MAWK_RUN_FPE_TESTS: fix unused-variable warning
2023-12-09
* AcSplit/CF_HEADER_RESOURCE: fix an incorrectly-placed parenthesis
* AcSplit/CF_CURSES_TERMCAP: quiet a const-mismatch warning
* AcSplit/CF_KILLPG, AcSplit/CF_RESTARTABLE_PIPEREAD,
AcSplit/CF_TYPE_OUTCHAR:
reduce compiler warnings
2023-12-03
* AcSplit/CF_POSIX_SAVED_IDS: amend test to reduce compiler warnings
* AcSplit/CF_STRUCT_TERMIOS: amend fix to split up two assignments
* AcSplit/CF_LARGEFILE:
reduce compiler warnings in configure check, noting that the "transient"
feature referred to is handled via the __REDIRECT macro.
* AcSplit/CF_FUNC_VSSCANF:
reduce compiler warnings in configure-check by ensuring the va_list variable
is initialized.
* AcSplit/CF_CHECK_GETENV:
sprintf parameter for %p should be void*, per gcc warning
* AcSplit/CF_CHECK_WCWIDTH_GRAPHICS:
fix signed/unsigned warning for sscanf()
* AcSplit/CF_LINK_DATAONLY: reduce no-previous-prototype warnings
* AcSplit/CF_UNSIGNED_LITERALS:
reduce unused-variable warnings in configure check
* AcSplit/CF_LINK_DATAONLY, AcSplit/CF_WITH_VERSIONED_SYMS:
reduce warnings for no previous prototype
* AcSplit/CF_LINK_FUNCS: reduce writable-string warnings
* AcSplit/CF_BOOL_DECL, AcSplit/CF_CHECK_WCHAR_H,
AcSplit/CF_FUNC_OPENPTY, AcSplit/CF_ISASCII, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_STRUCT_SIGACTION, AcSplit/CF_STRUCT_TERMIOS,
AcSplit/CF_TYPE_SIGACTION, AcSplit/CF_WCHAR_TYPE:
reduce unused-variable warnings in configure check
2023-12-02
* config.guess: 2023-10-19
From: streaksu <streaksu@protonmail.com>
config.guess: recognize ironclad OS
* config.guess (*:Ironclad:*:*): Recognize.
* doc/config.guess.1: Regenerate.
* testsuite/config-guess.data (x86_64-unknown-ironclad): New entry.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2023-12-02
From: streaksu <streaksu@protonmail.com>
config.sub: recognize ironclad OS
* config.sub ($os == ironclad*): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (x86_64-ironclad): New entry.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* AcSplit/CF_WITH_DESKTOP_CATEGORY: spelling fix
2023-12-01
* AcSplit/CF_FUNC_STRFTIME: improve call on AC_DEFINE
* AcSplit/CF_FUNC_STRFTIME: ci-aclocal generated header
* AcSplit/CF_XKB_BELL_EXT:
correct parameter type for XkbBell() in configure check (patch by Florian
Weimer, Redhat #2251945)
* AcSplit/CF_LASTLOG, AcSplit/CF_MKSTEMP, AcSplit/CF_TTY_GROUP:
fix a const-string warning
* AcSplit/CF_CONST_X_STRING: fix a redefinition warning
* AcSplit/CF_FUNC_TGETENT, AcSplit/CF_POSIX_SAVED_IDS, AcSplit/CF_SVR4:
use "#error"
2023-11-25
* AcSplit/CF_WITH_TYPE: fix quoting
* AcSplit/CF_WITH_TYPE: ci-aclocal generated header
2023-11-23
* AcSplit/CF_TTY_GROUP, AcSplit/CF_WITH_ICON_NAME,
AcSplit/CF_WITH_ICON_SYMLINK, AcSplit/CF_WITH_ICON_THEME,
AcSplit/CF_WITH_MAN2HTML, AcSplit/CF_WITH_XPM:
modify usage message, bracketing optional values
2023-11-22
* AcSplit/CF_WITH_ADA_LIBNAME, AcSplit/CF_WITH_EXPORT_SYMS,
AcSplit/CF_WITH_LIBTOOL_OPTS, AcSplit/CF_WITH_LIB_BASENAME,
AcSplit/CF_WITH_PKG_CONFIG_LIBDIR, AcSplit/CF_WITH_VERSIONED_SYMS,
AcSplit/CF_WITH_SYMLINK, AcSplit/CF_WITH_LOCALE_ALIAS,
AcSplit/CF_WITH_ENCODINGS_DIR:
modify usage message, bracketing optional values
2023-10-30
* config.sub: 2023-10-29
From: Adam Joseph <adam@westernsemico.com>
config.sub: recognize *-*-uefi
This commit causes gnu-config to recognize the *-*-uefi triples.
These triples describe binaries which use the PE object file format
and UEFI calling "convention" (which is actually a family of
completely unrelated calling conventions, one per CPU architecture):
- https://uefi.org/specs/UEFI/2.9_A/02_Overview.html#detailed-calling-convention
- https://uefi.org/specs/UEFI/2.9_A/02_Overview.html#detailed-calling-convention-1
- https://uefi.org/specs/UEFI/2.9_A/02_Overview.html#detailed-calling-convention-2
- https://wiki.osdev.org/UEFI#Calling_Conventions
This is in contrast to the *-w64-mingw32 triple of GNU-EFI, which
uses the ELF format and cdecl calling convention:
- https://wiki.osdev.org/UEFI#Calling_Conventions
Because *-*-uefi uses a different object file format (everywhere)
and calling convention (everywhere except x86_32) from
*-w64-mingw32, the resulting binaries cannot be linked against each
other. This is the primary justification for introducing a new
triple.
Since gnu-config does not yet have triples for these builds, we
might as well use the same ones already selected by LLVM. It
considers uefi to be an operating system:
- https://reviews.llvm.org/D131594
* config.sub (*-*-uefi): Recognize.
* testsuite/config-sub.data (i686-unknown-uefi, x86_64-unknown-uefi,
aarch64-unknown-uefi): New entries.
Signed-off-by: Adam Joseph <adam@westernsemico.com>
X-Disclaimer: This commit shall not be construed as the author's approval of the UEFI boondoggle in any way, shape or form.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-10-28
* config.sub: 2023-10-28
From: John Ericson <list@JohnEricson.me>
Recognize the VideoCore 4 processor
This is supported in this GCC fork. [1] It has been packaged in Nixpkgs
since 2019. [2]
Their change to `config.sub` [3] is similar, but also includes a `vc4`
-> `vc4-unknown-elf` shorthand; I don't see any reasons to add any new
legacy 1-component short-hands, so I just skipped that part.
[1]: https://github.com/itszor/vc4-toolchain
[2]: https://github.com/NixOS/nixpkgs/pull/72657
[3]: https://github.com/itszor/gcc-vc4/commit/91278afabe61808c0b02078c078e758d69f94aff
* config.sub (vc4-*): Recognize.
* testsuite/config-sub.data (vc4-unknown-none-elf): New entry.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
modify --with-pkg-config-libdir option to make "libdir" the default,
as documented, rather than "auto" (report by Branden Robinson).
* AcSplit/CF_ETIP_DEFINES, AcSplit/CF_PKG_CONFIG,
AcSplit/CF_SHARED_OPTS, AcSplit/CF_WITH_ADA_COMPILER,
AcSplit/CF_WITH_ADA_INCLUDE, AcSplit/CF_WITH_ADA_LIBNAME,
AcSplit/CF_WITH_ADA_OBJECTS, AcSplit/CF_WITH_ADA_SHAREDLIB,
AcSplit/CF_WITH_EXPORT_SYMS, AcSplit/CF_WITH_LIBTOOL_OPTS,
AcSplit/CF_WITH_VERSIONED_SYMS, AcSplit/CF_WITH_X11_RGB:
improve usage-messages (patch by Branden Robinson)
2023-10-21
* AcSplit/CF_ABI_DEFAULTS:
add cf_dft_widec and cf_dft_opaque_curses for changing defaults of the
configure options for enabling widec and opaqueness.
2023-10-19
* config.guess: 2023-10-19
From: Alfred Persson Forsberg <cat@catcream.org>
config.guess: detect LLVM-libc via features.h
* config.guess (__LLVM_LIBC__): Recognize.
* doc/config.guess.1: Regenerate.
Signed-off-by: Alfred Persson Forsberg <cat@catcream.org>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2023-10-19
From: Alfred Persson Forsberg <cat@catcream.org>
config.sub: add LLVM-libc support
* config.sub (*-linux-llvm): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data: Add x86_64-linux-llvm.
Signed-off-by: Alfred Persson Forsberg <cat@catcream.org>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-10-02 Sven.Joachim
* AcSplit/CF_WITH_WTMP_PATH: fix typo
2023-09-15
* config.sub: 2023-09-15
From: Billy Laws <blaws05@gmail.com>
config.sub: recognise ARM64EC machine type
ARM64EC is a custom ABI for AArch64 that allows for interoperability
with x86_64 compiled code. While technically just an ABI, it is treated
as its own machine type, with triples in the format arm64ec-*.
* config.sub (arm64ec): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (arm64ec-pc-mingw32, arm64ec-windows): New
entries.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-09-06
* AcSplit/CF_XOPEN_SOURCE:
add pattern for uClibc-ng to CF_XOPEN_SOURCE (report/patch by Waldemar Brodkorb)
2023-09-04
* AcSplit/CF_WITH_UTMP_PATH, AcSplit/CF_WITH_WTMP_PATH: New file.
2023-08-23
* config.sub: 2023-08-23
From: Urs Janßen <urs@tin.org>
config.sub: allow aarch64c-unknown-freebsd
config.guess says aarch64c-unknown-freebsd14.0 (cfarm240.cfarm.net,
an Arm Morello SoC, quad-core aarch64 Neoverse N1-based CPU
implementing CHERI), so let config.sub allow it.
* config.sub (aarch64c): Recognize.
* doc/config.sub.1: Regenerate.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-08-22
* config.guess: 2023-08-22
From: Dmitry V. Levin <ldv@altlinux.org>
config.guess: invoke "uname -p" from PATH for non-arm FreeBSD
Starting with commit afe1fa96bf32, "uname -p" from PATH is invoked in
case of FreeBSD on arm, while in other FreeBSD cases it was invoked
using a full pathname as "/usr/bin/uname -p". Fix this inconsistency
and invoke "uname -p" from PATH for all FreeBSD cases. This also allows
to test non-arm FreeBSD cases.
* config.guess (*:FreeBSD:*:*): Invoke "uname -p" from PATH.
* doc/config.guess.1: Regenerate.
* testsuite/config-guess.data (x86_64-unknown-freebsd5.2,
i586-unknown-freebsd7.0): Reintroduce the tests removed by commit
68873f3c11c6.
2023-08-17
* config.guess: 2023-08-17
From: Bruno Haible <bruno@clisp.org>
config.guess: Detect Android (as opposed to GNU/Linux)
Here's a patch to recognize Android environments.
Such environments are "apps" with POSIX-like tools. Today, the most frequently
used one is Termux [1][2][3]; on devices with Android versions before 5.0
one can use Terminal-IDE [4][5].
config.sub already supports this environment:
$ sh config.sub armv7l-linux-androideabi
armv7l-unknown-linux-androideabi
I've built many GNU packages in this environment, with the following recipe:
CONFIG_SHELL=$PREFIX/bin/sh; export CONFIG_SHELL
CC="clang -ferror-limit=0" CXX="clang++ -ferror-limit=0"; export CC CXX
./configure --host=armv7l-linux-androideabi --prefix=$HOME/local
The Termux people have compiled or ported more than 1000 packages as well [6].
But the requirement to pass the --host parameter each time is an annoyance.
Without it, based only on the results of uname, config.guess guesses
$ sh config.guess
armv7l-unknown-linux-gnueabi
and many configuration results are wrong (because Android has many functions
in libc without declaring them in the .h files, depending on the so-called
"Android API level"), leading to many compilation errors.
With the attached patch, it produces
$ sh config.guess
armv7l-unknown-linux-androideabi
The patch does not include an addition to the config.guess test suite, since
the uname values are:
$ uname -m
armv7l
$ uname -r
4.19.127
$ uname -s
Linux
$ uname -v
#1 SMP PREEMPT Tue Apr 4 16:54:58 IST 2023
$ uname -p
unknown
which maps to armv7l-unknown-linux-gnueabi.
[1] https://github.com/termux/termux-app
[2] https://f-droid.org/en/packages/com.termux/
[3] https://wiki.termux.com/wiki/Main_Page
[4] https://en.wikibooks.org/wiki/Android/Terminal_IDE
[5] http://www.spartacusrex.com/terminalide.htm
[6] https://github.com/termux/termux-packages/tree/master/packages
* config.guess (Linux|GNU|GNU/*): Detect Android.
* doc/config.guess.1: Regenerate.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-08-16
* config.sub: 2023-08-16
From: Adam Joseph <adam@westernsemico.com>
config.sub: add javascript-*-ghcjs
GHC has been using a custom triple "javascript-unknown-ghcjs" for
their (non asm.js, non wasm) javascript-emitting kernel-less target.
This triple is a bit of an oddball, so the support for it is highly
restricted in order to discourage further proliferation of the
javascript "cpu" or ghcjs "operating system", which are valid only
in combination with each other.
* config.sub (javascript-*-ghcjs): Allow.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (javascript-ghcjs,
javascript-unknown-ghcjs): New entries.
Link: https://gitlab.haskell.org/ghc/ghc/-/commit/6636b670233522f01d002c9b97827d00289dbf5c
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-08-07
* config.sub: 2023-08-07
From: John Ericson <list@JohnEricson.me>
config.sub: Systematize parsing of machine code formats
Instead of treating them as OSes, we treat them as their own category.
This is modeled on what LLVM does with its `ObjectFormatType` enum [1],
advancing my long-running project of trying to nudge GNU config and LLVM
towards each other, taking the best ideas of both.
Currently, my emphasis is just on code cleanup. There are just a few
tests for newly supported changes that fall out of this. But down the
road, this also opens the door to parsing configs with more than 4
components, like [2].
[1]: https://llvm.org/doxygen/classllvm_1_1Triple.html#a83e907e55fa50e093caa96a0aff96201
[2]: https://github.com/llvm/llvm-project/blob/a18266473be1439d324059afa0e8b124f0466428/llvm/unittests/TargetParser/TripleTest.cpp#L1873C50-L1873C77
added in
https://github.com/llvm/llvm-project/commit/28b82bc39ef076527c07718724913f70b5d60b69
* config.sub: Save machine code format name separately from the OS name.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (arm-unknown-none-aout,
arm-unknown-none-pe): New entries.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2023-07-31
* config.sub: 2023-07-31
From: Maciej W. Rozycki <macro@orcam.me.uk>
config.sub: Handle arbitrary MIPS CPU names
GNU binutils support the selection of the default MIPS subtarget via the
configuration triplet, e.g. `mips64octeon+el-unknown-linux-gnu' builds a
Linux/GNU 64-bit MIPS (n32 ABI) little-endian configuration with the CPU
set to Octeon+ by default. However `config.sub' rejects such a triplet
and indeed it only lets through a random choice of ones people submitted
changes for to support.
There is a large number of MIPS CPU configurations, 118 at the moment,
that GNU binutils know, so rather than adding them individually and then
hoping it will be kept up to date from now on accept any `mips*' pattern
for the machine part, just as we already do for a few of other targets.
* config.sub: Allow any `mips*' CPU rather than listing a choice
individually.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data: Add test cases.
2023-07-28 Sven.Joachim
* AcSplit/CF_MAN_PAGES:
add assignment to provide default value for TERMINFO_DIRS in ncurses manpages
2023-07-20
* config.guess: 2023-07-20
From: Dmitry V. Levin <ldv@altlinux.org>
config.guess (aarch64:Linux:*:*): Detect 32-bit ABI
On x86_64 Linux, a C compiler producing 32bit code causes the guessed
CPU to be changed to i686.
Adapt this approach for aarch64 Linux, i.e. change the guessed CPU
to armv8l when a compiler producing arm eabi code is detected.
* config.guess (aarch64:Linux:*:*): Test for the 32-bit ABI.
* doc/config.guess.1: Regenerate.
Suggested-by: David Allsopp <david.allsopp@metastack.com>
2023-07-19
* config.guess, config.sub: 2023-07-19
From: Paul Iannetta <piannetta@kalray.eu>
Add KVX MPPA detection
* config.guess (kvx:Linux:*:*, kvx:cos:*:*, kvx:mbr:*:*): Recognize.
* config.sub (kvx): Recognize cpu.
(cos*, mbr*): Recognize os.
* testsuite/config-guess.data: Add tests for kvx targets.
* testsuite/config-sub.data: Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
2023-07-13
* config.sub: 2023-07-13
From: Luis Campos <luisc@jfrog.com>
config.sub: Add Apple tvOS and watchOS
These are distinct from "ios". They are also technically Darwin, so while
something like "aarch64-apple-darwin" could be used when targeting these,
on Apple-silicon based systems there could be cases where `--host` and
`--build` have the same value, and a ./configure script may determine that
we are not cross building, causing it to try to execute test programs for
the target OS that will not run on macOS.
These are functionally equivalent to iOS, and targets with "-tvos" and
"-watchos" are already used by clang.
* config.sub (tvos*, watchos*): Recognize.
* testsuite/config-sub.data (arm64-apple-tvos, arm64-apple-tvos10.0.0,
arm64-apple-watchos, arm64-apple-watchos5.0): New tests.
* doc/config.sub.1: Regenerate.
2023-07-04
* config.sub: 2023-07-04
From: John Ericson <list@JohnEricson.me>
config.sub: Accept $cpu-$vendor-none-{coff,elf}
These are not real OSes, they are object file formats. There is a
longstanding tradition of using them for embedded/freestanding
programming, so it makes sense to parse them with `kernel=none`.
(I have a WIP future patch that systematizes parsing these non-OSes a
bit more. That also opens the door to parsing a 5th component as LLVM
can do.)
This change unblocks an issue we've been having with Nixpkgs (see
https://github.com/NixOS/nixpkgs/issues/165836 for the longer version).
* config.sub (none-coff*, none-elf*): Recognize.
* testsuite/config-sub.data (arm-unknown-none-coff,
arm-unknown-none-elf, riscv64-company-none-elf): New tests.
* doc/config.sub.1: Regenerate.
2023-06-26
* config.sub: 2023-06-26
From: John Ericson <list@JohnEricson.me>
config.sub: Accept LLVM-style $cpu-$vendor-windows-{gnu,msvc}
In older times, MinGW (GCC toolchain with modified windows headers) was
the only free software toolchain for Windows. But now, LLVM has support
both for MinGW ABI and Microsoft's own. The distinction matters for C++
more than C.
LLVM[1], Rust[2], and other projects have taken to differentiating these
two as `...windows-gnu` vs `...windows-msvc`. I think that makes a lot
of sense, as it correctly identifiers both their commonalities and their
differences.
A lot of MinGW-supporting software, most notably GCC itself, will
presumably continue to use configs like x86_64-pc-mingw32 and
i686-pc-mingw32. That's fine; this patch doesn't normalize them away
(like LLVM does) or remove them! If and when that software wants to
support the MSVC ABI without requiring MSVC itself, they can switch to
these newer configurations.
[1]: https://github.com/llvm/llvm-project/blob/a18266473be1439d324059afa0e8b124f0466428/llvm/unittests/TargetParser/TripleTest.cpp#L1907-L1951
[2]: https://github.com/rust-lang/rust/blob/36fb58e433c782e27dd41034284e157cf86d587f/compiler/rustc_target/src/spec/mod.rs#L1255-L1271
2023-06-23
* config.guess, config.sub: 2023-06-23
From: Paul Eggert <eggert@cs.ucla.edu>
Quote 'like this', not `like this'.
In 2012 the GNU Coding Standards changed to recommend quoting
'like this' or "like this" instead of `like this'.
Alter diagnostics and comments accordingly.
Use a more-consistent quoting style in config.sub diagnostics,
preferring 'like this' to "like this" as the former is more
resistant to shell metacharacters.
2023-06-03
* AcSplit/CF_INSTALL_OPTS: add CF_INSTALL_OPT_STRIP_PROG
* AcSplit/CF_INSTALL_OPT_STRIP_PROG: ci-aclocal generated header
* AcSplit/CF_LIB_RULES:
in distclean, remove "*.tmp" to catch file created by CF_INSTALL_OPT_STRIP_PROG
2023-05-21
* AcSplit/CF_INSTALL_MAN:
for symlink, use relative form for both source/target
* AcSplit/CF_INSTALL_MAN: add dependency on AC_LN_S
* AcSplit/CF_INSTALL_MAN:
handle symbolic link needed for reflex, e.g., "reflex++ -> reflex"
* AcSplit/CF_INSTALL_MAN:
provide sed commands for handling --program-transform-name, in byacc.
2023-05-19
* AcSplit/CF_INSTALL_MAN: ci-aclocal generated header
* AcSplit/CF_MANPAGE_FORMAT: provide substitutions for CF_INSTALL_MAN
2023-05-13
* AcSplit/CF_MAN_PAGES:
account for extra-suffix when deriving aliases from the manpage target
2023-05-06
* AcSplit/CF_WITH_REL_VERSION: make minor default to zero if not given
* AcSplit/CF_NUMBER_SYNTAX: improve check for positive-integer
2023-04-22
* COPYING.asc: sign
* COPYING: bump
* AcSplit/CF_LIB_RULES, AcSplit/CF_SRC_MODULES:
drop support for tack 1.07 (after 6 years)
2023-04-03
* AcSplit/CF_XOPEN_SOURCE:
amend default case to avoid undefining _XOPEN_SOURCE if _POSIX_C_SOURCE is
defined (report by A. Wilcox)
2023-03-05
* AcSplit/CF_GCC_VERSION, AcSplit/CF_GXX_VERSION:
improve sed expression, works with MinGW
* AcSplit/CF_XOPEN_SOURCE:
provide for MinGW(32|64) to use _DEFAULT_SOURCE
2023-02-26
* AcSplit/CF_TYPE_NFDS_T: ci-aclocal generated header
2023-02-25
* AcSplit/CF_FUNC_GETTIME: add check for clock_gettime()
2023-02-22
* AcSplit/CF_PROG_GPERF: ci-aclocal generated header
2023-02-20
* AcSplit/CF_CC_ENV_FLAGS:
change check for $cf_flags to eliminate $AWK, to make it simpler to port
both byacc and reflex
2023-02-18
* AcSplit/CF_NCURSES_CC_CHECK: use #error
* AcSplit/CF_FUNC_GETTIME: ci-aclocal generated header
* AcSplit/CF_XOPEN_SOURCE:
merge case for msys (i.e., msys2) into linux* for _DEFAULT_SOURCE
use #error
* AcSplit/CF_CLANG_COMPILER, AcSplit/CF_HASHED_DB_VERSION,
AcSplit/CF_INTEL_COMPILER, AcSplit/CF_POSIX_C_SOURCE,
AcSplit/CF_PREDEFINE, AcSplit/CF_SIGWINCH,
AcSplit/CF__XOPEN_SOURCE_BODY:
use #error
* AcSplit/CF_BOOL_SIZE, AcSplit/CF_C11_NORETURN, AcSplit/CF_CGETENT,
AcSplit/CF_CHECK_ENVIRON, AcSplit/CF_CHECK_ERRNO,
AcSplit/CF_CHECK_WCHAR_H, AcSplit/CF__XOPEN_SOURCE_HEAD:
use $ac_includes_default
2023-02-15
* AcSplit/CF_SRAND:
use $ac_includes_default in second compile-check, for better consistency
2023-02-10
* AcSplit/CF_FUNC_TGETENT:
in checking for full/partial tgetent, allow for ncursesw as well as the
corresponding tinfo/tinfow configurations.
2023-01-29
* AcSplit/CF_KILLPG: fix typo in cross-compile check (Debian #1029956)
2023-01-25
* AcSplit/CF_SETITIMER: ci-aclocal generated header
2023-01-24
* AcSplit/CF_SLANG_CPPFLAGS: corrected variable used in compile-check
* AcSplit/CF_CHECK_FUNCDECL:
reduce implicit-declaration warnings in configure-check
2023-01-22
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
work around very old pkg-config (Solaris 10, with 0.15.0) using PKG_CONFIG_PATH
(which it recognizes, to override PKG_CONFIG_LIBDIR which it does not).
2023-01-21
* AcSplit/CF_MAN_PAGES:
move calculation that ought to be in CF_MANPAGE_FORMAT there, and improve
naming convention for its calculated values.
* AcSplit/CF_MANPAGE_FORMAT:
move case-statement from CF_MAN_PAGES here, to make this more reusable.
* AcSplit/CF_MANPAGE_SYMLINKS: trim extra '$'
* AcSplit/CF_MAN_PAGES:
adjust check for aliases to work in out-of-tree builds (report by Sven Joachim)
* config.sub: 2023-01-21
From: Paul Eggert <eggert@cs.ucla.edu>
Fix config.sub spelling typo for "athlon"
This typo was introduced in commit
fc14426403b6106981d2b3a4666835df29203043
dated Tue Aug 28 09:59:08 2018 +1000.
2023-01-19
* AcSplit/CF_FIND_TDLIB: provide for finding td_lib using pkg-config
2023-01-16
* AcSplit/CF_HAVE_FUNCS: add semicolon, fix build
* AcSplit/CF_WITH_NCURSES_ETC: add cf_cv_screen to cache/substitutions
2023-01-15
* AcSplit/CF_WITH_PERL: add cast to quiet compiler-warning
* AcSplit/CF_MISSING_CHECK:
corrected prototype for XtToolkitInitialize()
2023-01-14
* AcSplit/CF_CRYPT_FUNC: appease clang's check for null-parameters
* AcSplit/CF_FP_ISREADY, AcSplit/CF_WCTYPE: fix unused variable warning
* AcSplit/CF_CHECK_ENVIRON: fix unused-variable warning
* AcSplit/CF_WITH_PERL: improve test-stub.
* AcSplit/CF_KILLPG: use prototype for handler
* AcSplit/CF_MISSING_CHECK: use prototype in test-declaration
* AcSplit/CF_TERMCAP_LIBS:
fix inconsistency between declaration and use.
2023-01-11
* AcSplit/CF_HAVE_FUNCS, AcSplit/CF_RE_COMP_FUNCS:
quiet implicit-declarations warning
* AcSplit/AM_LANGINFO_CODESET, AcSplit/CF_FUNC_LSTAT, AcSplit/CF_LOCALE,
AcSplit/CF_UTF8_LIB, AcSplit/CF_XOPEN_CURSES, AcSplit/CF_X_ATHENA,
AcSplit/CF_X_ATHENA_LIBS, AcSplit/CF_X_TOOLKIT:
use $ac_includes_default
2023-01-10
* AcSplit/CF_BUNDLED_INTL:
fix an inconsistency between $USE_INCLUDED_LIBINTL and $nls_cv_use_gnu_gettext
which caused build of tin to fail when the host had no libintl.h
2023-01-09
* AcSplit/CF_CURSES_TERMCAP: work around implicit-declaration warnings
* AcSplit/CF_CRYPT_FUNC:
reduce compiler warnings using $ac_includes_default plus fallback prototype
2023-01-08
* AcSplit/CF_FUNC_TGETENT:
use the same include/fallback in the test for partial support of tgetent,
to avoid implicit-declaration in that. remove redundant check for termcap.h
2023-01-07
* AcSplit/CF_ABI_DEFAULTS, AcSplit/CF_LIB_RULES,
AcSplit/CF_NCURSES_ABI_6, AcSplit/CF_SUBST_NCURSES_VERSION,
AcSplit/CF_WITH_ABI_VERSION:
added --with-abi-altered, using new symbol $cf_cv_abi_version for interim
* AcSplit/CF_WITH_ABI_ALTERED: ci-aclocal generated header
2023-01-05
* AcSplit/CF_POSIX_WAIT: use $ac_includes_default
* AcSplit/CF_TD_SIG_ARGS: use prototype for main
* AcSplit/CF_REGCMP_LIBS, AcSplit/CF_POSIX_VDISABLE:
use $ac_includes_default
* AcSplit/CF_ANSI_CPP, AcSplit/CF_WIDEC_SHIFT, AcSplit/CF_RE_COMP_FUNCS:
use prototype for main
* AcSplit/CF_REGEXPR_H_FUNCS, AcSplit/CF_REGEX_H_FUNCS,
AcSplit/CF_NCURSES_VERSION, AcSplit/CF_MIN_GETCCHAR, AcSplit/CF_KILLPG:
use $ac_includes_default
* AcSplit/CF_FUNC_TGETENT:
give tgetent a prototype, to help with bogus compiler warnings
* AcSplit/CF_FUNC_POLL, AcSplit/CF_FUNC_NANOSLEEP,
AcSplit/CF_FUNC_MEMMOVE, AcSplit/CF_FUNC_GETTTYNAM,
AcSplit/CF_FUNC_GETADDRINFO, AcSplit/CF_FUNC_CURSES_VERSION,
AcSplit/CF_FOPEN_BIN_R:
use $ac_includes_default
* AcSplit/CF_CRYPT_FUNC:
use $ac_includes_default, which _may_ provide a prototype for crypt(),
and it may/may not agree with this one (ymmv)
* AcSplit/CF_CHECK_WCWIDTH_GRAPHICS: use $ac_includes_default
* AcSplit/CF_BROKEN_TGETENT_STATUS:
give tgetent a prototype, to help with bogus compiler warnings
* AcSplit/CF_BOOL_SIZE, AcSplit/CF__CHECK_RUN, AcSplit/CF_TYPEOF_CHTYPE,
AcSplit/CF_TTY_GROUP, AcSplit/CF_SET_ERRNO, AcSplit/CF_STRUCT_LASTLOG,
AcSplit/CF_RESTARTABLE_PIPEREAD, AcSplit/CF_POSIX_SAVED_IDS,
AcSplit/CF_MKSTEMP, AcSplit/CF_MAWK_FIND_MAX_INT,
AcSplit/CF_LINK_FUNCS, AcSplit/CF_CHECK_GETENV:
use $ac_includes_default
2023-01-04
* AcSplit/CF_RESTARTABLE_PIPEREAD:
include <sys/wait.h> if it was found, to prototype wait()
Redhat #2158163
2023-01-01
* config.guess, config.sub: 2023-01-01
From: Dmitry V. Levin <ldv@altlinux.org>
Update copyright years
* config.guess: Update copyright years.
* config.sub: Likewise.
* testsuite/config-guess.sh: Likewise.
* testsuite/config-sub.sh: Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
2022-12-31
* AcSplit/CF_FIND_TDLIB:
add explicit check for "td" subdir of library-directory
* AcSplit/CF_FIND_TDLIB:
correct sed command for configuration without subdirs, which omitted a '/'
quote result $cf_td_lib_rules in test, in case it is empty
* AcSplit/CF_ALL_MAKEFILES: fix misspelled cache-variable name
2022-12-29
* AcSplit/CF_XOPEN_SOURCE:
amend the test for conflict between _XOPEN_SOURCE and _POSIX_C_SOURCE to
allow for the former to be undefined because unneeded.
2022-12-21
* AcSplit/CF_CURSES_TERMCAP: quiet an unused-variable warning
* AcSplit/CF_TERMCAP_LIBS: quiet an implicit-function warning
* AcSplit/CF_PROG_YACC:
reduce prototype-related warnings by prototyping yylex/yyerror with the most
common signature used in byacc/bison, which works with heirloom yacc.
* AcSplit/CF_TIOCGWINSZ, AcSplit/CF_SIG_CONST: use prototype for main()
* AcSplit/CF_FUNC_SYSTEM:
fixes for implicit function (adapted from patch by Florian Weimer)
* AcSplit/CF_FUNC_SYSTEM: use prototype for main()
* AcSplit/CF_FUNC_FORK:
fixes for implicit function (adapted from patch by Florian Weimer)
* AcSplit/CF_FUNC_FORK: use $ac_includes_default for consistency
* AcSplit/CF_COREFILE:
fixes for implicit function (adapted from patch by Florian Weimer)
* AcSplit/CF_COREFILE: use $ac_includes_default, for consistency
* AcSplit/CF_COREFILE, AcSplit/CF_FUNC_FORK, AcSplit/CF_FUNC_SYSTEM:
compiler-warning fixes from Florian Weimer forwarded by Urs
2022-12-06
* AcSplit/CF_FCNTL_VS_IOCTL:
include <sys/ioctl.h>, for systems which may define FIONREAD there (Redhat #2151353)
2022-12-04
* AcSplit/CF__INET_HEAD:
use $ac_includes_default to quiet implicit-function warnings with MinGW
* AcSplit/CF_BUILD_CC:
workaround for Lynx, which sets BUILD_CPPFLAGS to "#", commenting out part
of a makefile assignment.
2022-12-02
* AcSplit/CF_CURSES_HEADER, AcSplit/CF_CURSES_LIBS:
drop tgoto() from the sample code used for checking if a given header declares
curses functions, since the prototype for it in X/Open lives in term.h (tested
later in configure scripts, to resolve <term.h> vs <ncurses/term.h>, etc).
* AcSplit/CF_TERMCAP_LIBS:
make the try-link checks less likely to generate compiler warnings by
providing a suitable prototype for tgoto() and tigetstr(). The former
had no header-file in BSD curses, while both (for which there are other
configure checks) have prototypes in term.h (AT&T / X/Open / ncurses).
Since we only need to check for existence, using our own prototype is okay.
* AcSplit/CF_FUNC_SIGACTION: fix a compiler-warning
* AcSplit/CF_TERMIO_AND_CURSES:
fix one of the Lynx-specific issues using an optional parameter
* AcSplit/CF_TERMIO_AND_TERMIOS:
include stdio.h, to prototype putchar (report by Florian Weimer)
* AcSplit/CF_SRAND: prefer $ac_includes_default to just stdlib.h
* AcSplit/CF_GCC_ATTRIBUTES:
include stdio.h to help with getting proper attributes for printflike
2022-11-17
* AcSplit/CF_GCC_OPT_RDYNAMIC:
some configurations of clang support -rdynamic; allow but still warn.
2022-11-13
* AcSplit/AC_GNU_SOURCE, AcSplit/CF_CYGPATH_W:
ci-aclocal generated header
2022-11-05
* AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MAN_PAGES:
allow for bzip2 and xz compression in manpages
2022-10-23
* AcSplit/CF_MANPAGE_RENAMES: ignore if $MANPAGE_RENAMES is disabled
* AcSplit/CF_MAN_PAGES: defer existence-check for manpage-renames
2022-10-22
* AcSplit/CF_MANPAGE_RENAMES, AcSplit/CF_OUTPUT_MANPAGE_RENAMES:
build-fix
* AcSplit/CF_OUTPUT_MANPAGE_RENAMES: ci-aclocal generated header
* AcSplit/CF_MANPAGE_RENAMES: split-out CF_OUTPUT_MANPAGE_RENAMES
* AcSplit/CF_MANPAGE_RENAMES:
change man_db.renames to template, i.e., "man_db.renames.in", which is not
used directly during configuration.
2022-10-02
* AcSplit/CF_XOPEN_SOURCE:
add these GNU libc suffixes (report by Sven Joachim):
abi64
abin32
x32
2022-10-01
* AcSplit/CF_MANPAGE_RENAMES: make $FGREP configurable
2022-09-30
* AcSplit/CF_XOPEN_SOURCE:
add variants "gnueabi" and "gnueabihf" to get _DEFAULT_SOURCE special case
(report by Adam Sampson)
2022-09-24
* AcSplit/CF_BUILD_CC:
check if the build-compiler actually works, which is better than just checking
that it is not identical to the cross-compiler.
2022-09-17
* config.sub: 2022-09-17
From: Arsen Arsenović <arsen@aarsen.me>
config.sub: add linux-mlibc targets
Note: currently, there's no detection method for linux-mlibc in
config.guess, since there's no decided way to detect mlibc.
Signed-off-by: Arsen Arsenović <arsen@aarsen.me>
* config.sub (linux-mlibc*): Recognize.
* testsuite/config-sub.data (aarch64-linux-mlibc, riscv-linux-mlibc,
x86_64-linux-mlibc): New tests.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.guess: 2022-09-17
From: Arsen Arsenović <arsen@aarsen.me>
config.guess: support running on Managarm systems
Signed-off-by: Arsen Arsenović <arsen@aarsen.me>
* config.guess (x86_64:[Mm]anagarm:*:*, i?86:[Mm]anagarm:*:*,
*:[Mm]anagarm:*:*): Recognize.
(timestamp): Update.
* doc/config.guess.1: Regenerate.
* testsuite/config-guess.data: Add tests for managarm targets.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2022-09-17
From: Arsen Arsenović <arsen@aarsen.me>
config.sub: add managarm-{mlibc,kernel} targets
To achieve this, the `kernel' pseudo-OS value was added, intended to
identify targets requiring a kernel-specific compiler and base system.
Signed-off-by: Arsen Arsenović <arsen@aarsen.me>
* config.sub (managarm-*): Recognize.
(mlibc*): Likewise libc.
(timestamp): Update.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (aarch64-managarm, aarch64-managarm-kernel,
aarch64-managarm-mlibc, riscv-managarm, riscv-managarm-kernel,
riscv-managarm-mlibc, x86_64-managarm, x86_64-managarm-kernel,
x86_64-managarm-mlibc): New tests.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2022-09-10
* AcSplit/CF__XOPEN_SOURCE_BODY, AcSplit/CF__XOPEN_SOURCE_HEAD:
ci-aclocal generated header
* AcSplit/CF_XOPEN_SOURCE:
allow for case "linux-musl" by making the "linux*" case specific to the
GNU library, which lets it fall into the default case. In that case,
add a check in case (as done in musl) defining _POSIX_C_SOURCE turns
_XOPEN_SOURCE off (and recover from that by ignoring the _POSIX_C_SOURCE
definition).
* AcSplit/CF_TRY_XOPEN_SOURCE:
split-out CF__XOPEN_SOURCE_HEAD and CF__XOPEN_SOURCE_BODY
2022-08-27
* AcSplit/CF_LD_SEARCHPATH: use $LD in preference to "ld"
2022-08-25
* AcSplit/CF_CRYPT_FUNC: crypt() might return null - check for that
if crypt() exists, but does not work, reset its cache variable to ignore.
2022-08-20
* AcSplit/CF_CPP_OVERRIDE: ci-aclocal generated header
* AcSplit/CF_PROG_LINT:
provide dummy substitution for lint libraries, e.g., for cppcheck
2022-08-01
* config.guess, config.sub: 2022-08-01
From: Xiaotian Wu <wuxiaotian@loongson.cn>
Remove loongarchx32
loongarchx32 doesn't exist in the real world, so we can delete it.
* config.guess (loongarchx32:Linux:*:*): Remove.
(timestamp): Update.
* config.sub (loongarchx32): Remove.
(timestamp): Update.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data (loongarchx32): Remove.
* testsuite/config-sub.data (loongarchx32, loongarchx32-elf,
loongarchx32-linux, loongarchx32-loongson-elf): Likewise.
Link: https://loongson.github.io/LoongArch-Documentation/LoongArch-toolchain-conventions-EN.html
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2022-07-27
* AcSplit/CF_MATH_LIB: ensure that linking with the math-library works
2022-05-25
* config.guess: 2022-05-25
From: Alexander von Gluck IV <kallisti5@unixzen.com>
config.guess: Update Haiku guesses
* config.guess (ppc:Haiku:*:*, *:Haiku:*:*): Recognize.
(timestamp): Update.
* doc/config.guess.1: Regenerate.
* testsuite/config-guess.data: Add more test cases for Haiku.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2022-05-08
* config.guess: 2022-05-08
From: Bruno Haible <bruno@clisp.org>
config.guess (x86_64:Linux:*:*): Detect 32-bit ABI.
* config.guess (x86_64:Linux:*:*): Test for the 32-bit ABI. Don't assume that
__ILP32__ is a certain indicator for the x32 ABI; for GCC ≥ 9 it no longer is.
* doc/config.guess.1: Regenerate.
2022-03-17
* AcSplit/CF_WITH_BROTLI: ci-aclocal generated header
2022-02-24
* AcSplit/CF_WITH_IMAKE_CFLAGS: corrected quoting
2022-02-05
* COPYING.asc: regen
* COPYING: bump
* AcSplit/CF_MAN_PAGES: improve "trap" handling
2022-01-29
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
handle the case when either $PKG_CONFIG_LIBDIR or the option value has a
colon-separated list of directories (report by Rudi Heitbaum).
2022-01-28
* AcSplit/AM_WITH_NLS:
byacc's support for --version is recent (2021/08/02), but both byacc/bison
accept "-V" (report by Urs Janßen).
2022-01-27
* AcSplit/AM_WITH_NLS:
amend to work with the intl library bundled with tin 2.6.1
2022-01-22
* AcSplit/CF_ABI_DEFAULTS: initial changes for ncurses ABI 7
2022-01-19
* AcSplit/CF_WITH_IMAKE_CFLAGS:
provide a suitable default for PROJECTROOT, used in xterm for the default
location of luit.
2022-01-17
* AcSplit/CF_LOCALECONV: ci-aclocal generated header
2022-01-09
* config.guess: 2022-01-09
From: Idan Horowitz <idan.horowitz@gmail.com>
config.guess: recognize SerenityOS
* config.guess (*:SerenityOS:*:*): Recognize.
(timestamp): Update.
* doc/config.guess.1: Regenerate.
* testsuite/config-guess.data: Add test case for SerenityOS.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2022-01-03
* config.guess, config.sub: 2022-01-03
From: Bernhard Voelker <mail@bernhard-voelker.de>
Fix GPLv3 license headers to use a comma instead of semicolon
See: https://www.gnu.org/licenses/gpl-3.0.html#howto
Update license headers automatically using the following script:
$ git grep -l 'Foundation; either version 3' \
| xargs sed -i '/Foundation; either version 3/ s/n; e/n, e/'
* config.guess: Adjust via the above command.
(timestamp): Update.
* config.sub: Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2022-01-01
* config.guess, config.sub: 2022-01-01
From: Dmitry V. Levin <ldv@altlinux.org>
Update copyright years
* config.guess: Update copyright years.
* config.sub: Likewise.
* testsuite/config-guess.sh: Likewise.
* testsuite/config-sub.sh: Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
2021-12-25
* config.sub: 2021-12-25
From: Dmitry V. Levin <ldv@altlinux.org>
config.sub: alias armh to armv7l
ALT uses armh as an alias for armv7l-alt-linux-gnueabihf since 2012.
* config.sub (armh-unknown|armh-alt): Set cpu, vendor, and basic_os.
(timestamp): Update.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (armh, armh-alt-linux-gnueabihf): New tests.
2021-12-24
* config.sub: 2021-12-24
From: Dmitry V. Levin <ldv@altlinux.org>
config.sub: alias aarch64le to aarch64
Apparently, QNX reports aarch64 as aarch64le on little-endian machines.
* config.sub (aarch64le-*): Set cpu to aarch64.
(timestamp): Update.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (aarch64le-qnx): New test.
Reported-by: Elad Lahav <e2lahav@gmail.com>
Link: https://lists.gnu.org/archive/html/config-patches/2021-12/msg00009.html
2021-12-16
* AcSplit/CF_FIX_WARNINGS:
package-fix for cdk-perl: suppress the fixup for -Werror=format,
e.g., -Werror=format-security
2021-12-13 michael.baeuerle
* AcSplit/CF_FIND_CANLOCK:
amend package-name to match that released in libcanlock 3.2.0 (May 18, 2020)
2021-12-13
* config.sub: 2021-10-27
From: Dmitry V. Levin <ldv@altlinux.org>
config.sub: fix typo in timestamp
* config.sub: Fix timestamp.
* doc/config.sub.1: Regenerate.
Reported-by: Jordi Sanfeliu <jordi@fibranet.cat>
Fixes: a013aac61edfa2a03727521508286480010e7bf3
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-12-04
* AcSplit/CF_FUNC_GETTTYNAM: ci-aclocal generated header
2021-11-30
* config.guess: 2021-11-30
From: Andreas F. Borchert <github@andreas-borchert.de>
config.guess: x86_64-pc-solaris2.11 is not properly recognized
config.guess guesses Solaris 11 to run on a 32-bit platform
despite Solaris 11 no longer supporting any 32-bit platform.
See the following code at lines 434 to 445:
| SUN_ARCH=i386
| # If there is a compiler, see if it is configured for 64-bit objects.
| # Note that the Sun cc does not turn __LP64__ into 1 like gcc does.
| # This test works for both compilers.
| if test "$CC_FOR_BUILD" != no_compiler_found; then
| if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \
| (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \
| grep IS_64BIT_ARCH >/dev/null
| then
| SUN_ARCH=x86_64
| fi
| fi
If "cc" is installed, i.e. the Oracle Studio compiler, this one is
chosen for $CC_FOR_BUILD. This compiler, the gcc provided by Oracle
and also gcc bootstrapped from sources on that platform with a default
configuration will by default generate 32-bit binaries -- even on
a 64-bit platform. And __amd64 will not be defined for compilations
targeting a 32-bit platform. This is different from the corresponding
behaviour on GNU/Linux systems where the local platform is targeted by
default.
Thus, as long as you do not add "-m64" or if you have a custom-built
gcc which defaults to 64 bit, you will get 32-bit binaries on Solaris
despite living on a 64-bit platform.
* config.guess (i86pc:SunOS:5.*:* || i86xen:SunOS:5.*:*): Adapt the
test by adding the "-m64" flag. This will work properly for Solaris
10 as well (the last Solaris release that supported x86 32-bit
platforms).
* doc/config.guess.1: Regenerate.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-11-25
* AcSplit/CF_WITH_NO_LEAKS:
fix integration with "--with-valgrind" option
2021-11-20
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
handle case for "--without-pkg-config"
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
if no --with-pkg-config-libdir option is given, default to ${libdir}/pkgconfig
* AcSplit/CF_ENABLE_PC_FILES:
when validating $PKG_CONFIG_LIBDIR, do that on a copy, leaving the original
untouched to simplify manipulating ${prefix} in the makefile.
2021-11-07
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
add an explict "pkg-config" to package names, e.g., to handle Debian/testing
2021-10-30
* AcSplit/CF_CHECK_LIBSSP: ci-aclocal generated header
2021-10-27
* config.guess: 2021-10-27
From: Jordi Sanfeliu <jordi@fibranet.cat>
Recognize Fiwix
$ make check
cd testsuite && bash config-guess.sh && rm uname
PASS: config.guess checks (137 tests)
cd testsuite && bash config-sub.sh
PASS: config.sub checks (882 tests)
PASS: config.sub idempotency checks (819 tests)
PASS: config.sub canonicalise each config.guess testcase (137 tests)
* config.guess (i*86:Fiwix:*:*): Recognize.
* config.sub (fiwix*): Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data: Add a test case for Fiwix.
* testsuite/config-sub.data (i386-fiwix): New test.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2021-20-27
From: Jordi Sanfeliu <jordi@fibranet.cat>
Recognize Fiwix
$ make check
cd testsuite && bash config-guess.sh && rm uname
PASS: config.guess checks (137 tests)
cd testsuite && bash config-sub.sh
PASS: config.sub checks (882 tests)
PASS: config.sub idempotency checks (819 tests)
PASS: config.sub canonicalise each config.guess testcase (137 tests)
* config.guess (i*86:Fiwix:*:*): Recognize.
* config.sub (fiwix*): Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data: Add a test case for Fiwix.
* testsuite/config-sub.data (i386-fiwix): New test.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-10-26
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR: remove obsolete checks
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
redirect stderr, to handle older programs, and amend a regex to work with
Solaris 10.
2021-10-25
* AcSplit/CF_ENABLE_PC_FILES:
allow for an empty string in $PKG_CONFIG_LIBDIR
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
if there are no existing pkgconfig directories, choose one from the list
we were given ($libdir/pkgconfig if possible, otherwise the first in list)
2021-10-18
* config.sub: 2021-08-14
From: Kinshuk Dua <kinshukdua@gmail.com>
config.sub: Fix typo in comment
Fixes: 5e531d391852a54e7fab2d8ff55625fca514b305
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-10-17
* AcSplit/CF_ENABLE_PC_FILES, AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
$PKG_CONFIG_LIBDIR might be "none" rather than yes/no
2021-10-10
* AcSplit/CF_PKG_CONFIG: add comment
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
use the --debug output from pkg-config and pkgconf to determine the actual
search path, using that in preference to the ~2014 list derived from Debian.
also improved the check to ensure that PKG_CONFIG_LIBDIR is set to a single
directory rather than a list.
2021-09-05
* AcSplit/CF_REMOVE_CFLAGS:
handle multiple existing duplicates using a loop, and improve whitespace-trimming
* AcSplit/CF_APPEND_CFLAGS:
move the CF_ADD_CFLAGS inside the loop so that the purging of duplicates works
* AcSplit/CF_CHECK_CURSES_LIB:
save/restore $CFLAGS (like the existing $CPPFLAGS)
use CF_APPEND_CFLAGS for purging duplicate -D/-U options
* AcSplit/CF_REMOVE_CFLAGS:
corrected/simplified: the first version used an unset variable for the
actual update, and the case statement was not needed because BREs can handle
\(foo\)? expression
* AcSplit/CF_APPEND_CFLAGS: only -D and -U options should be removed
2021-09-04
* AcSplit/CF_MAKEFLAGS, AcSplit/CF_SHARED_OPTS, AcSplit/CF_NCURSES_LIBS,
AcSplit/CF_ADD_INCDIR, AcSplit/CF_BOOL_SIZE,
AcSplit/CF_CHECK_EXTERN_DATA, AcSplit/CF_DISABLE_ECHO,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_SIZECHANGE,
AcSplit/CF_WITH_PATH, AcSplit/CF_WITH_PATHLIST:
s/it's/it is/g
2021-08-28
* AcSplit/CF_APPEND_CFLAGS, AcSplit/CF_REMOVE_CFLAGS:
ci-aclocal generated header
* AcSplit/CF_NCURSES_CONFIG, AcSplit/CF_TRY_XOPEN_SOURCE,
AcSplit/CF_XOPEN_SOURCE:
filter out redefinition, e.g., changing _XOPEN_SOURCE from 500 to 600
2021-08-14
* config.sub: 2021-08-14
From: Nick Bowler <nbowler@draconx.ca>
config.sub: work around command assignment bug in some shells
When combining variable assignments with a shell command, some older
shells (notably heirloom-sh and presumably also Solaris 10 /bin/sh)
have a bug which causes the assignment to alter the current execution
environment whenever the command is a shell built-in. For example:
% dash -c 'x=good; x=bad echo >/dev/null; echo $x'
good
% jsh -c 'x=good; x=bad echo >/dev/null; echo $x'
bad
The config.sub script contains a few commands of the form:
IFS=- read ...
which triggers this bug, causing the IFS assignment to persist for the
remainder of the script. This can cause misbehaviour in certain cases,
for example:
% jsh config.sub i386-linux-gnu
config.sub: test: unknown operator gnu
% jsh config.sub i386-gnu/linux
sed: can't read s|gnu/linux|gnu|: No such file or directory
Invalid configuration `i386-gnu/linux': OS `' not recognized
* config.sub: Save and restore IFS explicitly to avoid shell bugs.
* doc/config.sub.1: Regenerate.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-08-11
* AcSplit/CF_WITH_PCRE2:
check for PCRE2regcomp, which Debian added to attemp to repair pcre2's
symbol conflicts.
2021-08-07
* AcSplit/CF_SHARED_OPTS: add case for haiku, from
https://github.com/haikuports/haikuports/blob/master/sys-libs/ncurses5/patches/ncurses-6.2.patchset
2021-08-04
* config.sub: 2021-08-04
From: Jeremy Soller <jackpot51@gmail.com>
config.sub: add Linux Relibc Target
$ make check
cd testsuite && bash config-guess.sh && rm uname
PASS: config.guess checks (136 tests)
cd testsuite && bash config-sub.sh
PASS: config.sub checks (881 tests)
PASS: config.sub idempotency checks (818 tests)
PASS: config.sub canonicalise each config.guess testcase (136 tests)
* config.sub (relibc*): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data (x86_64-linux-relibc): New test.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-07-17
* AcSplit/CF_LIB_RULES:
add USE_LIB_SUFFIX parameter to mk-1st.awk, to take into account the
"--with-extra-suffix" option
2021-07-10
* AcSplit/CF_XOPEN_CURSES:
don't make check for cchar_t specific to ncurses (X/Open Curses requires it)
also, if NCURSES_WIDECHAR happens to be set, e.g., as a side-effect, reject
the first try and fallthrough to prefer defining _XOPEN_SOURCE_EXTENDED
2021-07-06
* config.sub: 2021-07-06
From: Stephanos Ioannidis <root@stephanos.io>
config.sub: add Zephyr RTOS support
This adds the Zephyr RTOS targets in preparation for implementing the
Zephyr RTOS-specific toolchain support.
$ make check
cd testsuite && bash config-guess.sh && rm uname
PASS: config.guess checks (136 tests)
cd testsuite && bash config-sub.sh
PASS: config.sub checks (880 tests)
PASS: config.sub idempotency checks (817 tests)
PASS: config.sub canonicalise each config.guess testcase (136 tests)
* config.sub (zephyr*): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data: Add testcases for *-zephyr.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-07-05
* AcSplit/CF_WITH_IDNA:
check for idn2 library, use it in preference to idn
2021-07-03
* AcSplit/CF_LIB_RULES:
generate ".PHONY" directives if the make-program supports it
* config.sub: 2021-07-03
From: Ozkan Sezer <sezero@users.sourceforge.net>
config.sub: disable shellcheck SC2006 / SC2268 warnings
This is in line with the recent config.guess change in commit
12fcf67c9108f4c4b581eaa302088782f0ee40ea
* config.sub (shellcheck disable): Add SC2006,SC2268.
Suggested-by: Jacob Bachmeyer <jcb@gnu.org>
Signed-off-by: Ozkan Sezer <sezero@users.sourceforge.net>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2021-07-03
From: Ozkan Sezer <sezero@users.sourceforge.net>
config.sub: normalize the quoting in the `echo FOO | sed ...`
Some cases quote the argument to echo and some do not. At runtime
it probably does not matter because the substituted values will never
contain whitespace, but quoting them all would make shellcheck more
useful.
* config.sub: Consistently quote the argument of echo.
* doc/config.sub.1: Regenerate.
Suggested-by: Jacob Bachmeyer <jcb@gnu.org>
Signed-off-by: Ozkan Sezer <sezero@users.sourceforge.net>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-07-02
* config.sub: 2021-06-03
From: Ozkan Sezer <sezero@users.sourceforge.net>
config.sub: replace POSIX $( ) with classic ` ` throughout
This is in line with the recent config.guess change in commit
d70c4fa934de164178054c3a60aaa0024ed07c91.
The patch was generated using patch-6.gawk script introduced in that
commit.
* config.sub: Revert POSIX command substitutions to classic form.
Signed-off-by: Ozkan Sezer <sezero@users.sourceforge.net>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-06-19
* AcSplit/CF_GETOPT_HEADER: quiet a warning in config.log
2021-06-14
* AcSplit/CF_CURSES_EXIT: corrected function-name...
2021-06-13
* AcSplit/CF_WITH_NO_LEAKS:
guard against unexpected usage, e.g., "--without-no-leaks"
* AcSplit/CF_NO_LEAKS_OPTION:
guard against unexpected usage, e.g., "--without-valgrind"
2021-06-08
* AcSplit/CF_ALT_CHAR_SET, AcSplit/CF_FIONBIO, AcSplit/CF_FUNC_LSTAT,
AcSplit/CF_SRAND, AcSplit/CF_TM_GMTOFF, AcSplit/CF_UNION_WAIT:
quiet a configure-time warning
2021-06-07
* AcSplit/CF_CONST_X_STRING, AcSplit/CF_INPUT_METHOD,
AcSplit/CF_UTMP_UT_SESSION, AcSplit/CF__GRANTPT_BODY:
scan-build warnings
2021-06-04
* config.guess: 2021-06-03
From: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Recognize arc32
This is the 32-bit variant of ARCv3 ISA (which is not compatible with the
32-bit ARCv2 ISA)
| make check
| cd testsuite && bash config-guess.sh && rm uname
| PASS: config.guess checks (136 tests)
| cd testsuite && bash config-sub.sh
| PASS: config.sub checks (864 tests)
| PASS: config.sub idempotency checks (801 tests)
| PASS: config.sub canonicalise each config.guess testcase (136 tests)
* config.guess (arc32:Linux:*:*): Recognize.
* config.sub (arc32): Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data: Add a test case for arc32.
* testsuite/config-sub.data (arc32, arc*-elf): Add test cases.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2021-06-03 (repaired)
From: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Recognize arc32
This is the 32-bit variant of ARCv3 ISA (which is not compatible with the
32-bit ARCv2 ISA)
| make check
| cd testsuite && bash config-guess.sh && rm uname
| PASS: config.guess checks (136 tests)
| cd testsuite && bash config-sub.sh
| PASS: config.sub checks (864 tests)
| PASS: config.sub idempotency checks (801 tests)
| PASS: config.sub canonicalise each config.guess testcase (136 tests)
* config.guess (arc32:Linux:*:*): Recognize.
* config.sub (arc32): Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data: Add a test case for arc32.
* testsuite/config-sub.data (arc32, arc*-elf): Add test cases.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-05-26
* config.guess: 2021-05-24
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: manual fixups after previous automatic patch
The tool could not handle command substitutions that span lines, but
fortunately there were only two such substitutions in the script.
The test for which universe is active on Pyramid is rewritten into a
case block because it was the only use of a command substitution as an
argument to the test command, which would require quoting.
* config.guess: Rewrite "if" for Pyramid systems to "case".
* config.guess: 2021-05-24 (repaired)
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: replace POSIX $( ) with classic ` ` throughout
The previous replacement of backticks with POSIX command substitutions
was ill-considered and illogical: this script recognizes many archaic
machine types that probably never had POSIX shells, therefore it needs
to be able to run successfully under pre-POSIX shells.
This patch was generated using the included GNU Awk program.
* config.guess: Revert POSIX command substitutions to classic form.
* patch-6.gawk: Store the tool that produced the automated patch.
2021-05-25
* config.guess: 2021-05-24 (repaired)
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: manual fixup after previous automated patches
This patch provides the special handling for the GNU system. As these
were two small and unique edits, they were not included in the scripts.
This patch also cleans up other minor issues that must be addressed
before reverting to classic command substitutions and updates
"shellcheck" directives to account for changes in this script and the
change in "shellcheck" towards reporting individual portability issues.
* config.guess: 2021-05-24 (repaired)
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: automatic fixups after previous automated patch
This patch was generated using the following command:
sed -i config.guess \
-e '/="[^"]\+"\(-\|$\)/s/="\([^"([:space:])]\+\)"/=\1/' \
-e '/="[^"]\+"[[:alnum:]]/s/="\$\([^([:space:])]\+\)"/=${\1}/' \
-e \
'/\$(echo[^|]\+|/s/\([^[:space:]]\)[[:space:]]*|[[:space:]]*sed/\1 | sed/g'
* config.guess: Remove unneeded quotes in other variable assignments,
standardize spacing for "echo ... | sed" substitutions.
* config.guess: 2021-05-24 (repaired)
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: remove unneeded quotes and factor command substitutions
This is further cleanup and simplifies some constructs that can confuse
Emacs' syntax highlighting while generally reducing required quoting.
This patch was generated using the included GNU Awk program.
* config.guess: Remove unneeded variable quotes and factor out command
substitutions when setting GUESS.
* patch-3.gawk: Store the tool that produced the automated patch.
* config.guess: 2021-05-24 (repaired)
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: manual fixups after previous automatic patch
* config.guess: Adjust a few "leftover" cases that the tool could not
easily recognize and fixes comment indentation in a few other special
cases.
* config.guess: 2021-05-24 (repaired)
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: introduce intermediate variable with uname results
This will allow quoting to be significantly simplified in another
pass through the file.
* config.guess: Introduce GUESS variable to hold results of uname analysis.
* config.guess: 2021-05-24 (repaired)
From: Jacob Bachmeyer <jcb@gnu.org>
config.guess: use intermediate variable with uname results
This will allow quoting to be significantly simplified in another
pass through the file.
This patch was generated using the included GNU Awk program.
* config.guess: Use GUESS variable to hold results of uname analysis.
* patch-1.gawk: Store the tool that produced the automated patch.
2021-05-24
* config.guess: 2021-05-24 (repaired)
From: Dmitry V. Levin <ldv@altlinux.org>
config.guess: fix shellcheck warning SC2154
While, according to Plan 9 documentation, the environment variable
$cputype is set to the name of the kernel's CPU's architecture,
shellcheck warns that cputype is referenced but not assigned.
Be on the safe side and do not use cputype if it is not defined
or empty.
* config.guess (*:Plan9:*:*): Fix shellcheck warning SC2154.
* config.guess: 2021-05-24 (repaired)
From: Dmitry V. Levin <ldv@altlinux.org>
config.guess: remove redundant quotes in case commands
According to the GNU Autoconf Portable Shell Programming manual,
the Bourne shell does not systematically split variables and back-quoted
expressions, in particular on the right-hand side of assignments and in
the argument of 'case'.
The change is made automatically using the following command:
$ sed -E -i 's/(\<case )"(\$[^"]+)"( in\>)/\1\2\3/' config.guess
* config.guess: Simplify case commands by removing quotes around the
argument.
Suggested-by: Jacob Bachmeyer <jcb@gnu.org>
* config.guess: 2021-05-24 (repaired)
From: Dmitry V. Levin <ldv@altlinux.org>
config.guess: simplify exit status workaround on alphaev67-dec-osf5.1
Commit 29865ea8a5622cdd80b7a69a0afa78004b4cd311 introduced an exit trap
reset before exiting to avoid a spurious non-zero exit status on
alphaev67-dec-osf5.1. Simplify that code a bit by moving the exit trap
reset around.
* config.guess (alpha:OSF1:*:*): Reset exit trap earlier.
* doc/config.guess.1: Regenerate.
2021-05-19
* AcSplit/CF_UTF8_LIB:
putwc might be prototyped in wchar.h -- check for that
* AcSplit/CF_NCURSES_CONFIG:
quiet a configure-time const-mismatch warning
* AcSplit/CF__INTL_BODY:
quiet a configure-time warning about assigning pointer to integer
2021-05-01
* AcSplit/CF_XOPEN_SOURCE:
modify special-case for OpenBSD 6 (and up) to work around one of the places
where the compile-time environment is broken. This was reported before
OpenBSD 6.7:
https://www.mail-archive.com/bugs@openbsd.org/msg13200.html
but the ensuring discussion elicited no action.
* AcSplit/CF_CHECK_LIBTOOL_VERSION:
modify to prevent accidentally using the OpenBSD "libtool"
2021-04-30
* config.sub: 2021-04-30 (repaired)
From: Maciej W. Rozycki <macro@orcam.me.uk>
config.sub: Handle MIPS R3 and R5 ISA levels with CPU names
Complement binutils commit ae52f4830604 ("Add MIPS r3 and r5 support.")
and recognize MIPS CPU patterns for the R3 and R5 ISA levels, used by
GAS to set defaults.
* config.sub (mipsisa32r3, mipsisa32r3el, mipsisa32r5, mipsisa32r5el,
mipsisa64r3, mipsisa64r3el, mipsisa64r5, mipsisa64r5el): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data: Add test cases.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-04-21
* config.guess, config.sub: 2021-04-21 (repaired)
From: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Recognize arc64
This paves way for setting up arc64 software ecosystem.
$ make check
cd testsuite && bash config-guess.sh && rm uname
PASS: config.guess checks (136 tests)
cd testsuite && bash config-sub.sh
PASS: config.sub checks (853 tests)
PASS: config.sub idempotency checks (790 tests)
PASS: config.sub canonicalise each config.guess testcase (136 tests)
* config.guess (arc64:Linux:*:*): Recognize.
* config.sub (arc64): Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data: Add a test case for arc64.
* testsuite/config-sub.data (arc64, arc*-elf): Add test cases.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-04-18
* AcSplit/CF_LINK_FUNCS:
improve check by ensuring that the source file is closed before linking to target
2021-04-16
* config.guess: 2021-04-16 (repaired)
From: Purple Rain <purplerain@secbsd.org>
config.guess: add SecBSD support
* config.guess (*:SecBSD:*:*): Recognize.
* doc/config.guess.1: Regenerate.
* testsuite/config-guess.data: Add a test case.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* config.sub: 2021-04-16 (repaired)
From: Purple Rain <purplerain@secbsd.org>
config.sub: add SecBSD support
* config.sub (secbsd*): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data: Add x86_64-secbsd.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-04-03
* AcSplit/CF_DISABLE_LEAKS:
recent rewrite left "--enable-leaks" meaning the same as "--disable-leaks"
because it ignored $enableval
* AcSplit/CF_LINK_FUNCS:
workaround for NFS on FreeBSD-13rc4, which bumps the modification-time on
the link target. Also, make the test-function return distinct exit-codes
to make analysis of config.log easier.
2021-03-28
* AcSplit/CF_C11_NORETURN: add comment for autoheader
ensure that _Noreturn keyword is tested
add STDC_NORETURN to substitutions
2021-03-24
* AcSplit/CF_WITH_NC_ALLOC_H:
add checks for exit_curses and exit_terminfo, which are preferred
2021-03-22
* AcSplit/CF_C11_NORETURN:
add --enable-stdnoreturn option to make this feature optional, since several
programs outside ncurses use GCC_NORETURN and must be updated before the
feature can be used unconditionally. Also, those programs may have to
be built with older versions of ncurses.
2021-03-20
* AcSplit/CF_LIB_RULES: allow for c++ directory to do lint-rule
* AcSplit/CF_GCC_ATTRIBUTES: check for C11 _Noreturn keyword
* AcSplit/CF_C11_NORETURN: ci-aclocal generated header
2021-03-10
* config.sub: 2021-03-10 (repaired)
From: Idan Horo <idan.horowitz@gmail.com>
config.sub: Add support for SerenityOS
* config.sub (serenity*): Recognize.
* doc/config.sub.1: Regenerate.
* testsuite/config-sub.data: Add i386-serenity.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-03-05
* AcSplit/CF_LD_SEARCHPATH:
improve check for Solaris. add check for HP-UX
2021-01-26
* AcSplit/CF_WITH_PCRE2: add check for Fedora libpcre2-posix package
2021-01-25
* config.guess: 2021-01-25 (repaired)
From: Kalamatee <kalamatee@gmail.com>
config.guess: update AROS system detection
* config.guess: Recognize *:AROS:*:*.
* doc/config.guess.1: Regenerate.
* testsuite/config-guess.data: Add test cases.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-01-23
* AcSplit/CF_LIB_RULES:
run subdirectory-makes in a subprocess, to work around pmake, which is
guaranteed to misbehave when attempting to parallelize the make-process.
2021-01-19
* config.guess: 2021-01-19 (repaired)
From: M. Levinson <mlevins@users.sourceforge.net>
config.guess: fix shell variable quoting bug
* config.guess (*:NetBSD:*:*): Spell out the full sysctl command twice
instead of using a shell variable.
* doc/config.guess.1: Regenerate.
Fixes: 827c77253b396c07306927b2a4b794a3251c48eb
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-01-13
* AcSplit/CF_WITHOUT_X:
make X_CFLAGS and X_LIBS substitute, needed for luit
2021-01-10
* AcSplit/CF_ALL_MAKEFILES, AcSplit/CF_FIND_TDLIB, AcSplit/CF_ANSI_CPP,
AcSplit/CF_CHECK_REGEX, AcSplit/CF_CURSES_CCHAR_T,
AcSplit/CF_CURSES_MOUSE, AcSplit/CF_GMTOFF, AcSplit/CF_MAKE_AR_RULES,
AcSplit/CF_MAKE_INCLUDE, AcSplit/CF_PROGRAM_FULLPATH,
AcSplit/CF_PROGRAM_PREFIX, AcSplit/CF_REGCMP_FUNCS,
AcSplit/CF_REGEXPR_H_FUNCS, AcSplit/CF_REGEX_H_FUNCS,
AcSplit/CF_RE_COMP_FUNCS, AcSplit/CF_SIZE_T, AcSplit/CF_SRC_MAKEFILE,
AcSplit/CF_STAT_ST_BLOCKS, AcSplit/CF_STRUCT_SCREEN,
AcSplit/CF_TCAP_CURSOR, AcSplit/CF_TD_SIG_ARGS,
AcSplit/CF_TD_SRC_MODULES, AcSplit/CF_TD_TEST_MODULES,
AcSplit/CF_TOP_SRCDIR, AcSplit/CF_VARARGS, AcSplit/CF_WAIT:
quoting fixes for shellcheck
* AcSplit/CF_MAKE_DOCS:
suppress groff's hyphenation when making a ".txt" file
* AcSplit/CF_NEED_PROG: ci-aclocal generated header
2021-01-08
* AcSplit/CF_GXX_WARNINGS: s/GCC_VERSION/GXX_VERSION/
* AcSplit/CF_MAKE_PHONY: tweak glob expression, per shellcheck
2021-01-07
* config.sub: 2021-01-08 (repaired)
From: Peixing Xin <peixing.xin@windriver.com>
config.sub: recognize four-part configuration name for VxWorks
For example:
armv7m-wrs-vxworks-eabihf
armv7-wrs-vxworks-eabihf
i686-wrs-vxworks-simlinux
i686-wrs-vxworks-simwindows
powerpc-wrs-vxworks-spe
x86_64-wrs-vxworks-simlinux
x86_64-wrs-vxworks-simwindows
* config.sub: Recognize four-part configuration name for VxWorks.
* doc/config.guess.1: Regenerate.
* testsuite/config-sub.data: Add test cases.
Co-authored-by: John Ericson <git@JohnEricson.me>
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2021-01-06
* config.sub: 2021-01-07 (repaired)
From: Alan Modra <amodra@gmail.com>
config.sub: accept OS of eabi* and gnueabi*
Commit 5e531d391852 broke powerpc-eabivle:
$ ./config.sub powerpc-eabivle
Invalid configuration `powerpc-eabivle': OS `eabivle' not recognized
Also powerpc-eabisim and probably some arm configurations.
* config.sub: Accept OS of eabi* and gnueabi*.
* testsuite/config-sub.data: Add powerpc-eabisim and powerpc-eabivle.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
* AcSplit/CF_LINK_PREFIX:
naming-consistency, from vile's configure script
2021-01-05
* AcSplit/CF_DISABLE_RPATH_HACK:
prefer enable_rpath_hack to cf_disable_rpath_hack, for consistency
* AcSplit/CF_WITH_NO_LEAKS, AcSplit/CF_WITH_NC_ALLOC_H,
AcSplit/CF_DISABLE_LEAKS, AcSplit/CF_NO_LEAKS_OPTION:
prefer $enable_leaks to $with_no_leaks, for consistency
* AcSplit/CF_ENABLE_WARNINGS:
s/with_warnings/enable_warnings/, for consistency
* AcSplit/CF_AC213_COMPAT: ci-aclocal generated header
* AcSplit/CF_ENABLE_STRING_HACKS:
s/with_string_hacks/enable_string_hacks/, for consistency
* AcSplit/CF_INSTALL_OPT_S:
s/with_stripping/enable_stripping/, for consistency
* AcSplit/CF_MAN_PAGES: quoting fixes for shellcheck
2021-01-04
* AcSplit/CF_CHECK_CURSES_LIB, AcSplit/CF_CURSES_ACS_MAP,
AcSplit/CF_CURSES_CHECK_DATA, AcSplit/CF_CURSES_CHECK_TYPE,
AcSplit/CF_CURSES_WACS_MAP, AcSplit/CF_CURSES_WACS_SYMBOLS,
AcSplit/CF_TPUTS_PROTO, AcSplit/CF__CURSES_DATA,
AcSplit/CF_CFG_DEFAULTS, AcSplit/CF_NCURSES_ADDON,
AcSplit/CF_PROG_INSTALL:
quoting fixes for shellcheck
* AcSplit/CF_NCURSES_HEADER: use $ECHO_N and $ECHO_C from autoconf-252
* AcSplit/CF_FUNC_POLL: correct order of redirecting output of tty
* AcSplit/CF_CFG_DEFAULTS: use $ECHO_N, from autoconf-252
* AcSplit/CF_CHECK_GPM_WGETCH, AcSplit/CF_SUBST_NCURSES_VERSION:
use AC_PROG_EGREP
* AcSplit/CF_WITH_VERSIONED_SYMS: use AC_PROG_EGREP and AC_PROG_FGREP
* AcSplit/CF_CFG_DEFAULTS, AcSplit/CF_LIB_RULES, AcSplit/CF_LIB_SONAME:
use AC_PROG_FGREP
* AcSplit/AM_WITH_NLS, AcSplit/CF_SLANG_CPPFLAGS,
AcSplit/CF_TYPE_LONG_LONG:
quoting fixes for shellcheck
2021-01-03
* AcSplit/CF_LEX_VERSION, AcSplit/CF_MAKEFLAGS:
use AC_PROG_FGREP, per shellcheck warning
* AcSplit/CF_TTY_GROUP, AcSplit/CF_WITH_DESKTOP_CATEGORY:
use AC_PROG_EGREP, per shellcheck warning
* AcSplit/CF_GCC_ATTRIBUTES, AcSplit/CF_RPATH_HACK,
AcSplit/CF_WITH_MAN2HTML:
use AC_PROG_FGREP, per shellcheck warning
* AcSplit/CF_NCURSES_CONFIG, AcSplit/CF_NCURSES_LIBS:
change compile-check from tgoto to tigetstr, because tgoto is not prototyped
in ncurses' curses.h while tigetstr is prototyped. Lacking the prototype
can cause the configure check to fail with a recent clang configuration on
macOS, which turn on -Werror=implicit-function-declaration by default
(report by Dennis Preiser).
2021-01-02
* COPYING.asc: regen
* COPYING: bump
* AcSplit/CF_CHECK_CFLAGS:
actually undo the change to $CFLAGS on error, rather than clearing it.
* AcSplit/CF_WITH_CURSES_DIR, AcSplit/CF_CHECK_GETENV,
AcSplit/CF_CHECK_GPM_WGETCH, AcSplit/CF_CHECK_WCWIDTH_GRAPHICS,
AcSplit/CF_ENABLE_BROKEN_LINKER, AcSplit/CF_ETIP_DEFINES,
AcSplit/CF_GNAT_PROJECTS, AcSplit/CF_HASHED_DB_LIBS,
AcSplit/CF_LDFLAGS_STATIC, AcSplit/CF_LIB_SONAME,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MAN_PAGES,
AcSplit/CF_MAP_LIB_BASENAME, AcSplit/CF_PROG_GNAT,
AcSplit/CF_PROG_INSTALL, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_WITH_VERSIONED_SYMS:
quoting fixes for shellcheck
* AcSplit/CF__SED_TRIMBLANKS: ci-aclocal generated header
* AcSplit/AM_GNU_GETTEXT, AcSplit/AM_LC_MESSAGES,
AcSplit/AM_PATH_PROG_WITH_TEST, AcSplit/AM_WITH_NLS,
AcSplit/CF_BUILD_CC, AcSplit/CF_BUNDLED_INTL, AcSplit/CF_CHECK_SIZEOF,
AcSplit/CF_CHECK_TYPE, AcSplit/CF_COLOR_CURSES,
AcSplit/CF_CURSES_CHTYPE, AcSplit/CF_CURSES_CPPFLAGS,
AcSplit/CF_CURSES_LIBS, AcSplit/CF_CURSES_TERM_H,
AcSplit/CF_CURSES_UNCTRL_H, AcSplit/CF_CURS_PERFORMANCE,
AcSplit/CF_DEFINE_PROG, AcSplit/CF_FANCY_CURSES,
AcSplit/CF_FIND_IPV6_LIBS, AcSplit/CF_FIND_IPV6_TYPE,
AcSplit/CF_FIND_LIBRARY, AcSplit/CF_FUNC_CURSES_VERSION,
AcSplit/CF_FUNC_LSTAT, AcSplit/CF_FUNC_WAIT, AcSplit/CF_GNUTLS,
AcSplit/CF_LASTLOG, AcSplit/CF_LOCALE, AcSplit/CF_MERGE_EXTRA_CFLAGS,
AcSplit/CF_NCURSES_CPPFLAGS, AcSplit/CF_NCURSES_HEADER,
AcSplit/CF_NCURSES_LIBS, AcSplit/CF_NETLIBS, AcSplit/CF_NSS_COMPAT,
AcSplit/CF_OUR_MESSAGES, AcSplit/CF_PATH_PROG,
AcSplit/CF_PDCURSES_X11, AcSplit/CF_PROG_EXT,
AcSplit/CF_SLANG_CPPFLAGS, AcSplit/CF_SLANG_LIBS,
AcSplit/CF_SLANG_UNIX_DEFS, AcSplit/CF_SRAND, AcSplit/CF_SSL,
AcSplit/CF_SYSTEM_MAIL_FLAGS, AcSplit/CF_SYSV_UTMP,
AcSplit/CF_TERMIO_AND_CURSES, AcSplit/CF_TERMIO_AND_TERMIOS,
AcSplit/CF_TERM_HEADER, AcSplit/CF_TM_GMTOFF, AcSplit/CF_TRY_RPATH,
AcSplit/CF_TTYTYPE, AcSplit/CF_TYPE_LONG_LONG, AcSplit/CF_UNION_WAIT,
AcSplit/CF_UTMP, AcSplit/CF_UTMP_UT_HOST, AcSplit/CF_UTMP_UT_NAME,
AcSplit/CF_UTMP_UT_SESSION, AcSplit/CF_UTMP_UT_SYSLEN,
AcSplit/CF_UTMP_UT_XSTATUS, AcSplit/CF_UTMP_UT_XTIME,
AcSplit/CF_WAIT_HEADERS, AcSplit/CF_XOPEN_CURSES, AcSplit/CF_X_TOOLKIT:
quoting fixes for shellcheck
* AcSplit/CF_STRIP_G_OPT, AcSplit/CF_STRIP_O_OPT:
use CF__SED_TRIMBLANKS
* AcSplit/CF_ALT_CHAR_SET:
use curly braces to work around shellcheck confusion with shell array
2021-01-01
* AcSplit/CF_CURSES_TERM_H, AcSplit/CF_CURSES_UNCTRL_H:
quoting fixes for shellcheck
* AcSplit/CF_NCURSES_CONFIG: trim blanks
* AcSplit/AM_LANGINFO_CODESET, AcSplit/CF_CFG_DEFAULTS,
AcSplit/CF_GCC_WARNINGS, AcSplit/CF_INTEL_COMPILER,
AcSplit/CF_LIB_RULES, AcSplit/CF_LIB_SUFFIX,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MANPAGE_RENAMES,
AcSplit/CF_MAN_PAGES, AcSplit/CF_MIXEDCASE_FILENAMES,
AcSplit/CF_REGEX, AcSplit/CF_SHARED_OPTS, AcSplit/CF_WITH_LIB_PREFIX,
AcSplit/CF_WITH_PCRE2, AcSplit/CF_XOPEN_SOURCE,
AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
quoting fixes for shellcheck
* AcSplit/CF_GCC_ATTRIBUTES: *** empty log message ***
* AcSplit/CF_AR_FLAGS, AcSplit/CF_CFG_DEFAULTS,
AcSplit/CF_CHECK_GNAT_VERSION, AcSplit/CF_CHECK_GPM_WGETCH,
AcSplit/CF_CHECK_WCHAR_H, AcSplit/CF_CLANG_COMPILER,
AcSplit/CF_CONST_X_STRING, AcSplit/CF_CXX_AR_FLAGS,
AcSplit/CF_DIRS_TO_MAKE, AcSplit/CF_FIND_SUB_INCDIR,
AcSplit/CF_FIND_SUB_LIBDIR, AcSplit/CF_FUNC_OPENPTY,
AcSplit/CF_FUNC_VSSCANF, AcSplit/CF_GCC_WARNINGS,
AcSplit/CF_GETOPT_HEADER, AcSplit/CF_GNAT_GENERICS,
AcSplit/CF_GNAT_PROJECTS, AcSplit/CF_GNAT_SIGINT,
AcSplit/CF_GNAT_TRY_LINK, AcSplit/CF_GNAT_TRY_RUN,
AcSplit/CF_GPP_LIBRARY, AcSplit/CF_GXX_WARNINGS,
AcSplit/CF_HEADER_PATH, AcSplit/CF_INSTALL_OPT_P,
AcSplit/CF_LDFLAGS_STATIC, AcSplit/CF_LD_RPATH_OPT,
AcSplit/CF_LIBRARY_PATH, AcSplit/CF_LIB_PREFIX, AcSplit/CF_LIB_RULES,
AcSplit/CF_LIB_SONAME, AcSplit/CF_LIB_SUFFIX,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MANPAGE_RENAMES,
AcSplit/CF_MAN_PAGES, AcSplit/CF_MKSTEMP, AcSplit/CF_NCURSES_ABI_6,
AcSplit/CF_NCURSES_WITH_ABI_VERSION, AcSplit/CF_OBJ_SUBDIR,
AcSplit/CF_PATHSEP, AcSplit/CF_PKG_CONFIG, AcSplit/CF_PRG_RULES,
AcSplit/CF_PROG_CC_C_O, AcSplit/CF_PROG_GNAT, AcSplit/CF_PROG_INSTALL,
AcSplit/CF_REGEX, AcSplit/CF_RPATH_HACK_2, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_SIGWINCH, AcSplit/CF_SIZECHANGE, AcSplit/CF_SRC_MODULES,
AcSplit/CF_STDCPP_LIBRARY, AcSplit/CF_SUBST_NCURSES_VERSION,
AcSplit/CF_VERSION_INFO, AcSplit/CF_WITH_ABI_VERSION,
AcSplit/CF_WITH_LIBTOOL, AcSplit/CF_WITH_PATHLIST,
AcSplit/CF_XOPEN_SOURCE:
quoting fixes for shellcheck
* config.guess, config.sub: 2021-01-01 (repaired)
From: Dmitry V. Levin <ldv@altlinux.org>
Update copyright years
* config.guess: Update copyright years.
* config.sub: Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
2020-12-31
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_ADD_LIBDIR,
AcSplit/CF_ADD_SEARCHPATH, AcSplit/CF_ADD_SUBDIR_PATH,
AcSplit/CF_AR_FLAGS, AcSplit/CF_AUTO_SYMLINK, AcSplit/CF_CHAR_DEVICE,
AcSplit/CF_CURSES_FUNCS, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_CURSES_TERM_H, AcSplit/CF_CURSES_UNCTRL_H,
AcSplit/CF_DIRNAME, AcSplit/CF_FIND_LIBRARY, AcSplit/CF_FIND_LINKAGE,
AcSplit/CF_FP_ISREADY, AcSplit/CF_GCC_ATTRIBUTES,
AcSplit/CF_GCC_OPT_RDYNAMIC, AcSplit/CF_GCC_WARNINGS,
AcSplit/CF_HEADER_PATH, AcSplit/CF_INSTALL_OPT_O,
AcSplit/CF_INSTALL_OPT_P, AcSplit/CF_LEX_STATES,
AcSplit/CF_LEX_VERSION, AcSplit/CF_MAKEFLAGS, AcSplit/CF_MAKE_PHONY,
AcSplit/CF_MISSING_CHECK, AcSplit/CF_NCURSES_CC_CHECK,
AcSplit/CF_NCURSES_CONFIG, AcSplit/CF_NCURSES_HEADER,
AcSplit/CF_NCURSES_LIBS, AcSplit/CF_NCURSES_VERSION,
AcSplit/CF_PROG_PERL, AcSplit/CF_SIGWINCH, AcSplit/CF_STDIO_UNLOCKED,
AcSplit/CF_TRANSFORM_PROGRAM, AcSplit/CF_WITH_PATHLIST,
AcSplit/CF_WITH_PERL, AcSplit/CF_WITH_SYMLINK,
AcSplit/CF_CC_ENV_FLAGS, AcSplit/CF_CLANG_COMPILER,
AcSplit/CF_CONST_X_STRING, AcSplit/CF_ENABLE_NARROWPROTO,
AcSplit/CF_ENABLE_WARNINGS, AcSplit/CF_FIX_WARNINGS,
AcSplit/CF_FUNC_GRANTPT, AcSplit/CF_IMAKE_CFLAGS, AcSplit/CF_MATH_LIB,
AcSplit/CF_NO_LEAKS_OPTION, AcSplit/CF_PATH_SYNTAX,
AcSplit/CF_PROCFS_CWD, AcSplit/CF_TTY_GROUP,
AcSplit/CF_WITH_DESKTOP_CATEGORY, AcSplit/CF_WITH_MAN2HTML,
AcSplit/CF_ADD_CFLAGS, AcSplit/CF_CHECK_CACHE, AcSplit/CF_PATH_PROG,
AcSplit/CF_RPATH_HACK, AcSplit/CF_TRY_PKG_CONFIG,
AcSplit/CF_UTMP_GROUP, AcSplit/CF_WITH_ICON_THEME,
AcSplit/CF_X_ATHENA_CPPFLAGS:
quoting fixes for shellcheck
* AcSplit/CF_TTY_GROUP:
modify the test program to be more tolerant of the environment when calling
ttyname, falling back to "/dev/tty" when ttyname has no tty attached.
* config.guess, config.sub: 2020-12-31 (repaired)
From: Kito Cheng <kito.cheng@sifive.com>
Recognize riscv32be and riscv64be
Recently RISC-V community got patches big-endian support for binutils,
and we'd like to accept that, however before accepting that I think it
would be better to upstream config.sub and config.guess change here :)
It's my check result on Ubuntu 18.04:
$ make check
cd testsuite && bash config-guess.sh && rm uname
PASS: config.guess checks (131 tests)
cd testsuite && bash config-sub.sh
PASS: config.sub checks (830 tests)
PASS: config.sub idempotency checks (767 tests)
PASS: config.sub canonicalise each config.guess testcase (131 tests)
* config.guess (riscv32be:Linux:*:*, riscv64be:Linux:*:*): Recognize.
* config.sub (riscv32be, riscv64be): Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data: Add test cases for riscv32be, and riscv64be.
* testsuite/config-sub.data (riscv32be, riscv64be): Add test cases.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2020-12-27
* AcSplit/CF_WITH_DESKTOP_CATEGORY:
treat any category with two uppercase characters in a row as some distribution.
that handles GNOME, KDE, GDK and a lot of other false positives.
2020-12-26
* AcSplit/CF_WITH_MAN2HTML:
recent upgrade of groff caused it to overuse hyphenation at the expense
of readability. suppress the feature.
2020-12-06
* AcSplit/CF_TD_TEST_MAKEFILES:
add dependencies which VPATH should have provided
* AcSplit/CF_SRC_MAKEFILE: use srcdir, for out-of-tree builds
* AcSplit/CF_BUILD_DIR: ci-aclocal generated header
* AcSplit/CF_TD_SRC_MAKEFILES: use $(top_srcdir) rather than ..
* AcSplit/CF_TD_TEST_MAKEFILES: use $(top_srcdir), rather than ..
2020-12-05
* AcSplit/CF_GNAT_PROJECTS:
add symbols to allow commenting-out makefile rules using gnatmake or gprbuild
when one of those tools is absent
2020-12-03
* config.guess, config.sub: 2020-12-22 (repaired)
From: Xiaotian Wu <wuxiaotian@loongson.cn>
Recognize loongarch32, loongarch64, and loongarchx32
* config.guess (loongarch32:Linux:*:*, loongarch64:Linux:*:*,
loongarchx32:Linux:*:*): Recognize.
* config.sub (loongarch32, loongarch64, loongarchx32): Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
* testsuite/config-guess.data: Add test cases for loongarch32,
loongarch64, and loongarchx32.
* testsuite/config-sub.data (loongarch32, loongarch64, loongarchx32):
Add test cases.
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
2020-12-01
* config.sub: 2020-12-02 (repaired)
From: Dmitry V. Levin <ldv@altlinux.org>
config.sub: recognize thumbv7*
* config.sub (thumbv7*): Recognize.
* testsuite/config-sub.data (thumbv7): New test.
Reported-by: Karl Berry <karl@freefriends.org>
Link: https://lists.gnu.org/archive/html/config-patches/2020-12/msg00001.html
2020-11-30
* install-sh: use newer script from autoconf; as noted in
https://lists.gnu.org/archive/html/automake/2018-09/msg00005.html
the copyright dates are bogus.
2020-11-26
* AcSplit/CF_CLANG_COMPILER:
work around macOS's broken aliases for c89, etc., by transforming this to
clang's -std option, which lets me use other workarounds to get a usable
standard compiler.
2020-11-20
* AcSplit/CF_CLANG_COMPILER:
add the compiler flags needed for testing (report by Steve Pitman)
2020-11-19
* config.guess, config.sub: 2020-11-17 (repaired)
From: Dmitry V. Levin <ldv@altlinux.org>
.gitattributes: specify a custom git merge driver for the ChangeLog file
* config.guess, config.sub: 2020-11-19 (repaired)
From: Dmitry V. Levin <ldv@altlinux.org>
Update URLs of the latest version of config.guess and config.sub scripts
Prefer cgit URLs over gitweb as the former are usually served faster:
$ time -f %e wget -q 'https://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess'
1.06
$ time -f %e wget -q 'https://git.savannah.gnu.org/cgit/config.git/plain/config.guess'
0.73
* config.guess: Prefer cgit URLs over gitweb.
(timestamp): Update.
* config.sub: Likewise.
* doc/config.guess.1: Regenerate.
* doc/config.sub.1: Likewise.
2020-11-14
* AcSplit/CF_MIXEDCASE_FILENAMES:
when cross-compiling to darwin (macOS), assume the filesystem doesn't
support mixed-case (report by Eli Rykoff).
2020-11-06
* config.guess, config.sub: 2020-11-07 (repaired)
From: Ben Elliston <bje@gnu.org>
Update timestamps.
* config.sub: 2020-10-13 (repaired)
From: Ben Elliston <bje@gnu.org>
* config.sub, config.guess: Replace backtick `..` substitutions
with POSIX $(..) command substitutions throughout.
* Makefile (shellcheck): Don't exclude message SC2006.
* config.guess: 2020-10-22 (repaired)
From: Ben Elliston <bje@gnu.org>
* config.sub, config.guess: Replace backtick `..` substitutions
with POSIX $(..) command substitutions throughout.
* Makefile (shellcheck): Don't exclude message SC2006.
2020-10-31
* AcSplit/CF_CLANG_COMPILER:
split-out check for -Qunused-arguments to ensure that clang supports it.
IBM xlclang does not (report by Steven Pitman)
2020-10-24
* AcSplit/CF_BOOL_SIZE: in fallback, assume bool is unsigned
* AcSplit/CF_BOOL_SIZE:
use AC_CHECK_SIZEOF to get the actual size, which works when cross-compiling
2020-10-21
* config.guess: 2020-10-22
From: Rin Okuyama <rin@netbsd.org>
* config.guess (*:NetBSD:*:*): Handle aarch64eb.
* testsuite/config-guess.data: Add test cases.
Signed-off-by: Ben Elliston <bje@gnu.org>
2020-10-14
* config.sub: 2020-10-13
From: Ben Elliston <bje@gnu.org>
Fix whitespace problem in config.sub.
2020-10-13
* config.sub: 2020-10-13
From: Ben Elliston <bje@gnu.org>
* config.sub (i*86-pc-os2-emx): Recognise correctly.
* testsuite/config-sub.data: Add OS/2 tests to avoid regressions.
2020-10-11
* AcSplit/CF_CPP_DOES_COMMENTS: filter out error-messages
* AcSplit/CF_CPP_DOES_COMMENTS: ci-aclocal generated header
2020-10-04
* AcSplit/CF_WITHOUT_X:
cancel some related macros, to simplify script for cindent
2020-10-03
* AcSplit/CF_WITH_ENCODINGS_DIR: build-fixes for macOS
* AcSplit/CF_WITH_LOCALE_ALIAS:
make this less verbose, like the check for encodings directory
* AcSplit/CF_WITH_LOCALE_ALIAS:
OpenBSD puts locale.alias in a "share" directory rather than the conventional
"lib". Search both.
2020-09-26
* config.sub: 2020-09-08
From: Fabrice Fontaine <fontaine.fabrice@gmail.com>
* config.sub (uclinux-uclibc*): Fix detection.
* testsuite/config-sub.data: Add a test case to avoid regression.
Signed-off-by: Ben Elliston <bje@gnu.org>
* AcSplit/CF_REGEX: add gnurx to $LIBS on success
2020-09-20
* config.guess: 2020-09-19
From: Bruno Haible <bruno@clisp.org>
* config.guess: Don't use 'ldd --version' to determine the presence of
musl libc, as this fails on Alpine Linux 3.10.
Signed-off-by: Ben Elliston <bje@gnu.org>
2020-09-12
* AcSplit/CF_REGEX:
change special case for mingw's systre/tre library to check for dependencies
2020-09-08
* AcSplit/CF_MAKE_PHONY: ci-aclocal generated header
* AcSplit/CF_CRYPT_FUNC, AcSplit/CF_KILLPG,
AcSplit/CF_RESTARTABLE_PIPEREAD:
assume standard C
2020-09-07
* config.sub: 2020-09-08
From: Elad Lahav <e2lahav@gmail.com>
* config.sub: Fix regression in QNX recognition.
* testsuite/config-sub.data: Add some test cases.
Signed-off-by: Ben Elliston <bje@gnu.org>
2020-09-01
* AcSplit/CF_MERGE_EXTRA_CFLAGS: RCS_BASE
2020-08-29
* AcSplit/CF_GCC_ATTRIBUTES: g++ (i.e., $GXX) should be checked
* AcSplit/CF_ENABLE_WARNINGS:
ncurses (any library) should check for attributes whether or not warnings
are checked.
2020-08-28
* AcSplit/CF_WITH_SCREEN_PDCURSES: ci-aclocal generated header
* AcSplit/CF_GCC_WARNINGS:
don't reset $EXTRA_CFLAGS, to work with updated CF_ENABLE_WARNINGS, etc.
* AcSplit/CF_FIX_WARNINGS:
ensure EXTRA_CFLAGS is substituted, and fix quoting of progress-messages
* AcSplit/CF_FIX_WARNINGS: ci-aclocal generated header
* AcSplit/CF_CLANG_COMPILER, AcSplit/CF_ENABLE_WARNINGS:
trim spurious gcc -Werror options which interfere with configure checks
2020-08-22
* AcSplit/CF_SHARED_OPTS: add quoting
* AcSplit/CF_SHARED_OPTS:
add cases for HPE NonStop platforms (Randall S. Becker)
* AcSplit/CF_SHARED_OPTS:
for the NetBSD case, use gcc's "-shared" option for the non-rpath case
(like the rpath case) since recent compiler configurations provide a
different version of the crt begin/end objects when using "-shared"
(report by Rajeev V. Pillai)
2020-08-16
* config.guess, config.sub: 2020/08/17
* AcSplit/CF_SHARED_OPTS:
adapt Maarten's change to strip runtime-library options for msvc linker
2020-07-30
* AcSplit/CF_MAWK_FIND_MAX_INT: provide a definition for MAXULONG
also, minor test-cleanup
* AcSplit/CF_MAWK_MAX__INT_PROGRAM: provide a definition for MAX__ULONG
also, make the variables static, to work around buggy/unspecified behavior
in gcc-6 (if not static, the loop doesn't exit because "0 > 0"!).
* AcSplit/CF_SRAND: quiet some unused-variable warnings
2020-07-25 anonymous.maarten
* AcSplit/CF_SHARED_OPTS:
build-fix for msvc (see https://github.com/mirror/ncurses/pull/1)
2020-07-12
* config.guess: 2020-07-12
2020-07-11 anonymous.maarten
* AcSplit/CF_SHARED_OPTS: specify subsystem for Visual Studio linking
2020-07-11 juergen
* AcSplit/CF_REGEX: add systre for MinGW libs
2020-07-10
* config.sub: 2020-07-10
2020-07-04
* AcSplit/CF_GNAT_GENERICS: allow for gnat 10.x to 99.x
2020-06-02
* AcSplit/CF_FUNC_TGETENT:
allow the test-compile to use termcap.h, noting that only a couple of
special cases (which might conflict with each other) use this header.
2020-05-31
* AcSplit/CF_GNAT_GENERICS:
allow for generics with gnat-10 (to gnat-99)
2020-05-23 anonymous.maarten
* AcSplit/CF_SHARED_OPTS:
for ncurses compiled with Visual Studio, generate a script "mk_prog.sh"
which translates the linker options in the $LINK_PROGS variable to
Visual Studio options.
See https://github.com/mirror/ncurses/pull/1
"Minimum required patches for building ncurses libraries on Visual Studio #1"
2020-05-23
* AcSplit/CF_CHECK_GNAT_VERSION: allow for gnat 10.x to 99.x
2020-05-22
* AcSplit/CF_GCC_OPT_RDYNAMIC:
allow clang to support -rdynamic since current versions (checking MacOS
where this was a problem in 2014) appear to work. This probably happened
as a side-effect of the other BSDs deprecating gcc in favor of clang,
which hadn't had much effect until 2018-2019.
2020-04-04
* AcSplit/CF_ADD_CXXFLAGS: ci-aclocal generated header
* AcSplit/CF_ADD_CFLAGS: document parameters
* AcSplit/CF_CHECK_FVISIBILITY:
redo this as a configure-check, making it useful for C++ also
2020-04-04 anonymous.maarten
* AcSplit/CF_AR_FLAGS: add case for msvc
* AcSplit/CF_CHECK_FVISIBILITY: ci-aclocal generated header
* AcSplit/CF_LIB_PREFIX: add case for msvc
* AcSplit/CF_LIB_SUFFIX: add case for msvc (Microsoft Visual C++)
* AcSplit/CF_LIB_RULES, AcSplit/CF_SHARED_OPTS:
add case for msvcdll (Microsoft Visual C++ DLL)
2020-03-26
* AcSplit/CF_CURSES_EXIT: ci-aclocal generated header
2020-03-19
* AcSplit/CF_CURSES_CHTYPE, AcSplit/CF_SIZECHANGE,
AcSplit/CF_STRUCT_TERMIOS:
fix an uninitialized-variable warning
* AcSplit/CF_CURSES_CHTYPE, AcSplit/CF_CURSES_TERM_H,
AcSplit/CF_CURSES_UNCTRL_H, AcSplit/CF_CURSES_WACS_MAP,
AcSplit/CF_CURSES_WACS_SYMBOLS, AcSplit/CF_CURSES_WGETPARENT,
AcSplit/CF_LARGEFILE, AcSplit/CF_MBSTATE_T,
AcSplit/CF_NL_LANGINFO_1STDAY, AcSplit/CF_SIZECHANGE,
AcSplit/CF_STRUCT_TERMIOS, AcSplit/CF_TERM_HEADER,
AcSplit/CF_UNION_WAIT, AcSplit/CF_WIDEC_CURSES, AcSplit/CF_XOPEN_CURSES:
cleanup unnecessary unused-variable warnings
2020-03-10
* AcSplit/AM_LANGINFO_CODESET, AcSplit/CF_CHECK_ERRNO,
AcSplit/CF_CONST_X_STRING, AcSplit/CF_FUNC_TGETENT,
AcSplit/CF_GCC_ATTRIBUTES, AcSplit/CF_INPUT_METHOD,
AcSplit/CF_LASTLOG, AcSplit/CF_POSIX_SAVED_IDS, AcSplit/CF_POSIX_WAIT,
AcSplit/CF_REGEX, AcSplit/CF_SIGWINCH, AcSplit/CF_SIG_ATOMIC_T,
AcSplit/CF_STRUCT_LASTLOG, AcSplit/CF_TERMIOS_TYPES,
AcSplit/CF_TERMIO_C_ISPEED, AcSplit/CF_TTY_GROUP,
AcSplit/CF_TYPE_CC_T, AcSplit/CF_TYPE_FD_SET, AcSplit/CF_UTMP,
AcSplit/CF_UTMP_UT_HOST, AcSplit/CF_UTMP_UT_NAME,
AcSplit/CF_UTMP_UT_SESSION, AcSplit/CF_UTMP_UT_SYSLEN,
AcSplit/CF_UTMP_UT_XSTATUS, AcSplit/CF_UTMP_UT_XTIME,
AcSplit/CF_XKB_BELL_EXT, AcSplit/CF_XKB_KEYCODE_TO_KEYSYM,
AcSplit/CF_X_ATHENA, AcSplit/CF_X_FONTCONFIG, AcSplit/CF_X_FREETYPE,
AcSplit/CF__GRANTPT_BODY, AcSplit/CF__GRANTPT_HEAD:
cleanup unnecessary unused-variable warnings
* AcSplit/CF_X_TOOLKIT:
AC_PATH_XTRA is supposed to set X_PRE_LIBS and X_EXTRA_LIBS;
use that to simplify this check
2020-03-07
* AcSplit/CF_LIB_RULES, AcSplit/CF_SRC_MODULES:
Allow for overriding the basename of a library
* AcSplit/CF_MAP_LIB_BASENAME, AcSplit/CF_WITH_LIB_BASENAME:
ci-aclocal generated header
2020-03-03
* AcSplit/CF_WITHOUT_X: ci-aclocal generated header
2020-02-29 anonymous.maarten
* AcSplit/CF_SHARED_OPTS: add a case for msys2+msvc
* AcSplit/CF_WITH_PCRE2:
correct header filename: the pcre2-posix library has a "-", but not the header
2020-02-27
* AcSplit/CF_NCURSES_CONFIG: improve workaround for "-W" option
2020-02-25
* AcSplit/CF_NCURSES_CONFIG:
ncurses 6.2 config/pkg scripts may give -W linker options which are mishandled
by Apple's c89/c99 wrappers for clang. Just filter out that special case to
allow applications to build (possibly not link, since Apple's wrapper precludes
that possibility).
2020-02-08
* AcSplit/CF_CURSES_CHECK_DATA: split-out CF__CURSES_DATA
* AcSplit/CF__CURSES_DATA: use const to quiet warnings in config.log
* AcSplit/CF_ENABLE_BROKEN_LINKER: ci-aclocal generated header
* AcSplit/CF_LINK_DATAONLY: for consistency, ac-subst "BROKEN_LINKER"
* AcSplit/CF__CURSES_DATA: ci-aclocal generated header
2020-02-07
* AcSplit/CF_CURSES_CHECK_DATA:
workaround to make tack run with cygwin: the broken-linker feature confused
this test because it tried taking the address of a function return-value.
The workaround tests if the symbol is #define'd, and in that case settles
for just being able to cast the result as a void*. Thus, a check for
boolnames casts an address to an address. Some additional work is needed
to deal with curses-data which is not already an address with nonzero value.
2020-01-18
* AcSplit/CF_SIG_ATOMIC_T, AcSplit/CF_MAWK_RUN_FPE_TESTS,
AcSplit/CF_NSS_COMPAT, AcSplit/CF_RESTARTABLE_PIPEREAD:
spelling-fixes
2020-01-17
* AcSplit/CF_RESTARTABLE_PIPEREAD: fix typo in comment
2020-01-16
* AcSplit/CF_X_ATHENA_CPPFLAGS:
fix the last change - CPPFLAGS was not updated properly
2020-01-11
* COPYING.asc: sign
* COPYING: bump
* AcSplit/CF_CONST_X_STRING: improve bracketing of save/restore
* AcSplit/CF_X_ATHENA_LIBS, AcSplit/CF_X_ATHENA_CPPFLAGS:
when test-compiling, ensure we're taking $X_CFLAGS into account
* AcSplit/CF_RESTORE_XTRA_FLAGS, AcSplit/CF_SAVE_XTRA_FLAGS:
ci-aclocal generated header
* AcSplit/CF_CONST_X_STRING, AcSplit/CF_SYSV, AcSplit/CF_XKB_BELL_EXT:
when test-compiling, ensure we're taking $X_CFLAGS into account
2020-01-09
* AcSplit/CF_BUNDLED_CANLOCK, AcSplit/CF_FIND_CANLOCK:
ci-aclocal generated header
2020-01-05
* AcSplit/CF_GCC_WARNINGS: suppress -Winline (too noisy in gcc9)
ignore clang-specific options when c89/c99 wrappers would treat as error
* AcSplit/CF_LEX_CHAR_CLASSES, AcSplit/CF_LEX_OPTIONS,
AcSplit/CF_LEX_POINTER, AcSplit/CF_LEX_STATES:
cleanup MacOS conftest.dSYM
2020-01-01
* AcSplit/CF_FIND_PCRE: ci-aclocal generated header
* AcSplit/CF_BUNDLED_PCRE:
use CF_FIND_PCRE for the normal --with-pcre case (no option value) in tin,
and use that to decide whether to fallback to the bundled pcre library.
2019-12-31
* AcSplit/AM_MULTIBYTE_ABLE, AcSplit/CF_CHECK_NESTED_PARAMS,
AcSplit/CF_COMPTYPE, AcSplit/CF_COREFILE, AcSplit/CF_CPP_CONCATS,
AcSplit/CF_CPP_EXPANDS, AcSplit/CF_DEFAULT_SHELL,
AcSplit/CF_DEFINE_STRING, AcSplit/CF_FUNC_SYSTEM,
AcSplit/CF_PATH_EDITOR, AcSplit/CF_PATH_MAILBOX,
AcSplit/CF_PATH_MAILER, AcSplit/CF_POSIX_JC, AcSplit/CF_PROG_SUM_R,
AcSplit/CF_PW_GECOS, AcSplit/CF_SIG_ARGS, AcSplit/CF_SIG_CONST,
AcSplit/CF_SYS_SELECT_TIMEVAL, AcSplit/CF_TERMIOS,
AcSplit/CF_TIOCGWINSZ, AcSplit/CF_WITH_DFTENV,
AcSplit/CF_WITH_PROGRAM, AcSplit/CF_WITH_VALUE,
AcSplit/CF_FCNTL_VS_IOCTL, AcSplit/CF_POSIX_VDISABLE,
AcSplit/CF_WITH_ENCODINGS_DIR, AcSplit/CF_NL_LANGINFO_1STDAY,
AcSplit/CF_FUNC_VASPRINTF, AcSplit/CF_GNUTLS, AcSplit/CF_PDCURSES_W32,
AcSplit/CF_FOPEN_BIN_R, AcSplit/CF_WITH_X11_RGB:
appease autoheader-252
* AcSplit/CF_CHECK_GNAT_VERSION, AcSplit/CF_GNAT_GENERICS,
AcSplit/CF_GNAT_PROJECTS, AcSplit/CF_GNAT_VERSION, AcSplit/CF_PROG_GNAT:
correct name, per cache-variable naming convention
* AcSplit/CF_HELP_MESSAGE:
for the moment, null this macro if one tries to use autoconf-2.69
* AcSplit/CF_PROG_CC:
humor autoconf-2.69 (its dependency check is broken)
2019-12-26
* AcSplit/CF_TRIM_LDFLAGS: ci-aclocal generated header
2019-12-22
* AcSplit/CF_PIC_HACK: ci-aclocal generated header
2019-12-20
* config.guess: 2019-12-21
2019-12-18
* AcSplit/CF_SRC_MAKEFILE:
for platforms where "make" doesn't know about archive rules, sleep a second
after updating the archive to help "make" know that it may be older than
the next object to be compiled.
* AcSplit/CF_MAKE_AR_RULES: update description
2019-12-17
* AcSplit/CF_MAKE_AR_RULES:
make the verbose-message a little clearer, and also improve the file to
be compiled to make it unlikely it will generate warnings which might
cause the test to fail. Tested with NetBSD's non-POSIX make-program...
2019-12-11
* COPYING.asc: sign
* COPYING: bump
2019-12-06
* AcSplit/CF_RCS_SCCS:
change SCCS_PATH to ADMIN_PATH, adding a new check for "sccs" wrapper
since its location in Debian package is obscure.
* AcSplit/CF_CURSES_CCHAR_T: fix typo in comment
2019-11-28
* AcSplit/CF_FUNC_MBSTOWCWIDTH: RCS_BASE
2019-11-20
* AcSplit/CF_PROG_LINT:
provide a suitable default LINT_OPTS for cppcheck
2019-11-02
* AcSplit/CF_ADD_LIBS:
the change to filter out duplicates caused this to append rather than prepend.
revise to fix that.
2019-09-26
* AcSplit/CF_LD_SEARCHPATH: ci-aclocal generated header
* AcSplit/CF_LD_SEARCHPATH:
ensure that the result is nonempty, to avoid problems in scripts
2019-09-21
* AcSplit/CF_SRC_MODULES: add a missing dependency on gnat
* AcSplit/CF_CHECK_GNAT_VERSION: ci-aclocal generated header
* AcSplit/CF_PROG_GNAT:
modify to show where the gnat tools are, and their versions
* AcSplit/CF_GNAT_VERSION: split-out CF_CHECK_GNAT_VERSION
2019-09-11
* config.guess, config.sub: 2019-09-10
2019-09-07
* AcSplit/CF_WITH_ADA_LIBNAME: ci-aclocal generated header
* AcSplit/CF_GCC_WARNINGS:
move checks for icc/clang into CF_GCC_VERSION
* AcSplit/CF_GCC_VERSION:
move checks to distinguish icc/clang from gcc from the macro which handles
the --enable-warnings option, to make this work without-warnings for the
inline-checks.
* AcSplit/CF_C_INLINE:
require CF_GCC_VERSION to make checks for icc/clang work when no
--enable-warnings option was used
2019-07-31
* AcSplit/CF_FUNC_VASPRINTF: ci-aclocal generated header
2019-07-21
* AcSplit/CF_TERMIOS_TYPES: ci-aclocal generated header
2019-07-18
* AcSplit/CF_X_EXT_DOUBLE_BUFFER: ci-aclocal generated header
2019-06-30
* AcSplit/CF_WITH_PATH_PROG: ci-aclocal generated header
* AcSplit/CF_PATH_PROG: provide for $PATH parameter
* config.sub: 2019-06-30
2019-06-23
* AcSplit/CF_CHECK_GETENV: ci-aclocal generated header
2019-06-16
* AcSplit/CF_GCC_WARNINGS:
suppress the check for CONST_X_STRING unless we're looking for X
2019-06-10
* config.guess: 2019-06-10
2019-05-22
* config.sub: 2019-05-22
2019-04-08
* AcSplit/CF_CONST_X_STRING: ci-aclocal generated header
* AcSplit/CF_GCC_WARNINGS:
factor out workaround for XTSTRINGDEFINES as CF_CONST_X_STRING
2019-03-30
* AcSplit/CF_FOPEN_BIN_R: ci-aclocal generated header
2019-03-28
* config.sub: 2019-03-28
2019-03-23
* AcSplit/CF_SIGWINCH, AcSplit/CF_X_TOOLKIT:
fix typo in comment (Larry Hynes)
2019-03-07
* config.guess: 2019-03-07
2019-01-26
* AcSplit/CF_WITH_DFTENV:
eliminate a "dnl" which caused a "case" statement to be joined to the preceding
line when using autoconf-2.52+patches.
2019-01-14
* config.guess: 2019-01-15
2019-01-04
* config.sub: 2019-01-05
2019-01-02
* config.guess: 2019-01-03
2018-12-31
* AcSplit/CF_POSIX_VISIBLE: RCS_BASE
* AcSplit/CF_POSIX_C_SOURCE, AcSplit/CF_XOPEN_SOURCE:
add/use CF_POSIX_VISIBLE
2018-12-10
* AcSplit/CF_GNU_SOURCE:
The check for _DEFAULT_SOURCE should apply to "recent" Cygwin (since early 2016),
and except for "NEWLIB" vs "GLIBC" in the test, acts the same if I pretend
that "newlib" is the GNU C library. Without this, the check falls through
to the _XOPEN_SOURCE test, which breaks the pseudoterminal checks for xterm.
2018-12-04
* AcSplit/CF_VA_COPY: fix typos (report by Urs Jansen)
2018-12-03
* AcSplit/CF_VA_COPY:
add another fallback check, for __builtin_va_copy(), which could be used
with AIX xlc in c89 mode (report by Urs Jansen).
* AcSplit/CF_VA_COPY:
add fallback check if neither va_copy/__va_copy is supported, to try
copying the pointers for va_list, or as an array (prompted by discussion
with Urs Jansen).
2018-12-01
* config.guess, config.sub: 2018-11-28
2018-11-10
* AcSplit/CF_PROG_GNAT:
workaround for Fedora, whose gprconfig lacks "GNATGCC"
* AcSplit/CF_PROG_GNAT:
Debian9's gprconfig for "C" gives misleading results since it has a stub
for "LLVM" (clang) and a more complete set including "GNATGCC". The former
does not work with the C files in ncurses's Ada95 tree. Work around this
by special-casing a selection for GNATGCC.
2018-11-07 Urs.Jansen
* AcSplit/CF_XOPEN_SOURCE:
add midnightbsd to special cases with freebsd
2018-10-20
* AcSplit/CF_WITH_VERSIONED_SYMS:
allow option's value to be a relative pathname
2018-08-29
* config.guess, config.sub: 2018-08-29
2018-08-18
* AcSplit/CF_SHARED_OPTS: suppress $LD_RPATH_OPT if --disable-rpath
* AcSplit/CF_LD_RPATH_OPT: suppress check if --disable-rpath
* AcSplit/CF_INSTALL_OPTS: add CF_INSTALL_OPT_P
* AcSplit/CF_INSTALL_OPT_P: ci-aclocal generated header
* AcSplit/CF_INSTALL_OPT_S: improve help-text
* AcSplit/CF_LIB_RULES: pass ReRanlib value to mk-1st.awk
2018-08-13
* COPYING.asc: sign
* COPYING: bump
2018-08-12
* AcSplit/CF_FUNC_GRANTPT:
amend the posix_openpt vs grantpt change to work with BSDs
2018-07-29
* AcSplit/CF_CC_ENV_FLAGS:
putting preprocessor flags in CFLAGS also is a nuisance, which can be
addressed in the same way.
2018-07-21
* AcSplit/CF_WITH_ADA_SHAREDLIB:
disable the shared-library feature if GNAT projects are unsupported, e.g.,
if gprbuild is not available.
2018-07-14
* AcSplit/CF_WITH_PCRE2:
modified to work with Debian's (mis-numbered) pcre3 package.
2018-06-28
* AcSplit/CF_FUNC_GRANTPT:
add special-case to avoid using OpenBSD's nonfunctional POSIX-interface.
2018-06-27
* AcSplit/CF_WITH_MAN2HTML:
use sed to work around non-POSIX tail utility
2018-06-26
* AcSplit/CF_FUNC_GRANTPT:
if posix_openpt was found, prefer that since xterm relies upon the features
of grantpt which would be cancelled if openpty was found.
2018-06-20
* AcSplit/CF_PDCURSES_X11:
use CF_ADD_CFLAGS rather than direct assignment to CPPFLAGS
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_BUNDLED_INTL,
AcSplit/CF_CURSES_CPPFLAGS, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_FIND_LINKAGE, AcSplit/CF_GCC_WARNINGS,
AcSplit/CF_GNU_SOURCE, AcSplit/CF_LARGEFILE,
AcSplit/CF_NCURSES_CONFIG, AcSplit/CF_POSIX_C_SOURCE,
AcSplit/CF_PROG_EXT, AcSplit/CF_SIZECHANGE,
AcSplit/CF_TRY_XOPEN_SOURCE, AcSplit/CF_XOPEN_CURSES,
AcSplit/CF_X_ATHENA_CPPFLAGS:
use CF_APPEND_TEXT rather than direct assignment to CPPFLAGS
2018-06-16
* AcSplit/CF_GNU_SOURCE:
using "recent" changes for _DEFAULT_SOURCE, follow original goal of getting
a usable value for _XOPEN_SOURCE, to reduce need for _GNU_SOURCE (Debian
#900987)
* AcSplit/CF_XOPEN_SOURCE:
pass current _XOPEN_SOURCE parameter directly to CF_GNU_SOURCE
* config.guess: 2018-05-19
* config.sub: 2018-05-24
2018-06-08
* AcSplit/CF_SIZECHANGE: address an old fixme-comment
* AcSplit/CF_STRUCT_TERMIOS:
consistently check for sys/ioctl.h and sys/termio.h
* AcSplit/CF_SIZECHANGE:
drop a 1997 to-do for HPUX 10.x (which was tested long ago), and add a URL
pointing to my comments on the history of ttysize vs winsize.
2018-05-24
* AcSplit/CF_YACC_ERROR:
fixes to work with clang, which noticed (unlike gcc) that yaccError()
was returning a value, as well as to eliminate a missing cast to quiet warning.
2018-04-30
* AcSplit/CF_WITH_ICON_THEME: comment
2018-02-21
* AcSplit/CF_TRY_RPATH, AcSplit/CF__CHECK_RUN:
ci-aclocal generated header
* AcSplit/CF_MAKEFLAGS:
make a failure message less obvious by redirecting it to config.log
This is useful for NetBSD, whose implementation of "make" puts spurious
text into $MAKEFLAGS
* AcSplit/AM_WITH_NLS: whitespace
* AcSplit/CF_CHECK_SIZEOF:
workaround for broken configurations where AC_TRY_RUN fails
2018-02-05
* config.sub: 2018-01-15
* config.guess: 2018-01-26
2018-01-16
* AcSplit/CF_PROG_GNAT:
fix a choice between gnatmake and gprconfig when gprconfig is missing
2018-01-15
* AcSplit/CF_PROG_GNAT:
check for gprconfig's name for the Ada and C compilers
2018-01-14
* AcSplit/CF_GNAT_PROJECTS:
ensure that directory for checks is created before using it
* AcSplit/CF_PROG_GNAT:
add check for gprconfig to work around tool breakage in Debian 9+
2018-01-08
* AcSplit/CF_SRC_MAKEFILE:
add $(ARFLAGS) to generated library rule, overlooked with OpenBSD and Debian
because their make-program supports POSIX "library" rules. The make-program
provided by NetBSD 7.1 does not support the feature.
2018-01-07
* AcSplit/CF_PROG_GROFF:
test for mandoc, which is provided in lieu of nroff on occasion
* AcSplit/CF_WITH_MAN2HTML:
use the $TBL_PATH and $NROFF_PATH values from CF_PROG_GROFF, in case we have
only mandoc.
2018-01-04
* AcSplit/CF_BUILD_CC:
add clang, c99 and c89 as programs to check for build-cc
2018-01-03
* AcSplit/CF_CHECK_CURSES_LIB: integrate with pkg-config
* AcSplit/CF_NCURSES_CONFIG:
only check for term.h when the curses library check was successful
* AcSplit/CF_CURSES_FUNCS: quiet compiler warning in configure checks
* AcSplit/CF_CURSES_CHECK_DATA:
quiet compiler-warning in configure checks
2017-12-31
* AcSplit/CF_CURSES_LIBS:
set $cf_nculib_root as a side-effect, needed for pthreads test in ncurses-examples
2017-12-30
* AcSplit/CF_SHARED_OPTS:
consistently add $CFLAGS to $MK_SHARED_LIB symbol in configure script when the
latter happens to use the C compiler rather than directly using the loader
(report by Jeffrey Walton).
2017-12-26
* AcSplit/CF_NCURSES_CONFIG:
modify search-list for ncurses*-config to prefer my "dev" version by adding
a variable $NCURSES_CONFIG_SUFFIX to transform ncurses6-config into
ncurses6dev-config since Fedora27 has an ncurses5-config and ncurses6-config
concurrently.
2017-12-24
* AcSplit/CF_WITH_MAN2HTML:
updated to use my improved man2html script by default
2017-12-02
* AcSplit/CF_SHARED_OPTS:
move ${LDFLAGS} from makefile into the definition of MK_SHARED_LIB to help
work around problems in GNU linker if $LDFLAGS happens to use the "-pie"
option (prompted by patch by Yogesh Prasad). Also fix a couple of unquoted
"$@" tokens.
2017-11-25
* AcSplit/CF_WITH_X11_RGB: RCS_BASE
* config.guess: 2017-11-07
* config.sub: 2017-11-23
2017-10-16
* AcSplit/CF_MAWK_FIND_MAX_INT, AcSplit/CF_MAWK_MAX__INT_PROGRAM:
quiet strict gcc warning
2017-09-29
* AcSplit/CF_ENABLE_WARNINGS: pass parameter to CF_GCC_WARNINGS
2017-09-23
* AcSplit/CF_CC_ENV_FLAGS:
for tin, which still uses 2.13, quote "subst" to avoid conflict between
m4's builtin and awk's built-in
2017-09-07
* AcSplit/CF_CURSES_LIBS:
reorganize to handle the case where initscr is a macro
2017-08-12
* AcSplit/CF_MAN_PAGES, AcSplit/CF_WITH_LIBTOOL:
use SIGQUIT(3) rather than SIGTRAP(5) in trap commands
2017-08-08
* config.guess: 2017-08-08
2017-07-29 David.Michael
* AcSplit/CF_WITH_PCRE2: RCS_BASE
* AcSplit/CF_WITH_PCRE: whitespace fixes
2017-07-28
* AcSplit/CF_CURSES_CHECK_DATA:
change the first parameter to be a list (one or more items) rather than a
single item, and add a second parameter to allow quitting the loop if a
match is found.
2017-07-26
* AcSplit/CF_LIB_RULES, AcSplit/CF_SRC_MODULES:
build-fix for tack add-on
2017-07-23
* AcSplit/CF_NCURSES_CONFIG: add check for term.h when using pkg-config
* AcSplit/CF_LINK_DATAONLY: gcc warning
2017-07-11
* AcSplit/AM_WITH_NLS: build-fix for lynx
2017-07-10
* AcSplit/CF__INTL_BODY:
ifdef'd the GNU-specific symbol to allow configuring with Solaris gettext
* AcSplit/AM_WITH_NLS:
revised to allow configuring with Solaris gettext, and if neither GNU gettext
nor Solaris gettext is available, warn and disable the NLS feature.
2017-07-02
* AcSplit/CF_PROG_CPP_COMMENTS: ci-aclocal generated header
2017-05-10
* AcSplit/CF_FUNC_CURSES_VERSION, AcSplit/CF_FUNC_GETADDRINFO,
AcSplit/CF_SET_ERRNO:
main-prototype
2017-05-09
* AcSplit/CF_NCURSES_VERSION: use prototype for main
2017-04-29
* config.sub: 2017-04-02
* AcSplit/CF_WITH_GPM:
in the previous fixes, overlooked a case where $with_gpm would be leftover
as "maybe", which breaks callers that only know about "yes" and "no"
2017-03-18
* config.sub: 2017-02-07
* config.guess: 2017-03-05
2017-02-25
* AcSplit/CF_APPEND_TEXT: ci-aclocal generated header
* AcSplit/CF_CC_ENV_FLAGS:
correct two problems, the first reported by Ross Burton:
a) revise the expression generating $cf_flags to actually work when there is
more than one option.
b) correct the parameter passed to CF_ADD_CFLAGS, which was the whole list
rather than the current parameter.
* AcSplit/CF_ADD_CFLAGS:
improve formatting of generated lists using CF_APPEND_TEXT
2017-02-11
* AcSplit/CF_SHARED_OPTS: workaround for PGI compiler
* AcSplit/CF_GXX_VERSION:
workaround for PGI compiler, which fails at pretending to be g++
2017-01-21
* COPYING.asc: sign
* COPYING: bump
* AcSplit/CF_MKSTEMP, AcSplit/CF_LINK_FUNCS: quiet strict gcc warning
* AcSplit/CF_CHECK_WCHAR_H:
FreeBSD wchar.h doesn't include wctype.h; do that to quiet warning in config.log
* AcSplit/CF_BOOL_SIZE, AcSplit/CF_CHECK_GPM_WGETCH,
AcSplit/CF_CPP_PARAM_INIT, AcSplit/CF_FUNC_MEMMOVE,
AcSplit/CF_FUNC_NANOSLEEP, AcSplit/CF_LIB_SONAME,
AcSplit/CF_LINK_DATAONLY, AcSplit/CF_MATH_LIB, AcSplit/CF_PROG_CC_C_O,
AcSplit/CF_TYPEOF_CHTYPE:
quiet strict gcc warning
* AcSplit/CF_CGETENT, AcSplit/CF_LARGEFILE:
workaround for clang exit-code vs warnings
2016-12-31
* config.guess, config.sub:
2017-01-01, see again misuse of copyright dates...
* AcSplit/CF_WITH_LIBTOOL:
fix a few places in configure/build scripts where DESTDIR and rpath were
combined (report by Thomas Klausner).
2016-11-20
* config.sub: 2016-11-19
* config.guess: 2016-10-02
2016-10-08
* AcSplit/CF_ENABLE_STRING_HACKS:
add the HAVE_STRLCAT definition when AC_CHECK_FUNC succeeds
2016-09-10
* AcSplit/CF_ENABLE_STRING_HACKS:
extended to find BSD library, e.g., with Debian
2016-09-05
* AcSplit/CF_SRAND: eliminate a compiler warning with Solaris C
2016-08-29
* AcSplit/CF_CC_ENV_FLAGS:
amend the last change to move only the preprocessor, optimization and warning
flags to CPPFLAGS and CFLAGS, leaving the residue in CC. That happens to
work for gcc's various "model" options, but may require tuning for other
compilers (report by Sven Joachim).
2016-08-27
* AcSplit/CF_XOPEN_SOURCE:
add "uclinux" to list of Linux's (patch by Yann E. Morin)
* AcSplit/CF_CC_ENV_FLAGS:
improve split between compiler and options, prompted by report where user
had "ccache" before the compiler
* config.guess: 2016-05-25
* config.sub: 2016-08-25
2016-08-10
* AcSplit/CF_XOPEN_SOURCE:
use _GNU_SOURCE for cygwin headers, tested with cygwin 2.3, 2.5 (patch by
Corinna Vinschen).
2016-08-05
* AcSplit/CF_GNU_SOURCE:
recent glibc (Debian 2.23-4 for example) has misordered ifdef/checks for
new symbol _DEFAULT_SOURCE, producing warning messages when only _GNU_SOURCE
is defined. Add a followup check to define _DEFAULT_SOURCE.
2016-06-25
* AcSplit/CF_CC_ENV_FLAGS:
amend that last change to leave non-preprocessor options in "$CC" (but still
copy them to "$CFLAGS" since that's where they should be)
2016-06-20
* config.sub: 2016-06-20
2016-05-28
* AcSplit/CF_WITH_GPM:
if --with-gpm=maybe (default), allow for linkage problems, e.g., Red Hat's
problems with PIE/PIC
* AcSplit/CF_WITH_XINERAMA: ci-aclocal generated header
2016-05-22
* AcSplit/CF_PROG_LINT:
add cpplint to programs to use; drop ad hoc tdlint and alint.
2016-05-21
* AcSplit/CF_CC_ENV_FLAGS:
don't limit the check to -I, -U and -D options, since the added options can
include various compiler options before and after preprocessor options.
* AcSplit/CF_PROG_CC_C_O:
provide for passing compiler options to this check because some users have
broken configurations requiring compiler options to do anything
2016-05-16
* config.guess: 2016-05-15
2016-04-22
* AcSplit/CF_NCURSES_PTHREADS: eliminate double message
2016-04-21
* AcSplit/CF__ADD_SHLIB_RULES:
improve scripting for ldconfig to avoid warning messages if no permissions
2016-04-15
* AcSplit/CF_GNUTLS:
for lynx, add check for gnutls_protocol_set_priority
2016-02-20
* AcSplit/CF_WITH_NCURSES_ETC: add check for ncurses pthreads
* AcSplit/CF_NCURSES_PTHREADS: ci-aclocal generated header
* AcSplit/CF_LD_RPATH_OPT:
change FreeBSD to use -Wl,-rpath rather than -rpath option. According to
FreeBSD #178732, either works since FreeBSD 4.x; however scons does not
accept anything except the -Wl,-rpath form.
2016-02-08
* AcSplit/CF_NL_LANGINFO_1STDAY: RCS_BASE
2016-01-25
* config.guess, config.sub: 2016-01-01
2015-12-19
* AcSplit/CF_NCURSES_WITH_ABI_VERSION: correct comment
* AcSplit/CF_CHECK_WCWIDTH_GRAPHICS: ci-aclocal generated header
2015-12-12
* AcSplit/CF_XOPEN_CURSES: fix typo
2015-12-10
* AcSplit/CF_XOPEN_CURSES:
improve check, making it define NCURSES_WIDECHAR if we happen to be building
on a platform (such as OSX) where we do not define _XOPEN_SOURCE or
_XOPEN_SOURCE_EXTENDED
2015-11-07
* AcSplit/CF_LIB_RULES:
fix logic for $cf_awk_program to avoid adding rule for resulting.map if
no changes were needed.
2015-11-01
* AcSplit/CF_ENABLE_PC_FILES:
allow build of pc-files without a validly-configured library directory,
leaving it up to the makefile to compensate
2015-10-31
* AcSplit/CF_LIB_RULES:
modify the editing-script which generates resulting.map file, to kill a
label just before a (now) empty section left by deleting a "_*" pattern.
This happens with FreeBSD using clang. To do this, the script now uses awk.
2015-10-24
* AcSplit/CF_WITH_VERSIONED_SYMS: fix quoting for m4
* config.guess: 2015-10-21
* config.sub: 2015-08-20
* AcSplit/CF_WITH_VERSIONED_SYMS:
avoid ncurses build error on ppc64 architecture
the nm tool is reporting _is_missing in Data section for ppc64.
(by comparison ppc64le is reporting it in Text section)
(patch by Michel Normand)
2015-10-17 KO.Myung-Hun
* AcSplit/CF_LIB_PREFIX, AcSplit/CF_STDCPP_LIBRARY,
AcSplit/CF_WITH_LIBTOOL, AcSplit/CF_XOPEN_SOURCE:
build-fixes for OS/2
2015-10-10
* AcSplit/CF_AR_FLAGS:
binutils ar supports a so-called "deterministic" option. Bleeding-edge
Fedora and a few others build "ar" to use this. It breaks makefiles using
archive-rule (such as ded).
* AcSplit/CF_FUNC_POLL:
add a fallback to checking with posix_openpt(), in case someone's running
the configure script in batch mode.
2015-09-28
* AcSplit/CF__DEFINE_SHLIB_VARS:
add escape to fix assignment to LIB_BASENAME
2015-09-26
* AcSplit/CF_LIB_RULES:
remove resulting.map on "distclean" (since it is produced by configure
script) rather than by "clean".
2015-09-12
* AcSplit/CF_STDCPP_LIBRARY:
updated the test-program, omitting ".h" in the include to match "recent"
versions of stdc++, and also add a check to see if the library already
exists in the compiler's standard configuration. The $CXXLIBS variable
can be preset to take advantage of this additional check.
* AcSplit/CF_FUNC_TGETENT, AcSplit/CF_STDCPP_LIBRARY,
AcSplit/CF_FUNC_OPENPTY, AcSplit/CF_FUNC_DLSYM:
add braces around calls to CF_ADD_LIB which follow a test "&&" operator since
the macro uses multiple statements
2015-08-22
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
adapted patch by Joshua Root to allow configuring the pkg-config library
directory even if pkg-config is not found.
2015-08-20
* AcSplit/CF_WITH_MAN2HTML: use configured shell
2015-08-15
* AcSplit/CF_SHARED_OPTS:
cleanup change for --with-shlib-version in 20000706 overlooked disallowing
a simple --without-shlib-version (report by Tommy Alex)
2015-08-05
* AcSplit/CF_LIB_RULES, AcSplit/CF_MAKEFLAGS,
AcSplit/CF_MANPAGE_RENAMES, AcSplit/CF_MAN_PAGES,
AcSplit/CF_SHARED_OPTS:
use $SHELL consistently
2015-07-07
* AcSplit/CF_NCURSES_CONFIG: PKG_CONFIG may simply be unset - fix
2015-07-06
* AcSplit/CF_WITH_LOCALE_ALIAS, AcSplit/CF_WITH_ENCODINGS_DIR:
cosmetic fix
2015-07-04
* AcSplit/CF_MAKE_DOCS: integrate with CF_WITH_MAN2HTML
* AcSplit/CF_PROG_GROFF: add check for symmetry with nroff
omit redundant AC_SUBST
2015-06-27
* AcSplit/CF_LIB_RULES:
tweak for removing deprecated ABI5 symbols when building ABI6
2015-06-06
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR: tidy a checking-message
* AcSplit/CF_WITH_ABI_VERSION: add case to downgrade from 6.0 to 5.9
* AcSplit/CF_NCURSES_ABI_6:
make this message consistent with a similar one
* AcSplit/CF_NCURSES_CONFIG:
amend change for cross-compiling to make it succeed, but show "maybe" in log
* AcSplit/CF_ABI_DEFAULTS:
add filter-syms and with-lp64 option defaults
* AcSplit/CF_NCURSES_WITH_ABI_VERSION: ci-aclocal generated header
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR: fix typo
* AcSplit/CF_WITH_ABI_VERSION: handle conversion from 5.x to 6.0
2015-05-28
* AcSplit/CF_NCURSES_CONFIG: fix typo
2015-05-25
* AcSplit/CF_ADD_INCDIR:
fix an infinite loop when the include-directory to be added does not exist
(Debian #786436)
* AcSplit/CF_NCURSES_CONFIG:
fix for cross-compiling, could just assume ncurses but be consistent and
assume it is not - otherwise why am I running a test-program? (Debian #786436)
2015-05-15
* AcSplit/CF_CURSES_TERMCAP, AcSplit/CF_GCC_OPT_RDYNAMIC,
AcSplit/CF_INSTALL_OPT_O, AcSplit/CF_STDIO_UNLOCKED,
AcSplit/CF_TYPE_OUTCHAR:
cosmetic-fixes
2015-05-10
* AcSplit/AM_LC_MESSAGES, AcSplit/CF_ADD_OPTIONAL_PATH,
AcSplit/CF_ARG_OPTION, AcSplit/CF_BUNDLED_INTL,
AcSplit/CF_CURSES_LIBS, AcSplit/CF_WITH_NO_LEAKS,
AcSplit/CF__ADD_SHLIB_RULES, AcSplit/CF__DEFINE_LIB_TARGET,
AcSplit/CF__INTL_BODY, AcSplit/jm_GLIBC21:
coding-style
2015-05-09
* AcSplit/CF_CHECK_SIZEOF:
fix parameter substitution, in case $ac_cv_type_XX or $ac_cv_sizeof_XX are
set to null (report by Andrew Arensburger, on lynx-dev)
2015-05-03
* AcSplit/CF_WITH_MAN2HTML: fix quoting in comment
2015-05-02
* AcSplit/CF_ABI_DEFAULTS: RCS_BASE
* AcSplit/CF_MAWK_RUN_FPE_TESTS: cosmetic fix
* config.guess: 2015-03-04
* config.sub: 2015-03-08
2015-04-30
* AcSplit/CF_LIB_RULES:
fixup for OSX shared libraries, which have a double-suffix
2015-04-26
* AcSplit/CF_NCURSES_CONFIG: integrate with checks in CF_NCURSES_ADDON
* AcSplit/CF_NCURSES_ADDON:
integrate with pkg-config check in CF_NCURSES_CONFIG
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
suppress checking-for message if --without-pkg-config is used
* AcSplit/CF_PKG_CONFIG:
suppress warning message if --without-pkg-config is used
2015-04-25
* AcSplit/CF_WITH_NCURSES_ETC: add --with-screen option
* AcSplit/CF_CHECK_CURSES_LIB: ci-aclocal generated header
2015-04-24
* AcSplit/CF_NCURSES_CONFIG:
modify to check for ".pc" files first to allow those with a working pkg-config
to use them.
2015-04-23
* AcSplit/CF_NCURSES_CONFIG:
accommodate changes for lynx's "--with-screen" option by assuming no
version, letting that come from ${cf_ncuconfig_root}. To do this, add
a versionless case at the beginning of the list to check for *-config's
* AcSplit/CF_CURSES_HEADER:
change the order of the header-checks, so that subdir-includes are done
last. This might break some platforms where subdirs are needed for ncurses,
but without it, my changes to lynx to allow "--with-screen=ncursesw6" will
not work because this would otherwise find _other_ subdirs not in the
include-directories listed by ncursesw6-config
2015-04-18
* AcSplit/CF_CURSES_HEADER, AcSplit/CF_PROG_EXT: whitespace
* AcSplit/AM_LANGINFO_CODESET, AcSplit/CF_ADD_LIBDIR,
AcSplit/CF_CHECK_ERRNO, AcSplit/CF_CHECK_EXTERN_DATA,
AcSplit/CF_C_INLINE, AcSplit/CF_DISABLE_ECHO, AcSplit/CF_FIND_LINKAGE,
AcSplit/CF_FUNC_OPENPTY, AcSplit/CF_FUNC_VSSCANF,
AcSplit/CF_GNAT_VERSION, AcSplit/CF_HASHED_DB, AcSplit/CF_LARGEFILE,
AcSplit/CF_LDFLAGS_STATIC, AcSplit/CF_LIB_PREFIX,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MANPAGE_RENAMES,
AcSplit/CF_MAN_PAGES, AcSplit/CF_PATH_SYNTAX, AcSplit/CF_PROG_EGREP,
AcSplit/CF_PROG_GNAT, AcSplit/CF_PROG_INSTALL,
AcSplit/CF_PROG_LDCONFIG, AcSplit/CF_REGEX, AcSplit/CF_SIZECHANGE,
AcSplit/CF_SYS_TIME_SELECT, AcSplit/CF_WITH_PTHREAD:
whitespace/indent
2015-04-17
* AcSplit/CF_WITH_NCURSES_ETC: cosmetic change
* AcSplit/CF_TPUTS_PROTO: whitespace
* AcSplit/CF_AWK_BIG_PRINTF, AcSplit/CF_BOOL_SIZE,
AcSplit/CF_CFG_DEFAULTS, AcSplit/CF_CXX_AR_FLAGS,
AcSplit/CF_DISABLE_LIBTOOL_VERSION, AcSplit/CF_ENABLE_PC_FILES,
AcSplit/CF_FIND_SUB_INCDIR, AcSplit/CF_FIND_SUB_LIBDIR,
AcSplit/CF_FIXUP_ADAFLAGS, AcSplit/CF_FUNC_VSSCANF,
AcSplit/CF_GNAT_GENERICS, AcSplit/CF_GNAT_PROJECTS,
AcSplit/CF_GNAT_VERSION, AcSplit/CF_GPP_LIBRARY,
AcSplit/CF_GXX_WARNINGS, AcSplit/CF_HASHED_DB,
AcSplit/CF_LDFLAGS_STATIC, AcSplit/CF_LIB_PREFIX,
AcSplit/CF_LIB_RULES, AcSplit/CF_LIB_SUFFIX, AcSplit/CF_LIB_TYPE,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MANPAGE_RENAMES,
AcSplit/CF_MANPAGE_SYMLINKS, AcSplit/CF_MAN_PAGES,
AcSplit/CF_NCURSES_ABI_6, AcSplit/CF_NUMBER_SYNTAX,
AcSplit/CF_OBJ_SUBDIR, AcSplit/CF_PROG_INSTALL,
AcSplit/CF_PROG_LDCONFIG, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_STDCPP_LIBRARY, AcSplit/CF_VERSION_INFO,
AcSplit/CF_WITH_LIBTOOL, AcSplit/CF_WITH_LIBTOOL_OPTS,
AcSplit/CF_WITH_PATHLIST, AcSplit/CF_WITH_PKG_CONFIG_LIBDIR,
AcSplit/CF_WITH_VERSIONED_SYMS:
eliminate unnecessary "#(vi" markers in autoconf macros by using "(" to begin
case-statement cases (suggested by Jens Schweikhardt)
2015-04-15
* AcSplit/AM_GNU_GETTEXT, AcSplit/AM_PATH_PROG_WITH_TEST,
AcSplit/AM_WITH_NLS, AcSplit/CF_ADD_OPTIONAL_PATH,
AcSplit/CF_CURSES_CPPFLAGS, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_CURSES_TERM_H, AcSplit/CF_CURSES_UNCTRL_H,
AcSplit/CF_CURS_TOUCHLINE, AcSplit/CF_FIND_IPV6_LIBS,
AcSplit/CF_FIND_IPV6_TYPE, AcSplit/CF_GNUTLS, AcSplit/CF_HEADER_PATH,
AcSplit/CF_LIBRARY_PATH, AcSplit/CF_MAKEFLAGS,
AcSplit/CF_NCURSES_HEADER, AcSplit/CF_NCURSES_LIBS,
AcSplit/CF_NETLIBS, AcSplit/CF_NSS_COMPAT, AcSplit/CF_PROG_EXT,
AcSplit/CF_SIZECHANGE, AcSplit/CF_SLANG_CPPFLAGS,
AcSplit/CF_SLANG_LIBS, AcSplit/CF_SRAND, AcSplit/CF_SRAND_PARSE,
AcSplit/CF_SSL, AcSplit/CF_STRUCT_TERMIOS,
AcSplit/CF_SYSTEM_MAIL_FLAGS, AcSplit/CF_TAR_OPTIONS,
AcSplit/CF_TERMCAP_LIBS, AcSplit/CF_TERM_HEADER, AcSplit/CF_WITH_IDNA:
eliminate unnecessary "#(vi" markers in autoconf macros by using "(" to begin
case-statement cases (suggested by Jens Schweikhardt)
2015-04-12
* AcSplit/CF_ADD_CFLAGS, AcSplit/CF_CC_ENV_FLAGS,
AcSplit/CF_ENABLE_NARROWPROTO, AcSplit/CF_FUNC_GRANTPT,
AcSplit/CF_FUNC_TGETENT, AcSplit/CF_GCC_ATTRIBUTES,
AcSplit/CF_GCC_WARNINGS, AcSplit/CF_IMAKE_CFLAGS,
AcSplit/CF_INTEL_COMPILER, AcSplit/CF_LD_RPATH_OPT,
AcSplit/CF_MIXEDCASE_FILENAMES, AcSplit/CF_NO_LEAKS_OPTION,
AcSplit/CF_PATHSEP, AcSplit/CF_PATH_SYNTAX, AcSplit/CF_PKG_CONFIG,
AcSplit/CF_POSIX_C_SOURCE, AcSplit/CF_REGEX, AcSplit/CF_RPATH_HACK_2,
AcSplit/CF_TRIM_X_LIBS, AcSplit/CF_TTY_GROUP, AcSplit/CF_UTMP_UT_NAME,
AcSplit/CF_WITH_APP_CLASS, AcSplit/CF_WITH_DESKTOP_CATEGORY,
AcSplit/CF_WITH_ICON_NAME, AcSplit/CF_WITH_ICON_SYMLINK,
AcSplit/CF_WITH_ICON_THEME, AcSplit/CF_WITH_IMAKE_CFLAGS,
AcSplit/CF_WITH_MAN2HTML, AcSplit/CF_WITH_PCRE,
AcSplit/CF_XBOOL_RESULT, AcSplit/CF_XOPEN_SOURCE, AcSplit/CF_X_ATHENA,
AcSplit/CF_X_FONTCONFIG, AcSplit/CF_X_FREETYPE, AcSplit/CF_X_TOOLKIT:
eliminate unnecessary "#(vi" markers in autoconf macros by using "(" to begin
case-statement cases (suggested by Jens Schweikhardt)
* AcSplit/CF_LIB_RULES, AcSplit/CF_LIB_SUFFIX, AcSplit/CF_SRC_MODULES:
modified for ncurses --with-extra-suffix option
2015-03-02
* AcSplit/CF_ENABLE_NARROWPROTO:
change default for Cygwin to wide prototypes
2015-02-15
* AcSplit/CF_XINERAMA:
make testable result the same with/without pkg-config
* AcSplit/CF_XINERAMA: ci-aclocal generated header
* AcSplit/CF_WITH_MAN2HTML:
revised/extended the checks for man2html options to permit deprecating
groff's broken html conversion. This uses man2html 3.1, which also adds
a "-index" option.
2015-02-10
* COPYING.asc: sign
* COPYING: bump
* AcSplit/CF_AWK_BIGSTRING: prefer test to expr for Minix 3.2
* AcSplit/CF_WIDEC_SHIFT: prefer test to expr, for Minix 3.2
2015-01-21
* AcSplit/CF_WITH_VERSIONED_SYMS:
check if the loader can use wildcards selectively to declare "_" symbols
local after declaring some global. Set WILDCARD_SYMS to reflect test.
* AcSplit/CF_LIB_RULES: incorporate new symbol WILDCARD_SYMS
* AcSplit/CF_SHARED_OPTS, AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
tidy comments
2015-01-10
* AcSplit/CF_ENABLE_PC_FILES:
add a check in case PKG_CONFIG_LIBDIR is not found
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
add check for libdata directory, used in some older variants with FreeBSD
* AcSplit/CF_SHARED_OPTS: from mail by Baptiste Daroussin, this change:
https://svnweb.freebsd.org/ports/head/devel/ncurses/files/patch-aclocal.m4?view=markup&pathrev=376487
avoids using LD directly, for help with cross-compiling. The diff went past
that, by dropping the -Bshareable option. Based on manpage and testing with
FreeBSD 4.9 / 10.0, that option is redundant if -shared is given.
2015-01-03
* AcSplit/CF_LIB_RULES:
improve sed-script; there's more than "NCURSES_" prefixes to account for.
2015-01-02
* AcSplit/CF__DEFINE_SHLIB_VARS: typo
* AcSplit/CF_WITH_APP_DEFAULTS: add paths for OSX
2014-12-22
* AcSplit/CF_WITH_MAN2HTML: ci-aclocal generated header
2014-12-20
* AcSplit/CF_WITH_EXPORT_SYMS:
don't inherit an EXPORT_SYMS value (no need to, yet)
* AcSplit/CF_WITH_VERSIONED_SYMS:
introduce an intermediate ".map" file to allow ncurses to modify the
file with a sed-script
* AcSplit/CF_LIB_RULES:
provide a way to specify abi-version in ".map" files
2014-12-13
* AcSplit/CF_ENABLE_PC_FILES:
allow pc-files to be configured when pkg-config is missing
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
if pkg-config is missing, try to fill in some of the directory location.
* AcSplit/CF_PKG_CONFIG: warn if pkg-config is missing
2014-12-04
* AcSplit/CF_SUBDIR_PATH:
change this to a loop, adding /usr/pkg (NetBSD) and /opt/local (Mac OSX)
2014-11-30
* AcSplit/CF_GNUTLS:
add configure check for gnutls_rnd, not present in 2.8.6
drop -lcrypt if gnutls_rnd is found, to reduce libraries used.
drop -lgnutls-extra, not needed with 2.8.6 or later, e.g. Debian 6
* AcSplit/CF_CHECK_SSL_X509:
partial fix: check for the correct header if this is gnutls
2014-11-15
* AcSplit/CF_WITH_EXPORT_SYMS:
split-out EXPORT_SYMS variable to avoid conflicting use of $LIBTOOL_OPTS
* AcSplit/CF_WITH_LIBTOOL: use top_srcdir
* AcSplit/CF_WITH_VERSIONED_SYMS:
use top_srcdir to find package directory
attempt to support Solaris (won't happen right away, but this is a start)
* AcSplit/CF_DISABLE_LIBTOOL_VERSION:
when using -version-info (the native libtool mode), get the ABI_VERSION
value from $VERSION (which comes from CF_VERSION_INFO)
2014-11-02
* AcSplit/CF_WITH_LIBTOOL_OPTS: append to LIBTOOL_OPTS
* AcSplit/CF_WITH_EXPORT_SYMS, AcSplit/CF_WITH_VERSIONED_SYMS:
ci-aclocal generated header
* AcSplit/CF_WITH_SHARED_OR_LIBTOOL:
incorporate --with-export-syms and --with-versioned-syms options
2014-09-20
* AcSplit/CF_LIB_RULES:
honor install.libs in misc-directory even if install.data is not used
* AcSplit/CF_SRC_MODULES: misc-directory always has a makefile
2014-09-19
* AcSplit/CF_INCLUDE_DIRS:
remove the (1997-era) chunk which sets -I/usr/include, etc., since it seems
that none of my programs rely upon it. However, the report which pointed
this out (Gentoo #522586) is itself relying upon a defective ebuild.
2014-09-01
* AcSplit/CF_XOPEN_SOURCE:
add case for UnixWare (report/discussion with Mark Ryan)
* config.sub: 2014-07-28
2014-08-02
* AcSplit/CF_GNATPREP_OPT_T: ci-aclocal generated header
2014-07-26
* AcSplit/CF_INCLUDE_DIRS:
when srcdir isn't ".", add srcdir to the include-options after ".",
just like the include vs srcdir/include
2014-07-23
* AcSplit/CF_XOPEN_SOURCE:
Minix3.2 ifdef's the POSIX.1-2001 functions inside _NETBSD_SOURCE, even
though it was released 2012-02-29 - appease it. At the same time, turn
on the verbose flag to show that most platforms need platform-specific
define's to get POSIX (sic).
2014-07-22
* AcSplit/CF_GETOPT_HEADER:
add NEED_GETOPT_H definition, since that's the only interesting special case
* AcSplit/CF_ADD_CFLAGS, AcSplit/CF_CHECK_CFLAGS, AcSplit/CF_WITH_PERL:
workaround for ash-shell
2014-07-21
* AcSplit/CF_INSTALL_OPTS, AcSplit/CF_INSTALL_OPT_O,
AcSplit/CF_INSTALL_OPT_S:
ci-aclocal generated header
* AcSplit/CF_AUTO_SYMLINK:
MINIX3.2's Ash dequotes too early to pass tokens without guarding...
* AcSplit/CF_WITH_SYMLINK: tweak needed for MINIX3.2 Ash
* AcSplit/CF_LEX_STATES:
Assume a POSIX shell for the "test" built-in, to work around deficiency of
Ash, whose "expr" built-in lacks the inequality-comparisons.
2014-07-20
* COPYING.asc: regen
* COPYING: bump
2014-07-19
* AcSplit/CF_CURSES_FUNCS, AcSplit/CF_CURSES_CHECK_DATA:
improve workaround for weak-linkage, seems to fix tests with NetBSD 6.1
2014-07-13
* AcSplit/CF_X_TOOLKIT: add workaround for breakage in XQuartz upgrades
* AcSplit/CF_ADD_LIBS:
workaround in CF_X_TOOLKIT uses pkgconfig, whose files generally are
using incomplete dependencies - in turn introducing lots of duplication.
filter out the duplicates.
2014-07-12 Steve.Wall
* AcSplit/CF_X_ATHENA: add --with-Xaw3dxft option
2014-07-12
* AcSplit/CF_PROG_CC, AcSplit/CF_WITH_ICON_THEME: trim-blanks
2014-07-02
* AcSplit/CF_GCC_OPT_RDYNAMIC:
ignore clang - because it does not warn about unsupported options, it cannot
be detected for its support of (or lack of) -rdynamic
2014-06-21
* AcSplit/CF_LIB_SUFFIX:
it is not necessary or even a good idea to make AIX's shared-library
suffixes ".a" -- make it ".so" like everyone else (including the system's
shared libraries).
2014-06-04
* AcSplit/CF_ACVERSION_CHECK:
work around another gratuitous incompatibility introduced in 2.69 (reported
by Ross Burton, openembedded.org)
2014-06-04 openembedded.org
* AcSplit/CF_POPEN_TEST: fix for autoheader
2014-06-01
* AcSplit/CF_DISABLE_GNAT_PROJECTS: RCS_BASE
* AcSplit/CF_GNAT_PROJECTS:
optionally suppress use of the GNAT project-files
2014-05-31
* AcSplit/CF_WITH_ADA_SHAREDLIB: revert
* AcSplit/CF_GNAT_PROJECTS:
revert, except for the chunk about compiler options which appears to be
unnecessary.
2014-05-24 Nicolas.Boulenguez
* AcSplit/CF_GNAT_PROJECTS, AcSplit/CF_WITH_ADA_SHAREDLIB:
Only consider using a project for shared libraries
2014-05-10
* AcSplit/CF_MAKEFLAGS:
workaround for GNU make 4.0 incompatibility with previous releases.
2014-04-19
* config.sub: 2014-04-03
* config.guess: 2014-03-23
2014-04-12
* AcSplit/CF_HASHED_DB_VERSION: allow db6.x
* AcSplit/CF_FIND_SUB_INCDIR, AcSplit/CF_FIND_SUB_LIBDIR:
ci-aclocal generated header
* AcSplit/CF_HASHED_DB:
extend parameter to easily select FreeBSD port-version
2014-03-17
* AcSplit/CF_INTEL_COMPILER:
cleanup the -no-gcc option which was leftover from testing - prcs does not
build with this option.
* AcSplit/CF_PROG_MAKEINFO: ci-aclocal generated header
2014-02-09
* AcSplit/CF_XOPEN_SOURCE:
for Solaris (tested with gcc/g++ 3.4.3 on Solaris 10 and gcc/g++ 4.5.2 on
Solaris 11), suppress the followup check for defining _XOPEN_SOURCE because
it is not needed, as well as because g++ 4.7.3 (no package, used in Sage
for Solaris 10) has some unspecified header breakage which is triggered by
the duplicate definition.
2014-01-11
* AcSplit/CF_SSL:
fix check for ssl cflags (seems that pkg-config was returning a " ", which
confused my later check for /usr/include/openssl).
2014-01-05
* AcSplit/CF_WITH_ENCODINGS_DIR: Fedora can have the pathname backwards
* AcSplit/CF_MAKE_DOCS:
protect $2 from substitution, for luit's "$(manext)"
2014-01-01
* config.guess, config.sub: 2014-01-01 (fake)
* AcSplit/CF_YACC_ERROR:
in cproto 4.7h2, I changed param of yaccError() to const, but overlooked
the actual function used for configuring - fix.
2013-12-26
* AcSplit/CF_SET_MATH_LIB_VERSION: ci-aclocal generated header
* config.guess: 2013-11-29
2013-12-06 Urs.Jansen
* AcSplit/AM_MULTIBYTE_ABLE: compiler-warning fix
2013-12-05
* AcSplit/CF_ANSI_CPP, AcSplit/CF_CURSES_CCHAR_T,
AcSplit/CF_CURSES_MOUSE, AcSplit/CF_GMTOFF,
AcSplit/CF_PROGRAM_FULLPATH, AcSplit/CF_PROGRAM_PREFIX,
AcSplit/CF_REGCMP_FUNCS, AcSplit/CF_REGEXPR_H_FUNCS,
AcSplit/CF_REGEX_H_FUNCS, AcSplit/CF_RE_COMP_FUNCS, AcSplit/CF_SIZE_T,
AcSplit/CF_STAT_ST_BLOCKS, AcSplit/CF_STRUCT_SCREEN,
AcSplit/CF_TCAP_CURSOR, AcSplit/CF_TD_SIG_ARGS, AcSplit/CF_VARARGS,
AcSplit/CF_WAIT:
appease autoheader(-252) for td_lib
2013-11-19
* AcSplit/CF_GCC_WARNINGS:
the new gcc options are not recognized by clang - suppress when using clang
* AcSplit/CF_CLANG_COMPILER: typo in comment
2013-11-16
* AcSplit/CF_GXX_WARNINGS:
add checks for -Wextra,-Wignored-qualifiers,-Wlogical-op
* AcSplit/CF_GCC_WARNINGS:
add checks for gcc's -Wdeclaration-after-statement and -Wextra (which came
into ncurses from xterm in 2007 -- seems stable)
2013-11-03
* AcSplit/CF_WITH_SHARED_OR_LIBTOOL:
fix for Cdk, which needs substitutions for shlibdir and MAKE_DLLS symbols
* AcSplit/CF_SHARED_OPTS:
add variables to support cross-compiling to MinGW, etc.
* AcSplit/CF_CURSES_UNCTRL_H: separate warning/cache messages
2013-10-28
* AcSplit/CF_FUNC_GETOPT, AcSplit/CF_FUNC_POPEN:
ci-aclocal generated header
2013-10-25
* AcSplit/CF_POPEN_TEST:
improve check by assuming that if popen or pclose was #define'd, that we
don't want to provide a prototype
* config.sub: 2013-10-01
2013-10-14
* AcSplit/CF_ADA_INCLUDE_DIRS: fix typo (report by Roumen Petrov)
2013-10-12
* AcSplit/CF_ADA_INCLUDE_DIRS, AcSplit/CF_INCLUDE_DIRS:
allow for empty $prefix
2013-10-08
* AcSplit/CF__INET_HEAD: ci-aclocal generated header
* AcSplit/CF_MIXEDCASE_FILENAMES, AcSplit/CF_NETLIBS:
generalize to mingw64
* AcSplit/CF_INET_ADDR:
modify (for lynx) to split-out CF__INET_HEAD, and allow checking for winsock
* AcSplit/CF_ADD_SUBDIR_PATH:
tweak for empty $prefix, e.g., for my cross-compiling to mingw
2013-10-07
* AcSplit/CF_PDCURSES_W32: ci-aclocal generated header
2013-09-21
* AcSplit/CF_SHARED_OPTS: build-fix for DragonFly and FreeBSD
2013-09-07
* config.sub: 2013-09-05
* config.guess: 2013-06-10
* AcSplit/CF_XOPEN_SOURCE, AcSplit/CF_OBJ_SUBDIR,
AcSplit/CF_GNAT_PROJECTS:
coding-style
2013-09-07 Alexey.Pavlov
* AcSplit/CF_CFG_DEFAULTS, AcSplit/CF_GNAT_PROJECTS,
AcSplit/CF_LIB_RULES, AcSplit/CF_LIB_SUFFIX,
AcSplit/CF_MIXEDCASE_FILENAMES, AcSplit/CF_OBJ_SUBDIR,
AcSplit/CF_SHARED_OPTS, AcSplit/CF_WITH_LIBTOOL,
AcSplit/CF_XOPEN_SOURCE:
add msys / msysdll to known host/platform types
2013-09-02
* AcSplit/CF_WITH_SHARED_OR_LIBTOOL:
after calling CF_LIB_SUFFIX, update LIB_SUFFIX
* AcSplit/CF_WITH_SHARED_OR_LIBTOOL: substitute $LIBTOOL_MAKE
2013-09-01
* AcSplit/CF_RPATH_HACK:
use sort and uniq rather than sort -u, to work with HPUX 11.11, etc.
2013-08-03
* AcSplit/CF_LIB_RULES, AcSplit/CF_SRC_MODULES:
implement --with-db-install for ncurses
* AcSplit/CF_SHARED_OPTS: resync
2013-07-27
* AcSplit/CF_SHARED_OPTS:
change the symbol from top_builddir to rel_builddir to avoid possible
conflict with ncurses.
* AcSplit/CF_SHARED_OPTS:
reconsidered LOCAL_LIBDIR, since it is clutter
* AcSplit/CF_SHARED_OPTS: tidy
* AcSplit/CF_SHARED_OPTS: add RM_SHARED_OPT symbol
* AcSplit/CF__DEFINE_SHLIB_VARS, AcSplit/CF__INIT_SHLIB_RULES:
add RM_SHARED_OPTS
* AcSplit/CF_TOP_BUILDDIR:
allow optional parameter to override the use of absolute pathnames
* AcSplit/CF_SHARED_OPTS:
use LOCAL_LIBDIR symbol and $top_builddir to parameterize this better for
cross-compiling cdk and dialog.
* AcSplit/CF_SHARED_OPTS:
modify NetBSD and DragonFly to use ${CC} rather than ${LD} for consistency
with the --enable-rpath option.
2013-07-22
* AcSplit/CF__ADD_SHLIB_RULES: fix inconsistent target
* AcSplit/CF_WITH_SHARED_OR_LIBTOOL:
add variables to integrate to CF__ADD_SHLIB_RULES
* AcSplit/CF__ADD_SHLIB_RULES:
cancel unneeded $cf_dirname stuff - only ncurses would need that
* AcSplit/CF__DEFINE_LIB_TARGET, AcSplit/CF__DEFINE_SHLIB_VARS,
AcSplit/CF__INIT_SHLIB_RULES:
ci-aclocal generated header
* AcSplit/CF_RPATH_HACK: show "no" if nothing happens
2013-07-09
* AcSplit/CF_TRIM_X_LIBS:
change delimiter, quote the value to trim in case it contains slashes or
spaces.
* AcSplit/CF_ADD_LIB_AFTER:
use in xterm #295 exposed problem with -Wl,xxx options
2013-07-06
* AcSplit/CF_TRIM_X_LIBS: ci-aclocal generated header
* AcSplit/CF_X_FONTCONFIG:
put -lfontconfig (and any dependencies) after -lXft
* AcSplit/CF_TRY_PKG_CONFIG:
set variables for consistent usage of this macro
* AcSplit/CF_X_ATHENA:
trim extra X libraries after updating lists, to work with ld --as-needed
option which in effect uses only the first mention of the library.
If that does not follow everything that depends on the library, ld
will silently fail to resolve symbols.
* AcSplit/CF_LIB_SUFFIX:
remove the CF_VERBOSE - creates too much clutter
* AcSplit/CF_SHARED_OPTS:
add --enable-auto-import in mingw case - needed for ncurses
2013-06-16
* AcSplit/CF_VERSION_INFO:
change message to make it clearer that $VERSION is intended to be the ABI version
* AcSplit/CF_WITH_SHARED_OR_LIBTOOL: fix quoting
* AcSplit/CF__ADD_SHLIB_RULES:
fixes for quoting, etc - works for cdk now
* AcSplit/CF_LIB_SUFFIX:
move "normal" case to end, and make that match everything, to allow for
other terms such as "static".
2013-06-15
* AcSplit/CF_WITH_SHARED_OR_LIBTOOL, AcSplit/CF__ADD_SHLIB_DEFS,
AcSplit/CF__ADD_SHLIB_RULES:
ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS:
remove unnecessary dependency upon CF_SUBST_NCURSES_VERSION
* AcSplit/CF_LIB_SUFFIX:
remove unnecessary dependency on CF_SUBST_NCURSES_VERSION
* AcSplit/CF_CURSES_SETBEGYX: appease autoheader
2013-05-27
* AcSplit/CF_WITH_ICON_NAME:
add AC_SUBST so this option can be used by itself
* AcSplit/CF_WITH_ICON_SYMLINK: ci-aclocal generated header
2013-05-14 Brendan.O-Dea
* AcSplit/CF_WITH_PERL: improve lookup for xsubpp
2013-04-17
* AcSplit/CF_FUNC_GRANTPT:
amend the check for openpty() to prefer it over posix_openpt(), to fix
problems with terminal-mode initialization on BSD systems.
* AcSplit/CF_WITH_ICON_THEME:
change warning to verbose message - usually I don't want the warning
2013-04-13
* AcSplit/CF_CFG_DEFAULTS: mingw also uses ncurses
* AcSplit/CF_CPP_STATIC_CAST:
clang 3.2 needs default constructor declared
* AcSplit/CF_BOOL_SIZE: clang 3.2 needs type for main()
2013-04-06
* AcSplit/CF_CHECK_LIBTOOL_VERSION, AcSplit/CF_FORGET_TOOL,
AcSplit/CF_LIBTOOL_VERSION:
ci-aclocal generated header
* AcSplit/CF_WITH_LIBTOOL:
modify to work around OS X's unrelated "libtool" program
2013-03-24
* AcSplit/CF_LIB_RULES:
fix special case when libtool is used - the $cf_shared_suffix symbol is
empty and that caused a sed expression to fail, making c++/Makefile's
LIBRARIES symbol to become empty.
2013-03-23
* AcSplit/CF_LIB_RULES:
refactored to make the differences for mingw/Linux more manageable, and used
that to help bridge shared libraries for mingw/Linux/OSX.
* AcSplit/CF_LIB_RULES:
build-fix for OS X to handle the CXX-shared feature (report by Christian Ebert)
2013-03-09
* AcSplit/CF_LIB_RULES: typo
* AcSplit/CF_LIB_RULES:
mods to permit building libncurses++ as shared library
2013-03-04
* AcSplit/CF_ACVERSION_CHECK:
add Dave Becket's change from byacc package to work around breakage introduced
in 1999 by Demaille and compounded in 2007 by Blake. This is so the Debian
packagers can more/less use "newer" autoconf.
2013-02-16
* AcSplit/CF_WITH_ENCODINGS_DIR:
define an empty string if we found no encodings.dir file
2013-02-10
* AcSplit/CF_INTEL_COMPILER, AcSplit/CF_XOPEN_SOURCE:
$host_os needs AC_CANONICAL_HOST
* COPYING.asc: sign
* COPYING: bump
* config.sub, config.guess: 2013-02-04
2013-02-09
* AcSplit/CF_CURSES_LIBS:
modify to allow external script to set $cf_term_lib and/or $cf_curs_lib
* AcSplit/CF_MAN_PAGES:
ensure that the manpage substitutions have at least "no default value"
2013-02-02
* AcSplit/CF_MAN_PAGES: add TERMINFO_DIRS symbol
2013-01-28
* AcSplit/CF_WITH_ENCODINGS_DIR: ci-aclocal generated header
2013-01-26
* AcSplit/CF_SHARED_OPTS:
add case for hpux8.07 (report by Dennis Grevenstein)
* AcSplit/CF_ENABLE_STRING_HACKS: align help-message
* AcSplit/CF_WITH_SYSTYPE: ci-aclocal generated header
2013-01-02
* AcSplit/CF_PROG_GROFF: ci-aclocal generated header
* AcSplit/CF_MAKE_DOCS:
apply workaround from xterm #288 for html output
2013-01-01
* AcSplit/CF_WITH_DESKTOP_CATEGORY:
improve this a little more by making the $4 really optional.
* AcSplit/CF_WITH_DESKTOP_CATEGORY:
add checks for failure of the "auto" default setting. If building on
a server without a desktop (which is what packagers might do...) there
are no desktop files to inspect. Also, since none of the first parameters
does exactly what I need for a fallback for this macro, provide that as
the 4th parameter. From report by Julien Cristau.
2012-12-22
* AcSplit/CF_SHARED_OPTS:
move assignment to EXTRA_CFLAGS of -no-cpp-precomp into compile-check.
This was added 2001/12/02 for Darwin, and reportedly works only for Apple's
compiler. However it works for the MacPorts configuration as well.
2012-11-10
* AcSplit/CF_INCLUDE_DIRS:
modify so that this puts new variables before $CPPFLAGS, in case someone
has added -I options which conflict (report by Vassili Courzakis).
2012-11-09
* AcSplit/AM_GNU_GETTEXT:
remove dependencies on AC_ISC_POSIX, AC_C_CONST and replace AC_PROG_CC
with CF_PROG_CC (essentially discarding pre-ANSI stuff).
2012-11-08
* AcSplit/CF_ALT_CHAR_SET, AcSplit/CF_BOOL_DEFS,
AcSplit/CF_COLOR_CURSES, AcSplit/CF_CURS_PERFORMANCE,
AcSplit/CF_CURS_TOUCHLINE, AcSplit/CF_DEFINE_PROG,
AcSplit/CF_FANCY_CURSES, AcSplit/CF_FIONBIO,
AcSplit/CF_FUNC_GETADDRINFO, AcSplit/CF_FUNC_LSTAT,
AcSplit/CF_FUNC_SIGACTION, AcSplit/CF_GNUTLS, AcSplit/CF_INET_ADDR,
AcSplit/CF_NCURSES_BROKEN, AcSplit/CF_NETLIBS, AcSplit/CF_NGROUPS,
AcSplit/CF_NSS_COMPAT, AcSplit/CF_REMOVE_BROKEN, AcSplit/CF_SET_ERRNO,
AcSplit/CF_SLANG_CPPFLAGS, AcSplit/CF_SLANG_UNIX_DEFS,
AcSplit/CF_SOCKS, AcSplit/CF_SOCKS5, AcSplit/CF_SSL,
AcSplit/CF_SYSTEM_MAIL_FLAGS, AcSplit/CF_TERMIO_AND_CURSES,
AcSplit/CF_TERMIO_AND_TERMIOS, AcSplit/CF_TM_GMTOFF,
AcSplit/CF_TTYTYPE, AcSplit/CF_TYPE_LONG_LONG,
AcSplit/CF_WIDEC_CURSES, AcSplit/CF_WITH_IDNA:
add third parameter for autoheader
2012-10-31
* AcSplit/CF_SRAND: fixes for autoheader
document parameter
2012-10-27
* AcSplit/CF_MAWK_RUN_FPE_TESTS:
check for cross-compiling, provide a suitable message
2012-10-25
* AcSplit/CF_MAWK_CHECK_SIZE_T, AcSplit/CF_MAWK_FPE_SIGINFO,
AcSplit/CF_MAWK_RUN_FPE_TESTS:
add 3rd parameter for AC_DEFINE
2012-10-18
* AcSplit/CF_GCC_VERSION:
"\|" alternation is an ERE feature; work around for sed's BRE's
* AcSplit/CF_WITH_PATHLIST:
if the result expands to ${prefix} or ${exec_prefix}, autoconf treats those
specially and will have "NONE" for their value until config.status works.
check for this case and suppress the recent change for "eval", needed for
usage in vile.
2012-10-06
* AcSplit/CF_GCC_VERSION:
amend workarounds to accommodate Debian's modification of version-message
* AcSplit/CF_CGETENT, AcSplit/CF_CPP_PARAM_INIT,
AcSplit/CF_CPP_STATIC_CAST, AcSplit/CF_CXX_IOSTREAM_NAMESPACE,
AcSplit/CF_ENABLE_STRING_HACKS, AcSplit/CF_FUNC_NANOSLEEP,
AcSplit/CF_FUNC_TERMIOS, AcSplit/CF_FUNC_VSSCANF,
AcSplit/CF_GPP_LIBRARY, AcSplit/CF_ISASCII, AcSplit/CF_LINK_FUNCS,
AcSplit/CF_STRUCT_SIGACTION, AcSplit/CF_STRUCT_TERMIOS,
AcSplit/CF_TYPEOF_CHTYPE, AcSplit/CF_TYPE_SIGACTION,
AcSplit/CF_WITH_GPM, AcSplit/CF_WITH_SYSMOUSE,
AcSplit/CF_LINK_DATAONLY, AcSplit/CF_WITH_PTHREAD,
AcSplit/CF_CURSES_ACS_MAP, AcSplit/CF_CURSES_CHECK_TYPE,
AcSplit/CF_FUNC_CURSES_VERSION, AcSplit/CF_GETOPT_HEADER,
AcSplit/CF_NCURSES_EXT_FUNCS, AcSplit/CF_NETBSD_FORM_H,
AcSplit/CF_NETBSD_MENU_H, AcSplit/CF_PDCURSES_X11,
AcSplit/CF_WCHAR_TYPE:
add 3rd parameter to AC_DEFINE's to allow autoheader to run
* AcSplit/CF_DISABLE_ECHO: correct help message
* AcSplit/CF_BUILD_CC:
change dependency on CF_PROG_EXT to do this only for the 2.13-compat builds
* AcSplit/CF_PROG_CC:
ncurses uses the optional parameter of AC_PROG_CC - fix.
also (not just ncurses) show the gcc-version using CF_GCC_VERSION.
* AcSplit/CF_LIB_RULES:
s/CF_MFLAGS/TOP_MFLAGS/ to avoid confusing with my-autoconf macros.
* AcSplit/CF_ANSI_CC_CHECK, AcSplit/CF_CC_INIT_UNIONS,
AcSplit/CF_CHECK_FD_SET, AcSplit/CF_CURSES_TERMCAP,
AcSplit/CF_FP_ISREADY, AcSplit/CF_FUNC_DLSYM,
AcSplit/CF_HEADER_RESOURCE, AcSplit/CF_HEADER_SELECT,
AcSplit/CF_LOCALE, AcSplit/CF_RESTARTABLE_PIPEREAD,
AcSplit/CF_SIZECHANGE, AcSplit/CF_STDIO_UNLOCKED,
AcSplit/CF_TYPE_OUTCHAR, AcSplit/CF_VA_COPY, AcSplit/CF_WCTYPE,
AcSplit/CF_WITH_PERL, AcSplit/AM_WITH_NLS, AcSplit/CF_BUNDLED_INTL,
AcSplit/CF_CURSES_CHTYPE, AcSplit/CF_CURSES_TERM_H,
AcSplit/CF_CURSES_UNCTRL_H, AcSplit/CF_CURSES_WACS_MAP,
AcSplit/CF_CURSES_WACS_SYMBOLS, AcSplit/CF_CURSES_WGETPARENT,
AcSplit/CF_FUNC_WAIT, AcSplit/CF_LARGEFILE, AcSplit/CF_MBSTATE_T,
AcSplit/CF_NCURSES_CONFIG, AcSplit/CF_NCURSES_CPPFLAGS,
AcSplit/CF_NCURSES_HEADER, AcSplit/CF_NCURSES_VERSION,
AcSplit/CF_PROG_EXT, AcSplit/CF_TERM_HEADER, AcSplit/CF_UNION_WAIT,
AcSplit/CF_UTF8_LIB, AcSplit/CF_VERSION_INFO, AcSplit/CF_WITH_NO_LEAKS:
add 3rd parameter to AC_DEFINE's to allow autoheader to run
* config.sub: 2012-08-18
* config.guess: 2012-09-25
2012-10-04
* AcSplit/CF_FUNC_MEMMOVE, AcSplit/CF_FUNC_TGETENT, AcSplit/CF_LASTLOG,
AcSplit/CF_PATH_PROG, AcSplit/CF_POSIX_SAVED_IDS,
AcSplit/CF_POSIX_WAIT, AcSplit/CF_REGEX, AcSplit/CF_SIG_ATOMIC_T,
AcSplit/CF_STRUCT_LASTLOG, AcSplit/CF_SYSV_UTMP,
AcSplit/CF_TERMIO_C_ISPEED, AcSplit/CF_TTY_GROUP,
AcSplit/CF_TYPE_CC_T, AcSplit/CF_TYPE_FD_SET, AcSplit/CF_UTEMPTER,
AcSplit/CF_UTMP, AcSplit/CF_UTMP_UT_HOST, AcSplit/CF_UTMP_UT_NAME,
AcSplit/CF_UTMP_UT_SESSION, AcSplit/CF_UTMP_UT_SYSLEN,
AcSplit/CF_UTMP_UT_XSTATUS, AcSplit/CF_UTMP_UT_XTIME,
AcSplit/CF_WITH_PCRE, AcSplit/CF_XKB_BELL_EXT, AcSplit/CF_X_FREETYPE,
AcSplit/CF_TYPE_FD_MASK, AcSplit/CF_WITH_XPM, AcSplit/CF_XBOOL_RESULT,
AcSplit/CF_X_TOOLKIT:
add 3rd parameter to AC_DEFINE for autoheader
* AcSplit/CF_CHECK_TYPE, AcSplit/CF_ENABLE_TRACE,
AcSplit/CF_FUNC_GRANTPT, AcSplit/CF_FUNC_POLL, AcSplit/CF_SVR4,
AcSplit/CF_SYSV, AcSplit/CF_SYS_TIME_SELECT:
add 3rd parameter to AC_DEFINE's to allow autoheader to run
* AcSplit/CF_WITH_APP_CLASS, AcSplit/CF_WITH_APP_DEFAULTS: quote-name
2012-10-03
* AcSplit/CF_PROG_CC:
moved AC_PROG_CC_STDC call here, and conditionally wrap CF_ANSI_CC_REQD
to provide consistent functionality with 2.13/2.52 scripts.
* AcSplit/CF_ACVERSION_CHECK, AcSplit/CF_ACVERSION_COMPARE:
quote macro-name
* AcSplit/CF_PROG_CC: ci-aclocal generated header
* AcSplit/CF_ANSI_CC_CHECK:
split-out that last change as separate CF_CC_ENV_FLAGS
* AcSplit/CF_CC_ENV_FLAGS: ci-aclocal generated header
2012-10-03 Adrian.Bunk
* AcSplit/CF_MKSTEMP:
add parameter to AC_DEFINE_UNQUOTED() to allow it to be recognized by autoheader
2012-10-02 Adrian.Bunk
* AcSplit/CF_CHECK_CACHE, AcSplit/CF_DISABLE_LEAKS,
AcSplit/CF_MIXEDCASE_FILENAMES, AcSplit/CF_NO_LEAKS_OPTION:
add parameter to AC_DEFINE_UNQUOTED() to allow it to be recognized by autoheader
2012-10-02
* AcSplit/CF_GCC_ATTRIBUTES:
modify so that autoheader is able to see the definitions provided by this
macro.
2012-09-30
* AcSplit/CF_GCC_ATTRIBUTES:
use AC_DEFINE_UNQUOTED() rather than appending to confdefs.h, since long-ago
concern about the ability to pass-through parameterized macros appears to be
not a problem, testing with 2.13 and 2.52
2012-09-29
* AcSplit/CF_PATHSEP: add message to show the actual value used.
* AcSplit/CF_LIB_RULES:
s/LIBS_TO_MAKE/Libs_To_Make/ to quiet autoconf-252 warning
* AcSplit/CF_WITH_PATH:
fix after changes in ncurses to use --datarootdir
* AcSplit/CF_WITH_PATHLIST:
fix eval after changes in ncurses to use datarootdir
2012-09-28
* AcSplit/CF_XKB_QUERY_EXTENSION, AcSplit/CF_XKB_KEYCODE_TO_KEYSYM:
include <X11/Xlib.h> in test-program
* AcSplit/CF_XKB_KEYCODE_TO_KEYSYM, AcSplit/CF_XKB_QUERY_EXTENSION:
ci-aclocal generated header
2012-09-04
* AcSplit/CF_FIND_TDLIB: CF_LIB_PREFIX is no longer AC_DEFUN'd
2012-09-03
* AcSplit/CF_WITH_XPM: change default - everyone has this library
2012-08-25
* AcSplit/CF_WITH_ICON_NAME: ci-aclocal generated header
2012-08-11
* AcSplit/CF_MAN_PAGES: use datarootdir
2012-08-10
* AcSplit/CF_CHECK_SIZEOF: ci-aclocal generated header
2012-08-07
* AcSplit/CF_WITH_ICON_THEME:
Solaris 10 /bin/sh objects to assignment of unquoted value containing blanks
2012-08-04
* AcSplit/CF_ENABLE_PC_FILES:
limit the path-syntax check to the case where the --enable-pc-files option
is selected.
* AcSplit/CF_ENABLE_PC_FILES:
use the less-restrictive CF_PATH_SYNTAX check for the library directory,
noting that almost all valid uses are unchanged.
2012-07-27
* AcSplit/CF_WITH_ICON_THEME:
allow ICON_NAME to be pre-set, e.g., in vile using --with-icon-name=XXX
2012-07-25
* AcSplit/CF_WITH_APP_DEFAULTS: update EXTRA_INSTALL_DIRS
2012-07-24
* AcSplit/CF_WITH_ICON_THEME:
move the case for size as part of filename after the suffixes, in case
there's a need to use that in svg or png filename. also improve verbose
mode.
2012-07-22
* AcSplit/CF_WITH_XPM: ci-aclocal generated header
* AcSplit/CF_WITH_ICONDIR, AcSplit/CF_WITH_PIXMAPDIR:
update EXTRA_INSTALL_DIRS, needed for vile
2012-07-07
* AcSplit/CF_CURSES_LIBS: check for tinfo library, which may be present
* AcSplit/CF_TERMCAP_LIBS:
look for tinfo library ahead of curses/ncurses
2012-06-30
* AcSplit/CF_LIB_RULES: add parameters to mk-0th.awk call
2012-06-26 Thierry.Reding
* AcSplit/CF_WITH_ICON_THEME:
If an icon cannot be found in the current directory, the suffix is no
longer added twice before the icon is looked up again in the source
directory.
2012-06-21
* AcSplit/CF_WITH_ICON_THEME:
add 4th parameter, for non-desktop configurations.
* AcSplit/CF_WITH_ICONDIR: use "no" as default.
* AcSplit/CF_WITH_PIXMAPDIR: use ${datadir} in defaults
2012-06-20
* AcSplit/CF_WITH_ICON_THEME:
add ICON_NAME variable, to use in desktop files
* AcSplit/CF_WITH_ICON_THEME:
use the basename of the icon in derived mapping, not its full path
2012-06-19
* AcSplit/CF_PATH_SYNTAX:
also match (for substitution) variables such as ${includedir}
* AcSplit/CF_WITH_ICON_THEME, AcSplit/CF_WITH_PIXMAPDIR:
ci-aclocal generated header
* AcSplit/CF_WITH_APP_DEFAULTS: rename the output variable to $APPSDIR
* AcSplit/CF_WITH_ICONDIR: rename the output variable to $ICONDIR
change the default value and search list to the "icons" directory for consistency.
2012-06-16
* AcSplit/CF_GCC_WARNINGS: add check for clang
* AcSplit/CF_GXX_WARNINGS: check for clang compiler
* AcSplit/CF_CLANG_COMPILER: ci-aclocal generated header
* AcSplit/CF_C_INLINE: clang does not support this option - fix
* AcSplit/CF_GXX_VERSION:
show the correct compiler name in checking-version-of message
2012-06-12
* AcSplit/CF_WITH_APP_DEFAULTS:
implement "auto" option, to search well-known locations
2012-06-11
* AcSplit/CF_WITH_ICONDIR:
recognize "auto" as special case to look for existing directory in well-knownn
locations.
* AcSplit/CF_FUNC_GRANTPT, AcSplit/CF_SYSV, AcSplit/CF_X_FREETYPE:
trim whitespace
2012-06-09
* AcSplit/CF_FUNC_POLL: revised based on Debian #676461
2012-06-08
* AcSplit/CF_FUNC_POLL:
fix the case where stdin is redirected, e.g., in rpm/dpkg builds (reported
in Debian #676461). Also correct return-check, which could have been
confused by typing on the keyboard while the check was running.
2012-06-04
* AcSplit/CF_XBOOL_RESULT: ci-aclocal generated header
2012-05-07
* AcSplit/CF__GRANTPT_BODY:
add an ifdef'd chunk #5 for testing the 'tty' (rather than 'pty') and
moved the important 'pty' chunk before opening the slave device as a
simple way of making the test program lose any possible race between
parent/child processes in initializing both sides of the pseudoterminal
(report by Christian Weisgerber).
* AcSplit/CF_FUNC_GRANTPT: add another step to the test-program.
2012-04-29
* AcSplit/CF_MAKE_DOCS: ci-aclocal generated header
2012-03-31
* AcSplit/CF_GCC_WARNINGS:
ncurses change to (try to) use gnatgcc exposed gnatgcc 2.8.1 on my Sarge
system (versus 3.3.5 for the normal gcc). The 2.8.1's pointer-arithmetic
checks fell afoul of gcc's misuse of void* in string.h; work around by
excluding that check for pre-3.x compilers.
* AcSplit/CF_FIXUP_ADAFLAGS: ci-aclocal generated header
2012-02-26
* COPYING.asc, COPYING: bump
2012-02-25
* config.guess, config.sub: 2012-02-10
* AcSplit/CF_LIB_RULES, AcSplit/CF_LIB_SUFFIX, AcSplit/CF_SHARED_OPTS:
add case for mingw
* AcSplit/CF_ENABLE_STRING_HACKS: ci-aclocal generated header
2012-02-18 Pierre.Labastie
* AcSplit/CF_ETIP_DEFINES:
change break in inner loop to exit from both loops so that the test exits
properly on the first success (reported on ncurses-bug mailing list).
2012-02-13
* AcSplit/CF_MKSTEMP:
move existence-check for mkstemp out of the AC_TRY_RUN, to help with
cross-compiles
* AcSplit/CF_KILLPG:
move uses of AC_FUNC_SETPGRP and AC_FUNC_GETPGRP here since they need
special treatment for cross-compiling.
2012-01-29
* AcSplit/CF__GRANTPT_HEAD:
several improvements test on various Unix hosts
* AcSplit/CF__GRANTPT_BODY:
several improvements tested on various Unix hosts
* AcSplit/CF_FUNC_GRANTPT:
several improvements tested with AIX, HPUX, IRIX64, Tru64 and Solaris.
2012-01-27
* AcSplit/CF_FUNC_GRANTPT, AcSplit/CF__GRANTPT_BODY,
AcSplit/CF__GRANTPT_HEAD:
modify to check for Solaris and HPUX streams features
2012-01-25
* AcSplit/CF_FUNC_GRANTPT:
add a workability check for grantpt and indirectly posix_openpt, which
may be broken in Mac OS X 10.7 Lion (report by Kirby Zhou).
* AcSplit/CF__GRANTPT_BODY, AcSplit/CF__GRANTPT_HEAD:
ci-aclocal generated header
* AcSplit/CF_FUNC_POLL:
poll() does not work for devices (including ptys) on Darwin. Adapt check
from curl (in turn from glib2) to extend the testcase.
2012-01-21
* AcSplit/CF_GNAT_VERSION:
FreeBSD port for gnat uses different version convention.
* AcSplit/CF_WITH_LIB_PREFIX: ci-aclocal generated header
* AcSplit/CF_LIB_RULES:
made CF_LIB_PREFIX a "define", which requires different handling
* AcSplit/CF_LIB_PREFIX:
use "define" to avoid autoconf moving this macro.
2012-01-13
* AcSplit/CF_WITH_NCURSES_ETC:
change this from AC_DEFUN to define, to ensure that pdcurses is handled
in the proper order
* AcSplit/CF_WITH_NCURSES_ETC: ci-aclocal generated header
2012-01-12
* AcSplit/CF_FUNC_GRANTPT:
modify check to exclude Darwin 9.x and lower from using posix_openpt.
Bug report for xterm by Christian Ebert for 9.8.0 says screensize does
not work for the pty.
2012-01-07
* AcSplit/CF_REGEX: change the regex library used for mingw
* AcSplit/CF_SRC_MODULES:
followup fix for ".pc" files to use different variable.
* AcSplit/CF_XOPEN_SOURCE:
fix redefinition warning with Mac OS X (Darwin) of _XOPEN_SOURCE
fix missing typedef for fd_mask used by xterm with OpenBSD 4.9 by defining _BSD_SOURCE
2011-12-31
* AcSplit/CF_SRC_MODULES:
modified this (and the template script) so that $DFT_ARG_SUFFIX is applied
here, and also use $TICS_NAME, so that the naming for the pc-files is
consistent with whatever ticlib and termlib happen to be named.
2011-12-17
* AcSplit/CF_LIB_RULES:
add param to mk-1st to tell if we're using term-driver.
this is needed to add ncurses to tinfo library
2011-12-12
* AcSplit/CF_TYPE_CC_T: ci-aclocal generated header
2011-12-10
* AcSplit/CF_WITH_PTHREAD:
add check for pthread_muxexattr_settype(), which is not provided by Solaris 2.6
* AcSplit/CF_SRC_MODULES:
add tic-lib and term-lib to $PC_MODULES_TO_MAKE, restoring behavior from
before the previous change.
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR:
correct comparison, when pkg-config is not found
* AcSplit/CF_ENABLE_PC_FILES:
correct comparison, when there is no pkg-config found
* AcSplit/CF_XOPEN_SOURCE:
fix conflict for IRIX64 between _SGI_SOURCE and _XOPEN_SOURCE
2011-12-03
* AcSplit/CF_CPP_PARAM_INIT: main has to return int
* AcSplit/CF_ETIP_DEFINES: workaround for generated ncurses_dll.h
2011-11-28
* AcSplit/CF_FUNC_GRANTPT: add check for ptsname.
2011-11-26 Mike.Frysinger
* AcSplit/CF_SRC_MODULES: add PC_MODULES_TO_MAKE variable
2011-11-26
* AcSplit/CF_NCURSES_CONFIG, AcSplit/CF_PDCURSES_X11:
use AC_CHECK_TOOLS if available
2011-11-12
* AcSplit/CF_WITH_PTHREAD:
add check for libc_r, for old FreeBSD systems
* AcSplit/CF_XOPEN_SOURCE:
workarounds for header inconsistencies with OpenBSD and MirBSD
2011-10-30
* AcSplit/CF_TRY_XOPEN_SOURCE: RCS_BASE
* AcSplit/CF_XOPEN_SOURCE:
fix for IRIX64 to try if _XOPEN_SOURCE can be defined even if the platform
specific symbol is defined. This could happen for other platforms, so make
it general.
* AcSplit/CF_AWK_BIG_PRINTF:
suppress stderr in each use of $AWK, to quiet message from IRIX64 nawk.
2011-10-29
* AcSplit/CF_CHECK_WCHAR_H, AcSplit/CF_CXX_AR_FLAGS,
AcSplit/CF_CXX_IOSTREAM_NAMESPACE:
ci-aclocal generated header
* AcSplit/CF_XOPEN_SOURCE:
remove the solaris workaround for wide-character includes since it is
incompatible with gcc 4.6, and can be done in a different manner.
2011-10-28
* AcSplit/CF_X_FONTENC:
modify to add $X_LIBS when updating $LIBS, since otherwise configure-checks
for library functions will fail after this test succeeds
2011-10-26
* AcSplit/CF_WITH_LOCALE_ALIAS:
add /usr/openwin/lib/locale, for Solaris
2011-10-22
* AcSplit/CF_SRC_MODULES: only add the Ada95 makefile if gnat works.
* AcSplit/CF_PROG_GNAT:
ensure $cf_cv_prog_gnat_correct is set even when there's no gnat
2011-10-21
* AcSplit/CF_FUNC_GRANTPT: ci-aclocal generated header
* AcSplit/CF_X_FONTENC:
don't error-out if the libraries are not found; calling script can choose
better alternatives
2011-10-17
* AcSplit/CF_CURSES_WGETPARENT: fix typo
* AcSplit/CF_CURSES_WGETPARENT: ci-aclocal generated header
2011-10-03 Urs.Jansen
* AcSplit/CF_NETLIBS:
add check for inet_ntoa() in nsl library, fix needed for Control Data 4680-MP
with EP/IX 2.2.1AA in SysV mode.
2011-09-24
* AcSplit/CF_LDFLAGS_STATIC:
improve check for broken linkers; some (noticed with Debian 5.0) simply
ignore the "-dynamic" used in the previous check.
2011-09-10
* AcSplit/CF_ENABLE_PC_FILES:
ensure $enable_pc_files has a value even if there's no pkg-config found
* AcSplit/CF_WITH_PKG_CONFIG_LIBDIR: ci-aclocal generated header
* AcSplit/CF_MAN_PAGES: improved --with-curses-h and --disable-override
* AcSplit/CF_ENABLE_PC_FILES: split-out CF_WITH_PKG_CONFIG_LIBDIR
2011-09-04
* AcSplit/CF_SVR4:
cygwin still misuses SVR4 symbol in X header - force this to fail on that platform.
2011-08-29
* AcSplit/CF_X_FREETYPE: add --with-freetype-config option
2011-08-11
* AcSplit/CF_BROKEN_TGETENT_STATUS: ci-aclocal generated header
2011-08-09
* AcSplit/CF_CURSES_LIBS, AcSplit/CF_FUNC_TGETENT:
check for otermcap (OpenBSD)
2011-08-07
* AcSplit/CF_CGETENT:
add check for const params used in NetBSD flavor of cgetent()
2011-08-06
* AcSplit/CF_XOPEN_SOURCE:
on NetBSD, define _NETBSD_SOURCE to ensure that vsscanf and strdup get
prototyped - seems that something in unistd.h triggers this need.
* AcSplit/CF_ENABLE_PC_FILES:
wrap invocation of "arch" in parentheses to help redirect stderr with NetBSD
which has no "arch" and uses csh as the default shell.
2011-07-30
* AcSplit/CF_SHARED_OPTS:
add new variable RPATH_LIST which can extend the list of library directories
2011-07-17
* AcSplit/CF_X_ATHENA_LIBS: force exit from outer loop when done.
* AcSplit/CF_X_ATHENA_LIBS:
DragonFlyBSD's pkgsrc combines Xaw6 and Xaw7 into a single package, and
doesn't provide a plain "libXaw.so" symlink. Work around by testing for
the 6/7 versions.
* AcSplit/CF_LD_RPATH_OPT, AcSplit/CF_SHARED_OPTS:
add case for DragonFlyBSD
2011-07-16
* AcSplit/CF_ENABLE_PC_FILES: typo
* AcSplit/CF_ENABLE_PC_FILES:
64-bit platforms may put pc-files in /usr/lib64/pkgconfig
* AcSplit/CF_ENABLE_PC_FILES: ci-aclocal generated header
2011-07-14
* AcSplit/CF_FUNC_TGETENT:
termcap, e.g,. with Slackware, does not know about the "screen" variants
such as "screen-bce.xterm-new", and dumps core (confusing this test).
Filter those out.
2011-07-08
* AcSplit/CF_WITH_APP_CLASS: ci-aclocal generated header
2011-07-04
* AcSplit/CF_X_FONTCONFIG:
if CF_X_FREETYPE failed, do not compound it by attempting to fix the package.
2011-07-02
* AcSplit/CF_WITH_LIBTOOL, AcSplit/CF_LDFLAGS_STATIC,
AcSplit/CF_LIB_SUFFIX:
extend aix case to 7.x
* AcSplit/CF_XOPEN_SOURCE: extend aix case to aix 7.x
* AcSplit/CF_SHARED_OPTS:
extend aix case to aix 7.x, tested with cc on 7.1
* AcSplit/CF_WITH_DESKTOP_CATEGORY: filter out XFCE patterns
* AcSplit/CF_UNDO_CFLAGS: ci-aclocal generated header
* AcSplit/CF_FUNC_TGETENT:
prefix "$cf_cv_lib_tgetent" in tests, since it may begin with "-".
2011-07-01
* AcSplit/CF_ANSI_CC_CHECK:
do a sanity check on $CC; occasional users add options directly to it rather
than follow the standards by setting $CFLAGS and/or $CPPFLAGS.
2011-06-28
* AcSplit/CF_WITH_LIBTOOL:
modify LIB_CREATE, adding $(LIBS). This is needed by applications that
dynamically load the library, in contrast to the dialog program (which
is itself linked using the $(LIBS) list explicitly). (report/analysis
by xDog Walker).
2011-06-12
* AcSplit/CF_SRAND:
the compile-checks for <bsd/random.h> vs <bsd/stdlib.h> needed another level
to verify that the headers will really work.
2011-06-08
* AcSplit/CF_TERMCAP_LIBS:
modify the workaround for old AIX system (from 1997 or before) to not
set $cf_cv_termlib if it succeeds in finding initscr in libcurses.
Doing that confused a check in tin's configure.in which looks at that
variable to decide if termcap/terminfo library was found. The check
for tgoto in libtermcap on the next line is the actual check which is
needed, and should fail if there is no termcap library installed.
2011-06-04
* AcSplit/CF_CFG_DEFAULTS, AcSplit/CF_LD_RPATH_OPT,
AcSplit/CF_PROG_LDCONFIG, AcSplit/CF_SHARED_OPTS:
add mirbsd
2011-05-28
* AcSplit/CF_SHARED_OPTS:
add case for interix (report/patch by Markus Duft)
* AcSplit/CF_SRAND:
add compile-checks for bsd/stdlib.h and bsd/random.h to select the correct
one needed for arc4random()
* AcSplit/CF_WITH_ZLIB: check for zError after finding the library
2011-05-14
* AcSplit/CF_CURSES_FUNCS:
workaround for bug in gcc 4.2.1 (FreeBSD 8.1) which caused part of test
program to be omitted, i.e., when it saw two return-statements in a row
it omitted the _first_ one. Also add expression to pointer check to
help FreeBSD's linker decide it should be validated. Just an assignment
was not enough.
2011-05-08
* AcSplit/CF_ACVERSION_CHECK:
correct order of then/else for the 2.5x branch.
2011-05-01
* AcSplit/CF_CURSES_HEADER:
change order for curses.h / ncurses.h pairs to put ncurses.h first, which
will tend to provide the same #define's as in CF_NCURSES_HEADER (report
by Dennis Preiser).
2011-04-29
* AcSplit/CF_PKG_CONFIG:
retesting new CF_PKG_CONFIG with lynx, that works (did not work with tin).
2011-04-23
* AcSplit/CF_TPUTS_PROTO: build-fixes from HPUX and Solaris machines
* AcSplit/CF_TPUTS_PROTO: ci-aclocal generated header
2011-04-22
* AcSplit/CF_DISABLE_DESKTOP: add comment about parameters
* AcSplit/CF_WITH_DESKTOP_CATEGORY: ci-aclocal generated header
2011-04-21
* AcSplit/CF_WITH_APP_DEFAULTS, AcSplit/CF_WITH_ICONDIR,
AcSplit/CF_DISABLE_DESKTOP:
ci-aclocal generated header
2011-04-17
* AcSplit/CF_PKG_CONFIG:
revert the CF_ACVERSION check, since it does not (yet) work with both 2.13/2.52
2011-04-16
* AcSplit/CF_STRUCT_SIGACTION, AcSplit/CF_STRUCT_TERMIOS:
relax ncurses-oriented change to make this work as fixup after CF_XOPEN_SOURCE
2011-04-14
* AcSplit/CF_ACVERSION_COMPARE:
need extra quoting around the inner ifelse's
* AcSplit/CF_ACVERSION_CHECK, AcSplit/CF_ACVERSION_COMPARE:
ci-aclocal generated header
* AcSplit/CF_PKG_CONFIG:
use CF_ACVERSION_CHECK to select the AC_PATH_TOOL for "recent" autoconf,
and AC_PATH_PROG for "older". For this purpose, 2.52 is recent.
2011-04-09
* AcSplit/CF_CURSES_FUNCS: add check for unctrl.h
* AcSplit/CF_CURSES_UNCTRL_H: ci-aclocal generated header
* AcSplit/CF_CURSES_TERM_H:
modify to avoid spurious check for <curses.hterm.h> if there is no ncurses
version. Look for ncurses's term.h anyway, to work around breakage by packagers
who separate ncurses' header files.
* AcSplit/CF_CURSES_CPPFLAGS, AcSplit/CF_CURSES_LIBS:
make checks for special libraries on hpux10 and sunos4 optional
2011-04-01
* config.sub: 2011-04-01
2011-03-31
* AcSplit/CF_MAKEFLAGS:
filter out GNU make's entering/leaving messages. This only appeared when
using the macro in a dpkg script, though it should have in other cases.
2011-03-27
* AcSplit/CF_GNAT_SIGINT: ci-aclocal generated header
* AcSplit/CF_NCURSES_ADDON:
fix some portability issues ("+" modifier in sed, and quoting)
2011-03-23
* AcSplit/CF_GNAT_PROJECTS:
add a configure check to ensure that gnat project-files work.
also (as a side-effect) check if the library feature works.
* AcSplit/CF_GNAT_GENERICS: ensure this follows CF_GNAT_VERSION
* AcSplit/CF_GNAT_VERSION:
split-out CF_GNAT_PROJECTS and CF_GNAT_GENERICS so they can be used in the
more expected order - after verifying that gnat actually works.
2011-03-20
* COPYING.asc, COPYING: bump
2011-03-19
* AcSplit/CF_GNAT_TRY_LINK, AcSplit/CF_GNAT_TRY_RUN:
gnat can leave files named like b~conftest.ali
2011-03-06
* AcSplit/CF_CURSES_LIBS:
Debian Lenny has no "arch" command (though more recent releases fix this).
2011-02-20
* AcSplit/CF_XOPEN_SOURCE:
workaround for cygwin to get ncurses' configure script to define _XOPEN_SOURCE_EXTENDED
(cygwin's features.h doesn't do anything, so it needs a crutch).
2011-02-18 Thierry.Reding
* AcSplit/CF_PKG_CONFIG:
use AC_PATH_TOOL rather than AC_PATH_PROG for pkg-config search, to work
with cross-compiler environments.
2011-02-13
* AcSplit/CF_RPATH_HACK:
add a check for libraries not found, e.g., from suppressed functionality
of gcc in linking from /usr/local/lib, and add a -L option to help work
around this.
* AcSplit/CF_DISABLE_RPATH_HACK: fix garbled message
* AcSplit/CF_X_ATHENA_LIBS:
make link-test use header and parameterized test-function.
2011-02-02
* config.guess: 2011-01-01
2011-01-30
* AcSplit/CF_CURSES_MOUSE:
avoid reporting misleading test-result if we're using ncurses.
2011-01-18
* AcSplit/CF_XOPEN_CURSES:
fix typo in the workaround for _XOPEN_SOURCE_EXTENDED
2011-01-16
* AcSplit/CF_CURSES_LIBS:
replace check for /lib with /lib64 and/or /lib, depending on architecture.
2011-01-15
* AcSplit/CF_PDCURSES_X11:
only define HAVE_XCURSES if we found xcurses.h, since pkgsrc alters the
script and configuration to make it show curses.h
* AcSplit/CF_CURSES_CHECK_DATA:
work around BSD linker's weak linkage by running a program to test the case
where the symbol is present in the library, but not in the header.
* AcSplit/CF_NETBSD_FORM_H, AcSplit/CF_NETBSD_MENU_H,
AcSplit/CF_CURSES_WACS_SYMBOLS:
ci-aclocal generated header
* AcSplit/CF_CURSES_WACS_MAP:
make this less stringent, to allow NetBSD curses to pass.
2011-01-02
* AcSplit/CF_VERSION_INFO:
For dialog, modify to handle two parameters rather than one. That allows me
to rename the package without renaming the #define'd symbols for the version.
2010-11-29
* AcSplit/CF_LEX_VERSION:
remove extra '\' character in backreference (report by Mark McKinstry).
2010-11-28
* AcSplit/CF_ENABLE_TRACE: fill in default-value for $with_trace
2010-11-26 Doug.Kaufman
* AcSplit/CF_WITH_IDNA:
add LIBICONV as a library-dependency, needed for mingw
2010-11-20
* AcSplit/CF_NCURSES_CPPFLAGS:
move part of the CF_ADD_INCDIR to CF_WITH_CURSES_DIR
* AcSplit/CF_NCURSES_LIBS:
move part of the CF_ADD_LIBDIR to CF_WITH_CURSES_DIR
* AcSplit/CF_WITH_CURSES_DIR:
move the calls to CF_ADD_INCDIR and CF_ADD_LIBDIR for the curses-directory
here, from CF_NCURSES_CPPFLAGS and CF_NCURSES_LDFLAGS, so it will work
even with the default checking, e.g., no --with-ncurses, etc.
2010-11-19
* AcSplit/CF_X_TOOLKIT: add check for missing X11 dependency, see
https://bugs.launchpad.net/bugs/677129
2010-11-13
* AcSplit/CF_NCURSES_EXT_FUNCS: typo
* AcSplit/CF_GNAT_GENERICS, AcSplit/CF_GNAT_PROJECTS:
ci-aclocal generated header
* AcSplit/CF_GNAT_VERSION:
split-out CF_GNAT_GENERICS and CF_GNAT_PROJECTS
2010-11-09
* AcSplit/CF_X_ATHENA:
add a check for Xmu dependency before looking in other places
* AcSplit/CF_X_FONTCONFIG: formatting
* AcSplit/CF_X_TOOLKIT: remove redundant message
* AcSplit/CF_X_TOOLKIT:
add a check to see whether the Shell.h feature of Xt is usable before
looking in other places for ICE and SM libraries.
2010-11-08
* AcSplit/CF_ADD_LIB_AFTER:
make the caller supply "-l", so it can use a variable containing more than
one library
* AcSplit/CF_X_ATHENA, AcSplit/CF_X_FONTCONFIG, AcSplit/CF_X_TOOLKIT:
use CF_ADD_LIB_AFTER
* AcSplit/CF_ADD_LIB_AFTER, AcSplit/CF_X_FONTCONFIG:
ci-aclocal generated header
* AcSplit/CF_X_FREETYPE:
xterm doesn't need XRENDERFONT substituted, only defined.
* AcSplit/CF_X_ATHENA:
workaround for more pc-file breakage (removing the Xmu dependency from
Xaw's --libs output).
* AcSplit/CF_X_TOOLKIT:
Xt's dependencies may be in "ice" and "sm" packages.
give those a try.
2010-10-23
* AcSplit/CF_PATH_PROG, AcSplit/CF_PROGRAM_FULLPATH,
AcSplit/CF_PROGRAM_PREFIX, AcSplit/CF_LIB_RULES,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_WITH_PATHLIST,
AcSplit/CF_WITH_LIBTOOL, AcSplit/CF_BUNDLED_INTL,
AcSplit/CF_ALT_CHAR_SET, AcSplit/CF_BOOL_DEFS,
AcSplit/CF_COLOR_CURSES, AcSplit/CF_CURSES_ACS_MAP,
AcSplit/CF_CURSES_CCHAR_T, AcSplit/CF_CURSES_CHECK_DATA,
AcSplit/CF_CURSES_CHECK_TYPE, AcSplit/CF_CURSES_CHTYPE,
AcSplit/CF_CURSES_LIBS, AcSplit/CF_CURSES_MOUSE,
AcSplit/CF_CURSES_SETBEGYX, AcSplit/CF_CURSES_TERM_H,
AcSplit/CF_CURSES_WACS_MAP, AcSplit/CF_CURS_PERFORMANCE,
AcSplit/CF_CURS_TOUCHLINE, AcSplit/CF_FANCY_CURSES,
AcSplit/CF_FUNC_CURSES_VERSION, AcSplit/CF_MIN_GETCCHAR,
AcSplit/CF_NCURSES_ADDON, AcSplit/CF_NCURSES_BROKEN,
AcSplit/CF_NCURSES_EXT_FUNCS, AcSplit/CF_NCURSES_LIBS,
AcSplit/CF_NCURSES_VERSION, AcSplit/CF_STRUCT_SCREEN,
AcSplit/CF_TCAP_CURSOR, AcSplit/CF_TERMIO_AND_CURSES,
AcSplit/CF_TERM_HEADER, AcSplit/CF_TTYTYPE, AcSplit/CF_WIDEC_CURSES,
AcSplit/CF_XOPEN_CURSES, AcSplit/CF__CURSES_HEAD,
AcSplit/CF_ANSI_CC_CHECK, AcSplit/CF_ANSI_CPP,
AcSplit/CF_AWK_BIGSTRING, AcSplit/CF_COREFILE, AcSplit/CF_CRYPT_FUNC,
AcSplit/CF_CURSES_FUNCS, AcSplit/CF_FUNC_FORK,
AcSplit/CF_FUNC_GETADDRINFO, AcSplit/CF_FUNC_SYSTEM,
AcSplit/CF_GCC_ATTRIBUTES, AcSplit/CF_GCC_WARNINGS,
AcSplit/CF_GNAT_VERSION, AcSplit/CF_KILLPG, AcSplit/CF_MAKEFLAGS,
AcSplit/CF_MAKE_AR_RULES, AcSplit/CF_MAKE_INCLUDE,
AcSplit/CF_MAKE_TAGS, AcSplit/CF_MAWK_PROG_GCC, AcSplit/CF_MSG_LOG,
AcSplit/CF_POSIX_VDISABLE, AcSplit/CF_REGCMP_LIBS,
AcSplit/CF_REGEXPR_H_FUNCS, AcSplit/CF_REGEX_H_FUNCS,
AcSplit/CF_RESTARTABLE_PIPEREAD, AcSplit/CF_RE_COMP_FUNCS,
AcSplit/CF_SET_ERRNO, AcSplit/CF_SIG_CONST, AcSplit/CF_SRC_MAKEFILE,
AcSplit/CF_TD_CONFIG, AcSplit/CF_TD_SIG_ARGS,
AcSplit/CF_TD_TEST_MAKEFILES, AcSplit/CF_TIOCGWINSZ,
AcSplit/CF_WIDEC_SHIFT, AcSplit/CF_WITH_INSTALL_PREFIX,
AcSplit/CF_X_ATHENA, AcSplit/CF_YACC_ERROR, AcSplit/CF_GXX_VERSION,
AcSplit/CF_MAN_PAGES, AcSplit/CF_WITH_PATH:
change obsolete ${name-value} to standard ${name:-value}
* AcSplit/CF_LDFLAGS_STATIC:
some gcc configurations, e.g., Fedora 12 and up, are mildly broken since they
do not allow switching between static/dynamic libraries. Add a check for that.
2010-09-28
* config.guess: 2010-09-24
2010-09-26
* AcSplit/CF_SLANG_UNIX_DEFS:
add checks for two slang-specific typedefs used in quieting warnings for lynx
2010-09-23
* AcSplit/CF_TYPE_LONG_LONG:
revise test-compile to capture stderr, look for warning messages in case
compiler (as does gcc) warns about unsupported long-long but returns no
error status.
2010-09-19
* AcSplit/CF_BUNDLED_INTL:
add --with-textdomain option, to use with lynx-dev package
2010-09-13
* AcSplit/CF_LINK_PREFIX: ci-aclocal generated header
2010-09-10
* config.sub: 2010-09-11
2010-09-06
* AcSplit/CF_GCC_OPT_RDYNAMIC:
QNX' gcc 2.95.2 ignores the "dynamic", does not support that option.
2010-09-04
* AcSplit/CF_LIB_RULES: fix a typo in the last change.
* AcSplit/CF_SRC_MODULES:
implement --without-tests option in ncurses configure script
2010-08-14
* AcSplit/CF_LIB_RULES:
add cases for library-name transformation for "t" and "tw" to handle the
configurations where tic- and tinfo-libraries are built as shared libraries.
* AcSplit/CF_MAWK_PROG_GCC, AcSplit/CF_CHECK_GPM_WGETCH,
AcSplit/CF_GCC_WARNINGS, AcSplit/CF_GNAT_TRY_LINK,
AcSplit/CF_GNAT_TRY_RUN, AcSplit/CF_GXX_WARNINGS,
AcSplit/CF_LIB_SONAME, AcSplit/CF_MKSTEMP, AcSplit/CF_PROG_CC_C_O,
AcSplit/CF_PROG_LN_S:
fix for Mac OS X (compiler makes conftest.dSYM directory)
2010-08-13
* AcSplit/CF_WITH_PERL:
also redefine 'main' when including perl headers, since old perl versions
had a prototype for it which conflicts with autoconf's "main()"
* AcSplit/CF_WITH_PERL:
drop 2 details from the test-compile which are not in Perl 5.004 (the XSINIT_t
and aTHX_ symbols).
2010-08-07
* AcSplit/CF_REGEX:
Building on/for MinGW, there's more than one variation of the name to use
when linking. Add "-lregex.dll" as a possibility, since that seems to be
the most recent. Don't add the older "-lgnurx.dll".
2010-07-24
* AcSplit/CF_PROG_LN_S: ci-aclocal generated header
* AcSplit/CF_MAN_PAGES: modified to use results from CF_PROG_LN_S
* AcSplit/CF_MANPAGE_SYMLINKS: modify to work with CF_PROG_LN_S
* AcSplit/CF_LIB_RULES: pass LN_S value to mk-1st.awk
* AcSplit/CF_ENABLE_TRACE: ci-aclocal generated header
2010-07-23
* AcSplit/CF_DISABLE_LEAKS: add YY_NO_LEAKS symbol recognized by byacc
2010-07-08
* AcSplit/CF_NCURSES_CONFIG: correct comment about version
2010-07-03
* AcSplit/CF_FIND_TDLIB:
td_lib.mk is now in the datadir, e.g., /usr/share/td
* AcSplit/CF_ADD_SUBDIR_PATH:
workaround - if $prefix was not mkdir'd yet, no directories were added.
* AcSplit/CF_GNAT_VERSION:
add substitution to distinguish between gnatmake's which do/don't support
project files
2010-06-26
* AcSplit/CF_WITH_ADA_COMPILER, AcSplit/CF_WITH_ADA_INCLUDE,
AcSplit/CF_WITH_ADA_OBJECTS, AcSplit/CF_WITH_ADA_SHAREDLIB:
add progress messages
2010-06-21
* AcSplit/CF_WITH_DBMALLOC, AcSplit/CF_WITH_DMALLOC:
quote params of ifelse()
2010-06-20
* AcSplit/CF_GNUTLS: further fixes for with/without gnutls-compat
* AcSplit/AM_WITH_NLS, AcSplit/CF_CURSES_LIBS, AcSplit/CF_GNUTLS,
AcSplit/CF_INET_ADDR, AcSplit/CF_NCURSES_CONFIG,
AcSplit/CF_NCURSES_LIBS, AcSplit/CF_NETLIBS, AcSplit/CF_NSS_COMPAT,
AcSplit/CF_PDCURSES_X11, AcSplit/CF_SOCKS5, AcSplit/CF_SSL,
AcSplit/CF_TERMCAP_LIBS, AcSplit/CF_UTF8_LIB:
use CF_ADD_LIBS for consistency
* AcSplit/CF_GNUTLS:
amend the change for using pkg-config; the ".pc" file for gnutls provides
no useful information for the --cflags option, so it is still necessary to
check for the subdir feature of the -I options.
2010-06-19
* AcSplit/CF_WITH_ADA_SHAREDLIB: ci-aclocal generated header
* AcSplit/CF_GNAT_VERSION: add AC_SUBST's for variables set here
* AcSplit/CF_ADD_ADAFLAGS, AcSplit/CF_GNAT_PRAGMA_UNREF,
AcSplit/CF_PROG_GNAT, AcSplit/CF_WITH_ADA_COMPILER,
AcSplit/CF_WITH_ADA_INCLUDE, AcSplit/CF_WITH_ADA_OBJECTS:
ci-aclocal generated header
* AcSplit/AM_GNU_GETTEXT:
drop $MKINSTALLDIRS, use "mkdir -p" consistently.
2010-06-14
* AcSplit/CF_TRY_PKG_CONFIG:
add verbose output, to help see where the libraries come from.
* AcSplit/CF_X_ATHENA: use CF_TRY_PKG_CONFIG again, to reduce clitter
* AcSplit/CF_X_TOOLKIT:
workaround for broken ".pc" file for X Toolkit, which omits the ICE library.
2010-06-05
* AcSplit/CF_SHARED_OPTS:
assignment in case for aix was improperly quoted
* AcSplit/CF_SHARED_OPTS:
for Solaris, prefix -xcode=32 / -K PIC over -xcode=13 / -K pic
* AcSplit/CF_WITH_PCRE:
libpcre package on Centos 5.4 uses different entrypoint for libpcreposix
* AcSplit/CF_WITH_PCRE:
not all pcre packages show the posix library with --libs option
2010-06-04
* AcSplit/CF_FUNC_TGETENT: typo
2010-06-02
* AcSplit/CF_GRANTPT: ci-aclocal generated header
* AcSplit/CF_FUNC_TGETENT: typo
* AcSplit/CF_ADD_LIBS: RCS_BASE
* AcSplit/CF_ADD_LIB, AcSplit/CF_FUNC_TGETENT, AcSplit/CF_LD_RPATH_OPT,
AcSplit/CF_TRY_PKG_CONFIG, AcSplit/CF_UTEMPTER, AcSplit/CF_WITH_PCRE,
AcSplit/CF_X_ATHENA, AcSplit/CF_X_ATHENA_LIBS, AcSplit/CF_X_EXT,
AcSplit/CF_X_FREETYPE, AcSplit/CF_X_TOOLKIT:
use CF_ADD_LIB/CF_ADD_LIBS
2010-06-01
* AcSplit/CF_WITH_LOCALE_ALIAS: add case for OpenSolaris location
2010-05-29
* AcSplit/CF_FIND_LINKAGE, AcSplit/CF_FUNC_DLSYM,
AcSplit/CF_FUNC_OPENPTY, AcSplit/CF_GPP_LIBRARY, AcSplit/CF_HASHED_DB,
AcSplit/CF_HASHED_DB_LIBS, AcSplit/CF_LIB_SONAME, AcSplit/CF_MATH_LIB,
AcSplit/CF_REGEX, AcSplit/CF_STDCPP_LIBRARY, AcSplit/CF_WITH_PTHREAD:
use CF_ADD_LIB
* AcSplit/CF_ADD_LIB: RCS_BASE
* AcSplit/CF_WITH_LOCALE_ALIAS: NetBSD uses /usr/X11R7
2010-05-26
* AcSplit/CF_X_ATHENA: use pkg-config to find the library if possible
* AcSplit/CF_X_ATHENA_CPPFLAGS: rename variable for consistency
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_ADD_LIBDIR,
AcSplit/CF_CHECK_ENVIRON, AcSplit/CF_GNUTLS, AcSplit/CF_MATH_LIB,
AcSplit/CF_MAWK_DEFINE, AcSplit/CF_PREDEFINE, AcSplit/CF_WCHAR_TYPE:
quote params of ifelse()
* AcSplit/CF_X_TOOLKIT: use pkg-config if available
* AcSplit/CF_X_EXT: *** empty log message ***
* AcSplit/CF_ADD_CFLAGS, AcSplit/CF_ARG_OPTION, AcSplit/CF_CHECK_ERRNO,
AcSplit/CF_IMAKE_CFLAGS, AcSplit/CF_INTEL_COMPILER,
AcSplit/CF_PATHSEP, AcSplit/CF_PATH_SYNTAX, AcSplit/CF_POSIX_C_SOURCE,
AcSplit/CF_TRY_PKG_CONFIG, AcSplit/CF_WITH_IMAKE_CFLAGS,
AcSplit/CF_WITH_PATH, AcSplit/CF_XOPEN_SOURCE,
AcSplit/CF_X_ATHENA_CPPFLAGS, AcSplit/CF_X_ATHENA_LIBS:
quote params of ifelse()
2010-05-25
* AcSplit/CF_TRY_PKG_CONFIG, AcSplit/CF_X_EXT:
ci-aclocal generated header
* AcSplit/CF_WITH_PCRE:
improve the pkg-config check using CF_TRY_PKG_CONFIG
* AcSplit/CF_X_ATHENA: split-out CF_X_EXT
* AcSplit/CF_WITH_PCRE:
use pkg-config for libpcre, if available (tested with Fedora 12)
2010-05-22
* AcSplit/CF_MKSTEMP:
add check for cache variable set by AC_CHECK_FUNC, whose naming convention
differs from the CF_xxx macros (report by Pierre Labastie).
2010-05-20
* AcSplit/CF_AR_FLAGS:
allow $ARFLAGS to override the choice of ar-flags, in particular check if
a given choice is part of the current $ARFLAGS
2010-05-15
* AcSplit/CF_MAN_PAGES: use "mkdir -p" rather than mkdirs.sh
* AcSplit/CF_DISABLE_LIBTOOL_VERSION: ci-aclocal generated header
* AcSplit/CF_WITH_LIBTOOL: rename for consistency
* AcSplit/CF_LIB_RULES: use $LIBTOOL_VERSION
* AcSplit/CF_WITH_LIBTOOL: use CF_ENABLE_LIBTOOL_VERSION
2010-05-05
* AcSplit/CF_GNUTLS, AcSplit/CF_SSL:
eliminate duplicate libraries when configuring with pkg-config
* AcSplit/CF_FIND_LINKAGE: simplify save/restore of $LIBS
* AcSplit/CF_WITH_IDNA:
use $cf_cv_find_linkage_idn to allow CF_FIND_LINKAGE to update $CPPFLAGS and $LIBS
* AcSplit/CF_SOCKS:
use $cf_cv_find_linkage_socks variable set in CF_FIND_LINKAGE to allow that
macro to also update $CPPFLAGS, $LIBS
* AcSplit/CF_FIND_LINKAGE:
add assignments to cf_cv_header_path_$3 and cf_cv_library_path_$3 for the
case where the function is found in predefined libraries, to work with logic
in CF_GNUTLS
* AcSplit/CF_WITH_IDNA:
remove redundant update of CFLAGS, LIBS since CF_FIND_LINKAGE does this
* AcSplit/CF_HEADER_PATH:
revert part of last change: lynx configure script needs search for variations
in the predefined paths to pick up /usr/include/gnutls
2010-05-01
* AcSplit/CF_C_INLINE: show the compiler name
2010-04-28
* AcSplit/CF_CURSES_HEADER:
add parameter to allow looking for specific subdirectory ncurses/ncursesw/etc
* AcSplit/CF_NCURSES_CONFIG:
use CF_CURSES_HEADER to pick out the particular flavor of ncurses.h, e.g.,
<ncursesw/curses.h>
* AcSplit/CF_XOPEN_CURSES:
workaround for (temporary) problem with ncurses headers, which did not account
for the fact that _XOPEN_SOURCE_EXTENDED may be defined in a system header.
2010-04-24
* AcSplit/CF_GCC_WARNINGS:
move -W and -Wall into the list to check, since c89 alias for gcc complains
about these options.
* AcSplit/CF_GCC_VERSION: ignore stderr, e.g., from c89 alias for gcc
2010-04-21
* AcSplit/CF_HEADER_PATH:
don't search for variations of everything in the current include-path
* AcSplit/CF_WITH_IDNA: update $LIBS also
* AcSplit/CF_UTF8_LIB: use cache variable for consistency
* AcSplit/CF_SOCKS: use cache-variable for consistency
* AcSplit/CF_FIND_LINKAGE, AcSplit/CF_SUBDIR_PATH: whitespace
2010-04-20
* AcSplit/CF_TYPE_LONG_LONG: ci-aclocal generated header
* AcSplit/CF_SSL:
workaround for broken openssl pkg-config scripts (need a followup check to
get the openssl subdirectory for includes)
2010-04-17
* AcSplit/CF_RPATH_HACK_2:
bugfix for last change- "break" from a case/esac is incorrect.
* AcSplit/CF_RPATH_HACK_2: improve filtering by:
a) using $cf_rpath_list (from calling macro) to find directories to
ignore, e.g., /lib64 on 64-bit platforms.
b) check for repeated appends of the given rpath option.
* AcSplit/CF_RPATH_HACK:
look for ldd. If we find it, try to use it to obtain a list of standard
directories used.
2010-04-11
* AcSplit/CF_DISABLE_RPATH_HACK: ci-aclocal generated header
2010-04-05
* AcSplit/CF_RPATH_HACK_2:
add double-quotes on the substitution of $LD_RPATH_OPT, needed for FreeBSD
2010-04-03
* AcSplit/CF_MAKE_TAGS: typo
* AcSplit/CF_FIND_LINKAGE:
must set cache for library_file in the special case where no directory
search is made.
* AcSplit/CF_WITH_IDNA: remove redundant assignment to LIBS
* AcSplit/CF_SLANG_UNIX_DEFS:
move a chunk from CF_SLANG_CPPFLAGS, modify to work with slang2 package in pkgsrc
* AcSplit/CF_SLANG_CPPFLAGS:
parameterize to work with "slang2" header-path in pkgsrc
* AcSplit/CF_SLANG_LIBS:
parameterize to work with "slang2" library name in pkgsrc
* AcSplit/CF_MAKE_TAGS: fill in default value if program is not found
* AcSplit/CF_NCURSES_EXT_FUNCS: ci-aclocal generated header
* AcSplit/CF_MAKE_TAGS:
add check for exctags and exetags, prefer to ctags and etags to work around
pkgsrc (NetBSD) renaming.
2010-04-02
* AcSplit/CF_RPATH_HACK, AcSplit/CF_RPATH_HACK_2:
ensure that EXTRA_LDFLAGS is substituted
2010-03-28
* AcSplit/CF_SHARED_OPTS: fix for OpenSolaris with rpath
* AcSplit/CF_XOPEN_SOURCE:
add solaris2.11 (OpenSolaris) special case. ncurses needs __EXTENSIONS__
for vsscanf() function, but must define _XOPEN_SOURCE to override
_XOPEN_SOURCE_EXTENDED for wide-character includes.
* AcSplit/CF_WITH_IDNA:
add the include- and lib-directories, e.g., for Solaris which puts the
headers in /usr/include/idn
* AcSplit/CF_LIBRARY_PATH:
reset target at beginning, to avoid symbol conflict
* AcSplit/CF_RPATH_HACK_2: make this less verbose
2010-03-27
* AcSplit/CF_RPATH_HACK: no longer depends on CF_SHARED_OPTS
* AcSplit/CF_SHARED_OPTS: more fixes (NetBSD 5.02 is working again)
* AcSplit/CF_SHARED_OPTS:
separate $LD_RPATH_OPT (available option) from the part used to satisfy
--enable-rpath
* AcSplit/CF_LD_RPATH_OPT: typo
* AcSplit/CF_LD_RPATH_OPT, AcSplit/CF_RPATH_HACK_2:
ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS:
refactor, splitting out CF_LD_RPATH_OPT which makes CF_RPATH_HACK reusable
* AcSplit/CF_ENABLE_RPATH: rename variable for consistency
* AcSplit/CF_RPATH_HACK:
revise to use CF_RPATH_OPT macro's LD_RPATH_OPT, as well as moving redundant
chunk into new macro CF_RPATH_HACK_2
* AcSplit/CF_REGEX:
split-up the nested macros to make this simpler to change.
2010-03-23
* AcSplit/CF_FIND_TDLIB:
modify to allow td_lib to be a subdirectory of the calling application, to
simplify RPM script.
2010-03-21
* AcSplit/CF_FIND_LINKAGE:
add check between system (default) and explicity paths, where we can find
the entrypoint in the given library.
2010-03-20
* AcSplit/CF_GCC_OPT_RDYNAMIC: ci-aclocal generated header
2010-03-13
* AcSplit/CF_DISABLE_LEAKS:
modify so that --enable-leaks will not define NO_LEAKS
* AcSplit/CF_MAN_PAGES: supply a missing "g" on a substitute-command
2010-03-04
* AcSplit/CF_VA_COPY:
use AC_TRY_LINK rather than AC_COMPILE, since IRIX64 can compile w/o c99,
but will not link.
2010-03-02
* AcSplit/CF_VA_COPY: ci-aclocal generated header
2010-02-27
* AcSplit/CF_NCURSES_ADDON: fix typo
fill in non-config case
* AcSplit/CF_NCURSES_ADDON: ci-aclocal generated header
2010-02-26
* AcSplit/CF_ADA_INCLUDE_DIRS:
modify ACPPFLAGS to support Ada95 as separate tarball, and slightly different
directory structure.
2010-01-30
* COPYING.asc, COPYING: bump
2010-01-27
* AcSplit/CF_AUTO_SYMLINK:
use $SHELL in case version.sh's permissions are incorrect
* AcSplit/CF_TRANSFORM_PROGRAM: add quotes
2010-01-26
* AcSplit/CF_WITH_SYMLINK: align comment-text
* AcSplit/CF__INIT_TRANSFORM: ci-aclocal generated header
* AcSplit/CF_WITH_SYMLINK: use CF__INIT_TRANSFORM
* AcSplit/CF_WITH_SYMLINK: change to make more generic
* AcSplit/CF_AUTO_SYMLINK, AcSplit/CF_TRANSFORM_PROGRAM:
ci-aclocal generated header
2010-01-25
* AcSplit/CF_WITH_SYMLINK: ci-aclocal generated header
* AcSplit/CF_WITH_X_DESKTOP_UTILS: make comment generic
2010-01-24
* AcSplit/CF_CURSES_TERM_H:
modify to handle cases such as cygwin where packager has installed curses.h
and term.h in different directories, e.g., to wedge in a termcap library.
2010-01-17
* AcSplit/CF_BUNDLED_INTL: add comment for GT_YES/GT_NO substitutions
* AcSplit/CF_HEADER_PATH, AcSplit/CF_INCLUDE_PATH:
always clear target variable before appending
2010-01-09
* AcSplit/CF_REMOVE_DEFINE:
allow the third parameter (the symbol name) to be a variable.
* AcSplit/CF_XOPEN_SOURCE:
use CF_ADD_CFLAGS to consistently add (and remove duplicates) for -D's to
CPPFLAGS
* AcSplit/CF_POSIX_C_SOURCE:
use CF_ADD_CFLAGS to consistently add (and remove duplicates) for -D's to
CPPFLAGS.
* AcSplit/CF_ADD_CFLAGS:
when adding a -D, use CF_REMOVE_DEFINE to remove previous occurrences,
to avoid redefinition warnings.
2010-01-07
* AcSplit/CF_XOPEN_SOURCE:
restore the fix for hpux11, lost in last checkin
2010-01-04
* AcSplit/CF_FUNC_TGETENT:
modify the check for termcap.h to only reject it if no full tgetent was
found (which would correspond to ncurses' termcap.h). This allows matching
termcap.h on older FreeBSD's which provided a custom tgetent().
2009-12-31
* config.guess:
2009-12-30 (however copyright date is for the next year)
* config.sub: 2009-12-31 (however, copyright date is for the next year)
2009-12-30
* AcSplit/CF_XOPEN_SOURCE:
add special case for Darwin to use its extensions (patch by Dennis Preiser)
2009-12-24
* AcSplit/CF_NETLIBS: add check for -lnetwork, from tin
* AcSplit/CF_BUNDLED_PCRE:
suppress sourcing pcre's version.sh, from tin
* AcSplit/AM_MULTIBYTE_ABLE: updated check, from tin
* config.sub, config.guess: 2009-12-13
2009-12-20
* AcSplit/CF_MAWK_RUN_FPE_TESTS:
re-create test-program for running the second check
2009-12-19
* AcSplit/CF_GPP_LIBRARY, AcSplit/CF_MATH_LIB, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_STDCPP_LIBRARY, AcSplit/CF_MAWK_MATHLIB,
AcSplit/CF_MAWK_RUN_FPE_TESTS, AcSplit/CF_REGEX:
put new libraries before $LIBS
* AcSplit/CF_REGEX:
check for regular expression library before doing AC_TRY_LINK's since the
default runtime libraries may not have regular expression support.
2009-12-18
* AcSplit/CF_MAWK_RUN_FPE_TESTS:
improve check for NaN, and write progress to config.log
* AcSplit/CF_MAWK_FPE_SIGINFO: minor cleanup, write to config.log
* AcSplit/CF_MAWK_FIND_MAX_INT: trim blanks
2009-12-15
* AcSplit/CF_MAWK_FPE_SIGINFO, AcSplit/CF_MAWK_RUN_FPE_TESTS:
fix comment
2009-12-14
* AcSplit/CF_MAWK_FPE_SIGINFO: change test-variable to a cache variable
* AcSplit/CF_MAWK_RUN_FPE_TESTS:
pass configure #define's to compile of fpe_check, to setup sigaction/signal choice.
* AcSplit/CF_MAWK_FPE_SIGINFO, AcSplit/CF_MAWK_RUN_FPE_TESTS:
eliminate redundant use of CF_MAWK_FPE_SIGINFO in CF_MAWK_RUN_FPE_TESTS
* AcSplit/CF_MAWK_CHECK_FUNC:
use the unquoted define, otherwise the shell variable name is #define'd
2009-12-13
* AcSplit/CF_DISABLE_ECHO:
modify to put tab outside the macro, so the substitution marker would always
be prefixed in the makefile with a tab, like other commands (patch by
Jonathan Nieder).
2009-12-12
* COPYING: update copyright date
2009-12-08
* AcSplit/CF_LEX_VERSION:
if no lex is found, autoconf uses ":", which is not reasonable. replace
that by "lex", to ensure a build-failure if the output of lex is really
needed (patch by Brendan O'Dea).
2009-11-20
* AcSplit/CF_GNUTLS:
investigation of Debian #555579 noticed that gnutls depends on gcrypt library
rather than crypt (though I seem to recall that library was renamed at some
point). Just rename it now, perhaps fix as needed later.
2009-11-19
* config.guess, config.sub: update from
http://savannah.gnu.org/projects/config
2009-10-16
* AcSplit/CF_RCS_SCCS: add SVN
2009-10-15
* AcSplit/CF_CHECK_FUNCDECL:
fix cut/paste typo: "$ac_func" was used where "$2" should be.
2009-10-14
* AcSplit/CF_WITH_X_APP_DEFAULTS, AcSplit/CF_WITH_X_ICONDIR:
use $(name) form rather than ${name} for default-value, to work on Tru64
2009-10-08
* AcSplit/CF_WITH_X_APP_DEFAULTS, AcSplit/CF_WITH_X_DESKTOP_UTILS,
AcSplit/CF_WITH_X_ICONDIR:
ci-aclocal generated header
2009-10-06
* AcSplit/CF_STDIO_UNLOCKED: fix typo
* AcSplit/CF_STDIO_UNLOCKED:
changes for diffstat, which doesn't care about putc_unlocked, but does give
a warning on HPUX and Tru64 for lack of prototype
2009-09-17
* AcSplit/CF_MAWK_RUN_FPE_TESTS: allow build outside source-tree
2009-09-03
* AcSplit/CF_LEX_VERSION:
ensure that we start looking for version on the second token (still), which
must begin with a numeric digit. Then trim off following tokens. This works
for reflex's latest change, e.g.
reflex 2.5.4 (20090902)
2009-08-31
* AcSplit/CF_GETOPT_HEADER: fix redefinition error for msys
* AcSplit/CF_GETOPT_HEADER: currently, Linux puts getopt() in unistd.h
2009-08-27
* AcSplit/CF_WITH_IDNA: add libs too
* AcSplit/CF_WITH_IDNA: ci-aclocal generated header
2009-08-25
* config.guess, config.sub: 2009-08-19
2009-08-24
* AcSplit/CF_LEX_STATES:
modify to determine the number of states supported in the lex implementation
2009-08-12
* AcSplit/CF_X_FONTENC: fontenc library normally requires z library
on conventional X systems, I also need the macros from AC_PATH_XTRA - fix
* AcSplit/CF_PROG_LINT: add splint and lclint
2009-08-11
* AcSplit/CF_GCC_ATTRIBUTES:
that last change needs a case-statement to prevent extra #define's
2009-07-26
* AcSplit/CF_ENABLE_WARNINGS: add CF_GCC_ATTRIBUTES
* AcSplit/CF_MAWK_FIND_MAX_INT: typo
* AcSplit/CF_MAWK_FIND_MAX_INT: compute MAXINT
* AcSplit/CF_MAWK_MAX__INT_PROGRAM: compute MAX__UINT
2009-07-23
* AcSplit/CF_GCC_ATTRIBUTES:
add #define's for GCC_PRINTFLIKE and GCC_SCANFLIKE to output.
* AcSplit/CF_MAWK_RUN_FPE_TESTS: don't add()
* AcSplit/CF_MAWK_CHECK_FUNC, AcSplit/CF_MAWK_CHECK_HEADER,
AcSplit/CF_MAWK_FPE_SIGINFO, AcSplit/CF_MAWK_RUN_FPE_TESTS:
obsoleting CF_MAWK_DEFINE
* AcSplit/CF_MAWK_CC_FEATURES: obsoleting CF_MAWK_DEFINE2()
* AcSplit/CF_MAWK_CHECK_SIZE_T: revise to use AC_DEFINE_UNQUOTED
* AcSplit/CF_MAWK_FIND_MAX_INT:
revise to use AC_DEFINE_UNQUOTED rather than using defines.out
* AcSplit/CF_MAWK_MAX__INT_PROGRAM: write to conftest.out
2009-07-16
* AcSplit/CF_XOPEN_SOURCE:
HPUX 11.23 (at least) will declare mbstate_t in wchar.h only if an intermediate
#define is triggered, which happens only when _XOPEN_SOURCE is defined to 500.
* AcSplit/CF_MIN_GETCCHAR: add #define for MIN_GETCCHAR
* AcSplit/CF_CURSES_CHECK_DATA:
make this check for symbol in library if it's not declared in header
* AcSplit/CF_CURSES_FUNCS: split-out CF__CURSES_HEAD
* AcSplit/CF_CURSES_CHECK_DATA: RCS_BASE
* AcSplit/CF__CURSES_HEAD, AcSplit/CF_MIN_GETCCHAR:
ci-aclocal generated header
2009-07-14
* AcSplit/CF_MAWK_FPE_SIGINFO, AcSplit/CF_MAWK_RUN_FPE_TESTS:
fix to work with cygwin
2009-07-12
* AcSplit/CF_MAWK_MAINTAINER: update mawk maintainer email
2009-07-05
* AcSplit/CF_MAWK_MATHLIB:
fix typo that made this always fail if log isn't in -lm, but is found some
other place.
2009-04-04
* AcSplit/CF_LDFLAGS_STATIC:
add Darwin to the list where mixed -static/-dynamic doesn't work:
(gcc 4.0.1)
* AcSplit/CF_LEX_CHAR_CLASSES: amend message
* AcSplit/CF_LEX_OPTIONS, AcSplit/CF_LEX_POINTER:
ci-aclocal generated header
* AcSplit/CF_LEX_STATES:
move the MAKE_LEX variable out of here, to make it more configurable
2009-03-30
* AcSplit/CF_WITH_INSTALL_PREFIX:
remove extra '$' (report by Dagobert Michelsen)
2009-03-28
* AcSplit/CF_NCURSES_WRAP_PREFIX: ci-aclocal generated header
* AcSplit/CF_CURSES_ACS_MAP: use CF_NCURSES_WRAP_PREFIX
* AcSplit/CF_CURSES_ACS_MAP: add ncwrap_acs_map (Charles Wilson)
* AcSplit/CF_WITH_LIBTOOL:
better fix for adding -no-undefined (by setting a variable which can be
added as needed to libtool invocation). Also, allow libtool versions to
end with a lowercase a-z letter.
2009-03-21
* AcSplit/CF_WITH_LIBTOOL: typo
* AcSplit/CF_WITH_LIBTOOL:
add -no-undefined option for Cygwin, MinGW, UWIN and AIX (report by Charles Wilson)
2009-03-14
* AcSplit/CF_LDFLAGS_STATIC:
Solaris 10 (and presumably later versions) do not have a static libc,
and gcc gets confused when linking tic (even though I switch back to
dynamic after linking -
LIBS_TIC = -Bstatic -L../lib -lncurses -L../lib -lncurses -Bdynamic
2009-02-24
* AcSplit/CF_WITH_PERL:
give warning, but do not error out to avoid the nuisance of a failed build
due to gcc error when it sees "main()" after encountering perl's prototype
for main().
2009-02-13
* AcSplit/CF_XKB_BELL_EXT:
configure check needed Intrinsic.h since X's headers as a rule are not
self-contained. Add a check while I'm here to ensure that XkbBell()
can be linked, since that's what I'm using.
2009-02-07
* AcSplit/CF_AR_FLAGS: tweak to avoid changing line-count
2009-02-01
* AcSplit/CF_SSL: add check for -ldl
* AcSplit/CF_GCC_WARNINGS: correct a comment for icc9
allow warning for implicit conversion of 64-bit stuff
* AcSplit/CF_SSL:
eliminate duplicate -lssl/-lcrypto from pkg-config by moving the CF_ADD_LIBDIR
and CF_ADD_INCDIR to the non pkg-config logic.
2009-01-30
* AcSplit/CF_SSL: fixes to build lynx --with-ssl using msys (for mingw)
2009-01-25
* AcSplit/CF_X_FREETYPE: fix overrun progress message
* AcSplit/CF_PKG_CONFIG: parenthesis tweaks
* AcSplit/CF_X_FREETYPE: use CF_PKG_CONFIG
tidy up messages a little
2009-01-18
* AcSplit/CF_ADD_LIBDIR: better fix - I'd removed the wrong "$"
* AcSplit/CF_ADD_INCDIR: remove spurious "$" from ifelse parameter
2009-01-17
* AcSplit/CF_SHARED_OPTS:
add a -shared option to MK_SHARED_LIB where we have a -Bsharable
This fixes (for OpenBSD at least) a dangling reference to "main" in libncurses
2009-01-11
* AcSplit/CF_ADD_SEARCHPATH, AcSplit/CF_PATH_PROG,
AcSplit/AM_PATH_PROG_WITH_TEST, AcSplit/CF_MANPAGE_FORMAT,
AcSplit/CF_PATHSEP, AcSplit/CF_PROGRAM_FULLPATH,
AcSplit/CF_PROGRAM_PREFIX, AcSplit/CF_WITH_PATHLIST:
change variable from PATHSEP to PATH_SEPARATOR
2009-01-11 Doug.Kaufman
* AcSplit/AM_WITH_NLS:
provide iconv library in $LIBS when checking for intl library
2009-01-11
* AcSplit/CF_ADD_LIBDIR:
remove spurious dollar-sign, report by José H. Espinosa on lynx-dev.
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_X_ATHENA_CPPFLAGS,
AcSplit/CF_PDCURSES_X11, AcSplit/CF_NCURSES_CONFIG,
AcSplit/CF_BUNDLED_INTL:
append (not prepend) to CFLAGS/CPPFLAGS
2009-01-07
* AcSplit/CF_AR_FLAGS: ac 2.13 uses single param for AC_SUBST()
2009-01-06
* AcSplit/CF_FIND_TDLIB: append (not prepend) to CFLAGS/CPPFLAGS
* AcSplit/CF_CURSES_STYLE: append (not prepend) to CFLAGS/CPPFLAGS
* AcSplit/CF_X_FREETYPE, AcSplit/CF_X_FONTENC, AcSplit/CF_INCLUDE_DIRS,
AcSplit/CF_FUNC_ALLOCA, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_CURSES_CPPFLAGS, AcSplit/CF_ADD_CFLAGS:
append (not prepend) to CFLAGS/CPPFLAGS
2009-01-03
* AcSplit/CF_SHARED_OPTS: fix for Solaris vs xcode
* AcSplit/CF_SHARED_OPTS:
export the LD_RPATH_OPT, to use in ncurses's gen-pkgconfig.in
2009-01-01
* AcSplit/CF_AR_FLAGS: completed first cut, for testing.
* AcSplit/CF_LINK_DATAONLY: lynx - change options name to $ARFLAGS
* AcSplit/CF_AR_FLAGS, AcSplit/CF_PROG_AR, AcSplit/CF_PROG_RANLIB:
ci-aclocal generated header
2008-12-29
* AcSplit/CF_NETLIBS:
add a case for mingw, which needs winsock/winsock2/ws2_32 dll
2008-12-27
* AcSplit/CF_WITH_LIBTOOL: tabs
* AcSplit/CF_LDFLAGS_STATIC: add case for AIX 5 and AIX 6
* AcSplit/CF_LIB_SUFFIX, AcSplit/CF_SHARED_OPTS:
add case for AIX 5 and AIX 6 using gcc
* AcSplit/CF_XOPEN_SOURCE:
ncurses patch for AIX mentions AIX 6.1 - add aix6 to case for _ALL_SOURCE
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_AWK_BIG_PRINTF,
AcSplit/CF_SRC_MODULES:
tabs
2008-12-25
* AcSplit/CF_GNUTLS, AcSplit/CF_SSL: add comment
2008-12-24
* AcSplit/CF_FIND_LINKAGE:
add a verbose message, to tell what we're searching for.
* AcSplit/CF_PKG_CONFIG:
add a --with option, to allow this to be disabled (per lynx-dev complaint
that pkg-config was found in ports, and did not have a way to suppress it,
to build solely from base libraries).
2008-12-18
* config.sub: 2008-12-11
* config.guess: 2008-12-19
2008-12-15
* AcSplit/CF_SSL:
correct second check of the $1 parameter, which broke configuration without
the pkg-config file.
2008-12-13
* AcSplit/CF_LIB_RULES:
for ncurses, if I use --without-manpages, then the man/Makefile will not
be executed. Move the distclean action to delete edit_man.* to the parent
directory.
* AcSplit/CF_SRC_MODULES:
duh - overlooked first assignment to SRC_SUBDIRS when implementing the
"--without-manpages" option.
* AcSplit/CF_XOPEN_SOURCE:
add FreeMiNT ncurses support, from GenToo #250454
* AcSplit/CF_PROG_LDCONFIG:
build-fix for dragonfly, from NetBSD patch for ncurses port
2008-12-11
* AcSplit/CF_GNUTLS:
hmm - unsure if Debian simply does not package the gnutls-openssl script
for pkg-config, but leave it as a possibility
* AcSplit/CF_GNUTLS: looks ok - pkg-config checks.
* AcSplit/CF_CHECK_SSL_X509:
make this a "define" to avoid odd problems including it from CF_SSL and CF_GNUTLS
2008-12-09
* AcSplit/CF_GNUTLS: start integrating pkg-config
* AcSplit/CF_GNUTLS: reindent'd
* AcSplit/CF_SSL: make this properly verbose
* AcSplit/CF_SSL: revised to use pkg-config, if available
2008-12-07
* AcSplit/CF_HEADER_PATH:
lynx - someone reported a problem using --with-ssl=XXX, and I realized that
I'd added the XXX to CPPFLAGS, which wouldn't be tested in CF_SUBDIR_PATH.
* AcSplit/CF_LIBRARY_PATH:
lynx: add-back values from LDFLAGS/LIBS which happen to be -L's, in case
an option (such as --with-ssl=XXX) set the environment variables.
2008-11-08
* AcSplit/CF_LIB_RULES: implement "--without-manpages" option
2008-11-01
* AcSplit/CF_WITH_NC_ALLOC_H: let's use --disable-leaks instead.
* AcSplit/CF_WITH_NC_ALLOC_H:
make this use --with-no-leaks option (since I changed CDK's configure script
to use that macro).
2008-10-30
* AcSplit/CF_CURSES_LIBS:
add pdcurses library to search list, used in CDK by Gregory Sharp (from
patch 7 Jul 2007).
2008-10-25
* AcSplit/CF_SHARED_OPTS: add case for QNX 6 (works for me)
2008-10-18
* AcSplit/CF_LDFLAGS_STATIC:
have to cancel these symbols for OS/2 EMX as well, since the -static
conflicts with -Zcrtdll
2008-10-17
* AcSplit/CF_UTF8_LIB:
rewrote using CF_FIND_LINKAGE; the problem is that the BSDs put packages
where gcc is not likely to find them, so for any unusual libraries we need
to add -I and -L options.
2008-10-04
* AcSplit/CF_AWK_BIG_PRINTF:
AIX 5.1's nawk choked on printf's bigger than the threshold, writing zero
bytes. That fooled my check for record length (since nawk found no records).
Replace the second stage with a check for number of records and record length.
2008-09-20
* AcSplit/CF_LIB_RULES:
another tweak - this built both cfg-shared and cfg-thread on cygwin
* AcSplit/CF_LIB_RULES:
fixes for cygwin and cfg-thread (which builds libtic, libtinfo and libncurses).
The problem was that the "t" was appended after the $(ABI_VERSION).
2008-09-13
* AcSplit/CF_RPATH_HACK:
don't remove the -L option, since we may have static libraries in the same
directory.
* AcSplit/CF_SHARED_OPTS:
paste-in chunk which seems to make openbsd do rpath, adapted from linux
* AcSplit/CF_LIB_SUFFIX: tabsd
* AcSplit/CF_LIB_RULES: add cf_depsuf params to CF_LIB_SUFFIX calls.
* AcSplit/CF_LIB_PREFIX: reformatted
* AcSplit/CF_LIB_SUFFIX:
merge in the 3rd parameter used in ncurses, so we set both lib- and dep-suffixes
That's so I can move a chunk of cygwin-specific stuff out of configure.in
* AcSplit/CF_ENABLE_RPATH: ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS:
for openbsd 2.x on, build shared libraries with a release-version appended.
OpenBSD doesn't have symbolic links in /usr/lib, but does append the
release xx.yy to each ".so" filename.
2008-09-09
* AcSplit/CF_MAWK_CHECK_HEADER: use CF_UPPER for case-conversion].
* AcSplit/CF_MAWK_ADD_NO: don't use translit - again a porting problem
* AcSplit/CF_MAWK_DEFINE:
first parameter is now unquoted, since it is a variable set in CF_UPPER
* AcSplit/CF_MAWK_CHECK_FUNCS, AcSplit/CF_MAWK_CHECK_HEADERS:
formatting
* AcSplit/CF_MAWK_CHECK_FUNC: use CF_UPPER
* AcSplit/CF_MAWK_CHECK_FUNC, AcSplit/CF_MAWK_CHECK_HEADER: add quotes
* AcSplit/CF_MAWK_CHECK_FUNCS, AcSplit/CF_MAWK_CHECK_HEADERS:
don't use CF_MAWK_REPEAT_IT, since it's likely to be a problem porting to
newer autoconf's
* AcSplit/CF_MAWK_ADD_NO, AcSplit/CF_MAWK_CC_FEATURES,
AcSplit/CF_MAWK_CHECK_FPRINTF, AcSplit/CF_MAWK_CHECK_FUNC,
AcSplit/CF_MAWK_CHECK_FUNCS, AcSplit/CF_MAWK_CHECK_HEADER,
AcSplit/CF_MAWK_CHECK_HEADERS, AcSplit/CF_MAWK_CHECK_LIMITS_MSG,
AcSplit/CF_MAWK_CHECK_SIZE_T, AcSplit/CF_MAWK_CONFIG_H_HEADER,
AcSplit/CF_MAWK_CONFIG_H_TRAILER, AcSplit/CF_MAWK_DEFINE,
AcSplit/CF_MAWK_DEFINE2, AcSplit/CF_MAWK_FIND_MAX_INT,
AcSplit/CF_MAWK_FIND_SIZE_T, AcSplit/CF_MAWK_FPE_SIGINFO,
AcSplit/CF_MAWK_GET_CONFIG_USER, AcSplit/CF_MAWK_MAINTAINER,
AcSplit/CF_MAWK_MATHLIB, AcSplit/CF_MAWK_MAX__INT_PROGRAM,
AcSplit/CF_MAWK_OUTPUT_CONFIG_H, AcSplit/CF_MAWK_PROG_GCC,
AcSplit/CF_MAWK_PROG_YACC, AcSplit/CF_MAWK_REPEAT_IT,
AcSplit/CF_MAWK_RUN_FPE_TESTS, AcSplit/CF_MAWK_SET_IF_UNSET,
AcSplit/CF_MAWK_XADD_NO:
ci-aclocal generated header
2008-09-08
* AcSplit/CF_SHARED_SONAME:
change cf_shared_soname to cf_cv_shared_soname (a cache variable)
* AcSplit/CF_SHARED_OPTS:
rename $cf_shared_soname to $cf_cv_shared_soname
2008-09-06
* AcSplit/CF_SHARED_OPTS: fix for gcc on IRIX64
* AcSplit/CF_AWK_BIG_PRINTF: ci-aclocal generated header
* AcSplit/CF_AWK_BIGSTRING: rewrote, adding a special case for mawk.
* AcSplit/CF_AWK_BIGSTRING: ci-aclocal generated header
2008-08-30
* config.guess: 2008-04-14 from
http://savannah.gnu.org/projects/config
* config.sub: 2008-06-16 from
http://savannah.gnu.org/projects/config
2008-08-23
* AcSplit/CF_WITH_PTHREAD:
change from AC_CHECK_LIB to AC_TRY_LINK, since OSF/1 5.1 (at least) does
symbol mangling, e.g., puts "__" in the front of the symbol name in the
library. An explicit link makes the test work.
* AcSplit/CF_WITH_LOCALE_ALIAS:
rewrote to make the default value "smart" (based on a list)
2008-08-22
* AcSplit/CF_CHECK_ENVIRON: fix comment
* AcSplit/CF_CHECK_ENVIRON: ci-aclocal generated header
* AcSplit/CF_CHECK_ERRNO: implement a second parameter
2008-08-21
* AcSplit/CF_X_FONTENC:
second cut of rewrite - needs some testing, but should be minor fixes.
2008-08-20
* AcSplit/CF_X_FONTENC:
start rewriting this so it will work w/o pkg-config
it's not complete - $withval could be "yes" if not given.
but I did fix the name of the library (fontenc, not Xfont).
2008-08-16
* AcSplit/CF_WEAK_SYMBOLS: ci-aclocal generated header
2008-08-11
* AcSplit/CF_WITH_PERL:
rename LINK_PREFIX to avoid conflict(?) with old Purify hack.
add a link-test, in case the library is actually missing (casualty of packagers)
* AcSplit/CF_WITH_PERL: ci-aclocal generated header
2008-08-04
* AcSplit/CF_HASHED_DB_LIBS: allow $LIBS to override
2008-07-27
* AcSplit/CF_XOPEN_SOURCE: Date: Tue, 1 Jul 2008 22:29:26 +0300
From: Hasso Tepper <hasso@estpak.ee>
Subject: xterm problem in DragonFly BSD
To: dickey@invisible-island.net
[-- Attachment #1 --]
[-- Type: text/plain, Encoding: 7bit, Size: 0.5K --]
Xterm doesn't compile in latest DragonFly BSD:
../charproc.c: In function 'in_put':
../charproc.c:3242: error: 'fd_mask' undeclared (first use in this function)
../charproc.c:3242: error: (Each undeclared identifier is reported only once
../charproc.c:3242: error: for each function it appears in.)
*** Error code 1
It worked in older versions because fd_mask wasn't guarded by __BSD_VISIBLE.
Attached patch seems to reasonable to apply and is tested to be working by
many DragonFly BSD users.
-------TD: of course this is used by other programs (will see).
2008-07-26
* AcSplit/CF_GCC_WARNINGS:
give up on -Winline for a while - gcc 4.3.1 appears to have no workable
settings that make ncurses's inlining work w/o warnings.
2008-07-05
* AcSplit/CF_LDFLAGS_STATIC:
suppress gcc's own static/dynamic flags on AIX, since the latest (gcc 4.2.4)
does not link test/bs anymore. That is because bs.c uses getlogin, which
isn't picked up properly. For details, I added -Wl,b,noquiet to the end
of the link-line. Here's the properly-working linkage:
2008-06-14
* AcSplit/CF_SHARED_OPTS: Date: Tue, 10 Jun 2008 05:46:38 +0000
From: Poor Yorick <org.gnu.bug-ncurses@pooryorick.com>
Subject: undefined symbol "main" when compiling libXpm,
which depends on ncurses
To: bug-ncurses@gnu.org
X-Mailer: WebMail 2.55
For the record: compiling ncurses on SunOS 5.9, with gcc-4.2.2 and, it was
necessary to patch ncurses install with a '-shared' flag. Otherwise,
libncurses(w).so.5.6 was linked as an executable with a "main" symbol, causing
a subsequent install of any software package depending on libncurses to fail,
complaining about undefined "main" symbol:
-------TD: tww's using gcc 3.4.3, which works fine w/o the "-shared" option.
It is specific to gcc. While reading Sun's cc manpage, I see this note
about the PIC options:
-KPIC
(SPARC) Obsolete. You should not use this option. Use
-xcode=pic32 instead. For a complete list of obsolete
options and flags, see the C User's Guide.
(x86) -KPIC is identical to -Kpic on x86 architectures.
-Kpic
(SPARC) Obsolete. You should not use this option. Use
-xcode=pic13 instead. For a complete list of obsolete
options and flags, see the C User's Guide.
(x86) Generate position-independent code for use in
shared libraries (small model). Permits references to,
at most, 2**11 unique external symbols
So I changed to use the -xcode option.
2008-04-13
* AcSplit/CF__SSL_BODY: let USE_GNUTLS_FUNCS override OpenSSL check
* AcSplit/CF__SSL_HEAD:
let USE_GNUTLS_FUNCS override OpenSSL compat header
* AcSplit/CF_GNUTLS:
q/d hack to use presence of 2nd parameter to make this use only gnutls functions
via a wrapper/stub, for lynx to co-exist with the restrictive GPLv3 relicensing
of gnutls-openssl library.
2008-04-12
* AcSplit/CF_FUNC_OPENPTY:
FreeBSD and OpenBSD have the same function - but use different header files.
* AcSplit/CF_FUNC_OPENPTY: ci-aclocal generated header
2008-03-29
* AcSplit/CF_WITH_LIBTOOL: from bug-ncurses:
From: Michael Haubenwallner
Subject: minor issues when building ncurses-5.6 with libtool
Date: Thu, 20 Mar 2008 17:52:58 +0100
Hi,
there are two minor issues (depending on libtool and compiler used).
1) The libtool object files are still called '.o' (should be '.lo').
2) libtool might not detect that it should use tag 'CC'.
Can't remember the actual error messages, but those two have shown up on
aix when using CC=powerpc-ibm-aix5.3.0.0-gcc.
Attached patch fixes them.
Thanks!
/haubi/
ncurses-5.6-libtool.patch.
* AcSplit/CF_WITH_PTHREAD: ci-aclocal generated header
2008-03-25
* AcSplit/CF_TYPE_FD_MASK: fix to make the CSRG_BASED ifdef work
* AcSplit/CF_TYPE_FD_MASK: ci-aclocal generated header
* AcSplit/CF_TYPE_FD_SET: minor cleanup
2008-03-23
* AcSplit/CF_IMAKE_CFLAGS, AcSplit/CF_LEX_CHAR_CLASSES,
AcSplit/CF_LEX_STATES, AcSplit/CF_X_TOOLKIT:
use AC_MSG_WARN (new name works with 2.13)
* AcSplit/CF_ANSI_CC_REQD, AcSplit/CF_ARG_WITH, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_FIND_LIBRARY, AcSplit/CF_FIND_TDLIB,
AcSplit/CF_MAKE_AR_RULES, AcSplit/CF_MAKE_INCLUDE,
AcSplit/CF_NCURSES_HEADER, AcSplit/CF_NSS_COMPAT,
AcSplit/CF_PDCURSES_X11, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_SLANG_CPPFLAGS, AcSplit/CF_SOCKS, AcSplit/CF_SOCKS5,
AcSplit/CF_SYS_NAME, AcSplit/CF_WITH_DFTENV, AcSplit/CF_WITH_GPM,
AcSplit/CF_WITH_PROGRAM, AcSplit/CF_X_MOTIF, AcSplit/CF_X_OPENLOOK:
use AC_MSG_ERROR (new name, works with 2.13)
* AcSplit/CF_X_ATHENA_LIBS, AcSplit/CF_CHECK_CACHE,
AcSplit/CF_PATH_SYNTAX:
use AC_MSG_ERROR
2008-03-14
* config.sub: 2008-03-08 (from tin beta)
* config.guess: 2008-03-12 (from tin beta)
2008-02-09
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_ADD_LIBDIR:
building with autoconf 2.13, get warnings - use CF_VERBOSE instead
2008-01-30
* AcSplit/CF_NSS_COMPAT: RCS_BASE
2008-01-25
* AcSplit/CF_UTMP_UT_SYSLEN: RCS_BASE
* AcSplit/CF_UTMP: add CF_UTMP_UT_SYSLEN - see
http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=4045962
2008-01-06
* AcSplit/CF_CHECK_SSL_X509:
TG insists that he's got a const or something like that in the
prototype for (X509_STORE_CTX_get_error() - tweak the check.
* AcSplit/CF_CHECK_TYPE:
Thorsten Glaser patch to use AC_CHECK_TYPE as/is with autoconf 2.13
* AcSplit/CF_SSL:
patch to check X509 support after establishing libraries - Thorsten Glaser
2007-12-17
* AcSplit/CF_FUNC_ICONV:
tweak to make cf_cv_func_iconv set to "yes" if I do find the function in
"-liconv". Noticed on cygwin that I had to do this (a shame that cygwin
doesn't provide locale support).
2007-12-01
* AcSplit/CF_HASHED_DB_VERSION:
for grins, look for db5 - I see Debian has a package
* AcSplit/CF_HASHED_DB_LIBS: look also for -ldb-3, etc.
* AcSplit/CF_RPATH_HACK:
don't try to update LDFLAGS if there's no applicable rpath-hack
2007-11-26
* AcSplit/CF_LEX_VERSION:
lex on AIX (even 5.1) doesn't stop on error from illegal option - it starts
reading the data from stdin. Adjust test to avoid hanging.
2007-11-24
* AcSplit/CF_RPATH_HACK:
put the custom paths in front of the others in case we're starting with the
system -L/usr/lib
* AcSplit/CF_RPATH_HACK: ci-aclocal generated header
* AcSplit/CF_HASHED_DB: typo
* AcSplit/CF_HASHED_DB:
add optional param to specify the include- and lib-directories
2007-08-11
* AcSplit/CF_C_INLINE:
don't try to apply the --param to Intel compiler (it complains about a -fp
option and gives occasional error returns though not on _this_ test).
2007-07-30
* AcSplit/CF_ADD_INCDIR: typo
* AcSplit/CF_ADD_LIBDIR:
combine check on $LDFLAGS and $LIBS just in case
* AcSplit/AM_ICONV: corrected check for libiconv.*
* AcSplit/CF_CHECK_SSL_X509: q/d hack to make this work...
* AcSplit/CF_ADD_INCDIR:
use the check-for-duplicates from CF_ADD_LIBDIR
2007-07-29
* AcSplit/CF_FIND_LINKAGE: in the default case, must update $LIBS too
* AcSplit/AM_ICONV, AcSplit/CF_ADD_SEARCHPATH,
AcSplit/CF_FIND_IPV6_LIBS, AcSplit/CF_NCURSES_CPPFLAGS:
use CF_ADD_INCDIR
* AcSplit/AM_WITH_NLS: use CF_ADD_INCDIR
don't need check for empty param before CF_ADD_LIBDIR
* AcSplit/CF_WITH_BZLIB, AcSplit/CF_WITH_ZLIB:
rewrote using CF_FIND_LINKAGE
* AcSplit/CF_FIND_LINKAGE: use CF_ADD_INCDIR
* AcSplit/CF_SOCKS5: use CF_ADD_OPTIONAL_PATH
Considered using CF_FIND_LINKAGE, but decided not to, since this is used
rarely (in lynx), and because the logic for finding socks5 versus the
socks4 prefix is fragile.
* AcSplit/CF_ADD_OPTIONAL_PATH: ci-aclocal generated header
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_ADD_LIBDIR:
ensure that the parameter is non-empty
* AcSplit/CF_SOCKS: rewrote using CF_FIND_LINKAGE
* AcSplit/CF_FIND_LINKAGE:
tweak messages to avoid leaving unneeded "CF_" strings in the configure script
* AcSplit/AM_ICONV, AcSplit/CF_GNUTLS, AcSplit/CF_SSL:
use CF_ADD_OPTIONAL_PATH
* AcSplit/CF_ADD_INCDIR:
amend a comment (I don't have a nice way to express the lobotomization of
gcc done for BSD ports...)
* AcSplit/CF_CHECK_SSL_X509: ci-aclocal generated header
* AcSplit/CF_GNUTLS: misc fixes - got this working with lynx.
* AcSplit/CF_FIND_HEADER, AcSplit/CF_SSL, AcSplit/CF_WITH_BZLIB,
AcSplit/CF_WITH_ZLIB:
use CF_ADD_INCDIR
* AcSplit/CF_NCURSES_CC_CHECK:
parameter-2 wasn't used in one case where the usual-parameter $cf_header
was - fix.
* AcSplit/CF_CURSES_LIBS: detab'd
* AcSplit/AM_WITH_NLS: use CF_ADD_LIBDIR
* AcSplit/CF_ADD_LIBDIR:
rewrote - don't simply suppress tokens that are not directories, but also
filter out /usr/lib and check for directories already in $LDFLAGS.
change the variable from $LIBS to $LDFLAGS, but allow it to be overridden
for macros that collect the options in a separate variable.
* AcSplit/AM_ICONV, AcSplit/CF_ADD_SEARCHPATH, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_FIND_LINKAGE, AcSplit/CF_GNUTLS, AcSplit/CF_NCURSES_LIBS,
AcSplit/CF_SOCKS, AcSplit/CF_SOCKS5, AcSplit/CF_SSL:
use CF_ADD_LIBDIR
* AcSplit/CF_ADD_SUBDIR_PATH: make a message clearer
* AcSplit/CF_VERBOSE:
add a call to CF_MSG_LOG so the same information goes to config.log
* AcSplit/CF_ADD_SUBDIR_PATH: ci-aclocal generated header
* AcSplit/CF_FIND_LINKAGE:
put the CF_VERBOSE in the loops _before_ the test, to make it simpler to
see what was tested before the "found" message.
* AcSplit/CF_SUBDIR_PATH: split-out CF_ADD_SUBDIR_PATH
also change order of tests to put /usr, $prefix first to lessen user surprise
* AcSplit/CF_GCC_ATTRIBUTES, AcSplit/CF_GCC_WARNINGS:
use $as_me if defined
* AcSplit/CF_MSG_LOG:
use autoconf 2.5x's $as_me variable if defined, to put the script name in the log.
make the line-number formatted like the other configure-lineno stuff.
2007-07-28
* AcSplit/CF_FIND_LINKAGE:
AC_TRY_LINK doesn't add optional libraries - modified to work around
* AcSplit/CF_GNUTLS:
second cut - fix some typos, make the check for -lgnutls-openssl work.
* AcSplit/CF_SSL:
separate the checks for <openssh/xxx> versus -I/usr/include
* AcSplit/CF_SSL:
add 7th param to CF_FIND_LINKAGE to pass in additional libraries
* AcSplit/CF_FIND_LINKAGE: add 7th parameter for additional libraries
* AcSplit/CF_GNUTLS:
rewrite/simplify using CF_FIND_LINKAGE and CF_ADD_SEARCHPATH
* AcSplit/CF_SSL: rewrote to use CF_FIND_LINKAGE and CF_ADD_SEARCHPATH
* AcSplit/AM_ICONV: typo
* AcSplit/CF_FIND_LINKAGE:
mods to support CF_SSL, which uses "openssl" module name as 6th parameter.
The module name is not necessarily the same as the library name.
2007-07-26
* AcSplit/CF__SSL_BODY, AcSplit/CF__SSL_HEAD:
ci-aclocal generated header
* AcSplit/AM_WITH_NLS: get the -L option from CF_FIND_LINKAGE data
* AcSplit/CF_FIND_LINKAGE:
to compile the test-code in the AC_TRY_LINK, we need the CPPFLAGS from
the AC_TRY_COMPILE
* AcSplit/AM_WITH_NLS:
broaden/simplify the search using CF_FIND_LINKAGE
also cleanup the untidy indents to make this readable...
* AcSplit/AM_ICONV:
modified CF_FIND_LINKAGE to generate per-library cache names, so we don't
have to worry about symbol conflicts.
* AcSplit/CF_FIND_LINKAGE:
ensure that the associated variables for header and library paths are
unique, making them into cache variables suffixed by the library name.
* AcSplit/CF__INTL_BODY, AcSplit/CF__INTL_HEAD:
ci-aclocal generated header
* AcSplit/CF_ADD_SEARCHPATH, AcSplit/CF__ICONV_BODY: comment
* AcSplit/AM_ICONV:
make it clearer that the second part of this macro is using a cache variable,
and discard the overly verbose recitation of the derived prototype.
* AcSplit/CF_FIND_LINKAGE:
add CF_HEADER_PATH also - considered merging the two more closely as in
CF_ADD_SEARCHPATH but for the moment too complex.
* AcSplit/CF_ADD_SEARCHPATH, AcSplit/CF__ICONV_BODY,
AcSplit/CF__ICONV_HEAD:
ci-aclocal generated header
* AcSplit/AM_ICONV:
rewrote, splitting out CF__ICONV_HEAD, CF__ICONV_BODY, CF_ADD_SEARCHPATH
(actually from CF_SSL), CF_FIND_LINKAGE
2007-07-04
* AcSplit/CF_HEADERS_SH: ci-aclocal generated header
* config.sub: 2007/6/28
* config.guess: 2007/5/17
2007-06-27
* AcSplit/CF_GCC_WARNINGS:
take "$1" out of the Intel list, since it's unreliable
2007-06-18
* README.asc, COPYING.asc: RCS_BASE
* README: make the scope of COPYING clearer
2007-06-17
* COPYING, README: RCS_BASE
2007-06-02
* AcSplit/CF_LARGEFILE:
add definition for _FILE_OFFSET_BITS to fix redefinition problems on Solaris
2.6 and 7 (patch by Peter O'Gorman).
2007-05-24
* AcSplit/CF_IMAKE_CFLAGS:
bug report for xterm with Solaris sed noted that the script I'd
written didn't pass through the quoted stuff for X.Org's version.
Tested with Heirloom sed (since it was simpler than getting
an imake configuration on Solaris to match X.Org), and found two
main problems:
a) some bogus bug prevented "[^- ]" from matching where "[^ -]" does.
b) sed didn't like the nested \(...\) I was using - didn't even assign
it a different \3 or something like that. It just didn't match.
So I filtered out the tabs first, then wrote a loop in the sed script to
convert "embedded" spaces in a parameter value to tabs (since that seems
to be the safest character for a magic replacement). Then I could simplify
the name=value part to an expression that didn't need any nesting of \(...\).
Finally restore the embedded blank (should be one!).
2007-05-13
* AcSplit/CF_WITH_PATH:
if the "real" default value can be empty, don't insist that the result is
nonempty
2007-05-05
* AcSplit/CF_STDIO_UNLOCKED: ci-aclocal generated header
* AcSplit/CF_KILLPG: use $cf_cv_main_return
2007-04-28
* AcSplit/CF_LDFLAGS_STATIC: typo
* AcSplit/CF_C_INLINE, AcSplit/CF_CHECK_GPM_WGETCH,
AcSplit/CF_LDFLAGS_STATIC:
ci-aclocal generated header
* AcSplit/CF_ANSI_CPP, AcSplit/CF_RESTARTABLE_PIPEREAD,
AcSplit/CF_COREFILE, AcSplit/CF_CRYPT_FUNC, AcSplit/CF_CURSES_FUNCS,
AcSplit/CF_FUNC_CURSES_VERSION, AcSplit/CF_FUNC_FORK,
AcSplit/CF_FUNC_GETADDRINFO, AcSplit/CF_FUNC_SYSTEM,
AcSplit/CF_NCURSES_VERSION, AcSplit/CF_POSIX_VDISABLE,
AcSplit/CF_REGCMP_LIBS, AcSplit/CF_REGEXPR_H_FUNCS,
AcSplit/CF_REGEX_H_FUNCS, AcSplit/CF_RE_COMP_FUNCS,
AcSplit/CF_SET_ERRNO, AcSplit/CF_SIG_CONST, AcSplit/CF_TD_SIG_ARGS,
AcSplit/CF_TIOCGWINSZ, AcSplit/CF_WIDEC_SHIFT, AcSplit/CF_YACC_ERROR:
use $cf_cv_main_return
2007-04-22
* AcSplit/CF_SRAND_PARSE: ci-aclocal generated header
* AcSplit/CF_SRAND:
support the OpenBSD flavor of arc4random(), noting that it is asymmetric
without a good reason -lynx
2007-04-08
* AcSplit/CF_WITH_LIBTOOL: LIB_LINK has to contain ${CC} now
* AcSplit/CF_WITH_LIBTOOL_OPTS: typo
* AcSplit/CF_WITH_LIBTOOL_OPTS: ci-aclocal generated header
* AcSplit/CF_WITH_LIBTOOL: add LIBTOOL_OPTS
this is inserted as a MODE-ARG in LIB_CREATE and LIB_LINK.
To make this work, LIB_LINK now has ${CC} (so each package must be updated).
This is for dialog - will apply to cdk, ncurses...
2007-03-31
* AcSplit/CF_MAN_PAGES:
use mkdirs.sh in ncurses (that's what I maintain in my-autoconf)
2007-03-25
* mkdirs.sh: Date: Mon, 19 Mar 2007 07:20:21 +0100
From: Harald van D??k <truedfx@gentoo.org>
Subject: Gentoo bug #171348: dialog fails parallel install
To: dickey@invisible-island.net
[-- PGP output follows (current time: Sun Mar 25 18:22:13 2007) --]
gpg: Signature made Mon Mar 19 02:20:44 2007 EDT using DSA key ID 2C28B87C
gpg: Can't check signature: public key not found
[-- End of PGP output --]
[-- The following data is signed --]
Hi,
When installing dialog with 'make -j# install', it's possible to get an
error when several processes try to create the same directory at the
same time. This can happen because mkdir fails when a directory already
exists, and the directory can be created between the test ! -d and the
mkdir in mkdirs.sh. This can easily be solved by changing mkdir to mkdir
-p, which silently returns if the directory already exists. If there are
systems you support that do not provide the -p option to mkdir, an
alternative approach is to ignore mkdir's return value, and run test -d
afterwards. A third option is to say you don't support parallel
installs, but I'm hoping you don't take that approach. :)
Thanks,
Harald .
-------TD: I did the alternative approach, since I'm unsure about mkdir -p.
* AcSplit/CF_WITH_LIBTOOL:
duh - I broke the assignment to LIB_OBJECT in my ncurses fix, did not
notice til rebuilding dialog
* AcSplit/CF_MBSTATE_T:
include stdio.h as well since OSF/1's wchar.h uses __VA_LIST__ (from
/usr/include.dtk/wchar.h), but that symbol is defined in stdio.h
rather than stdarg.h !
2007-03-24
* AcSplit/CF_LIB_RULES:
add TermlibSuffix symbol to invocation of mk-1st.awk
2007-03-21
* AcSplit/CF_X_FREETYPE:
move that last chunk inside a relevant if-statement
* AcSplit/CF_X_FREETYPE:
add check for "newer" Xft functions, including those needed for clipping
in xterm.
2007-03-17
* AcSplit/CF_WITH_LIBTOOL: Date: Sun, 11 Mar 2007 21:33:48 +0300 (MSK)
From: Michail Vidiassov <master@iaas.msu.ru>
Subject: Mac OS X - libtool problems
To:
cc: Ncurses Mailing List <bug-ncurses@gnu.org>, sci-fi@hush.ai
Dear Thomas,
I do not compile ncurses via libtool, but just to have things fixed.
Apple provides libtool that is not command-line compatible with GNU libtool.
Apple provides GNU libtool as glibtool (/usr/bin/glibtool).
Package managment system Macports provides GNU libtool as
glibtool (/opt/local/bin/glibtool)
Its competitor Fink installs GNU libtool as /sw/bin/glibtool or
/sw/bin/libtool, I am not sure.
Ergo it looks like a good idea to look for glibtool, if search
for libtool fails, or libtool turns out not to be GNU libtool.
I am not sure.
May be the user MUST manually point configure to the correct libtool
using command-line option if anything is non-standart?
Sincerely, Michail
--------also:
libtool --version
libtool: unknown option character `-' in: --version
Usage: libtool -static [-] file [...] [-filelist listfile[,dirname]]
[-arch_only arch] [-sacLT]
Usage: libtool -dynamic [-] file [...] [-filelist listfile[,dirname]]
[-arch_only arch] [-o output] [-install_name name] [-compatibility_version
#] [-current_version #] [-seg1addr 0x#] [-segs_read_only_addr 0x#]
[-segs_read_write_addr 0x#] [-seg_addr_table <filename>]
[-seg_addr_table_filename <file_system_path>] [-all_load] [-noall_load]
libtool -V
Apple Computer, Inc. version cctools-622.5
Sincerely, Michail
* AcSplit/CF_MAN_PAGES:
change sed-script for @DATADIR@, etc., to do all instances per line.
2007-03-14
* AcSplit/CF_TTY_GROUP, AcSplit/CF_POSIX_SAVED_IDS,
AcSplit/CF_FUNC_TGETENT:
use $cf_cv_main_return
2007-03-13
* AcSplit/CF_UTMP_UT_HOST, AcSplit/CF_UTMP_UT_NAME,
AcSplit/CF_UTMP_UT_SESSION, AcSplit/CF_UTMP_UT_XTIME:
improved message to show utmp or utmpx as appropriate
2007-03-12
* AcSplit/CF_PROCFS_CWD:
tested on FreeBSD 6.0, got this to work there...
2007-03-11
* AcSplit/CF_PROCFS_CWD: ci-aclocal generated header
2007-02-25
* headers-sh: RCS_BASE
* AcSplit/CF_WITH_LIBTOOL:
correct typo (noticed when building dialog - didn't see this in ncurses?)
2007-02-24
* AcSplit/CF_SHARED_OPTS:
fixes to make cygdll to work with "cfg-shared --with-ticlib" (also worked
for just "cfg-shared" - did not try --with-termlib).
* AcSplit/CF_SHARED_OPTS:
mk_shared_lib.sh script will live in the top-level build-directoryu
* AcSplit/CF_LIB_RULES:
cygdll case doesn't use SHARED_LIB variable anymore
* AcSplit/CF_LIB_RULES:
use TINFO_ARG_SUFFIX for TermlibRoot, since that's closer to what I need
for termlib
* AcSplit/CF_MAN_PAGES:
reduce error-messages on uninstall by checking if the target exists
2007-02-17
* AcSplit/CF_SHARED_OPTS:
for cygwin, build a script (mk_shared_lib.sh) and use that to construct
the two output files.
* AcSplit/CF_LIB_RULES:
start changing to fix cygwin wrt ncurses termlib
that means I drop the $(IMPORT_LIB) symbol, and just build it as a side-effect
of building the SHARED_LIB
* AcSplit/CF_LIB_RULES: pass Relink to mk-1st.awk
* AcSplit/CF_LIB_RULES: add TermlibRoot to mk-1st.awk call
* AcSplit/CF_REMOVE_LIB: ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS: Date: Wed, 14 Feb 2007 07:37:40 +0300 (MSK)
From: Michail Vidiassov <master@iaas.msu.ru>
Subject: ncurses installation on Mac OS
To:
cc: Ncurses Mailing List <bug-ncurses@gnu.org>
Dear Thomas,
since you applied the following change
Rather than copying the build-tree's libraries, link the shared
objects into the install directory.
a problem surfaced on Mac OS X (whith --disable-libtool, --enable-shared),
installed libraries are linked against libraries in build locations,
not instal locations (as described at
http://lists.macosforge.org/pipermail/macports-users/2007-January/001342.html)
I suspect that DESTDIR in the following configure line breaks things.
MK_SHARED_LIB='${CC} ${CFLAGS} -dynamiclib -install_name
${DESTDIR}${libdir}/`basename $@` -compatibility_version ${ABI_VERSION}
-current_version ${ABI_VERSION} -o $@'
Since DESTDIR sets temporary location where things are installed before
being moved to a proper place by package management system, there is
no reason to include DESTDIR in their install_names (Mac OS analog of
rpath).
Sincerely, Michail
------TD: the relinking is going to be a problem (address that with an
option), while the ${DESTDIR} dates to Jason Evan's initial mods for
Darwin in 2001. I agree that it doesn't appear to be needed/wanted - removed.
2007-01-29
* AcSplit/CF_XOPEN_SOURCE: Debian Bug report logs - Debian #404111
xterm: FTBFS on GNU/kFreeBSD
From: Petr Salinger <Petr.Salinger@seznam.cz>
To: submit@bugs.debian.org
Subject: xterm: FTBFS on GNU/kFreeBSD
Date: Thu, 21 Dec 2006 20:32:18 +0100 (CET)
----TD: this is the same change as for CF_SHARED_OPTS
2007-01-28
* AcSplit/CF_LIB_RULES: correct typo that made
cfg-shared --with-termlib
not work (reported by Rong-En Fan)
2007-01-27
* AcSplit/CF_WITH_LIBTOOL:
oops - I constructed a sed-command that would delete the whole thing if
the first line was blank.
2007-01-13
* AcSplit/CF_LIB_RULES:
split out the loops for install/uninstall of header files as mk-1st.awk,
so I can configure whether to install tic's header files.
* AcSplit/CF_LIB_RULES:
extended this to handle ncurses' new --with-ticlib option
2006-12-30
* AcSplit/CF_WITH_LIBTOOL: typo
* AcSplit/CF_WITH_LIBTOOL:
some recent version of libtool adds a blank line in the version-message:
<<<<
Okay *now* I remember something about this version of libtool,
I think they changed the --version syntax just a bit:
>>>>
# /usr/local/bin/glibtool --version
ltmain.sh (GNU libtool) 1.5.23a (1.1220.2.416 2006/10/24 20:18:28)
Copyright (C) 2006 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
# _
<<<<
(report by "sci-fi@hush.ai" on bug-ncurses)
2006-12-24
* AcSplit/CF_MAN_PAGES:
rather than executing the big inline-sed stuff, generate a new script
man_alias.sed, and use that also when filtering the list of aliases.
That's needed since I changed "all" of the program references to use
the transformed names, e.g,. tic -> ntic.
2006-12-23
* AcSplit/CF_LIB_RULES:
make misc/Makefile in ncurses handle install.libs - use that for installing
the ncurses5-config script
* AcSplit/CF_SHARED_OPTS:
make openbsd2.* and up use the "newer" rule for shared libraries.
make freebsd3.1 use rpath - and added a "." to ensure I don't get caught by
FreeBSD 10.x
2006-12-22
* config.guess: 2006-12-22
2006-12-17
* AcSplit/CF_WITH_GPM: quiet the messages a little
2006-12-16
* AcSplit/CF_WITH_GPM: needed a non-empty field to make this work
* AcSplit/CF_WITH_GPM: closer - still working on providing a _value_
* AcSplit/CF_WITH_GPM:
oops - forgot to actually allow the option value to work.
* AcSplit/CF_WITH_GPM:
minor tidying up to allow --with-gpm to have a useful value
* AcSplit/CF_LIB_SONAME, AcSplit/CF_PROG_CC_C_O: use $cf_cv_main_return
* AcSplit/CF_DISABLE_LEAKS: don't add purify to ncurses
* AcSplit/CF_WITH_DBMALLOC, AcSplit/CF_WITH_DMALLOC: put back the test:
* AcSplit/CF_NO_LEAKS_OPTION:
quote the help-text, otherwise blanks are removed
* AcSplit/CF_NO_LEAKS_OPTION: better kludge for adding -g
* AcSplit/CF_DISABLE_LEAKS:
moved the #define for NO_LEAKS outside the option, so the contributing
options can provide a default value for it.
* AcSplit/CF_DISABLE_LEAKS: ncurses has another way to set -g option
* AcSplit/CF_BOOL_SIZE, AcSplit/CF_FUNC_MEMMOVE,
AcSplit/CF_FUNC_NANOSLEEP, AcSplit/CF_FUNC_POLL,
AcSplit/CF_HASHED_DB_LIBS, AcSplit/CF_LINK_DATAONLY,
AcSplit/CF_LINK_FUNCS, AcSplit/CF_MKSTEMP, AcSplit/CF_TYPEOF_CHTYPE:
typo
* AcSplit/CF_DISABLE_LEAKS: ci-aclocal generated header
2006-12-14
* AcSplit/CF_NO_LEAKS_OPTION: add a 4th param for CF_WITH_PURIFY
* AcSplit/CF_WITH_PURIFY: use CF_NO_LEAKS
* AcSplit/CF_WITH_DBMALLOC, AcSplit/CF_WITH_DMALLOC:
use CF_NO_LEAKS_OPTION
* AcSplit/CF_NO_LEAKS_OPTION, AcSplit/CF_WITH_NO_LEAKS,
AcSplit/CF_WITH_PURIFY, AcSplit/CF_WITH_VALGRIND:
ci-aclocal generated header
2006-12-10
* AcSplit/CF_MAIN_RETURN: ci-aclocal generated header
* AcSplit/CF_BOOL_SIZE, AcSplit/CF_FUNC_MEMMOVE,
AcSplit/CF_FUNC_NANOSLEEP, AcSplit/CF_FUNC_POLL,
AcSplit/CF_HASHED_DB_LIBS, AcSplit/CF_LINK_DATAONLY,
AcSplit/CF_LINK_FUNCS, AcSplit/CF_MKSTEMP, AcSplit/CF_TYPEOF_CHTYPE:
use $cf_cv_main_return
2006-12-09
* AcSplit/CF_SHARED_OPTS:
use ${CC} rather than ${LD} in two cases where "cc" passes the options
properly to "ld": IRIX64 and Solaris.
For Solaris, this is motivated by this reply from Casper Dik:
(on comp.unix.solaris)
Date: 04 Dec 2006 14:04:44 GMT
From: Casper H.S. Dik <Casper.Dik@Sun.COM>
Subject: Re: Help needed for migrating HPUX tool to Solaris
>Most people (well, perhaps not on this newsgroup) are aware that it
>doesn't work for the first case (not a shared library), because one must
>include the c runtime. Your comment does not provide any useful
>information on what "runtime glue" is needed to make a shared library
>which is lacking by using ld directly(*).
a.out and lib.so use the same glue but a.out adds one extra bit,
the call to main. The particular object files in question are
crti.o and crtn.o, they're used to bracket all .o files like this:
crti.o .... *.o .... crtn.o
crti.o contains an .init section with the _init function preamble
and a .fini section with the _fini function preamble
crtn.o contains an .init section with a function epilogue
and a .fini with the same.
It generally works without it except when you:
- use #pragma init
- use #pragma fini
- use __attribute__((constructor))
- use __attribute__((destructor))
and C++ code which needs static initalizers.
>(*) I've used it both ways, but generally directly. It's nice to hear
> from the experts that programs that I use every day - including
> the copy of tin that I'm using for posting - do not work.
It mostly works, but I remember one particular occasion when someone
change libc.so to contain a different crypt() an dhe then proceeded
to relink the library with "ld". The resulting library did not
call fflush() on exit causing no end to trouble.
Casper
--------- reviewing, I see Linux already did use "${CC}".
Checking, I see IRIX64 can do this also, though the sections do not add
and .init/.fini -- rather, it adds a ".conflict" section:
[18] | 4205076| 8|SECT |LOCL |DEFAULT |MIPS_DATA|.conflict
also some weak symbols
[32] | 4205092| 0|FUNC |WEAK |DEFAULT |UNDEF |attr_get
[33] | 4205092| 0|FUNC |WEAK |DEFAULT |UNDEF |attr_set
Finally - I see that nm's listing for libc symbols contains offset
values rather than 0's. I think that is an improvement.
* AcSplit/CF_MAN_PAGES:
rename "$target" to "$cf_target" and "$source" to "$cf_source".
Inspecting SuSE's ncurses package, I saw they (unknown changelog entry)
changed "$target" to "$t" (which is not good for grep's). I can see
the apparent reason is that someone thought it conflicted with "$target"
in other places in the script. But it doesn't (didn't), since this
chunk uses it only in a generated script. Renamed it anyway to avoid
seeing junk in the rpm package...
2006-12-08
* config.guess, config.sub: 2006-12-08
2006-11-30
* AcSplit/CF_X_ATHENA_LIBS: sigh...
* AcSplit/CF_X_ATHENA_LIBS: oops - missed that I had to reset $LIBS
(that didn't build on HPUX 10.20, probably because of command-length)
2006-11-29
* AcSplit/CF_X_ATHENA_LIBS:
CF_X_TOOLKIT adds $X_PRE_LIBS and $X_EXTRA_LIBS to $LIBS, so I don't want to
repeat that here.
* AcSplit/CF_X_TOOLKIT:
add $X_EXTRA_LIBS when completing the library list.
(this isn't the autoconf assumption, but does simplify building up other tests).
* AcSplit/CF_X_FREETYPE:
rewrote so I can nominally support pre-package versions of FreeType.
This worked for Debian 2.2 (may work for Redhat 7.2 - haven't tested if
I also need -lrender).
2006-11-18
* AcSplit/CF_SUBDIR_PATH:
quote $HOME as "$HOME" (prompted by report on lynx-dev of problems with cygwin by
"Veley" (Vepa Velmu?) (Vebazu?)
2006-11-12
* AcSplit/CF_PDCURSES_X11:
use xcurses-config if we have it, to fix rpath issues on *BSD
2006-10-29
* AcSplit/CF_CURSES_CONFIG: comment
2006-10-28
* AcSplit/CF_TERMCAP_LIBS:
move the calls to CF_NCURSES_CPPFLAGS and CF_NCURSES_LIBS into CF_NCURSES_CONFIG
* AcSplit/CF_NCURSES_CONFIG: fix one error
* AcSplit/CF_NCURSES_CONFIG:
first cut (not correct) using the ncurses5-config (or aliases)
* AcSplit/CF_NCURSES_CONFIG: corrected the use of ifelse()
* AcSplit/CF_IMAKE_CFLAGS:
use curly-braces ${symbol} for better reuse in shell-scripts
* AcSplit/CF_CURSES_CONFIG, AcSplit/CF_NCURSES_CONFIG:
ci-aclocal generated header
2006-10-21
* AcSplit/CF_SHARED_OPTS:
add CF_SHARED_SONAME for several places where I had the appropriate
basename construct
* AcSplit/CF_SHARED_SONAME:
ensure we have a useful default value for $cf_cv_shlib_version
* AcSplit/CF_SHARED_SONAME: ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS: make shared abi/rel work for netbsd 2.0
* AcSplit/CF_SHARED_OPTS:
reverse the cases for FreeBSD so I can easily make rpath work up to version 19.
* AcSplit/CF_MAKEFLAGS:
NetBSD 2.0 spits out a trailing blank on the testcase I made for the -k
option. That made it not handle the case ".*k)". I tried adding a
pattern for blanks after - but NetBSD's shell couldn't handle _that_
either. Dumbed it down by trimming the trailing blank(s) with a sed
script.
2006-10-15
* AcSplit/CF_TOP_BUILDDIR: ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS:
factor out symbol $(LOCAL_LIBDIR) so I can make an install-rule for the
test programs in ncurses that will work with the -rpath option.
2006-10-14
* AcSplit/CF_ADA_INCLUDE_DIRS, AcSplit/CF_BUILD_CC,
AcSplit/CF_GNAT_VERSION, AcSplit/CF_INCLUDE_DIRS,
AcSplit/CF_LIB_RULES, AcSplit/CF_MAKEFLAGS, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_WITH_LIBTOOL:
use curly-braces in makefile symbols so they can be reused in shell scripts
2006-10-07
* AcSplit/CF_SHARED_OPTS:
widened the FreeBSD check to version 6 (should invert the case to cover
versions 2,3). Fix for cygwin as well, to cover changes in awk files.
* AcSplit/CF_SHARED_OPTS:
another bit of breakage from the 2000 rewrite of -rpath
* AcSplit/CF_LIB_RULES:
move the directory-name out of the SHARED_LIB and IMPORT_LIB variables
to allow reuse in the install-rule
2006-09-23
* AcSplit/CF_LARGEFILE: Date: Thu, 21 Sep 2006 19:26:42 +0200
From: ThMO <thmo-13@gmx.de>
Subject: ncurses v5.5 - some problems
To: bug-ncurses@gnu.org
X-Mailer: Mozilla 4.08 [en] (X11; U; Linux 2.0.35 i686)
[-- Attachment #1 --]
[-- Type: text/plain, Encoding: 8bit, Size: 1.4K --]
Hello,
I've decided to upgrade ncurses from v4.2 -> v5.5 and there were
two problems:
? ./configure --disable-largefile
the symbol _LARGE_FILES gets defined, which shouldn't with the
above option
--------TD: oops the enable option is tested in one part of the macro only.
2006-09-16
* AcSplit/CF_PROG_AWK, AcSplit/CF_PROG_EGREP, AcSplit/CF_PROG_LINT:
ci-aclocal generated header
* AcSplit/CF_SUBST_NCURSES_VERSION: use CF_PROG_EGREP
2006-09-02
* AcSplit/CF_PATH_SYNTAX:
lynx (PG) noted that the result from this ignored the --prefix option.
fixed...
2006-08-20
* AcSplit/CF_WITH_LOCALE_ALIAS: typo
* AcSplit/CF_WITH_LOCALE_ALIAS: ci-aclocal generated header
* AcSplit/CF_SYSV: typo
(in comment)
* AcSplit/CF_PKG_CONFIG, AcSplit/CF_X_FONTENC:
ci-aclocal generated header
2006-08-19
* AcSplit/CF_SHARED_OPTS:
Redhat #199369 wanted to add $(CFLAGS) for sparc/64-bit; I noted that the
same issue applied to a few other places.
* AcSplit/CF_HASHED_DB_LIBS:
make this work for db4 (tested with db 4.2, 4.3)
* AcSplit/CF_HASHED_DB_LIBS: add a check for db3
* AcSplit/CF_HASHED_DB_VERSION: minor tidying
* AcSplit/CF_HASHED_DB_LIBS:
don't use the reference to O_CREAT|O_RDWR in db 1.85 test since application
should test for fcntl.h separately.
* AcSplit/CF_HASHED_DB, AcSplit/CF_HASHED_DB_LIBS,
AcSplit/CF_HASHED_DB_VERSION:
ci-aclocal generated header
2006-08-14
* config.sub: 2006-08-14
2006-08-06
* AcSplit/AM_PATH_PROG_WITH_TEST:
Date: Sun, 23 Jul 2006 22:09:51 -0700 (PDT)
From: Doug Kaufman <dkaufman@rahul.net>
Subject: [Lynx-dev] DJGPP patch for 2.8.6dev.18
To: lynx-dev@nongnu.org
This is a minor patch for two items that affect portability to DJGPP.
The first is a change to aclocal.m4 from "test -f" to "test -x". This
is important because if testing for "xgettext" and a file exists
called "xgettext.exe", "test -f" fails and "test -x" succeeds. Since
the line in aclocal.m4 is testing for an executable file, the change
shouldn't affect other platforms adversely. Without the patch, the
lynx configuration script doesn't find the utilities needed for NLS.
The change will be propagated to "configure" when autoconf is run.
------TD: I pointed out that "test -f" wasn't portable, and DK resubmitted
the change modifying the test to include the executable suffix, which
should work.
2006-08-05
* AcSplit/CF_MAKEFLAGS:
check for the (obvious) case where $(MAKEFLAGS) evaluates to nothing,
suppress the warning. I saw this with FreeBSD 6.0 "make"
2006-08-03
* AcSplit/CF_WITH_PATH:
xterm: change to not use "eval" since it breaks for the appsdir substitution
on AIX 5.x by trying to evaluate the "$(exec_prefix)" as "$exec_prefix, or
something like that.
2006-08-02
* AcSplit/CF_POSIX_SAVED_IDS:
HPUX doesn't have seteuid/geteuid, and its linker only warns - does not
error - when trying to link to those. Do a quick fix by trying to assign
the address to a pointer.
2006-07-22
* config.guess, config.sub: 2006-7-2
2006-06-21
* AcSplit/CF_POSIX_SAVED_IDS:
extend thiss to include the quasi-POSIX BSD's
* AcSplit/CF_SYSV: Date: Wed, 21 Jun 2006 12:21:34 +0200
From: Emanuele Giaquinta <e.giaquinta@glauco.it>
To:
Subject: Re: xterm utmp
Seems the termios change has a typo:
static STRUCT_TERMIO d_tio;
should be
static STRUCT_TERMIOS d_tio;
2006-06-20
* AcSplit/CF_SYSV:
xterm - on HPUX 10.20 with gcc, I wasn't able to build it because that
combination relies on including <termios.h> before <term.h> (though it
builds fine with cc). Will have to rebuild all platforms...
2006-06-18
* AcSplit/CF_POSIX_SAVED_IDS:
renamed the defined variable for consistency
* AcSplit/CF_POSIX_SAVED_IDS:
add a check for sysconf(), needed (probably everywhere except Linux)
* AcSplit/CF_POSIX_SAVED_IDS: add a comment
* AcSplit/CF_POSIX_SAVED_IDS: ci-aclocal generated header
2006-06-17
* AcSplit/CF_SUBST_IF:
I needed success/fail values for the HAVE_xxx usage.
Move the cache-variable (cf_cv_subst_XXX) back to CF_SUBST
* AcSplit/CF_SUBST:
moved the assignment to the cache variable back from CF_SUBST_IF
* AcSplit/CF_SUBST_IF: ci-aclocal generated header
* AcSplit/CF_SUBST: split-out CF_SUBST_IF
2006-06-03
* AcSplit/CF_PRG_RULES: ci-aclocal generated header
* AcSplit/CF_LIB_RULES:
add a few checks to allow directories to be optional, also some blank lines
for readability.
2006-05-29
* AcSplit/AM_PATH_PROG_WITH_TEST:
Date: Fri, 27 Jan 2006 01:22:54 -0800 (PST)
From: Doug Kaufman <dkaufman@rahul.net>
Subject: [Lynx-dev] DOS / PDCurses patch for dev.17
To: lynx-dev@nongnu.org
-------TD: comment said:
One of the macros in aclocal.m4 isn't portable, and so
didn't properly detect the DJGPP gettext executables. I think this
patch makes the macro more portable.
------- which in this case means that the script didn't expect a DOS drive
letter, and it didn't override $PATHSEP with $PATH_SEPARATOR.
2006-05-13
* AcSplit/CF_MAN_PAGES:
add ncurses major/minor/patch variables so I can expand them inside manpages.
(It's hard to see which one I have).
* config.sub: 2006-05-13
http://subversions.gnu.org/cgi-bin/viewcvs/config/config/
* config.guess: 2006-04-26
http://subversions.gnu.org/cgi-bin/viewcvs/config/config/
2006-04-17
* AcSplit/AM_MULTIBYTE_ABLE:
tin - Urs commented out the check for wscnlen()
2006-04-15
* AcSplit/CF_LIB_RULES:
change some names to make it a little simpler to spot spurious CF_xxx names.
2006-04-02
* AcSplit/CF_SIGWINCH: ci-aclocal generated header
* AcSplit/CF_XOPEN_SOURCE: comments
2006-03-12
* AcSplit/CF_STRUCT_LASTLOG: ci-aclocal generated header
2006-02-12
* AcSplit/CF_ENABLE_NARROWPROTO: it's "X.org", not "Xorg"
* AcSplit/CF_X_FREETYPE:
oops - noticed in CF_WITH_PCRE that $withval isn't set by the default of
a --with-OPTION. q/d fix by setting it to an empty value before the
existing AC_ARG_WITH's
* AcSplit/CF_WITH_PCRE: don't use $withval outside of the AC_ARG_WITH
2006-02-05
* AcSplit/CF_WITH_PCRE: fix the help-message
* AcSplit/CF_WITH_PCRE: ci-aclocal generated header
2006-02-04
* AcSplit/CF_CURSES_CPPFLAGS, AcSplit/CF_CURSES_LIBS:
don't use curses_colr on HPUX 11
2006-01-23
* AcSplit/CF_TTY_GROUP:
add --with-tty-group option, allow that to override environment
2006-01-22
* AcSplit/AM_WITH_NLS:
FreeBSD's ports system puts "everything" except for the base system in
/usr/local (actually not - there are some inconsistencies). Make the
configure script work in spite of that.
* AcSplit/CF_FIND_HEADER, AcSplit/CF_FIND_LINKAGE:
ci-aclocal generated header
2006-01-08
* AcSplit/CF_ENABLE_NARROWPROTO:
fix a typo - missing "cf_" from a variable name
2006-01-07
* AcSplit/CF_X_FREETYPE: Date: Fri, 6 Jan 2006 17:29:48 +0100
From: Emanuele Giaquinta <e.giaquinta@glauco.it>
Subject: xterm freetype fix
To:
[-- Attachment #1 --]
[-- Type: text/plain, Encoding: 7bit, Size: 0.3K --]
Hi Thomas,
Your last change to CF_X_FREETYPE breaks xterm on darwin; the problem
is that it is not enough to check that pkg-config exists, we must also
check that xft.pc exists too, which is not the case on darwin where we
must so revert to xft-config. The attached patch fixes it.
------TD looks ok, will test
2006-01-02
* AcSplit/CF_PATH_SYNTAX:
check for "no", which may be the value if "--without-XXX" was used.
2005-12-31
* AcSplit/CF_NCURSES_HEADER, AcSplit/CF_CURSES_HEADER,
AcSplit/CF_TERM_HEADER:
ci-aclocal generated header
* AcSplit/CF_NCURSES_CPPFLAGS:
split out CF_NCURSES_HEADER to simplify comparison with CF_CURSES_HEADER
use CF_TERM_HEADER
* AcSplit/CF_CURSES_CPPFLAGS:
split-out CF_CURSES_HEADER to simplify comparison with CF_NCURSES_HEADER
2005-12-26
* AcSplit/CF_WITH_NC_ALLOC_H:
CF_CURSES_FUNCS couldn't find _nc_free_and_exit (should revisit).
I worked around that by moving the check for _nc_free_and_exit here,
to use AC_CHECK_FUNCS
* AcSplit/CF_WITH_NC_ALLOC_H: ci-aclocal generated header
2005-11-24
* AcSplit/CF_X_FREETYPE: add parameter for pkg-config
* AcSplit/CF_X_FREETYPE:
add a configure check for pkg-config (the latest variation of FreeType
configuration wrappers). Fixes Redhat #173541. Jason Vas Dias forwarded
a patch by Nalin Dahyabhai but it was incorrect, so I did this one myself.
2005-11-19
* AcSplit/CF_SHARED_OPTS: Date: Mon, 07 Nov 2005 12:02:53 +0100
From: Andrea -XFox- Govoni <xfox@mac.com>
Subject: Re: Compile on MacOS
To: Christian Gennerat <cgennerat@cna-cefag.org>
Cc: bug-ncurses@gnu.org
--------TD
Andrea reported that there's an option for ld which changes the search
rules (see email). This fixes (sort of) ld's behavior which searches
for any shared library anywhere before looking at the -L options and
noticing that there's a static library in the -L directory. I adapted
this from a chunk that google showed me in
http://cvs.sourceforge.net/viewcvs.py/tcl/tclconfig/tcl.m4?rev=1.74
(which is interesting, since it contains code like this macro).
2005-11-13
* config.sub: from tin, 2005-11-13
2005-11-12
* config.guess: from tin 2005-11-11
2005-11-02
* AcSplit/CF_ENABLE_NARROWPROTO: ci-aclocal generated header
* AcSplit/CF_WITH_IMAKE_CFLAGS: split-out CF_ENABLE_NARROWPROTO
2005-10-15
* AcSplit/CF_XOPEN_SOURCE:
remove the special case for Darwin. Emanuele Giaquinta asserts that the
problems with this platform can be addressed (lynx and xterm).
2005-10-06
* AcSplit/CF_UTMP_GROUP: ci-aclocal generated header
2005-09-29
* AcSplit/CF_DEFAULT_SHELL: from tin: add mksh before jsh
,
2005-09-24
* AcSplit/CF_BUILD_CC:
Dan Jacobiwitz pointed out that I'd defined a fallback for $BUILD_CC that
overrode the AC_CHECK_TOOL() usage in --with-build-cc=XX (fixed).
I also made the macro more verbose.
* AcSplit/CF_CFG_DEFAULTS:
remove netbsd from the list of systems that install by default in /usr
* config.guess: 2005-9-19
* config.sub: 2005-7-8
2005-09-18
* AcSplit/CF_SIG_ATOMIC_T:
make this a little more stringent by ensuring it's for an extern type.
* AcSplit/CF_SYSV:
ensure we've got X11/Intrinsic.h before trying to include it.
* AcSplit/CF_FUNC_TGETENT:
I noticed in one of Nelson Beebe's logs that it was complaining about
a "buffer0" value which wasn't defined. That was improperly-quoted
text.
2005-09-17
* AcSplit/CF_NCURSES_ABI_6: ci-aclocal generated header
2005-09-05
* AcSplit/CF_SIG_ATOMIC_T: ci-aclocal generated header
* AcSplit/CF_LEX_CHAR_CLASSES:
oops - HPUX reported that DATE wasn't defined, but didn't complain about
IDENT. By reversing the two, I could see the other broke. Finally saw
that I had an unnecessary "%%" in the first line of the test case which
flex ignored, but HPUX lex did not.
* AcSplit/CF_LEX_CHAR_CLASSES:
dispense with the states (want to keep the test independent).
Tweaked the pattern for DATE so the generated brackets work - though I
don't see why it's counted different from IDENT. Verified that it does
work for 2.13, 2.52 and 2.59 (gives the same result).
* AcSplit/CF_LEX_CHAR_CLASSES:
ensure that the pattern is used in a rule, just in case lex doesn't bother
to look at it til then. Also, add [:blank:], since that seems to be a
problem area. Finally, fix the quotes - I thought I had to double each
square bracket, but it turns out not.
2005-09-04
* AcSplit/CF_LEX_CHAR_CLASSES: ci-aclocal generated header
2005-08-27
* AcSplit/CF_GCC_VERSION, AcSplit/CF_GXX_VERSION:
fix to work with output from cross-compilers that emit a platform name, e.g.,
i586-pc-msdosdjgpp-c++ (GCC) 3.3
The "586" would be interpreted as a version number (ncurses report by
Frank van Vugt <ftm.van.vugt@foxi.nl>, regarding a different compiler
arm-sa1100-linux-gnu-g++ (GCC) 4.0.1").
2005-08-24
* AcSplit/CF_TAR_OPTIONS:
Doug Kaufman (lynx) added $EXEEXT to the paths used for tar, etc.
But I'm pretty sure that on cygwin I'll not be getting the extension
from CF_PATH_PROG, so I duplicated the pattern, so it has with/without.
2005-08-21
* AcSplit/CF_POPEN_TEST, AcSplit/CF_YACC_ERROR:
we don't use ARGS() macro anymore
2005-08-13
* AcSplit/CF_GXX_WARNINGS: don't look for -Wno-undef for Intel compiler
2005-08-06
* AcSplit/CF_INTEL_COMPILER: show which type of compiler we're checking
* AcSplit/CF_GCC_WARNINGS:
use explicit-params call to CF_INTEL_COMPILER - I think that autoconf was
confused by the two different calls and was only doing one until I did this.
* AcSplit/CF_XOPEN_SOURCE:
tweak to make ncurses build with QNX - to get strdup, etc.
2005-07-25
* AcSplit/CF_GXX_VERSION:
make the pattern the same as for CF_GCC_VERSION
* AcSplit/CF_GXX_WARNINGS: add compiler-check for Intel-compiler
* AcSplit/CF_INTEL_COMPILER:
parameterized better to share between CF_GCC_WARNINGS and CF_GXX_WARNINGS
2005-07-23
* AcSplit/CF_GXX_WARNINGS:
suppress -Weffc++ warning in g++ 2.x (header files just aren't good enough)
* AcSplit/CF_CPP_STATIC_CAST: ci-aclocal generated header
2005-07-17
* AcSplit/CF_LARGEFILE:
oops - that is "struct dirent64", not "dirent64".
2005-07-14
* AcSplit/CF_POSIX_C_SOURCE: fix quoting (tested on cygwin)
* AcSplit/CF_XOPEN_SOURCE: use CF_PROG_CC_U_D
* AcSplit/CF_PROG_CC_U_D: ci-aclocal generated header
* AcSplit/CF_POSIX_C_SOURCE: use CF_CC_U_D_OPTIONS
2005-07-12
* AcSplit/CF_LARGEFILE:
hedge that check for dirent64 - it comes in Solaris header only, and seems
to be for a compiler model that I cannot build.
2005-07-09
* AcSplit/CF_XOPEN_SOURCE:
If we have to define _XOPEN_SOURCE, don't try to cancel a previous -D option
with a -U: Urs Jansen reported that the Intel compiler and Cray compilers
don't process -U/-D options left-right right, but apparently all -D's and
then all -U's.
* AcSplit/CF_POSIX_C_SOURCE:
Urs Jansen reported that the Intel compiler and the Cray compiler both do
not process -U and -D options left-to-right. That is
cc -Dxx -Uxx -Dxx=5
would result in a program that does not define xx. So I modified this to
not use the -U to cancel, but to strip the conflicting -D from $CFLAGS and
$CPPFLAGS
* AcSplit/CF_REMOVE_DEFINE:
add the pattern for removing the final option
make this more readable
* AcSplit/CF_REMOVE_DEFINE: ci-aclocal generated header
* AcSplit/CF_GNU_SOURCE:
don't need an AC_REQUIRE for CF_INTEL_COMPILER anymore
* AcSplit/CF_GCC_WARNINGS:
updated to incorporate options for the Intel 9.0 compiler running on 64-bit
platform (some additional nuisances regarding loss of precision).
2005-07-02
* AcSplit/CF_FUNC_NANOSLEEP: ci-aclocal generated header
2005-06-29
* AcSplit/CF_FUNC_ALLOCA: ci-aclocal generated header
* AcSplit/CF_GNU_SOURCE:
tin - long discussion with Urs regarding icc 9.0, decided to revert the
change that excluded Intel compiler from using -D_GNU_SOURCE, since it
seems icc is moving closer to gcc and it's a nuisance to try to duplicate
the ifdef's that _GNU_SOURCE sets. Also for some such as vasprintf(),
that has to be defined anyway.
2005-06-25
* AcSplit/CF_TYPEOF_CHTYPE:
split-out the code that invokes CF_UNSIGNED_LITERALS, move that to
configure.in (ncurses) so I can make --with-chtype=XXX option set the
$cf_cv_1UL option properly
* AcSplit/CF_LIB_SONAME:
I don't want the path - just the soname, since that's what dlsym() needs
* AcSplit/CF_LIB_SONAME, AcSplit/CF_WITH_GPM:
ci-aclocal generated header
* config.sub, config.guess:
from http://subversions.gnu.org/cgi-bin/viewcvs/config/config/
2005-06-21
* AcSplit/CF_BUNDLED_INTL:
reviewing changes for tin, I noticed that last change wasn't correct (it
didn't quote the parameter that was being edited, and the editing command
had two typos).
2005-06-20
* config.guess: 2005-05-27 (from tin)
* config.sub: 2005-06-02 (from tin)
* AcSplit/AM_ICONV: from tin: split the help message to two lines
2005-06-18
* AcSplit/CF_MAN_PAGES: typo
* AcSplit/CF_MANPAGE_RENAMES:
move edit_man.sed into top-level directory
* AcSplit/CF_MAN_PAGES:
moved the manlinks.sed and edit_man.* stuff into top directory to simplify
reuse for tack
2005-06-11
* AcSplit/CF_TYPEOF_CHTYPE:
simplified this - we need 31 bits for ncurses' chtype.
Dan Jacobowitz (Debian maintainer) pointed out that the 'x' variable was
not initialized, making the test always say "long". It turns out that the
test has been broken that way since I first implemented it in 1997.
Will probably rewrite the test, but decided first to repair it to make it
do what I intended. (Also remove the ifdef for WANT_BITS - I don't need
39 bits in any configuration).
2005-05-28
* AcSplit/CF_LARGEFILE:
add a workaround for ncurses' c++ headers, put the large-file definitions
into the Makefile
* AcSplit/CF_GCC_ATTRIBUTES:
change CF_ATTRIBUTE to cf_ATTRIBUTE to avoid nuisance of finding CF_xx in
configure script
* AcSplit/CF_LIB_RULES:
rename CF_ITEM to cf_ITEM to avoid nuisance of finding CF_xxx in configure script
* AcSplit/CF_SRC_MODULES: add TEST_ARG2 and TEST_DEP2, for Ada95 tree
2005-05-27
* AcSplit/CF_LARGEFILE:
add a check for struct dirent64, needed to appease icc on Linux.
tested with gcc on Linux also (at least it's not warning).
2005-05-22
* AcSplit/CF_LARGEFILE: ci-aclocal generated header
2005-05-15
* AcSplit/CF_XOPEN_SOURCE: trim whitespace from the last change
2005-04-16
* config.sub: 2005/2/10
* config.guess: 2005/3/24
* AcSplit/CF_XOPEN_SOURCE: Date: Fri, 15 Apr 2005 11:44:14 -0400
From: Chris Zubrzycki <beren@mac.com>
Subject: patch for new os x versions
To:
X-Mailer: Apple Mail (2.728)
[-- Attachment #1 --]
[-- Type: text/plain, Encoding: 7bit, Size: 0.1K --]
This patch is needed for building newer versions of os x. the c++
demos won't build on 10.4 with the latest patches.
[-- Attachment #2: ncurses.patch --]
2005-04-05
* AcSplit/CF_WITH_IMAKE_CFLAGS:
make a quick check on the result from CF_IMAKE_CFLAGS to see if we found
$IMAKE and set $IMAKE_CFLAGS (the first is necessary, the latter is only
very likely). If that fails, fall-through as if we're getting --disable-imake
* AcSplit/CF_IMAKE_CFLAGS:
skip over most of this if we don't have $IMAKE set
2005-04-02
* AcSplit/CF_GXX_WARNINGS: hmm (first version was probably for bcpp).
I rewrote this for ncurses, to try to use g++ to flush out problems with
its c++ binding, after finding that Solaris c++ complained (rightly) about
some anachronisms.
2005-02-06
* AcSplit/CF_WITH_IMAKE_CFLAGS:
repair the sed script for OSMINORVERSION to avoid trailing non-digits
add a special case for IRIX 5 & 6 to define SYSV and SVR4 (xterm kludge)
* AcSplit/CF_WITH_IMAKE_CFLAGS:
add a kludge for aix to define AIXV3 and AIXV4 if no imake is used.
* AcSplit/CF_XOPEN_SOURCE: define _ALL_SOURCE for AIX 4 and 5.
* AcSplit/CF_WITH_IMAKE_CFLAGS:
ran into problem using uname on AIX (which someone on autoconf list did
comment about recently). AIX's uname -r returns the minor version rather
than the major version. Use uname -v for the major version. I don't want
to own that type of script. However, config.guess' output is mostly usable,
giving the major/minor versions for most vendor systems - but not for Linux.
So if $host_os contains a {digit}.{digit}, I'll use that, otherwise use the
uname values.
2005-02-05
* AcSplit/CF_WITH_IMAKE_CFLAGS:
OSF1's uname -r returns "V4.0" for 4.0D, so I modified the sed script to
strip leading non-digits when getting a value for OSMAJORVERSION
* AcSplit/CF_IMAKE_CFLAGS:
typo - I'd left off the blank needed for one of the fix_cflags.sed expressions
2005-02-04
* AcSplit/CF_POSIX_C_SOURCE:
add -U_POSIX_C_SOURCE to avoid redefinition, e.g., with cygwin
* AcSplit/CF_XOPEN_SOURCE:
add an -U option to avoid redefinition warning with cygwin for -D_XOPEN_SOURCE
* AcSplit/CF_IMAKE_CFLAGS:
hmm - restore the complicated expression (needed for cygwin) and tune the
simpler expressions in fix_cflags.sed to (probably) work with Solaris
2005-02-03
* AcSplit/CF_POSIX_C_SOURCE:
add a check to avoid redefining _POSIX_SOURCE, e.g., with cygwin
* AcSplit/CF_IMAKE_CFLAGS:
simplify expression in fix_cflags.sed to work with pre-POSIX crap on Solaris.
2005-01-22
* AcSplit/CF_SRC_MODULES: eliminate PROG_ARGS in favor of TINFO_ARGS
2005-01-09
* AcSplit/CF_FUNC_TGETENT: typo in help message
* AcSplit/CF_FUNC_TGETENT:
address Gentoo bug #69926 by adding a --disable-full-tgentent option.
If disabled, this macro does not check for the termcap library, and
does not insist that the tgetent function return data in its buffer.
That makes it accept the first tgetent function it finds (which will
probably be from terminfo).
2005-01-01
* config.sub: 2004-11-30
* config.guess: 2004-11-12
* AcSplit/CF_SHARED_OPTS: Albert Chin added case for SGI with -rpath
* AcSplit/CF_CHECK_TYPE: ci-aclocal generated header
2004-12-03
* AcSplit/CF_GCC_WARNINGS: integrated the Intel compiler's warnings.
* AcSplit/CF_GNU_SOURCE:
don't try to set _GNU_SOURCE if we're using the Intel compiler.
May rethink that, but have been burned by inconsistent evaluation between
icc/gcc when this is set.
* AcSplit/CF_INTEL_COMPILER: ci-aclocal generated header
* AcSplit/CF_GCC_WARNINGS: check for Intel compiler.
If it's set, turn on all of its warnings.
2004-11-30
* AcSplit/CF_IMAKE_CFLAGS:
add a workaround to remove empty "-DPROJECTROOT= " string. It looks as if
all of the Unix vendors have that in their imake config, while XFree86 and
X.Org define a value.
2004-11-27
* AcSplit/CF_BUILD_CC:
add a sanity-check to ensure that --with-build-cc was used to specify a
different compiler from $CC when cross-compiling.
2004-11-23
* AcSplit/CF_FIND_LIBRARY: correct comment for $6
* AcSplit/CF_FUNC_ICONV:
set flag for $6 of CF_FIND_LIBRARY so we do not exit on failure to find iconv
* AcSplit/CF_XOPEN_SOURCE: Date: Sun, 21 Nov 2004 22:30:42 +0100
From: Georg Schwarz <georg.schwarz@freenet.de>
To: bug-ncurses@gnu.org
Cc: jlam@NetBSD.org
Subject: Ncurses 5.4 on IRIX 5.3 bug report
The following is a bug report for Ncurses 5.4 on IRIX 5.3:
ncurses/tinfo/lib_napms.c includes time.h to define struct timespec.
However, on
IRIX 5.3 that include file reads as follows:
#if (defined(_POSIX_4SOURCE) || defined(_SGI_SOURCE)) &&
!defined(_POSIX_SOURCE)
&& !defined(_XOPEN_SOURCE)
#if defined(_LANGUAGE_C) || defined(_LANGUAGE_C_PLUS_PLUS)
typedef struct timespec {
time_t tv_sec; /* seconds */
long tv_nsec; /* and nanoseconds */
} timespec_t;
....
now the problem is that ncurses defines _POSIX_SOURCE (and in addition does
not
define _SGI_SOURCE).
2004-11-03
* AcSplit/CF_FUNC_ICONV:
oops - my call to iconv_open() also was wrong; I omitted the dummy params.
2004-11-02
* AcSplit/CF_FUNC_ICONV:
add quotes so this works on Solaris, where we don't have the library
2004-10-25
* AcSplit/CF_GXX_WARNINGS: RCS_BASE
2004-10-21
* AcSplit/CF_FUNC_ICONV: ci-aclocal generated header
2004-10-17
* AcSplit/CF_POSIX_C_SOURCE: ci-aclocal generated header
* AcSplit/CF_XOPEN_SOURCE:
split-out CF_POSIX_C_SOURCE (which mistakenly reused the $cv_cv_xopen_source
cache variable, and had an error in its second compiler test which was not
noticed because of the first error!).
2004-09-18
* AcSplit/CF_LIB_RULES: eliminate the test -a option, using a case
* AcSplit/CF_LIB_RULES: Date: Sat, 4 Sep 2004 14:19:56 -0700 (PDT)
From: "James J. Ramsey" <jjramsey_6x9eq42@yahoo.com>
Subject: Patch for appending "w" in the right place on OS X
To: bug-ncurses@gnu.org
[-- Attachment #1 --]
[-- Type: text/plain, Encoding: 7bit, Size: 0.4K --]
This patch fixes aclocal.m4 and configure so that the
wide-character version of ncurses gets compiled as
libncursesw.5.dylib, instead of libncurses.5w.dylib. I
was able to compile and install ncursesw with this
patch.
(the chunk before mk-1st.awk didn't work since I'd made changes for libtinfo,
so I fixed that -TD)
2004-09-12
* AcSplit/CF_BUNDLED_INTL:
fix to generate intl/makefile rather than $srcdir/intl/makefile when the
configure --srcdir option is given.
* AcSplit/CF_OUR_MESSAGES:
fix to generate po/makefile rather than $srcdir/po/makefile when the
configure --srcdir option is given
2004-08-22
* AcSplit/CF_XOPEN_SOURCE:
Modified to ensure that _POSIX_C_SOURCE is defined with a value. Here's
the report:
From: Paul Green (Paul.Green@stratus.com)
Subject: Re: IS:: build of lynx on VOS
This is the only article in this thread
View: Original Format
Newsgroups:
comp.sys.stratus
Date: 2004-08-20 13:00:11 PST
Tao Diesel <stefan_ridinger(at)yahoo.de> wrote:
>
> building lynx failed with followed error:
>
> !gcc -I ./ -I ../ -I ../../ncurses/include -I
> ../WWW/Library/Implementation -I .
> +./src -c ../WWW/Library/Implementation/HTAABrow.c -o
> ../objects/HTAABrow.o -DHA
> +VE_CONFIG_H -D_POSIX_C_SOURCE -D_BSD_SOURCE -D_SVID_SOURCE
> -U__PROTOTYPES__ -O3
1. The POSIX standard requires that the _POSIX_C_SOURCE macro have a
value. Try the following control argument to gcc:
-D_POSIX_C_SOURCE=200112L
2. On VOS, you should use either -D_BSD_SOURCE or -D_SVID_SOURCE.
You should not use both of them. I generally define _SVID_SOURCE
as it has everything that _BSD_SOURCE has plus additional features.
Ditto for setting up your object library search paths.
Thanks
PG
2004-08-21
* AcSplit/CF_GNAT_TRY_LINK: RCS_BASE
* AcSplit/CF_GNAT_TRY_RUN: add $ADAFLAGS to test
add comments for parameters
2004-08-03
* AcSplit/CF_FCNTL_VS_IOCTL:
change parameter type of ioctl(f,FIONREAD,p) to an int, since that's what
works for IRIX64 64-bit (and other platforms, based on discussion on netbsd
seem to agree)
2004-07-23
* AcSplit/CF_GCC_WARNINGS:
add a parameter so I can (in vttest) add checks for less useful options such
as -Wno-unused.
* AcSplit/CF_WITH_WARNINGS:
pass a parameter to CF_GCC_WARNINGS for vttest's hack with -Wno-unused
2004-06-26
* AcSplit/CF_XOPEN_SOURCE: Date: Thu, 20 May 2004 13:15:42 -0700
From: Jeff Bailey <jbailey@raspberryginger.com>
Subject: Re: Bug#249214: FTBFS: Doesn't recognise that hurd-i386 is a
+glibc-based arch
To: dickey@his.com, 249214-quiet@bugs.debian.org
Cc: Jeff Bailey <jbailey@raspberryginger.com>,
249214-submitter@bugs.debian.org
On Thu, May 20, 2004 at 03:29:16PM -0400, Thomas Dickey wrote:
> > Here are patches to make hurd-i386 build again. the hurd terminal
> ...
> > --- aclocal.m4.bak 2004-05-15 23:02:22.000000000 -0400
> > +++ aclocal.m4 2004-05-15 23:02:40.000000000 -0400
> > @@ -3646,7 +3646,7 @@
> > irix6.*) #(vi
> > CPPFLAGS="$CPPFLAGS -D_SGI_SOURCE"
> > ;;
> > -linux*) #(vi
> > +gnu*|linux*) #(vi
>
> What is the whole output of config.guess?
i386-unknown-gnu0.3
Tks,
Jeff Bailey
--
Ignorance was bliss.
~.
* config.sub, config.guess: ncurses: update from subversions
2004-06-16
* AcSplit/CF_FUNC_DLSYM: ci-aclocal generated header
2004-06-15
* AcSplit/CF_X_ATHENA:
prompted by a gentoo bug report, modify the AC_ARG_WITH checks to assume
that someone might really do --without-Xaw3d and not like it if I just
assumed $withval was "yes"
2004-05-23
* AcSplit/CF_CHECK_CACHE:
Thorsten Glaser is apparently packaging lynx w/o the bundled config.guess,
putting that in $ac_aux_dir
* AcSplit/CF_SRAND:
from Thorsten Glaser: check for arc4random on MirBSD.
google says this is an OpenBSD function.
2004-04-27
* AcSplit/CF_NCURSES_LIBS: 2004-04-27 (2.8.6dev.2)
* modify CF_NCURSES_LIBS to only look for -lmytinfo when building with ncurses.
This works around FreeBSD's broken linker semantics when configuring for
ncursesw -TD
(for lynx)
2004-04-26
* AcSplit/CF_SSL:
change a variable name to match CF_GNUTLS a little better
* AcSplit/CF_GNUTLS:
modified (like CF_SSL) to be a little more cautious about looking for
the <gnutls/openssl.h> file, and to accommodate recent change (11/2003)
to move the openssl-compatibile functions from gnutls-extra to gnutls-openssl
* AcSplit/CF_SSL:
fix a conflict with system-installed openssl library and private ssleay
(I was picking up the headers for the former when looking at the latter).
2004-04-25
* AcSplit/CF_IMAKE_CFLAGS:
fix for cases when the IMAKE_CFLAGS and/or IMAKE_LOADFLAGS are empty.
* AcSplit/CF_IMAKE_CFLAGS:
hmm - this broke again when I added $(VENDORMANDEFS), since that contained
different escaping of the quotes. This seems to work, for xterm #187.
* AcSplit/CF_IMAKE_CFLAGS:
the second parameter to xmkmf should be the top of the build tree, not
the directory containing the cf-files.
* AcSplit/CF_ADD_CFLAGS:
fix the first expression for cf_tst_cflags by ensuring we match single
quote rather than just any character after the '='. Do a similar fix for
the "yes" part of the case statement, so we exit on a fragment such as
xxx"'
Finally, put any quoted (single-token) -D option into EXTRA_CPPFLAGS.
That seems to be consistent, though of course we don't handle things like
-Dxxx=\"foo\"
* AcSplit/CF_X_FREETYPE:
add --with-freetype-cflags and --with-freetype-libs options (actually because
I thought I could use that to override xft-config easily via an option to
cfg-xterm+, but that didn't work well, so I ended up overriding its cache
variables). The options still are useful.
* AcSplit/CF_IMAKE_CFLAGS:
if --verbose is set, show the $IMAKE_CFLAGS and $IMAKE_LOADFLAGS at the
end of this macro.
* AcSplit/CF_IMAKE_CFLAGS:
fix the order of setting CDPATH (does not seem to help on my current Debian
testing, but still better than it was)
* AcSplit/CF_X_TOOLKIT:
remove the checks for clix (obsolete) and move the call to AC_PATH_XTRA to
the beginning - which it essentially was anyway. Now it simply calls
AC_PATH_XTRA and ensures that I can get -lXt and -lX11. Reading the code,
it seems that the IRIX6/IRIX5 issue was taken care of in autoconf 2.13.
2004-04-24
* AcSplit/CF_IMAKE_CFLAGS:
xterm: modify the generated Makefile to use sed-scripts that effectively
recognize -D options that have non-numeric values. When those are found,
this forces quotes around the value.
See also CF_ADD_CFLAGS, which has to work around limitations in passing
backslashes and quotes around.
* AcSplit/CF_ADD_CFLAGS:
xterm: inspect each -D option to see if it looks like a quoted string.
We're expecting to see something like
-D'"xxx"'
from CF_IMAKE_CFLAGS -> CF_WITH_IMAKE_CFLAGS. In some cases this
-D'"xxx yyy"'
would be passed as
-D'"xxx
yyy"'
so we put it back together in the loop here. Crude, but it seems the only
way to cope with $(VENDOR_DEFINES) in the X makefiles.
* AcSplit/CF_IMAKE_CFLAGS:
I made a script (cfg-xorg) to run cfg-xterm+ within the X.Org build tree.
To make this work (since this script prefers xmkmf to imake), I have both
imake and xmkmf in its cfg-xorg's $PATH. One more detail is needed to get
going: xmkmf has to point to the config-files within the build tree rather
than in its final installed location. So I added a case to set cf_imake_opts
to add the build-tree's location to xmkmf as a parameter.
2004-02-28
* AcSplit/CF_WITH_DBMALLOC, AcSplit/CF_WITH_DMALLOC:
join lines in an AC_CHECK_LIB where a parameter was split across lines.
that made the resulting substitution split also (reported in tin 1.7.4).
2004-02-23
* config.sub: 2004-02-23 (from tin)
2004-02-17
* AcSplit/CF_BUILD_CC: add BUILD_CPP, BUILD_OBJEXT for lynx
* config.guess: 2004-02-16 (from tin)
2004-02-14
* AcSplit/CF_LIB_RULES:
don't use underscores in awk command-line variables
* AcSplit/CF_LIB_RULES:
modified to allow ncurses "--with-termlib" option to set the name of the
terminfo library.
* AcSplit/CF_SRC_MODULES:
"$with_termlib" can now have values other than yes/no.
* AcSplit/CF_PROG_CC_C_O: ci-aclocal generated header
* AcSplit/CF_LIB_RULES:
pass cc_c_o and cxx_c_o values to mk-2nd.awk, which uses those to decide
whether to use a cd-command or -c/-o options to get its result into the
object-directories
2004-01-30
* AcSplit/CF_CHECK_CACHE:
ncurses: revert TG's change since it breaks my build on FreeBSD 3.1
* AcSplit/CF_BOOL_DECL:
ncurses: fix for BeOS 4.5, which has bool declared inconsistently in two
locations. We can control whether we include <stdbool.h>, but not the
other (which is buried away in support headers for I/O).
2004-01-26
* AcSplit/CF_XOPEN_SOURCE: Thorsten Glaser adds a case for mirbsd
* AcSplit/CF_TAR_OPTIONS: ci-aclocal generated header
* AcSplit/CF_PATH_PROG:
tweak to allow a list of programs, use this in lynx to support selecting
among tar/star/pax
* AcSplit/AM_GNU_GETTEXT:
Thorsten Glaser added a check for a system copy of mkinstalldirs (note that
his check relied on using environment variables, so I fixed it).
* AcSplit/CF_CHECK_CACHE:
Thorsten Glaser noticed a standard variable $ac_config_sub to use.
* AcSplit/CF_XOPEN_SOURCE:
oops - defining _XOPEN_SOURCE breaks lynx on SCO (some problems immediately
with the network headers).
2004-01-24
* AcSplit/CF_WITH_BZLIB:
original bzip2 1.0.2 header <bzlib.h> does not include <stdio.h>, but uses
FILE*, so unless someone's fixed it (for a package), I have to add <stdio.h>
to the includes
2004-01-23
* AcSplit/AM_WITH_NLS:
lynx: suppress check for bison if we're not bundling intl/. The dependency
upon bison is contrived (not needed for the application), anyway.
* AcSplit/AM_WITH_NLS:
lynx: add a check to see if intl/ directory is packaged, and if not to
stop with an error.
2004-01-22
* AcSplit/CF_CHECK_IPV6:
lynx: treat non-Linux systems comparably to Linux by not exiting with an
error when getaddrinfo() check fails.
2004-01-19
* AcSplit/CF_BUILD_CC:
tin: add 2 parameters to allow me to pick up the proper $CPPFLAGS for build,
and ensure that the defaults are different when I'm cross-compiling.
* AcSplit/CF_BUILD_CC: ci-aclocal generated header
* AcSplit/CF_WITH_DBMALLOC, AcSplit/CF_WITH_DMALLOC:
tin: remove "test:" from usage message. Add optional parameter to allow
delaying adding the $LIBS and $CPPFLAGS updates til the end of the script.
* AcSplit/AM_MULTIBYTE_ABLE:
corrected the way cache variable interacts with the #define's (based on
warning message from autoconf 2.52).
2004-01-17
* AcSplit/CF_WCHAR_TYPE:
ncurses: Tru64 4.0d headers include files under /usr/include.dtk which
reference __VA_LIST__ symbol when <wchar.h> is included. So I included
<stdarg.h> - not enough, had to add <stdio.h> also. That made the test
work.
2004-01-16
* AcSplit/CF_WITH_LIBTOOL:
add ifdef'd invocation of AC_PROG_LIBTOOL with an example of how to turn
it on, and documented the reasons why not. Also improved the sed expression
for libtool's version - Debian's package adds some stuff in parentheses
which confused my script because there were digits before the actual version
number.
2004-01-14
* AcSplit/CF_BOOL_DECL:
ncurses: rewrote the check for stdbool.h, since HPUX has a <stdbool.h> that
defines bool only if the compiler is c99. In that case, we'll normally
fall-through to the non-bool case.
2004-01-12
* AcSplit/CF_XOPEN_SOURCE:
add special case for NetBSD (hp-testdrive's 1.6 doesn't like _XOPEN_SOURCE).
2004-01-11
* AcSplit/CF_XOPEN_SOURCE:
lynx: for IRIX64 (e.g., "irix6.*"), use -D_SGI_SOURCE rather than _XOPEN_SOURCE,
since defining the latter makes the IPv6 code fail to compile. I don't see
that its curses library really supports wide-characters anyway, though we
have something for mbstate_t.
2004-01-10
* AcSplit/CF_SRC_MODULES:
ncurses: tweak so ./progs will link with -ltinfo when --with-termlib is used.
* AcSplit/CF_LIB_RULES:
ncurses: modify so termlib can have its own extensions 'ext_tinfo', e.g.,
the version.c module which I moved there.
2004-01-06
* config.sub, config.guess: tin: 2004-01-05
2004-01-03
* AcSplit/CF_PROG_EXT:
ncurses - using AC_EXEEXT made a conflict between explicit suffix and the -Zexe
flag. Modified this to set $ac_cv_exeext to bypass the AC_EXEEXT macro
(which would set both -Zexe and EXEEXT=.exe).
* AcSplit/CF_OUR_MESSAGES:
fix sense of help-message - since the default is "yes", the help message
should show "--disable"
2003-12-31
* AcSplit/CF_X_FREETYPE:
fix -DRENDERFONT when there is no freetype, etc., broken by the cygwin/X fix.
* AcSplit/CF_X_FREETYPE:
current cygwin/X provides xft-config and freetype-config scripts. The former
declares -lXft and -lfontconfig (which are both needed for xterm with AA-fonts).
The latter only declares -lfreetype. So I added a check for both, preferring
the xft-config script. That got xterm to configure/build.
2003-12-30
* AcSplit/CF_SYSV:
back to Slackware 3.6, and found it did not flag any warnings. But adding
an include for <stdlib.h> got it to cooperate. No more "#ifdef linux".
* AcSplit/CF_X_FREETYPE:
rewrote this to use freetype-config (which seems to work for Debian 3.0r1
and current Debian testing - probably not for Redhat9). The problem was
that Debian/testing didn't configure Xft because there's a magic header
needed to compile the test-case. It seems that freetype-config supplies
the appropriate -I with "freetype-config --cflags".
2003-12-29
* AcSplit/CF_XOPEN_SOURCE:
testing on OpenBSD 2.8 I found the _XOPEN_SOURCE and _POSIX_C_SOURCE fixes broke ncursesw and xterm.
* AcSplit/CF_X_FREETYPE: remove redundant -lXrender
take the -lfontconfig out again, leaving a note for when I have to revisit.
* AcSplit/CF_SYSV:
xterm - require CF_SYS_ERRLIST and ensure that sys_errlist is declared.
That's needed for HPUX which provides it but does not declare it.
Change the include to Intrinsic.h from Xos.h since it appears that I can
get Linux (Debian 3.0testing) to error on the redefinition of wchar_t.
2003-12-27
* AcSplit/CF_SHARED_OPTS:
change FreeBSD 4.x, 5.x, to use soname. This differs a little from the
system's version (libncurses.so.5 versus my libncurses.so.5.3), but doesn't
seem to do any harm. I don't see a clean way to check if my library is
preferred to the system library though.
I also tried to get -rpath to work, but it didn't work consistently enough
for me to want it.
* AcSplit/CF_XOPEN_SOURCE:
modified the freebsd case so I can build ncursesw also. I did this on my
FreeBSD 5.1 - will have to retest on 4.x. Probably doesn't work for older
FreeBSD's.
* AcSplit/CF_XOPEN_SOURCE:
add definition for _BSD_TYPES (thought I'd done this, but apparently got
sidetracked with __EXTENSIONS__)
2003-12-20
* AcSplit/CF_MAN_PAGES:
always use tbl when creating formatted manpages.
optionally use tbl when installing unformatted manpages
* AcSplit/CF_MANPAGE_FORMAT:
add --with-manpage=catonly flavor, and be more rigorous about allowing
the various keywords for --with-manpage in any combination.
* AcSplit/CF_MAN_PAGES:
not quite a rewrite - this adds support for --with-manpage-format=catonly
by modifying the man/edit_man.sh script to recur as needed, so the recurrance
will process formatted manpages.
2003-12-19
* AcSplit/CF_SYSV: typo: dn/dnl
* AcSplit/CF_SYSV:
hmm - I couldn't force a symbol conflict for SIGCHLD/SIGCLD, so I reverted
part of my changes, just saying linux!=SYSV. The cygwin fix works...
2003-12-17
* AcSplit/CF_SYSV:
Slackware 3.6 didn't error on the sys/termio.h, so I added a check to see
if SIGCHLD is defined in signal.h
2003-12-14
* AcSplit/CF_SYSV:
define SYSV during the test, and include <X11/Xos.h> so we'll get the
conflicting sys_errlist[]. Documented this for cygwin. Will have to
re-test on older linux to be sure that the test breaks.
2003-12-13
* AcSplit/CF_MAN_PAGES:
modified to implement --with-manpage-aliases option combined with --without-manpage-symlinks
* AcSplit/CF_MANPAGE_SYMLINKS:
ncurses: add --with-manpage-aliases option
* AcSplit/CF_X_FREETYPE:
add -lfontconfig, needed on current Cygwin. Note that current Cygwin is
broken anyway since the fontconfig package contains only documentation,
no code (courtesy of Harold Hunt ;-)
* AcSplit/CF_SYSV:
don't define SYSV for __CYGWIN__ since this in turn leads to <X11/Xos.h>
declaring sys_errlist[], which differs from Cygwin's hack (the latter
adds a 'const'). Comparing with Linux, it seems that's a good way to
improve the check rather than rely on hardcoded ifdef's.
2003-12-06
* AcSplit/CF_XOPEN_SOURCE:
ncurses: make a big default case to handle my preferred _XOPEN_SOURCE definition.
* AcSplit/CF_XOPEN_SOURCE:
add __EXTENSIONS__ for solaris* (the scope of this macro is losing focus)
2003-11-29
* install-sh: copied from install.sh
* install.sh: nicer(?) version from tin
2003-11-26
* AcSplit/CF_XOPEN_SOURCE: add case for osf1 (4.x, 5.x)
2003-11-22
* AcSplit/CF_VERSION_INFO:
dialog: initially modify so I can read the "0.9b-20031002" version from
dialog's VERSION file. This macro will read that, plus the 3-column
text from cdk's VERSION file, making it a little more general than it was.
(Not all programs should use libtool). Also integrated the PACKAGE stuff
here, so I can define DIALOG_VERSION and DIALOG_PATCHDATE.
2003-11-08
* config.sub: this is 1.294
from http://subversions.gnu.org/cgi-bin/viewcvs/config/config/
* config.guess: this is 1.284 (should have a timestamp for 2003/11/3).
from http://subversions.gnu.org/cgi-bin/viewcvs/config/config/
2003-11-07
* AcSplit/CF_XOPEN_CURSES:
tack on a check for getbegyx(), since I'm using it and HPUX dropped compat
for older stuff such as getbegy()/getbegx().
* AcSplit/CF_XOPEN_SOURCE: ci-aclocal generated header
2003-11-06
* AcSplit/CF_NCURSES_CPPFLAGS:
provide CF_CURSES_CPPFLAGS so I can make this an equal dependency of macros
that use the header-variables
* AcSplit/CF_CURSES_TERM_H, AcSplit/CF_NCURSES_VERSION:
require CF_CURSES_CPPFLAGS
* AcSplit/CF_CURSES_CCHAR_T, AcSplit/CF_CURSES_CHTYPE,
AcSplit/CF_CURSES_FUNCS, AcSplit/CF_CURSES_LIBS, AcSplit/CF_TCAP_CURSOR:
require CF_CURSES_CPPFLAGS and use the derived header
* AcSplit/CF_CURSES_MOUSE, AcSplit/CF_STRUCT_SCREEN:
require CF_CURSES_CPPFLAGS and use the derived header variable
* AcSplit/CF_XOPEN_CURSES:
require CF_CURSES_CPPFLAGS and use the derived header.
2003-11-01
* AcSplit/CF_LIB_SUFFIX: Date: Mon, 27 Oct 2003 14:55:10 -0500
From: Jonathan Ward <Jonathan.Ward@hp.com>
Subject: Patch for HP-UX for Itanium support
To: bug-ncurses@gnu.org
The following is a patch for the Itanium releases of HP-UX. There are several
key features but the most notable is the library suffix is .so on Itanium
+libraries
not .sl as on PA-RISC libraries.
(omitted a chunk that put a modified-by comment at the top of aclocal.m4)
2003-10-23
* config.sub: 2003-08-18 (from autoconf 2.57g)
* config.guess: 2003-10-03 (from autoconf 2.57g)
2003-10-18
* AcSplit/CF_MAN_PAGES:
$srcdir seems to work for location of edit_man.sed
* AcSplit/CF_PROG_EXT:
use AC_EXEEXT, AC_OBJEXT (which are slated for removal in the next pass of
autoconf-incompatibilties ;-)
2003-10-11
* AcSplit/CF_IMAKE_CFLAGS:
check that we really did find imake or xmkmf, so the script doesn't run amok
2003-10-05
* AcSplit/CF_WITH_ZLIB, AcSplit/CF_WITH_BZLIB:
ci-aclocal generated header
* AcSplit/CF_SSL:
add USE_X509_SUPPORT definition to distinguish this from gnutls
* AcSplit/CF_GNUTLS:
add check for x509 support (which is very incomplete)
2003-09-21
* AcSplit/CF_X_FREETYPE: hmm - Xft is "documented"
2003-09-20
* AcSplit/CF_NUMBER_SYNTAX, AcSplit/CF_WITH_ABI_VERSION,
AcSplit/CF_WITH_REL_VERSION, AcSplit/CF_PROG_LDCONFIG:
ci-aclocal generated header
* AcSplit/CF_MIXEDCASE_FILENAMES:
if cross-compiling, rely on known cases
2003-09-17
* AcSplit/CF_X_FREETYPE:
add a couple of symbols needed as stubs to let xterm compile with slightly
different versions of Xft.
2003-09-14
* AcSplit/CF_SLANG_CPPFLAGS: add check for slang-utf8 package.
* AcSplit/CF_BUNDLED_INTL:
list some of the environment variables (at least the ones I use directly).
allow override of CONFIG_H symbol (needed for lynx, which renames it).
2003-09-08
* AcSplit/CF_BUNDLED_INTL: change "head -1" to "sed -e '2,$d'"
2003-09-06
* AcSplit/CF_GNAT_VERSION:
ncurses: replace "head -1" with "sed -e '2,$d'", to work when POSIXLY_CORRECT
is set. This also gets rid of an unnecessary use of $AWK, and ensures that
$cf_ada_make is set.
* AcSplit/CF_GCC_VERSION:
ncurses: replace "head -1" with "sed -e '2,$d'" to work when POSIXLY_CORRECT
is set.
* AcSplit/CF_GXX_VERSION, AcSplit/CF_WITH_LIBTOOL:
ncurses: replace "head -1" with "sed -e '2,$d'", to work when POSIXLY_CORRECT
is set.
2003-08-30
* AcSplit/CF_WITH_LIBTOOL:
ncurses: mods to make this work with libtool 1.5 -- it requires "options"
that are not really optional, for planned support of other languages than C.
2003-08-20
* AcSplit/CF_CURSES_FUNCS:
dialog: add a dependency on CF_CURSES_TERM_H, and include <ncursesw/term.h>
if that's what it found.
* AcSplit/CF_CURSES_TERM_H:
dialog: check for <ncursesw/term.h> if we found <ncursesw/curses.h>.
2003-07-26
* AcSplit/CF_PREDEFINE: ci-aclocal generated header
2003-07-19
* AcSplit/CF_LIB_RULES:
Charles Wilson - change prefix on cygwin dll's to "cyg" to avoid conflict
with other dll's that happen to have the same root name.
2003-07-12
* AcSplit/CF_CFG_DEFAULTS: Date: Tue, 8 Jul 2003 02:12:25 +0200
From: Robert Millan <zeratul2@wanadoo.es>
To: "Thomas E. Dickey" <dickey@invisible-island.net>
Subject: patch for GNU/FreeBSD support
Message-ID: <20030708001225.GA10771@aragorn>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="5mCyUwZo2JvN/JJP"
Content-Disposition: inline
Organisation: free as in freedom
User-Agent: Mutt/1.5.4i
Sender: Robert Millan <rmh@debian.org>
Status: RO
Content-Length: 1453
Lines: 48
--5mCyUwZo2JvN/JJP
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Hi!
If i'm not confused, you're the maintainer of ncurses, right?
I mailed you some time ago about your special autoconf templates
that need your hacked autoconf2.13 to be processed.
I finaly installed your hacked autoconf2.13 and tested my patches
for ncurses to support GNU/FreeBSD. My patch works fine and is
attached; all it does is adding our triplet in a pair of places
in aclocal.m4
(his change installs ncurses with --prefix=/usr by default for kfreebsd-gnu)
* AcSplit/CF_SHARED_OPTS: Date: Tue, 8 Jul 2003 02:12:25 +0200
From: Robert Millan <zeratul2@wanadoo.es>
To: "Thomas E. Dickey" <dickey@invisible-island.net>
Subject: patch for GNU/FreeBSD support
Message-ID: <20030708001225.GA10771@aragorn>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="5mCyUwZo2JvN/JJP"
Content-Disposition: inline
Organisation: free as in freedom
User-Agent: Mutt/1.5.4i
Sender: Robert Millan <rmh@debian.org>
Status: RO
Content-Length: 1453
Lines: 48
--5mCyUwZo2JvN/JJP
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Hi!
If i'm not confused, you're the maintainer of ncurses, right?
I mailed you some time ago about your special autoconf templates
that need your hacked autoconf2.13 to be processed.
I finaly installed your hacked autoconf2.13 and tested my patches
for ncurses to support GNU/FreeBSD. My patch works fine and is
attached; all it does is adding our triplet in a pair of places
in aclocal.m4
(hit change adds shared-library rules for kfreebsd-gnu)
* config.sub: 2003-07-04
ncurses - needs k*bsd*gnu
* config.guess: 2003-07-02
ncurses: needs k*bsd*gnu
2003-07-05
* AcSplit/CF_GCC_WARNINGS:
back-out version 11, which allowed this to set $CFLAGS if $GXX (g++) was
found. That doesn't work well when compiling with HPUX cc and g++, since
the gcc options leak into the cc compile.
2003-06-16
* AcSplit/CF_BUNDLED_INTL:
only update $VERSION if it's not set (it should be)
* AcSplit/CF_BUNDLED_INTL:
dialog: adapt this to use "my-gettext", which happens to have Makefile.in,
while po has makefile.inn (by making this macro look for Makefile.in or
makefile.in). Also set CONFIG_H and VERSION if they're not set
* AcSplit/CF_OUR_MESSAGES:
dialog: prefix the "po/" with "$srcdir/", in case I'd like to use --srcdir
2003-06-10
* AcSplit/CF_RCS_SCCS:
add /usr/local/lib/cssc and /usr/lib/cssc, since Debian puts the executables
for CSSC in /usr/lib/cssc
2003-06-07
* AcSplit/CF_WITH_LIBTOOL:
define LIB_PREP as token to substitute for 'ranlib' in makefiles rather
than 'RANLIB', so I don't accidentally change the $RANLIB that the configure
script is using, e.g., in the test for broken-linker.
* AcSplit/CF_SUBST_NCURSES_VERSION:
don't need to ifdef based on cf_cv_cc_bool_type any more
2003-06-05
* AcSplit/CF_CURSES_CPPFLAGS:
change the order that I add to $CPPFLAGS, putting the curses include-directory
first (noticed this while working on tin)
* AcSplit/CF_OUR_MESSAGES:
tin: move logic for MSF_DIR_MAKE from CF_BUNDLED_INTL (which had some
conflicting logic with INTLDIR_MAKE).
also added some comments.
* AcSplit/CF_BUNDLED_INTL:
tin: move some logic over to CF_OUR_MESSAGES that should have been there,
e.g., for MSG_DIR_MAKE. Add some comments.
2003-06-02
* config.sub: 2003-05-24 (from tin 1.5.18)
2003-05-24
* AcSplit/CF_GCC_WARNINGS:
"-Winline" is broken in gcc 3.3 because someone decided to turn on inlining
by default without adjusting the warning to be less intrusive when no explicit
inlining was requested for a function. (gcc developers choose to not use
compiler warnings, of course ;-)
* AcSplit/CF_WITH_WARNINGS:
make check for gcc/g++ consistent with CF_GCC_WARNINGS
* AcSplit/CF_GCC_VERSION:
make this work for gcc 3.3, which writes its version as "gcc (GCC) 3.3".
* AcSplit/CF_GXX_VERSION: typo (missing brackets)
* AcSplit/CF_GCC_VERSION, AcSplit/CF_GXX_VERSION:
ci-aclocal generated header
* AcSplit/CF_ENABLE_WARNINGS:
make check for gcc/g++ consistent with CF_GCC_WARNINGS
2003-05-22
* config.guess: 2003-05-22 (from tin 1.5.18)
2003-05-19
* AcSplit/CF_GCC_WARNINGS:
xterm: moved the -DXTSTRINGDEFINES to $CPPFLAGS so I don't get confused when
making a ".i" file.
2003-05-18
* AcSplit/CF_XKB_BELL_EXT: autoconf-252 reported misplaced parenthesis
2003-05-17
* AcSplit/CF_CURSES_ACS_MAP, AcSplit/CF_CURSES_WACS_MAP,
AcSplit/CF_FUNC_CURSES_VERSION:
use $cf_cv_ncurses_header
2003-05-04
* AcSplit/CF_WCTYPE:
don't define HAVE_WCTYPE until we've also found wctype() in a library.
cygwin has the header files, but no working code.
2003-05-02
* AcSplit/CF_LEX_VERSION: ci-aclocal generated header
2003-04-27
* AcSplit/CF_GNUTLS: look for "-lcrypto", not "-lcrypt"
2003-04-18
* AcSplit/CF_WITH_LIBTOOL: fix quoting on the ECHO_LT change
2003-04-17
* AcSplit/CF_WITH_LIBTOOL:
cdk: for CF_DISABLE_ECHO, reset the ECHO_LT variable if we're not using libtool
* AcSplit/CF_DISABLE_ECHO: cdk: add ECHO_LT, to quiet libtool
2003-04-13
* AcSplit/CF_GNUTLS: ci-aclocal generated header
2003-04-12
* AcSplit/CF_WITH_IMAKE_CFLAGS:
add the result to $CFLAGS, for the fallback, to match the CF_IMAKE_CFLAGS
behavior
* AcSplit/CF_IMAKE_CFLAGS: add comments for parameters
* AcSplit/CF_WITH_IMAKE_CFLAGS: ci-aclocal generated header
2003-04-05
* AcSplit/CF_YACC_ERROR: add some messages for the configure-output
2003-03-22
* AcSplit/CF_XKB_BELL_EXT: RCS_BASE
* AcSplit/CF_WCHAR_TYPE: ci-aclocal generated header
* AcSplit/CF_ETIP_DEFINES: fix includes needed in build
* AcSplit/CF_WITH_SYSMOUSE:
moved a "fi" to keep from emitting message when not on FreeBSD
added generated header
2003-03-19
* AcSplit/CF_TTY_GROUP: ci-aclocal generated header
* AcSplit/CF_TTY_GROUP:
xterm: the AC_TRY_RUN check won't run in batch mode (I knew that, but didn't
consider it important - Nelson Beebe did). Make it a little smarter and
more tolerant to accommodate.
2003-03-17
* AcSplit/CF_WCTYPE: ci-aclocal generated header
* AcSplit/CF_WCTYPE:
vile: the parameter to AC_SEARCH_LIBS must not include the "-l"
2003-03-01
* AcSplit/CF_INHERIT_SCRIPT: RCS_BASE
* AcSplit/CF_INHERIT_SCRIPT: ci-aclocal generated header
* AcSplit/CF_CURSES_CHECK_TYPE: RCS_BASE
* AcSplit/CF_CURSES_CHECK_TYPE: ci-aclocal generated header
* AcSplit/CF_CURSES_ACS_MAP, AcSplit/CF_CURSES_WACS_MAP,
AcSplit/CF_FUNC_CURSES_VERSION:
RCS_BASE
* AcSplit/CF_CURSES_ACS_MAP, AcSplit/CF_CURSES_WACS_MAP,
AcSplit/CF_FUNC_CURSES_VERSION, AcSplit/CF_UTF8_LIB:
ci-aclocal generated header
* AcSplit/CF_UTF8_LIB: ncurses/test: fix a caching issue with $LIBS
2003-02-22
* config.sub, config.guess: tin: 2003/2/22
2003-02-16
* AcSplit/CF_WCTYPE: add a check for -lw, needed for Solaris
* AcSplit/CF_X_ATHENA_LIBS: ci-aclocal generated header
* AcSplit/CF_X_ATHENA_LIBS:
oops - the script didn't really add $X_PRE_LIBS and $X_EXTRA_LIBS since I
used a macro that didn't use that parameter.
* AcSplit/CF_LOCALE: ci-aclocal generated header
* AcSplit/CF_LOCALE:
vile: modified the result so we can put a --with-locale option inside the
part where we're sure there's something to offer.
2003-02-02
* AcSplit/CF_HAVE_FUNCS: ci-aclocal generated header
* AcSplit/CF_HAVE_FUNCS: ncurses: add quotes around test $variable
2003-02-01
* AcSplit/CF_GPP_LIBRARY, AcSplit/CF_LINK_DATAONLY,
AcSplit/CF_LINK_FUNCS, AcSplit/CF_SHARED_OPTS:
ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS:
ncurses: use $GCC rather than $ac_cv_prog_gcc since the latter is a different
name in autconf 2.5x
* AcSplit/CF_GPP_LIBRARY:
ncurses: use $GXX rather than $ac_cv_prog_gxx since the latter is a different
name in autoconf 2.5x
* AcSplit/CF_LINK_DATAONLY:
ncurses: use $RANLIB rather than $ac_cv_prog_RANLIB since the latter is a
different name in autoconf 2.5x
* AcSplit/CF_LINK_FUNCS:
ncurses: use $cross_compiling rather than $ac_cv_prog_cc_cross, since the
latter is a different name in autoconf 2.5x
2003-01-25
* AcSplit/CF_WITH_SYSMOUSE: RCS_BASE
2003-01-16
* AcSplit/AM_MULTIBYTE_ABLE: ci-aclocal generated header
* AcSplit/AM_MULTIBYTE_ABLE: tin: whitespace
2003-01-12
* AcSplit/CF_NCURSES_CC_CHECK: ci-aclocal generated header
* AcSplit/CF_NCURSES_CC_CHECK:
lynx: undef HAVE_LIBUTF8_H in case it was already defined via CF_UTF8_LIB
2003-01-04
* AcSplit/CF_WITH_INSTALL_PREFIX: RCS_BASE
* AcSplit/CF_WITH_INSTALL_PREFIX: ci-aclocal generated header
2002-12-29
* AcSplit/CF_SUBDIR_PATH:
add: added some special chunks - most for symmetry - but to pick up the
/usr/local/ncurses/include/ncurses directory which I use in cfg-local on
my Linux boxes.
* AcSplit/CF_NCURSES_CPPFLAGS:
add: when attempting to port to OS/2, I found this macro didn't find my
header files -- because the -I options weren't being set. Also, the
CF_ADD_INCDIR adds both /usr/include/ncurses and /usr/include, so there's
no need for the special test to add each separately.
* AcSplit/CF_NCURSES_CPPFLAGS, AcSplit/CF_SUBDIR_PATH:
ci-aclocal generated header
* AcSplit/CF_ENABLE_WARNINGS, AcSplit/CF_WITH_DBMALLOC,
AcSplit/CF_WITH_DMALLOC:
RCS_BASE
* AcSplit/CF_ENABLE_WARNINGS, AcSplit/CF_WITH_DBMALLOC,
AcSplit/CF_WITH_DMALLOC:
ci-aclocal generated header
2002-12-27
* AcSplit/CF_FUNC_TGETENT: ci-aclocal generated header
* AcSplit/CF_FUNC_TGETENT:
xterm: workaround for running in 'screen', which may set $TERM to a value
that is not in termcap - noting that xterm prefers termcap if it can get
it, while screen is more easily fooled.
2002-12-26
* AcSplit/CF_X_ATHENA: ci-aclocal generated header
* AcSplit/CF_X_ATHENA: vile: add a check/option for XawPlus
2002-12-22
* AcSplit/CF_WCTYPE: RCS_BASE
* AcSplit/AM_MULTIBYTE_ABLE: tin: add calls to fputwc() and fwide()
* AcSplit/CF_NCURSES_LIBS: ci-aclocal generated header
* AcSplit/CF_NCURSES_LIBS:
vile: define HAVE_LIBNCURSES/HAVE_LIBNCURSESW as needed
2002-12-21
* AcSplit/CF_LIB_RULES, AcSplit/CF_MAN_PAGES, AcSplit/CF_PROG_INSTALL:
ci-aclocal generated header
* AcSplit/CF_LIB_RULES, AcSplit/CF_MAN_PAGES, AcSplit/CF_PROG_INSTALL:
ncurses: prefer '%' as sed-delimiter
* AcSplit/CF_NCURSES_CC_CHECK: RCS_BASE
* AcSplit/CF_NCURSES_CPPFLAGS:
ded: reverse order of curses.h, ncurses/curses.h search list (from tin).
split-out CF_NCURSES_CC_CHECK to simplify implementing a compile-time check for ncursesw
* AcSplit/CF_WITH_WARNINGS: ci-aclocal generated header
* AcSplit/CF_WITH_WARNINGS:
ded: fix some typos that made this silently fail
* AcSplit/CF_NCURSES_LIBS: ded: use '%' as sed-delimiter
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_DIRNAME, AcSplit/CF_GCC_ATTRIBUTES,
AcSplit/CF_PROG_EXT, AcSplit/CF_STRIP_G_OPT:
ci-aclocal generated header
* AcSplit/CF_ADD_INCDIR:
ded: filter out requests to add -I/usr/local/include when gcc 3.x barfs
* AcSplit/CF_GCC_ATTRIBUTES: ded: remove some commented-out stuff
* AcSplit/CF_DIRNAME, AcSplit/CF_PROG_EXT, AcSplit/CF_STRIP_G_OPT:
ded: use '%' as sed-delimiter
* AcSplit/CF_TD_SIG_ARGS: ci-aclocal generated header
* AcSplit/CF_HAVE_FUNCS, AcSplit/CF_TD_SIG_ARGS:
ded: rename $WithWarnings to $cf_opt_with_warnings to work with CF_WITH_WARNINGS
2002-12-14
* AcSplit/CF_GNAT_VERSION: ci-aclocal generated header
* AcSplit/CF_GNAT_VERSION:
only use the first line of gnatmake's version, to avoid showing trash from the help-message
* AcSplit/CF_LIB_RULES: rename target/host
2002-12-07
* AcSplit/CF_WITH_LIBTOOL: ci-aclocal generated header
* AcSplit/CF_WITH_LIBTOOL: integrated this (more/less) with ncurses.
* AcSplit/CF_GNAT_VERSION: report gnatmake's version for the log
allow for a 3.2 version (probably what gcc 3.2 does ;-)
2002-12-04
* config.sub, config.guess: 2002-11-30
2002-12-03
* AcSplit/CF_DEFAULT_SHELL: ci-aclocal generated header
* AcSplit/AM_MULTIBYTE_ABLE: tin: refine compile-time test
* AcSplit/CF_DEFAULT_SHELL: tin: whitespace
2002-11-30
* AcSplit/CF_ADA_INCLUDE_DIRS, AcSplit/CF_ADD_CFLAGS,
AcSplit/CF_INCLUDE_DIRS:
ci-aclocal generated header
* AcSplit/CF_ADA_INCLUDE_DIRS:
ncurses: when adding to $CPPFLAGS, put new -I stuff first
* AcSplit/CF_ADD_CFLAGS:
ncurses: when updating $CPPFLAGS, add new -I's at the beginning
* AcSplit/CF_INCLUDE_DIRS:
ncurses: when updating $CPPFLAGS, put new -I's at the beginning
2002-11-23
* AcSplit/CF_GCC_WARNINGS: BCPP: apply this to g++ also
* AcSplit/CF_GCC_WARNINGS: ci-aclocal generated header
2002-11-17
* AcSplit/AM_GNU_GETTEXT, AcSplit/CF_BUNDLED_INTL:
ci-aclocal generated header
* AcSplit/CF_BUNDLED_INTL:
lynx: add a 2nd param so we can have --enable-nls option rather than --disable-nls
* AcSplit/AM_GNU_GETTEXT:
lynx: pass a 4th parameter to AM_WITH_NLS, to make it use --enable-nls
2002-11-12
* AcSplit/CF_NCURSES_CPPFLAGS:
check for $cf_ncu_hdr_root equal to ncursesw also
* AcSplit/CF_XOPEN_CURSES: ci-aclocal generated header
* AcSplit/CF_XOPEN_CURSES:
tin: if we cannot compile even after adding _XOPEN_SOURCE_EXTENDED, mark it
as "unknown", so we can check this case later when we're sure it should work.
2002-11-10
* AcSplit/CF_WITH_CURSES_DIR: RCS_BASE
* AcSplit/CF_HEADER_PATH, AcSplit/CF_LIBRARY_PATH,
AcSplit/CF_WITH_CURSES_DIR:
ci-aclocal generated header
* AcSplit/CF_NCURSES_CPPFLAGS, AcSplit/CF_NCURSES_LIBS:
tin: add/use --with-curses-dir option
* AcSplit/CF_HEADER_PATH, AcSplit/CF_LIBRARY_PATH:
tin: rewrite/simplify using CF_SUBDIR_PATH
2002-11-09
* AcSplit/CF_SUBDIR_PATH: RCS_BASE
* AcSplit/AM_WITH_NLS: ci-aclocal generated header
* AcSplit/AM_WITH_NLS:
reintegrated with tin, add 4th param to allow tin to have --disable-nls,
and lynx to have --enable-nls
* AcSplit/AM_MULTIBYTE_ABLE: RCS_BASE
* AcSplit/CF_FIND_IPV6_LIBS: ci-aclocal generated header
* AcSplit/CF_FIND_IPV6_LIBS: tin: change "ipv6" to "IPv6"
* AcSplit/CF_DEFAULT_SHELL:
tin: urs added dash, csh, tcsh, as well as a check for absolute path
2002-11-07
* config.guess: 2002-10-21 (from tin 1.5.15)
2002-10-27
* AcSplit/AM_LC_MESSAGES, AcSplit/AM_PATH_PROG_WITH_TEST:
change # to dnl since ac2.13 would emit those after AC_REVISION
* AcSplit/jm_GLIBC21: change # to dnl to work properly with ac2.13
* AcSplit/AM_ICONV, AcSplit/AM_LANGINFO_CODESET, AcSplit/AM_LC_MESSAGES,
AcSplit/AM_PATH_PROG_WITH_TEST, AcSplit/CF_COLOR_CURSES,
AcSplit/CF_CURSES_CPPFLAGS, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_DEFINE_PROG, AcSplit/CF_FANCY_CURSES, AcSplit/CF_LASTLOG,
AcSplit/CF_NCURSES_VERSION, AcSplit/CF_PDCURSES_X11,
AcSplit/CF_TTYTYPE, AcSplit/CF_UTMP, AcSplit/CF_UTMP_UT_HOST,
AcSplit/CF_UTMP_UT_NAME, AcSplit/CF_UTMP_UT_SESSION,
AcSplit/CF_UTMP_UT_XTIME, AcSplit/jm_GLIBC21:
ci-aclocal generated header
* AcSplit/CF_UTF8_LIB:
lynx: add AC_DEFINE for HAVE_LIBUTF8_H to work with ncursesw
* AcSplit/AM_WITH_NLS:
lynx: change # to dnl so we don't get those lines after AC_REVISION
lynx: change option to --enable-nls rather than --disable-nls
* AcSplit/CF_BUNDLED_INTL:
lynx: if we're making the ./intl directory, define HAVE_LIBGETTEXT_H
lynx: define GT_YES and GT_NO symbols which were present in older versions of this macro
* AcSplit/AM_GNU_GETTEXT: lynx: put a dnl on the end
* AcSplit/AM_ICONV, AcSplit/AM_LANGINFO_CODESET,
AcSplit/CF_COLOR_CURSES, AcSplit/CF_CURSES_CPPFLAGS,
AcSplit/CF_CURSES_LIBS, AcSplit/CF_DEFINE_PROG,
AcSplit/CF_FANCY_CURSES, AcSplit/CF_LASTLOG, AcSplit/CF_LOCALE,
AcSplit/CF_NCURSES_VERSION, AcSplit/CF_PDCURSES_X11,
AcSplit/CF_TTYTYPE, AcSplit/CF_UTMP, AcSplit/CF_UTMP_UT_HOST,
AcSplit/CF_UTMP_UT_NAME, AcSplit/CF_UTMP_UT_SESSION,
AcSplit/CF_UTMP_UT_XTIME:
lynx: tidy up comments
2002-10-09
* AcSplit/CF_X_ATHENA_CPPFLAGS, AcSplit/CF_X_ATHENA_LIBS: RCS_BASE
* AcSplit/CF_ANSI_QSORT, AcSplit/CF_CHECK_FD_SET,
AcSplit/CF_HEADER_RESOURCE, AcSplit/CF_HEADER_SELECT,
AcSplit/CF_MISSING_CHECK, AcSplit/CF_TYPE_OUTCHAR,
AcSplit/CF_X_ATHENA_CPPFLAGS:
ci-aclocal generated header
* AcSplit/CF_ANSI_QSORT, AcSplit/CF_CHECK_FD_SET,
AcSplit/CF_HEADER_RESOURCE, AcSplit/CF_HEADER_SELECT,
AcSplit/CF_MISSING_CHECK, AcSplit/CF_TYPE_OUTCHAR:
vile: change "#if" to "#ifdef"
* AcSplit/CF_X_ATHENA:
vile: split-out CF_X_ATHENA_CPPFLAGS and CF_X_ATHENA_LIBS
2002-10-07
* AcSplit/CF_CHAR_DEVICE:
change sed delimiter from '@' to '%' for consistency (vile)
* AcSplit/CF_CHAR_DEVICE: ci-aclocal generated header
2002-09-22
* AcSplit/AM_WITH_NLS:
lynx: uses lowercase makefile.in, not Makefile.in
2002-09-21
* AcSplit/CF_MAN_PAGES:
ncurses: put the edit_man.sed script back in the "local" directory; otherwise
I cannot generate html-manpages
* AcSplit/CF_LIB_RULES: use '%' rather than '@' for sed-delimiter
2002-09-17
* AcSplit/CF_PATH_SYNTAX, AcSplit/CF_SLANG_LIBS, AcSplit/CF_SRAND:
ci-aclocal generated header
* AcSplit/CF_NCURSES_VERSION, AcSplit/CF_PATH_SYNTAX,
AcSplit/CF_SLANG_LIBS:
lynx: change sed delimiter to '%'
* AcSplit/CF_SRAND: lynx; change sed-delimiter to '%'
2002-09-07
* AcSplit/CF_MAN_PAGES:
the edit_man.sed script lives in the man-directory
2002-09-05
* config.sub: 2002-09-05 (from tin 1.5.15)
2002-08-31
* config.guess, config.sub: snapshot, for ncurses pre-release
2002-07-22
* AcSplit/CF_CURSES_SETBEGYX: ci-aclocal generated header
2002-07-19
* AcSplit/CF_VERSION_INFO, AcSplit/CF_WITH_LIBTOOL,
AcSplit/CF_WITH_WARNINGS:
RCS_BASE
* AcSplit/CF_VERSION_INFO: ci-aclocal generated header
2002-07-05
* AcSplit/CF_GCC_WARNINGS:
ded: add -Wundef since I want to use Tru64's strict checking, which is the same.
2002-06-29
* mkdirs.sh: update from ncurses:
oops - can't really use a range in a case-statement since locale stuff may
break it. Modified to be an explicit set of characterss.
2002-04-27
* AcSplit/CF_LIB_RULES:
fix an oops (I'd broken --with-termlib to make ccygwin dll's work)
2002-04-26
* config.sub: 2002/4/26
2002-04-21
* AcSplit/CF_HEADER_PATH:
testing on BeOS, I found that my checks for $HOME were looking in 'lib'
rather than 'include' (oops).
2002-04-17
* AcSplit/AM_GNU_GETTEXT: tin: from gettext 0.10.40
I merged in the rename of mkinstalldirs to mkdirs.sh
Do not see the "po" directory referenced here - it's moved to a different macro
* AcSplit/AM_PATH_PROG_WITH_TEST: tin: from gettext 0.10.40
I merged in my CF_PATHSEP change
* AcSplit/AC_ISC_POSIX, AcSplit/AM_ICONV, AcSplit/AM_LANGINFO_CODESET,
AcSplit/jm_GLIBC21:
RCS_BASE
* AcSplit/AC_ISC_POSIX, AcSplit/CF_BUNDLED_PCRE:
ci-aclocal generated header
* AcSplit/CF_BUNDLED_PCRE:
tin: Urs changed this so --with-pcre would simply add -lpcre to libs
* AcSplit/AM_LC_MESSAGES: tin: from gettext 0.10.40
* AcSplit/AM_WITH_NLS: tin: from gettext 0.10.40
this macro is the product of a sick mind.
2002-04-16
* AcSplit/CF_GMTOFF:
change to use the ifdef's from AC_HEADER_TIME (but still include <sys/types.h>)
* AcSplit/CF_GMTOFF, AcSplit/CF_RCS_SCCS: ci-aclocal generated header
* AcSplit/CF_RCS_SCCS: look in /usr/local/libexec/cssc for CSSC clone
2002-03-20
* config.guess: tin: 2002-03-20
2002-03-19
* config.sub: tin: 2002-03-07
2002-02-23
* AcSplit/CF_MAN_PAGES: add substitution for NCURSES_OSPEED
* AcSplit/CF_BOOL_SIZE, AcSplit/CF_DIRS_TO_MAKE, AcSplit/CF_OBJ_SUBDIR:
ci-aclocal generated header
* AcSplit/CF_LIB_RULES: fixes for cygwin dll's
* AcSplit/CF_DIRS_TO_MAKE:
ncurses: better logic (filter out duplicates)
* AcSplit/CF_BOOL_SIZE:
ncurses: if $cf_cv_type_of_bool is empty, make it 'unknown'
* AcSplit/CF_LINK_DATAONLY:
set $BROKEN_LINKER in addition to the #define, so we can use it in substitutions
* AcSplit/CF_OBJ_SUBDIR: special case for cygwin dll's
* AcSplit/CF_SHARED_OPTS: special cases for cygwin dll's
2002-01-23
* config.guess: 2002/1/23
2002-01-19
* AcSplit/CF_LIB_SUFFIX, AcSplit/CF_SHARED_OPTS:
add case for cygwin dll
* AcSplit/CF_LIB_SUFFIX: ci-aclocal generated header
* AcSplit/CF_LIB_RULES: first cut of changes for cygwin dll
* AcSplit/CF_LIBUTF8, AcSplit/CF_MBSTATE_T: RCS_BASE
* AcSplit/CF_MANPAGE_RENAMES:
add check for file in $srcdir/man, adjust param if found (makes this easier to use)
rename variable from $cf_manpage_renames to MANPAGE_RENAMES, for consistency
* AcSplit/CF_LIBUTF8, AcSplit/CF_MANPAGE_FORMAT,
AcSplit/CF_MANPAGE_RENAMES, AcSplit/CF_MANPAGE_SYMLINKS,
AcSplit/CF_MANPAGE_TBL, AcSplit/CF_MBSTATE_T, AcSplit/CF_SRC_MODULES:
ci-aclocal generated header
* AcSplit/CF_MAN_PAGES:
oops - refix my Debian fix from 2001/3/19, which dropped the manpage-renames.
also, rename $cf_manpage_xxx to $MANPAGE_xxx variables for consistency.
* AcSplit/CF_SRC_MODULES:
remove -L$(libdir), since someone pointed out it's no longer needed for HPUX
* AcSplit/CF_MANPAGE_FORMAT:
rename variable from $cf_manpage_form to MANPAGE_FORMAT, for consistency
* AcSplit/CF_MANPAGE_SYMLINKS:
rename variable from $cf_manpage_symlinks to MANPAGE_SYMLINKS
* AcSplit/CF_MANPAGE_TBL:
rename variable from $cf_manpage_tbl to $MANPAGE_TBL, for consistency
2002-01-13
* AcSplit/CF_CURSES_TERMCAP: ci-aclocal generated header
* AcSplit/CF_CURSES_TERMCAP:
corrected renamed config variable: $cf_cv_have_term_h -> $cf_cv_term_header
2002-01-12
* AcSplit/CF_LIB_RULES:
ncurses - remove conflicting files that may be gzip'd
* AcSplit/CF_CURSES_TERM_H: add check for <ncurses/term.h>
* AcSplit/CF_CURSES_TERMCAP:
changed $cf_cv_have_term_h variable to be normally a header filename, so
I can incorporate <ncurses/term.h>
* AcSplit/CF_CURSES_TERM_H: ci-aclocal generated header
* AcSplit/CF_TYPE_OUTCHAR:
expanded comment, added <ncurses/ncurses.h> and <ncurses/term.h>
2002-01-02
* config.sub: 2002/1/2
2002-01-01
* AcSplit/CF_UTMP_UT_HOST: delete a blank line on the end
2001-12-31
* AcSplit/CF_PDCURSES_X11:
use CF_CHECK_CFLAGS rather than CF_ADD_CFLAGS
2001-12-30
* AcSplit/CF_ADD_CFLAGS:
delay updating the $CFLAGS and $CPPFLAGS til the end so I can make the
macro conditionally verbose
* AcSplit/CF_CHECK_CFLAGS: RCS_BASE
* AcSplit/CF_CHECK_CFLAGS, AcSplit/CF_X_TOOLKIT:
ci-aclocal generated header
* AcSplit/CF_X_TOOLKIT:
vile: use CF_CHECK_CFLAGS rather than CF_ADD_CFLAGS
* AcSplit/CF_RESTARTABLE_PIPEREAD: ci-aclocal generated header
* AcSplit/CF_RESTARTABLE_PIPEREAD: vile: use AC_CACHE_VAL
* AcSplit/CF_KILLPG, AcSplit/CF_MAKEFLAGS: ci-aclocal generated header
* AcSplit/CF_CURSES_TERMCAP: use AC_CACHE_CHECK
* AcSplit/CF_CURSES_TERM_H, AcSplit/CF_KILLPG, AcSplit/CF_MAKEFLAGS,
AcSplit/CF_TYPE_OUTCHAR:
vile: use CF_CACHE_CHECK
* AcSplit/CF_CHECK_ERRNO:
change to use AC_CACHE_CHECK, which ultimately meant I cannot use eval's to
support a loop
* AcSplit/CF_CHECK_ERRNO, AcSplit/CF_CHECK_EXTERN_DATA,
AcSplit/CF_SYS_ERRLIST:
ci-aclocal generated header
* AcSplit/CF_SYS_ERRLIST:
don't use a loop on CF_CHECK_ERRNO (since I rewrote it to avoid using eval,
for vile)
* AcSplit/CF_CHECK_EXTERN_DATA:
vile: eliminated eval's used to support loop, since that got in the way of
using AC_CACHE_CHECK
* AcSplit/CF_ANSI_CC_CHECK: ci-aclocal generated header
* AcSplit/CF_ANSI_CC_CHECK, AcSplit/CF_ANSI_QSORT:
vile: use AC_CACHE_CHECK
2001-12-27
* AcSplit/CF_UTMP_UT_NAME: RCS_BASE
* AcSplit/CF_NGROUPS, AcSplit/CF_SYSV_UTMP, AcSplit/CF_UTMP_UT_XSTATUS:
ci-aclocal generated header
* AcSplit/CF_NGROUPS:
lynx: ac252 noted I should move an AC_DEFINE outside an AC_CACHE_CHECK
* AcSplit/CF_SYSV_UTMP, AcSplit/CF_UTMP_UT_HOST,
AcSplit/CF_UTMP_UT_SESSION, AcSplit/CF_UTMP_UT_XSTATUS,
AcSplit/CF_UTMP_UT_XTIME:
lynx: ac252 tells me not to require CF_UTMP
* AcSplit/CF_UTMP:
lynx: split-out CF_UTMP_UT_NAME so this compiles with ac252
2001-12-26
* AcSplit/CF_CURSES_CPPFLAGS:
lynx: ensure that we found a curses.h (or ncurses.h), since some people
don't know the difference between runtime and development
2001-12-18
* AcSplit/CF_CHECK_CACHE, AcSplit/CF_FUNC_VSSCANF,
AcSplit/CF_SUBST_NCURSES_VERSION:
ci-aclocal generated header
* AcSplit/CF_CHECK_CACHE: fix quoting for autoconf 2.52
* AcSplit/CF_SUBST_NCURSES_VERSION: make this work with autoconf 2.52
* AcSplit/CF_FUNC_VSSCANF:
ncurses: fix an incorrect value for cf_cv_func_vsscanf
2001-12-15
* config.sub: 2001-12-10
* config.guess: 2001-12-13
2001-12-11
* AcSplit/CF_REGEXPR_H_FUNCS, AcSplit/CF_REGEX_H_FUNCS:
ci-aclocal generated header
* AcSplit/CF_REGEXPR_H_FUNCS, AcSplit/CF_REGEX_H_FUNCS:
td_lib: add checks for cygwin variation
2001-12-09
* AcSplit/CF_PROG_PERL, AcSplit/CF_WITH_PATHLIST:
ci-aclocal generated header
* AcSplit/CF_MISSING_CHECK:
vile: escape the double-quotes in the AC_MSG_CHECKING message, which makes
this produce the same (quoted) message with autoconf 2.13 as with 2.52
* AcSplit/CF_PROG_PERL:
vile: modified the use of brackets to work with autoconf 2.52, which doesn't
allow an unbalanced square bracket to happen.
* AcSplit/CF_WITH_PATHLIST:
vile: omit some double quotes which don't appear to be needed, which don't
work when I generate the configure script with autoconf 2.52
2001-12-08
* AcSplit/CF_LIB_RULES, AcSplit/CF_SHARED_OPTS:
Jason Evan's changes for ncurses on Mac OS X
* AcSplit/CF_FUNC_VSSCANF: RCS_BASE
* AcSplit/CF_MANPAGE_RENAMES:
use explicit 'sh' to invoke 'mkinstalldirs'
* AcSplit/CF_MAN_PAGES: use explicit 'sh' to invokoe mkinstalldirs
2001-12-01
* AcSplit/CF_LIB_SUFFIX, AcSplit/CF_SHARED_OPTS: add case for darwin
* AcSplit/CF_BOOL_DECL: add check for stdbool.h
* AcSplit/CF_BOOL_SIZE: add check to accommodate stdbool.h
* AcSplit/CF_SUBST_NCURSES_VERSION:
add substitute for cf_cv_header_stdbool_h
* AcSplit/CF_BOOL_DECL, AcSplit/CF_CPP_VSCAN_FUNC:
ci-aclocal generated header
* AcSplit/CF_LIB_RULES, AcSplit/CF_SRC_MODULES:
mods to make generating terminfo database optional
* AcSplit/CF_CPP_VSCAN_FUNC:
ncurses: omit the ifdef that says we only know about g++
correct test, which was for vscanf(), not vsscanf()
2001-11-18
* AcSplit/CF_WIDEC_CURSES: add a check to define HAVE_MBSTATE_T
* AcSplit/CF_PATH_PROG, AcSplit/CF_WIDEC_CURSES:
ci-aclocal generated header
* AcSplit/CF_PATH_PROG:
lynx: don't define anything if we're using full-pathnames and we didn't find anything
* AcSplit/CF_PATH_SYNTAX: make the AC_ERROR optional, for lynx
2001-11-08
* AcSplit/CF_UTF8_LIB, AcSplit/CF_WIDEC_CURSES: RCS_BASE
* AcSplit/CF_MKSTEMP: ci-aclocal generated header
* AcSplit/CF_MKSTEMP:
improved check by doing it twice, to see if names are unique
2001-10-18
* AcSplit/CF_TERMCAP_LIBS: ci-aclocal generated header
* AcSplit/CF_NCURSES_CPPFLAGS:
lynx: add an optional parameter so I can implement --with-screen=ncursesw
* AcSplit/CF_NCURSES_LIBS:
lynx: add optional parameter so I can implement --with-screen=ncursesw
* AcSplit/CF_TERMCAP_LIBS:
lynx: allow parameter to be ncurses or ncursesw
* AcSplit/CF_HEADER_PATH:
lynx: put the /usr/local stuff first, to more closely match gcc's nonstandard
include rules, so we lessen the chance of picking up the wrong version of
curses.h
I did that to make <ncursesw/curses.h> work ok.
* AcSplit/CF_LIBRARY_PATH:
lynx: put the /usr/local stuff first, to more closely match gcc's rules.
This is mainly for symmetry with CF_HEADER_PATH, but may be needed to
pick up the "right" version of ncurses/curses
2001-10-11
* AcSplit/CF_FIND_IPV6_TYPE: ci-aclocal generated header
* AcSplit/CF_DEFAULT_SHELL: tin: Urs' improvements
* AcSplit/CF_FIND_IPV6_TYPE:
tin: oops Andreas Metzler <tin-dev@downhill.at.eu.org>
noticed spurious dnl's in the output which have been there since this macro
was written
2001-09-23
* AcSplit/CF_MKSTEMP: RCS_BASE
2001-09-20
* AcSplit/CF_PATH_SYNTAX:
vile: allow leading double-backslash for DOS-style pathnames
* AcSplit/CF_WITH_PATHLIST:
vile: fix an error in $IFS setting, and add an option to make it simpler
to force backslashes to be escaped when we're defining a string.
2001-08-15
* AcSplit/CF_SRAND:
lynx 2.8.5dev.2 : change order of srand48 vs srandom
* AcSplit/AM_WITH_NLS:
lynx 2.8.5dev.2 : fix an oops in the order of -lintl library to avoid
conflict with GNU gettext vs Solaris
2001-07-22
* AcSplit/CF_WITH_PROGRAM: ci-aclocal generated header
* AcSplit/CF_WITH_PROGRAM:
get rid of changequote() because it's broken in autoconf 2.50 & up (tin)
* AcSplit/CF_BUNDLED_PCRE, AcSplit/CF_DEFAULT_SHELL,
AcSplit/CF_DEFINE_STRING:
RCS_BASE
* AcSplit/CF_PATH_SYNTAX: allow "." as a pathname
fix quoting in the AC_ERROR parameter
(for tin).
* AcSplit/CF_CHECK_1_DECL, AcSplit/CF_DEFINE_STRING:
ci-aclocal generated header
* AcSplit/CF_CHECK_1_DECL: spelling - tin
2001-07-18
* AcSplit/CF_MAKEFLAGS:
oops (comparing with tin, I lose the SHELL assignment on the last checkin,
somehow)
2001-07-17
* AcSplit/CF_CURSES_FUNCS: ci-aclocal generated header
* AcSplit/CF_CURSES_FUNCS:
tweak an ifdef to avoid picking up Solaris' term.h when we're configuring
for ncurses (if we have <ncurses/curses.h>, we should have <ncurses/term.h>,
except of course on broken systems such as cygwin ;-).
--lynx
2001-07-15
* AcSplit/CF_MAKE_AR_RULES, AcSplit/CF_MAKE_INCLUDE,
AcSplit/CF_TOP_SRCDIR:
ci-aclocal generated header
* AcSplit/CF_MAKE_INCLUDE, AcSplit/CF_TOP_SRCDIR: override CDPATH
* AcSplit/CF_MAKE_AR_RULES: overrides for SHELL and CDPATH
(ded)
2001-07-11
* AcSplit/CF_STRERROR: RCS_BASE
* AcSplit/CF_STRERROR: ci-aclocal generated header
2001-06-26
* config.guess: tin 1.5.9
2001-06-22
* install-sh:
copied from install.sh, which uses ":" rather than "true", and allows a
tab in $IFS
* install.sh: tin: trim trailing blanks
2001-06-16
* AcSplit/CF_PROG_EXT:
add PROG_EXT in quotes, for ncurses' programs that check if they're renamed
* AcSplit/CF_MAKEFLAGS:
aha: set SHELL in the test-makefile to make warnings go away
* AcSplit/CF_CPP_VSCAN_FUNC: ugh: add strstream_cast case
2001-06-09
* config.sub: tin 1.5.9
2001-06-03
* AcSplit/CF_NCURSES_VERSION:
lynx: make this a little more paranoid wrt cleaning up tempfiles
2001-05-27
* AcSplit/CF_FIX_SLASHES: RCS_BASE
* AcSplit/CF_POSIX_JC:
autoconf 2.50 warned about an AC_DEFINE in a questionable location
* AcSplit/CF_FIX_SLASHES, AcSplit/CF_PATH_MAILER, AcSplit/CF_POSIX_JC:
ci-aclocal generated header
* AcSplit/CF_MAKEFLAGS:
fix a "no match" problem on OS/2 EMX by setting SHELL in the test-makefile.
That fixes a problem with the echo, which apparently did not strip single quotes
unless the shell was given.
* AcSplit/CF_PATH_MAILER: fix/workaround for OS/2 pathnames
2001-05-26
* AcSplit/CF_LOCKFILE: RCS_BASE
* AcSplit/CF_LOCKFILE: ci-aclocal generated header
2001-05-19
* AcSplit/CF_LIB_RULES:
fix for --disable-overwrite (it was only allowing overwrite of the first
library)
add "make depend" rule
* config.sub, config.guess: autoconf 2.50
2001-05-13
* AcSplit/CF_CURS_TOUCHLINE: RCS_BASE
* AcSplit/CF_CURS_TOUCHLINE: ci-aclocal generated header
* AcSplit/CF_NCURSES_VERSION:
define NCURSES as a side-effect (use for lynx)
* AcSplit/CF_CURSES_CPPFLAGS, AcSplit/CF_NCURSES_CPPFLAGS:
test for ncurses.h after curses.h
2001-05-02
* install.sh: better version, from tin 1.5.9, which cites X11R5
2001-04-28
* AcSplit/CF_X_FREETYPE: add XRENDERFONT ifdef
* AcSplit/CF_X_FREETYPE: ci-aclocal generated header
2001-04-21
* AcSplit/CF_DIRNAME:
ncurses: trim extra newlines, to avoid adding them in configure script
2001-04-18
* AcSplit/CF_CURSES_CPPFLAGS: oops - wrong $host_os case for sunos
* AcSplit/CF_CURSES_LIBS: oops - wrong $host_os value for sunos.
also, force -lcurses -ltermcap in that case (should incorporate real fixes)
2001-04-15
* AcSplit/CF_PROG_INSTALL: use CF_DIRNAME
* AcSplit/CF_SLANG_CPPFLAGS: ci-aclocal generated header
* AcSplit/CF_SLANG_CPPFLAGS: use CF_DIRNAME
* AcSplit/CF_NCURSES_VERSION: changed this to use AC_CACHE_CHECK
* AcSplit/CF_CURSES_LIBS: split-out the CF_CURSES_CPPFLAGS stuff
* AcSplit/CF_CURSES_CPPFLAGS: RCS_BASE
* AcSplit/CF_INCLUDE_PATH: ci-aclocal generated header
* AcSplit/CF_INCLUDE_PATH: use CF_DIRNAME
* AcSplit/CF_NCURSES_CPPFLAGS: nit
* AcSplit/CF_DIRNAME: RCS_BASE
* AcSplit/CF_NCURSES_CPPFLAGS:
ded: rewrite to handle the case where ncurses is not installed in any of
the standard locations, and we are using $CPPFLAGS to predefine its header
location. Also, change preference to look for ncurses.h rather than curses.h,
since it's less likely that we chose the wrong header in that case.
2001-04-07
* AcSplit/CF_CPP_VSCAN_FUNC:
ncurses: better check, don't need sstream header
* AcSplit/CF_CPP_VSCAN_FUNC: RCS_BASE
* AcSplit/CF_CPP_PARAM_INIT: ci-aclocal generated header
* AcSplit/CF_CPP_PARAM_INIT:
oops: why was I still checking if $CXX is 'yes'?
also, save/restore language to make this more reusable
2001-04-01
* AcSplit/CF_TERMCAP_LIBS:
change this to return a warning if not found, since slang (which uses this
macro) may not need it (lynx)
* AcSplit/CF_SRAND:
lynx 2.8.4 dev.18 - better check for range of rand(sic) functions
2001-03-18
* AcSplit/CF_MAN_PAGES: fix oops for Debian renames
2001-03-16
* config.guess: from autoconf 2.49e
2001-03-12
* config.sub: from autoconf 2.49e
2001-02-14
* AcSplit/CF_X_FREETYPE: RCS_BASE
* AcSplit/CF_IMAKE_CFLAGS:
cancel message from $CDPATH by making it an empty string.
* AcSplit/CF_IMAKE_CFLAGS: ci-aclocal generated header
2001-02-13
* config.sub, config.guess: 2000-2-13
2001-02-08
* AcSplit/CF_SRAND:
changed order - feedback indicated that the most portable code is the worst...
2001-02-03
* AcSplit/CF_MAN_PAGES:
eliminate changequote(), since it is broken in autoconf 2.49c
* AcSplit/CF_GNAT_VERSION:
eliminate changequote() since it is broken in autoconf 2.49c
* AcSplit/CF_SUBST_NCURSES_VERSION, AcSplit/CF_MANPAGE_FORMAT:
eliminate changequote(), since it is broken in autoconf 2.49c
2001-02-02
* AcSplit/CF_SRAND: RCS_BASE
* AcSplit/CF_UPPERCASE: autoconf 2.49c breaks changequote()
* AcSplit/CF_UPPERCASE: ci-aclocal generated header
* AcSplit/CF_SYSTYPE: autoconf 2.49c breaks changequote()
* AcSplit/CF_SYSTYPE: ci-aclocal generated header
* AcSplit/CF_SYSRELV: autoconf 2.49c breaks changequote()
* AcSplit/CF_SYSRELV: ci-aclocal generated header
* AcSplit/CF_PROG_INSTALL, AcSplit/CF_INCLUDE_PATH,
AcSplit/CF_LINK_DATAONLY, AcSplit/CF_XTERM_MODE:
autoconf 2.49c breaks changequote()
* AcSplit/CF_XTERM_MODE: ci-aclocal generated header
* AcSplit/CF_PATH_SYNTAX, AcSplit/CF_SLANG_CPPFLAGS,
AcSplit/CF_STRIP_O_OPT:
autoconf 2.49c breaks changequote()
* AcSplit/CF_STRIP_O_OPT: ci-aclocal generated header
* AcSplit/CF_X_TOOLKIT: autoconf 2.49c breaks changequote()
2001-01-31
* AcSplit/CF_CURSES_FUNCS:
braindead gcc 2.91.66 optimized away my test. nail it down.
2001-01-29
* AcSplit/CF_GCC_ATTRIBUTES, AcSplit/CF_GCC_WARNINGS,
AcSplit/CF_NCURSES_CPPFLAGS, AcSplit/CF_NCURSES_VERSION,
AcSplit/CF_UPPER:
autoconf-2.49c-breaks-changequote
* AcSplit/CF_UPPER: ci-aclocal generated header
2001-01-11
* AcSplit/CF_PATHSEP: allow system name to be os2-emx
* AcSplit/CF_PATHSEP: ci-aclocal generated header
* AcSplit/CF_LIB_PREFIX: allow system-name to be os2-emx
* AcSplit/CF_LIB_PREFIX: ci-aclocal generated header
* config.sub: Pavel Roskin's -os2 fix
* config.guess:
slightly newer version to get same-timestamp as config.sub for OS/2
* config.sub, config.guess: current, from subversions
2001-01-10
* AcSplit/CF_CURSES_LIBS:
oops: ensure that if we haven't checked for ncurses version, that this doesn't assume we found it
2001-01-05
* AcSplit/CF_CHAR_DEVICE: RCS_BASE
2001-01-01
* AcSplit/CF_SSL: RCS_BASE
* AcSplit/CF_SSL, AcSplit/CF_SOCKS, AcSplit/CF_SOCKS5:
ci-aclocal generated header
* AcSplit/CF_SOCKS:
lynx: expand use of the parameter to set $CPPFLAGS as well as $LIBS
* AcSplit/CF_SOCKS5:
lynx: expand use of the parameter to set $CPPFLAGS as well as $LIBS
also, add check to define INCLUDE_PROTOTYPES
2000-12-31
* AcSplit/CF_TERMIO_AND_CURSES: ci-aclocal generated header
* AcSplit/CF_FIND_IPV6_LIBS, AcSplit/CF_SOCKS5,
AcSplit/CF_TERMIO_AND_CURSES, AcSplit/CF_XOPEN_CURSES:
lynx: change $CFLAGS to $CPPFLAGS
* AcSplit/CF_PDCURSES_X11: lynx: use CF_ADD_CFLAGS
2000-12-28
* AcSplit/CF_ADD_CFLAGS: add gcc's -undef, -nostdinc options
add some logic to suppress repetition of previous token, which appears to
happen in xterm's configure script.
* AcSplit/CF_X_TOOLKIT:
use CF_ADD_CFLAGS rather than setting $CFLAGS directly
* AcSplit/CF_ANSI_CC_CHECK:
use CF_ADD_CFLAGS rather than setting $CFLAGS directly (I missed this on
the last change, would break on some machines that need $CPPFLAGS updated).
2000-11-18
* AcSplit/CF_CFG_DEFAULTS: ci-aclocal generated header
* AcSplit/CF_CFG_DEFAULTS:
ncurses: add cygwin to list of those installing ncurses under /usr
2000-11-04
* AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_WITH_PATHLIST: use CF_PATHSEP
* AcSplit/CF_ANSI_CC_CHECK:
just for tidyness, use CF_ADD_CFLAGS to distinguish between CFLAGS/CPPFLAGS
* AcSplit/CF_STRUCT_TERMIOS:
assume $CFLAGS may contain preprocessor options, but prefer $CPPFLAGS
* AcSplit/CF_STRUCT_TERMIOS: ci-aclocal generated header
* AcSplit/CF_TYPE_OUTCHAR: use $CPPFLAGS for preprocessor options
* AcSplit/CF_SIZECHANGE: ci-aclocal generated header
* AcSplit/CF_SIZECHANGE: use $CPPFLAGS for preprocessor options
* AcSplit/CF_PROG_EXT: split-out $CPPFLAGS for preprocessor options
* AcSplit/CF_MISSING_CHECK, AcSplit/CF_CURSES_TERMCAP,
AcSplit/CF_CURSES_LIBS:
use $CPPFLAGS for preprocessor options
2000-11-03
* AcSplit/CF_CHECK_HEADERS: ci-aclocal generated header
* AcSplit/CF_CHECK_HEADERS:
oops - don't use AC_TRY_COMPILE - use AC_TRY_CPP
* AcSplit/AM_PATH_PROG_WITH_TEST: add dependency on CF_PATHSEP
* AcSplit/CF_CHECK_HEADERS: RCS_BASE
2000-11-01
* AcSplit/CF_PROGRAM_FULLPATH, AcSplit/CF_PROGRAM_PREFIX:
ci-aclocal generated header
* AcSplit/CF_PROGRAM_PREFIX:
use CF_PATHSEP to make this work on OS/2 EMX
* AcSplit/CF_PROGRAM_FULLPATH:
use CF_PROG_EXT to make this search for files with exe suffix if needed
* AcSplit/CF_FIND_TDLIB: ci-aclocal generated header
* AcSplit/CF_FIND_TDLIB:
use CF_LIB_PREFIX to make this work on OS/2 EMX
* AcSplit/CF_PATHSEP: RCS_BASE
* AcSplit/CF_LIB_PREFIX:
add special case for OS/2's uname so I don't need to distribute config.guess
or config.sub with ded
* AcSplit/CF_PROGRAM_FULLPATH: use CF_PATHSEP to make this work on OS/2
2000-10-29
* AcSplit/CF_GNU_SOURCE: ci-aclocal generated header
* AcSplit/CF_GNU_SOURCE, AcSplit/CF_X_ATHENA:
xterm: change $CFLAGS to $CPPFLAGS
2000-10-25
* config.guess: ncurses: add-back os2 fix
* config.sub: ncurses: add-back the os2 fix
* config.sub: CVS version 2000-10-25
* config.guess: CVS 2000-10-23 version
* AcSplit/CF_SLANG_LIBS: lynx: add check for -lvideo on OS/2 EMX
* AcSplit/CF_PATH_SYNTAX:
lynx: correct missing changequote() around a bracket expression,
and make the error message clearer by adding the pathname which fails.
2000-10-20
* AcSplit/CF_MANPAGE_TBL: RCS_BASE
* AcSplit/CF_LIB_TYPE: ci-aclocal generated header
* AcSplit/CF_LIB_RULES: ncurses: add --without-curses-h
* AcSplit/CF_LIB_SUFFIX, AcSplit/CF_LIB_TYPE, AcSplit/CF_OBJ_SUBDIR:
ncurses: add model for libtool
* AcSplit/CF_SRC_MODULES: ncurses: add support for libtool
* AcSplit/CF_MAN_PAGES: ncurses: add support for manpage-tbl option
modified edit_man.sh to allow reuse for "make -f dist.mk manhtml"
2000-10-18
* AcSplit/CF_RECHECK_FUNC: ci-aclocal generated header
* AcSplit/CF_RECHECK_FUNC: ensure we only add a library to list $3 once
2000-10-07
* AcSplit/CF_SHARED_OPTS: add -rpath support for IRIX
* AcSplit/CF_CURSES_CHTYPE: ci-aclocal generated header
* AcSplit/CF_CURSES_CHTYPE: use $cf_cv_ncurses_header variable
2000-10-04
* AcSplit/CF_MAKE_TAGS, AcSplit/CF_MIXEDCASE_FILENAMES: RCS_BASE
* AcSplit/CF_LIB_RULES: add install.man rule for tack.
add $(AUTO_SRC) dependency for install.includes
do a basename of install-target, in case the given $SRC is not in the current directory
* AcSplit/CF_MAKE_TAGS, AcSplit/CF_MIXEDCASE_FILENAMES,
AcSplit/CF_SYS_TIME_SELECT, AcSplit/CF_TYPEOF_CHTYPE:
ci-aclocal generated header
* AcSplit/CF_CHECK_CACHE:
ncurses: allow a parameter so we can specify use of AC_CANONICAL_SYSTEM
rather than AC_CANONICAL_HOST, for cross-compiles.
* AcSplit/CF_CPP_PARAM_INIT:
ncurses: change to testing $CXX if it is "yes", to avoid conflict with people's
environments when they set $CXX to the actual compiler name.
* AcSplit/CF_ADA_INCLUDE_DIRS:
ncurses: change to testing if $GCC is "yes", rather than nonnull, avoid problems with people's environment settings.
* AcSplit/CF_TYPEOF_CHTYPE:
ncurses: change to using "#ifdef" thoughout, rather than "#if"
* AcSplit/CF_BOOL_SIZE:
ncurses: change to using "#ifdef" throughout rather than "#if"
* AcSplit/CF_SIZECHANGE, AcSplit/CF_SYS_TIME_SELECT:
ncurses: change to using "#ifdef" throughout, rather than "#if"
* AcSplit/CF_SHARED_OPTS:
ncurses: eliminate a redundant case for HPUX, check $GCC for equality to
"yes", and revise rpath support to (try to) eliminate relative paths.
* AcSplit/CF_MAN_PAGES:
ncurses: fix for configure --srcdir, since some files to install would be
in the current directory, and some would be in the $srcdir directory
* AcSplit/CF_GCC_ATTRIBUTES:
ncurses: test $GCC for "yes" rather than nonnull, since some people set the
compiler name that way.
* AcSplit/CF_INCLUDE_DIRS:
ncurses: test for $GCC set to "yes", in case someone set it to the compiler name.
* AcSplit/CF_GCC_WARNINGS:
ncurses: test for $GCC set to "yes", in case someone uses it to set the compiler name.
2000-09-28
* AcSplit/CF_FUNC_GETADDRINFO: ci-aclocal generated header
* AcSplit/CF_FUNC_GETADDRINFO: lynx 2.8.4dev.11
2000-08-31
* AcSplit/CF_CURSES_LIBS:
lynx: my juggling of ncurses/curses search order exposed a hole. If I
specified the curses library in $LIBS, that might depend on -ltermcap,
which then would not be found. Fixed by doing an AC_TRY_LINK explicitly.
2000-08-24
* AcSplit/CF_FUNC_GETADDRINFO:
lynx: add AC_DEFINE's for HAVE_GAI_STRERROR and HAVE_GETADDRINFO
* config.guess, config.sub: ncurses: add cases for OS/2 EMX
2000-08-19
* AcSplit/CF_LIB_RULES: change $(INSTALL_PREFIX) to $(DESTDIR)
* AcSplit/CF_MAN_PAGES: from ncurses 20000812 final - a few fixes
2000-08-12
* AcSplit/CF_MAN_PAGES:
parameterized the list of program-names to be transformed
* AcSplit/CF_BOOL_SIZE: boo! autoconf 2.49a changes meaning of AC_LANG
* AcSplit/CF_CGETENT, AcSplit/CF_FUNC_MEMMOVE, AcSplit/CF_ISASCII,
AcSplit/CF_STDCPP_LIBRARY, AcSplit/CF_STRUCT_SIGACTION,
AcSplit/CF_TYPE_SIGACTION:
ci-aclocal generated header
* AcSplit/CF_SIZECHANGE:
ncurses: autoconf 2.49a warns about AC_DEFINE within AC_CACHE_CHECK
* AcSplit/CF_LINK_FUNCS:
ncurses: autoconf 2.49a warns about AC_DEFINE within an AC_CACHE_CHECK
* AcSplit/CF_CGETENT, AcSplit/CF_FUNC_MEMMOVE, AcSplit/CF_GPP_LIBRARY,
AcSplit/CF_ISASCII, AcSplit/CF_LIB_RULES, AcSplit/CF_LINK_DATAONLY,
AcSplit/CF_SHARED_OPTS, AcSplit/CF_SRC_MODULES,
AcSplit/CF_STDCPP_LIBRARY:
ncurses: quote test-parameters
* AcSplit/CF_BOOL_DECL: ncurses: quote test-parameters
boo! autoconf 2.49a changes meaning of AC_LANG
* AcSplit/CF_STRUCT_SIGACTION, AcSplit/CF_STRUCT_TERMIOS,
AcSplit/CF_SYS_TIME_SELECT, AcSplit/CF_TYPE_SIGACTION:
quote 'test' parameter
2000-08-03
* config.guess, config.sub: lynx2.8.4dev.7
2000-07-30
* AcSplit/CF_TYPE_OUTCHAR: add a log message
* AcSplit/CF_BUNDLED_INTL:
tin: add a clause for SUB_MAKEFILE if $nls_cv_use_gnu_gettext is set.
2000-07-29
* AcSplit/CF_CURSES_LIBS:
dialog: suppress redundant check for $cf_curs_lib vs $cf_term_lib
* AcSplit/CF_ARG_MSG_ENABLE: RCS_BASE
* AcSplit/CF_ARG_MSG_ENABLE: ci-aclocal generated header
* AcSplit/CF_UPPER: improve using explicit range
2000-07-22
* AcSplit/CF_FUNC_TERMIOS: RCS_BASE
* AcSplit/CF_SIZECHANGE: add dependency on CF_STRUCT_TERMIOS
* AcSplit/CF_STRUCT_TERMIOS: add header checks
* AcSplit/CF_FUNC_TERMIOS: ci-aclocal generated header
2000-07-20
* AcSplit/CF_X_ATHENA: add /usr/local to search path, for OpenBSD
2000-07-19
* AcSplit/CF_CURSES_TERMCAP:
tin - define NEED_TERM_H if we can really use term.h rather than termcap.h, etc.
* AcSplit/CF_ARG_WITH, AcSplit/CF_PATH_EDITOR:
ci-aclocal generated header
* AcSplit/CF_PATH_EDITOR:
provide a more reasonable default for the help message
* AcSplit/CF_PATH_MAILER:
provide a more reasonable default for the help-message
* AcSplit/CF_ARG_WITH:
tin - added 5th parameter to provide _some_ default value for help message
2000-07-17
* config.sub: lynx - os390 changes
2000-07-16
* config.guess: lynx - os390 changes
2000-07-15
* AcSplit/CF_MAN_PAGES: mods to make this work with program-prefix
(very ugly)
2000-07-11
* AcSplit/CF_CURSES_LIBS:
lynx: if we've checked for ncurses version and found ncurses, prefer -lncurses
to avoid problems with misconfigured system where ncurses headers are visible,
but not library
2000-07-07
* AcSplit/CF_LIB_SUFFIX:
ncurses: move assumptions about *bsd using REL_VERSION suffix to CF_SHLIB_OPTS
* AcSplit/CF_LIB_RULES:
ncurses: use $cf_cv_shlib_version and $cf_cv_do_symlinks to handle the
special case of constructing $cf_suffix. Most of the shared-library
suffix logic is in mk-1st.awk, but there's a list of dependencies that
are generated here.
* AcSplit/CF_SHARED_OPTS:
ncurses: various fixes, mostly because I broke something before 5.0 which
dealt with the rel/abi symbolic links. This rearranges things so the
library name passed to mk-1st.awk doesn't have the version number tacked
onto the ".so" suffix - the main problem.
Also moved some logic out of CF_LIB_SUFFIX to work with that rearrangement.
2000-07-01
* AcSplit/CF_SHARED_OPTS: quote a '$' so m4 doesn't eat it
* AcSplit/CF_MAN_PAGES: sort/uniq the output from manlinks.sed
2000-06-29
* AcSplit/AM_WITH_NLS:
Julian Coleman <J.D.Coleman@newcastle.ac.uk> noticed that if the intl
directory is missing, this macro still tries to make the intl/libintl.h
link. Found that AC_LINK_FILES is evaluated at autoconf time, while I
want to suppress the links at configure time. Did this by resetting the
variables used as parameters to AC_LINK_FILES.
2000-06-20
* AcSplit/CF_CHECK_ERRNO: John Coyne <John.Coyne@CTBTO.ORG>
noted a syntax error (single quotes were not at the same level of brackets)
2000-06-18
* config.sub: based on 2000-06-13 snapshot
* config.guess: based on 2000-06-13 snapshot from gnu.org
2000-06-17
* AcSplit/CF_SHARED_OPTS:
make -fPIC vs -fpic choice at configure time based on whether the compiler
recognizes it
* AcSplit/CF_LIB_RULES:
oops - modify headers.sh so I don't write on the build directory.
2000-06-11
* AcSplit/CF_FIND_IPV6_LIBS, AcSplit/CF_FIND_IPV6_TYPE:
special case for 'solaris', from tin 1.5.5
* config.sub: merge my changes: hurd and os390
* config.sub: version distributed with pcre 3.2
* config.guess:
re-merge changes for os390 (note that one case is shadowed: "*:OS/390:*:*"),
elf32ppc (add asterisk)
and elf64_ia64
* config.guess: version with pcre 3.2
2000-06-10
* AcSplit/CF_MANPAGE_SYMLINKS: RCS_BASE
* config.guess: elf64_is64 target from Kevin Buettner
additional Rhapsody targets from Bernd Kuemmerlen <bkuemmer@mevis.de>
* config.sub:
forwarded by Urs JanBen - Bernd Kuemmerlen <bkuemmer@mevis.de>
to work with MacOS X (Rhapsody)
2000-05-31
* AcSplit/CF_SVR4: ci-aclocal generated header
* AcSplit/CF_SVR4: need a compile-check (I used termio with VLNEXT)
2000-05-29
* AcSplit/CF_POSIX_WAIT, AcSplit/CF_SVR4, AcSplit/CF_SYSV,
AcSplit/CF_TERMIO_C_ISPEED:
RCS_BASE
* AcSplit/CF_TTY_GROUP:
check for "not a tty" result if we're running this in batch mode.
also, add quotes that were missing in a test.
* AcSplit/CF_POSIX_WAIT, AcSplit/CF_SYSV, AcSplit/CF_TERMIO_C_ISPEED:
ci-aclocal generated header
2000-05-27
* AcSplit/CF_GNAT_VERSION: Juergen requires gnat 3.11 now
* AcSplit/CF_LIB_RULES: add "make libs"
add hook to suppress 'trace()' function
* AcSplit/CF_LIB_SUFFIX, AcSplit/CF_LIB_TYPE:
add LIB_SUFFIX variable, for libncursesw
* AcSplit/CF_SHARED_OPTS: add beos
* AcSplit/CF_MAN_PAGES:
add logic to generate symbolic links to manpages
* AcSplit/CF_MATH_LIB: ci-aclocal generated header
* AcSplit/CF_MATH_LIB: correct a mis-nested parenthesis
2000-05-20
* AcSplit/CF_SLANG_LIBS:
CF_FIND_LIBRARY takes extra param ($1 is duplicated)
* AcSplit/CF_SLANG_UNIX_DEFS: ci-aclocal generated header
* AcSplit/CF_SLANG_UNIX_DEFS:
lynx: ensure SLang_TT_Baud_Rate is declared extern
* AcSplit/CF_CURSES_FUNCS, AcSplit/CF_XOPEN_CURSES:
lynx: use $cf_cv_ncurses_header
2000-04-15
* AcSplit/CF_FUNC_GETADDRINFO: make quoting consistent
* AcSplit/AM_WITH_NLS:
mods from tin, to make choice of Makefile/makefile in po irrelevant.
also, fixes one of those GNU-bugs where it would refuse to build with non-GNU gettext,
by changing the $nls_cv_force_use_gnu_gettext assignment
2000-04-14
* AcSplit/CF_HEADER_RESOURCE: make quoting in AC_DEFUN consistent
* AcSplit/CF_HEADER_SELECT: make quoting of AC_DEFUN consistent
2000-04-13
* AcSplit/CF_FIND_IPV6_TYPE: get rid of -o option in test
* AcSplit/CF_FIND_IPV6_LIBS:
remove redundant assignments that are in the other macro
* AcSplit/CF_LIBRARY_PATH: add directories at '/', for SuSE
* AcSplit/CF_HEADER_PATH: look in $HOME, too
* AcSplit/CF_CHECK_IPV6, AcSplit/CF_FIND_IPV6_LIBS,
AcSplit/CF_FIND_IPV6_TYPE, AcSplit/CF_FUNC_GETADDRINFO:
RCS_BASE
* AcSplit/CF_FIND_LIBRARY:
add 2 params, to make this reusable from IPV6 tests.
drop that SuSE junk - it's not used here.
* AcSplit/CF_NCURSES_LIBS: added param to CF_FIND_LIBRARY
* AcSplit/AM_GNU_GETTEXT: ah - let him keep 'Makefile'
* AcSplit/CF_CHECK_IPV6, AcSplit/CF_FIND_LIBRARY:
ci-aclocal generated header
* AcSplit/CF_HEADER_PATH:
permute the second parameter before '/include', to allow for weird packages
such as socks5
* AcSplit/CF_LIBRARY_PATH:
permute the second parameter before '/lib', for weird packages
* AcSplit/CF_BUNDLED_INTL, AcSplit/CF_PW_GECOS: RCS_BASE
* AcSplit/CF_PW_GECOS: ci-aclocal generated header
2000-04-11
* AcSplit/CF_INPUT_METHOD: ci-aclocal generated header
* AcSplit/CF_INPUT_METHOD: initialize a couple of variables
* AcSplit/CF_TTY_GROUP:
modified to compute the group-name of the configurer's tty, and from that
infer if a 'tty' (or similar, such as 'terminal') group is being applied
to it.
2000-04-06
* AcSplit/CF_PDCURSES_X11:
testing with Lynx, I found that if I installed only the static libXcurses.a,
then I need Xt and Xaw also.
* AcSplit/CF_SLANG_UNIX_DEFS: RCS_BASE
* AcSplit/CF_CURSES_LIBS: gotta watch those late-friday-night patches
2000-03-26
* AcSplit/CF_SET_ERRNO: RCS_BASE
* AcSplit/CF_SET_ERRNO: ci-aclocal generated header
* AcSplit/AM_WITH_NLS: restore $(top_builddir) expression, used by lynx
2000-02-18
* AcSplit/CF_CURSES_LIBS:
from cdk: add case for HPUX 11.x for cur_colr library
* AcSplit/CF_GETOPT_HEADER, AcSplit/CF_PDCURSES_X11: RCS_BASE
* AcSplit/CF_CURSES_FUNCS:
allow for checks in XCurses (PDCurses X11 library)
* AcSplit/CF_GETOPT_HEADER: ci-aclocal generated header
2000-02-05
* AcSplit/CF_FUNC_POLL: RCS_BASE
* AcSplit/CF_FUNC_POLL: ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS:
updated Unixware shared-library rules to work for 7.1
2000-01-22
* AcSplit/CF_MATH_LIB:
make the test-function a parameter, defaulting to sin(x)
* AcSplit/CF_UTEMPTER: RCS_BASE
* AcSplit/CF_UTEMPTER: ci-aclocal generated header
2000-01-21
* AcSplit/CF_UTMP_UT_XSTATUS: RCS_BASE
* AcSplit/CF_UTMP: adds CF_UTMP_UT_XSTATUS in xterm #121
* AcSplit/CF_SYSV_UTMP: check for both setutent/setutxent, etc.
* AcSplit/CF_POSIX_VDISABLE, AcSplit/CF_SIZE_T:
ci-aclocal generated header
* AcSplit/CF_CHECK_ERRNO, AcSplit/CF_POSIX_VDISABLE, AcSplit/CF_SIZE_T:
duh: cosmetic changes
* AcSplit/CF_UTMP_UT_XTIME:
force an error if we can't assign to x.ut_xtime
2000-01-20
* AcSplit/CF_OUR_MESSAGES: ci-aclocal generated header
* AcSplit/CF_OUR_MESSAGES: if'd this inside $USE_NLS
* AcSplit/AM_WITH_NLS:
move/correct the chunk that generates po/makefile
move the chunk that links the intl.h header (so it's done only if we use NLS)
* AcSplit/AM_WITH_NLS:
Urs added an assignment to make $LIBS have -lintl (does not look right -fixme)
* AcSplit/AM_WITH_NLS:
temporary workaround for $(top_builddir), which is not defined by autoconf
2000-01-19
* AcSplit/AM_WITH_NLS: resync with drepper's serial #5
* AcSplit/AM_GNU_GETTEXT:
from tin 1.51, resync with drepper's current gettext macros:
add test for strdup, drop test for values.h
* AcSplit/CF_WITH_PROGRAM:
Urs Janben added a missing AC_DEFINE_UNQUOTED() for the last case, for tin.
2000-01-18
* AcSplit/CF_MATH_LIB:
change argument of sin() to a variable to work around report that gcc
evaluates sin(1.0) and replaces it with the constant result.
2000-01-17
* AcSplit/CF_LIB_PREFIX:
assign directly to LIB_PREFIX, retaining parameter for optional use.
(will obsolete the parameter)
2000-01-01
* AcSplit/CF_SHARED_OPTS:
from Debian ncurses 5.0 diffs, a better choice for CC_SHARED_OPTS for linux
* config.sub: from Debian ncurses 5.0 diffs, a guess for 'hurd'
1999-12-25
* AcSplit/CF_MATH_LIB: RCS_BASE
1999-12-23
* AcSplit/CF_NETLIBS: ci-aclocal generated header
* AcSplit/CF_NETLIBS:
from Urs Janben, for tin 1.5.1 - change the test for inet (-linet) to
use AC_CHECK_LIB to get rid of warning wrt AC_HAVE_LIBRARY being obsolete.
1999-12-12
* config.sub: from Kevin Buettner, recognize ia64 machine.
1999-11-30
* config.guess: added wildcard on elf32ppc, for lynx
* AcSplit/CF_PATH_SYNTAX: fix for OS/2 EMX, from Lynx
* mkdirs.sh: kludge for OS/2 EMX
1999-11-27
* AcSplit/CF_STDCPP_LIBRARY: drop check for math lib
* AcSplit/CF_GPP_LIBRARY:
drop check for math lib, since it's not really needed
* AcSplit/CF_SRC_MODULES:
tweaks to let me drop Ada95 directory by checking directly for enclosed makefile
1999-11-18
* AcSplit/CF_XOPEN_CURSES:
put the definition in $CFLAGS since we want to ensure it happens right
at the beginning of the compile
1999-11-11
* AcSplit/CF_TERMCAP_LIBS:
rewrote (for vile 8.4b) to use AC_TRY_LINK throughout so I avoid the
broken linker on FreeBSD
* AcSplit/CF_ADD_CFLAGS: RCS_BASE
* AcSplit/CF_CRYPT_FUNC: ci-aclocal generated header
* AcSplit/CF_CRYPT_FUNC:
nit - forgot to reset $LIBS in the second test.
1999-11-10
* AcSplit/CF_CRYPT_FUNC: RCS_BASE
* AcSplit/CF_TERMCAP_LIBS:
don't bother looking for termcap functions in libcurses on FreeBSD
* AcSplit/CF_PROG_PERL: trim trailing blanks
1999-10-30
* config.guess: add powerpc-apple
1999-10-28
* AcSplit/CF_UTMP_UT_SESSION: RCS_BASE
* AcSplit/CF_UTMP: add CF_UTMP_UT_SESSION
also, add definition for ut_name to ut_user
* AcSplit/CF_UTMP_UT_XTIME:
beyond existence check for ut_xtime, add check for missing #define
* AcSplit/CF_LASTLOG: comments
1999-10-24
* AcSplit/CF_LIB_RULES: add uninstall.data and uninstall.man rules
* AcSplit/CF_STDCPP_LIBRARY:
limit this to checking for variants of g++ (skip vendor's c++)
* AcSplit/CF_LIB_SUFFIX:
modify for netbsd: the suffix may be simply ".so", or ".so.$(REL_VERSION)",
depending on whether ld.elf_so is present
* AcSplit/CF_LINK_FUNCS:
oops - forgot to #define HAVE_LINK and HAVE_SYMLINK
* AcSplit/CF_GCC_WARNINGS: put the compiler name into the messages
(may not be 'gcc')
* AcSplit/CF_GCC_ATTRIBUTES:
put the compiler name into the messages - may not be 'cc'
* AcSplit/CF_GPP_LIBRARY:
took out the part of the test code that uses 'abs()', since (at least SCO's
copy of gcc 2.7.2.3) some versions of g++ may report ambiguity in the test
compile, making the test fail unexpectedly.
* AcSplit/CF_SHARED_OPTS:
treat netbsd differently from openbsd/freebsd (from patch).
fix openbsd/freebsd (I can test FreeBSD, anyway) to keep the REL_VERSION
on the library suffix (needed for FreeBSD 3.1, still - though someone
long ago told me it wouldn't be ;-).
corrections for SCO's shared libraries (what is -melf, anyway?)
modify linux to use $(CC) rather than gcc.
1999-10-16
* AcSplit/CF_GNU_SOURCE, AcSplit/CF_LASTLOG, AcSplit/CF_UTMP_UT_XTIME:
RCS_BASE
* AcSplit/CF_TYPE_FD_SET: add a dependency on X11/Xpoll.h
* AcSplit/CF_TYPE_FD_SET: ci-aclocal generated header
* AcSplit/CF_UTMP: expanded to handle some HPUX 10.x cases
* AcSplit/CF_FUNC_TGETENT: keep from combining ncurses' termcap.h !
* AcSplit/CF_SYSV_UTMP: linking is stronger than compiling
* AcSplit/CF_UTMP_UT_HOST:
make this use the definitions from CF_UTMP for include-files
1999-10-05
* AcSplit/CF_ANSI_CC_CHECK:
change from Urs Janssen for defining _HPUX_SOURCE in ANSI mode on HPUX
1999-09-24
* AcSplit/CF_TTY_GROUP, AcSplit/CF_SYSV_UTMP, AcSplit/CF_UTMP_UT_HOST:
RCS_BASE
* AcSplit/CF_UTMP: add sub-tests for struct members and functions
* AcSplit/CF_XTERM_MODE:
more refinements - don't really want to use the -o and -g options if s-bit
isn't set, since they're also a portability concern.
1999-09-21
* AcSplit/CF_XTERM_MODE: fixes for 'ls' that inverts the -g option
1999-09-15
* AcSplit/CF_XTERM_MODE: add a check for -g option
* AcSplit/CF_X_ATHENA: check for nonstandard locations
1999-09-11
* AcSplit/CF_CPP_PARAM_INIT: RCS_BASE
* AcSplit/CF_BOOL_SIZE:
add kludge for including builtin.h to avoid qt's builtin.h
* AcSplit/CF_ETIP_DEFINES:
huh - fix quoting for ETIP_NEEDS_xxx, which was wrong for a long time...
* AcSplit/CF_GPP_LIBRARY: smarter check for builtin.h
* AcSplit/CF_LIB_PREFIX:
oops - typo (masked in ncurses by redundant assignment)
1999-09-08
* AcSplit/CF_DISABLE_ECHO: Urs dropped the "test: " from help-message
* AcSplit/CF_DISABLE_ECHO: ci-aclocal generated header
* AcSplit/CF_MAKEFLAGS: whitespace
* AcSplit/CF_CHECK_1_DECL:
workaround for trash in glibc2's headers - if we're compiling with extra
includes (e.g., for tin and vile), double-check that we can compile properly
_with_ the headers.
1999-09-04
* AcSplit/CF_GPP_LIBRARY:
add check for HAVE_BUILTIN_H, so we don't accidentally pickup qt's builtin.h
on my FreeBSD 3.1
1999-09-01
* AcSplit/CF_PROG_EXT: cygwin needs .exe
* AcSplit/CF_STDCPP_LIBRARY, AcSplit/CF_GPP_LIBRARY: RCS_BASE
* AcSplit/CF_PROG_EXT:
add CXXFLAGS, which lets me build ncurses c++ demo (it needs to be built
with threaded code since the rest of the library is, so that break's
errno's definition otherwise)
* AcSplit/CF_SRC_MODULES:
change ncurses "--without-cxx" configure option so there's now also
a "--without-cxx-binding" option. That changes naming in this macro
* AcSplit/CF_LIB_PREFIX:
define LIB_PREFIX for substitution to handle special case in ncurses/c++
directory
1999-08-29
* AcSplit/CF_CHECK_FD_SET:
integrate more/less with the older CF_TYPE_FD_SET
* AcSplit/CF_CHECK_FD_SET, AcSplit/CF_HEADER_RESOURCE,
AcSplit/CF_HEADER_SELECT:
RCS_BASE
* AcSplit/CF_TERMCAP_LIBS:
check for curses library first - on terminfo systems that's the only way
to get color.
1999-08-22
* AcSplit/CF_INPUT_METHOD: RCS_BASE
1999-08-21
* AcSplit/CF_LINK_FUNCS: RCS_BASE
* AcSplit/CF_BOOL_DECL:
allow for bogus definitions of bool and false in <sys/types.h> and <stdio.h>
* AcSplit/CF_BOOL_SIZE:
corrected sign test for bool, since gcc incorrectly promoted -(unsigned) to
int.
1999-07-28
* AcSplit/CF_FUNC_SIGACTION: RCS_BASE
* AcSplit/CF_FUNC_SIGACTION: ci-aclocal generated header
1999-06-21
* AcSplit/CF_MAKEFLAGS:
correct (I had used 'set' as if this was csh!) and improve (by checking
the expansion of $(MAKEFLAGS) if assignments are passed through, like the
POSIX make program on IRIX64 6.2).
1999-06-19
* AcSplit/CF_SPEED_TYPE:
change default 'ospeed' type to 'unsigned' from 'short'
* AcSplit/CF_SPEED_TYPE: ci-aclocal generated header
1999-06-16
* AcSplit/CF_SRC_MODULES: Juergen renamed ada_include to src
* AcSplit/CF_BOOL_DECL:
change to allow additional test for 'bool' in CC by passing the cache
variable as an optional parameter
* AcSplit/CF_SUBST_NCURSES_VERSION: changed to test for 'bool' in CC
* AcSplit/CF_GNAT_VERSION: don't pass $(GENOBJS) - jpf
* AcSplit/CF_SHARED_OPTS: hurd (gnu) is the same as linux, here.
* AcSplit/CF_LIB_SUFFIX: hurd (gnu) is treated like linux
* AcSplit/CF_BOOL_SIZE:
tell which compiler we're testing - may be CC as well
1999-05-22
* AcSplit/CF_PROG_PERL: put brackets for consistency
1999-05-14
* AcSplit/CF_LIBRARY_PATH: from CDK, add /opt
* AcSplit/CF_HEADER_PATH: from CDK, add /opt to path
1999-05-08
* AcSplit/CF_CHECK_EXTERN_DATA: RCS_BASE
* AcSplit/CF_CHECK_ERRNO:
split-out CF_CHECK_EXTERN_DATA, to use in lynx's h_errno test
1999-05-07
* config.guess: xterm-101
* AcSplit/CF_LEX_STATES: RCS_BASE
* AcSplit/CF_LEX_STATES: ci-aclocal generated header
* AcSplit/CF_MAKEFLAGS:
prefer $(MAKEFLAGS) to $(MFLAGS), since the latter is more standard now,
and incidentally gmake passes '=' stuff along with it.
* AcSplit/CF_XTERM_MODE:
allow for possibility that xterm's installed via a symbolic link
* AcSplit/CF_ANSI_CC_CHECK, AcSplit/CF_GCC_WARNINGS: parenthesis
* config.sub: xterm-101 HP-UX 11.00 fixes
1999-03-30
* AcSplit/CF_OUR_MESSAGES: RCS_BASE
* AcSplit/CF_CHECK_FUNCDECL, AcSplit/CF_CHECK_FUNCDECLS: brackets
* AcSplit/CF_ARG_DISABLE, AcSplit/CF_ARG_ENABLE,
AcSplit/CF_CHECK_FUNCDECL, AcSplit/CF_CHECK_FUNCDECLS,
AcSplit/CF_EBCDIC, AcSplit/CF_INET_ADDR:
ci-aclocal generated header
* AcSplit/CF_CURSES_LIBS, AcSplit/CF_FIND_LIBRARY:
fix err in linux case (assignment to wrong variable)
* AcSplit/CF_PATH_PROG:
fix missing assignment, which left paths defined to empty strings when this
function was disabled
* AcSplit/CF_INET_ADDR: include <sys/types.h> in the test-compile
* AcSplit/AM_GNU_GETTEXT: make this work if the po directory is missing
* AcSplit/CF_EBCDIC: make this work with SunOS K&R C (and also IRIX C)
* AcSplit/CF_ARG_ENABLE: omit the wordy '(default: off)'
* AcSplit/CF_ARG_DISABLE: omit the wordy '(default: on)'
1999-03-13
* AcSplit/CF_POSIX_VDISABLE: comments
1999-01-24
* AcSplit/CF_INET_ADDR:
oops: forgot to reset $LIBS if I don't find anything to append
* AcSplit/CF_TERMCAP_LIBS:
make the cache-variable spelled consistently...
1999-01-23
* AcSplit/CF_SHARED_OPTS: mods to support --with-shlib-version
* config.guess: re-merge Mike Hopkirk's patch
1999-01-17
* AcSplit/CF_INET_ADDR: RCS_BASE
* AcSplit/CF_X_ATHENA:
Vikas pointed out on cygwin32 (static libraries?) that Xext must follow
Xmu - checked this with liborder and X11R6.3, seems right.
* config.sub, config.guess: changes from autoconf 2.13
1999-01-15
* AcSplit/CF_XOPEN_CURSES: RCS_BASE
* AcSplit/CF_STAT_ST_BLOCKS: ci-aclocal generated header
* AcSplit/CF_STAT_ST_BLOCKS:
oops: this didn't work because the test program had a typo
* AcSplit/CF_CURSES_FUNCS: use CF_XOPEN_CURSES
1998-12-25
* AcSplit/CF_POSIX_JC: RCS_BASE
1998-12-24
* AcSplit/CF_MAKE_INCLUDE:
don't use $CONFIG_SHELL, since it breaks something on CLIX
1998-12-19
* AcSplit/CF_LIB_SUFFIX, AcSplit/CF_SHARED_OPTS:
add gnu* to shared-library cases
* AcSplit/CF_SRC_MODULES: whitespace
1998-12-10
* AcSplit/CF_VARARGS: check for header files, so this works standalone
* AcSplit/CF_VARARGS: ci-aclocal generated header
* AcSplit/AM_GNU_GETTEXT: mods to let me disable nls properly
* AcSplit/CF_MAKE_INCLUDE: use CONFIG_SHELL for /bin/sh
1998-12-09
* install-sh: RCS_BASE
1998-11-21
* config.guess: fix for another CLIX configure, from Lynx
1998-11-18
* AcSplit/CF_EBCDIC: RCS_BASE
* AcSplit/AM_WITH_NLS: change default to --disable-nls
* AcSplit/CF_ALT_CHAR_SET, AcSplit/CF_SYSTEM_MAIL_FLAGS:
ci-aclocal generated header
* AcSplit/AM_GNU_GETTEXT: remove the po2tbl.sed script when done
* AcSplit/CF_ALT_CHAR_SET, AcSplit/CF_SLANG_CPPFLAGS,
AcSplit/CF_SYSTEM_MAIL_FLAGS:
whitespace
1998-11-16
* config.guess, config.sub: changes for os390
1998-11-14
* AcSplit/CF_MAN_PAGES: filter out .orig and .rej files
1998-11-07
* AcSplit/CF_CURSES_LIBS: fix for Suse Linux
1998-11-06
* AcSplit/CF_TERMIO_AND_CURSES, AcSplit/AM_GNU_GETTEXT,
AcSplit/AM_LC_MESSAGES, AcSplit/AM_PATH_PROG_WITH_TEST,
AcSplit/AM_WITH_NLS:
RCS_BASE
1998-10-17
* AcSplit/CF_MANPAGE_RENAMES, AcSplit/CF_MAN_PAGES:
make man-directory if it's not there
1998-10-12
* AcSplit/CF_IMAKE_CFLAGS:
fix so this works from another directory (made xterm build that way)
1998-10-10
* AcSplit/CF_PATH_SYNTAX, AcSplit/CF_WITH_PATHLIST: RCS_BASE
* AcSplit/CF_WITH_PATH: ci-aclocal generated header
* AcSplit/CF_WITH_PATH:
split-out CF_PATH_SYNTAX to use in CF_WITH_PATHLIST macro
* AcSplit/CF_FUNC_MEMMOVE:
make this standalone (no need for AC_CHECK_FUNC(memmove bcopy)
1998-10-04
* AcSplit/CF_SOCKS5: define USE_SOCKS4_PREFIX if we're using socks4
* AcSplit/CF_MAKE_INCLUDE:
from lynx, put parentheses around cd&&make, to appease bash
* AcSplit/CF_PATH_PROG: special hack for OS/2 EMX to adjust $IFS
1998-08-30
* AcSplit/CF_TERMCAP_LIBS:
add check for $LIBS having the termcap lib already
also add parameter that makes this look for ncurses
* AcSplit/CF_PROG_PERL: RCS_BASE
* AcSplit/CF_CC_OFFSETOF_CASES: ci-aclocal generated header
* AcSplit/CF_CC_OFFSETOF_CASES: whitespace
* AcSplit/CF_XTERM_MODE: RCS_BASE
1998-08-23
* AcSplit/CF_FUNC_TGETENT:
allow for the case where we've already got a termcap/terminfo library in $LIBS
* AcSplit/CF_TERMCAP_LIBS: don't mention slang
1998-08-22
* AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MANPAGE_RENAMES:
nit in help-message
1998-07-31
* AcSplit/CF_SOCKS: RCS_BASE
* AcSplit/CF_SOCKS5: first rewrite based on feedback from Brian Hauber
for lynx
1998-07-30
* AcSplit/CF_ALT_CHAR_SET: oops: missed a $mapname instance vs acs_map
* AcSplit/CF_SOCKS5:
rewrite based on bug report in lynx from Brian Hauber
1998-07-25
* config.guess: re-merged patch from Mike Hopkirk for Unixware
* config.guess: version from automake 1.3
1998-07-22
* AcSplit/CF_X_MOTIF, AcSplit/CF_X_OPENLOOK: RCS_BASE
* AcSplit/CF_X_MOTIF, AcSplit/CF_X_OPENLOOK:
ci-aclocal generated header
1998-07-18
* AcSplit/CF_ETIP_DEFINES, AcSplit/CF_GNAT_TRY_RUN,
AcSplit/CF_MANPAGE_FORMAT, AcSplit/CF_MANPAGE_RENAMES:
RCS_BASE
* AcSplit/CF_GNAT_TRY_RUN: ci-aclocal generated header
* AcSplit/CF_GNAT_VERSION: naming convention
* AcSplit/CF_MAN_PAGES:
split-out CF_MANPAGE_FORMAT and CF_MANPAGE_RENAMES
* AcSplit/CF_SRC_MODULES: tweaks for HPUX 10.x
* AcSplit/CF_SHARED_OPTS: tweaks for HPUX and OpenBSD
1998-07-01
* AcSplit/CF_CC_INIT_UNIONS, AcSplit/CF_CC_OFFSETOF_CASES: RCS_BASE
* AcSplit/CF_CC_INIT_UNIONS: ci-aclocal generated header
* AcSplit/CF_CURSES_CCHAR_T: RCS_BASE
* AcSplit/CF_CURSES_DATA: check for cchar_t
* AcSplit/CF_CURSES_CCHAR_T, AcSplit/CF_CURSES_DATA:
ci-aclocal generated header
* AcSplit/CF_CURSES_FUNCS:
use term.h if we have it, for terminfo functions
1998-06-24
* AcSplit/CF_ALT_CHAR_SET: oops: fix cache-reference
1998-05-26
* AcSplit/CF_IMAKE_CFLAGS:
typo (missing underscore in EXTRA_LOAD_FLAGS)
1998-05-23
* config.guess: patch from Mike Hopkirk to recognize Unixware
1998-05-16
* AcSplit/CF_SHARED_OPTS: add special cases for HPUX 10.x, SCO 5
* AcSplit/CF_SED_CONFIG_H: ci-aclocal generated header
* AcSplit/CF_SRC_MODULES:
hack so HPUX gets the -L$(libdir) first, but no other systems.
* AcSplit/CF_CHECK_CACHE, AcSplit/CF_MAN_PAGES, AcSplit/CF_SED_CONFIG_H,
AcSplit/CF_SPEED_TYPE:
trim trailing blanks
1998-05-15
* AcSplit/CF_CURSES_TERMCAP: whitespace
* AcSplit/CF_COLOR_CURSES: whitespace.
1998-05-06
* AcSplit/CF_IMAKE_CFLAGS: add logic to get IMAKE_LOADFLAGS
1998-04-27
* AcSplit/CF_ALT_CHAR_SET:
there's more than one possible array-name (OSF/1 uses _acs_map[])
* AcSplit/CF_NCURSES_BROKEN: ci-aclocal generated header
* AcSplit/CF_NCURSES_BROKEN, AcSplit/CF_TTYTYPE: fallback for curses.h
* AcSplit/CF_CURSES_LIBS:
check for cursesX before curses (for Ultrix, from lynx)
* AcSplit/CF_BOOL_DEFS, AcSplit/CF_CURS_PERFORMANCE:
ci-aclocal generated header
* AcSplit/CF_BOOL_DEFS, AcSplit/CF_CURS_PERFORMANCE,
AcSplit/CF_FANCY_CURSES:
fallback for curses.h
1998-04-25
* AcSplit/CF_SIZE_T: merge with autoconf's version : include <stdlib.h>
1998-04-18
* AcSplit/CF_ANSI_CC_CHECK:
add case for HPUX that uses +e option, since I don't think it used to
be supported
* AcSplit/CF_SHARED_OPTS: Juergen changed -fPIC to -fpic
* AcSplit/CF_FP_ISREADY: RCS_BASE
* AcSplit/CF_FP_ISREADY: ci-aclocal generated header
* AcSplit/CF_PROG_EXT: make this depend on CF_CACHE_CHECK
1998-04-01
* AcSplit/CF_ARG_WITH: New file.
* AcSplit/CF_PATH_MAILBOX, AcSplit/CF_WITH_DFTENV, AcSplit/CF_WITH_VALUE:
ci-aclocal generated header
* AcSplit/CF_WITH_DFTENV, AcSplit/CF_WITH_PROGRAM: use AC_ERROR
* AcSplit/CF_PATH_EDITOR, AcSplit/CF_PATH_MAILBOX,
AcSplit/CF_PATH_MAILER, AcSplit/CF_WITH_VALUE:
use CF_ARG_WITH
1998-03-21
* AcSplit/CF_SPEED_TYPE: mods to check for headers that define speed_t
* AcSplit/CF_SHARED_OPTS: mods to make -ltinfo chain properly on Linux
* AcSplit/CF_LIB_RULES:
switch order of cf_libs_to_make so I can force -ltinfo last
1998-03-19
* AcSplit/CF_CURSES_FUNCS:
we implicitly assume some of the _XOPEN_SOURCE_EXTENDED stuff
1998-03-05
* AcSplit/CF_CURSES_LIBS: fixes from Jonathan Sergent for lynx
1998-03-03
* AcSplit/CF_FIND_LIBRARY:
add special case for Linux systems that don't put everything in /usr/lib
* AcSplit/CF_COLOR_CURSES:
must allow for this macro being used w/o setting $cf_cv_ncurses_header
1998-03-01
* AcSplit/CF_PROG_EXT: RCS_BASE
1998-02-24
* config.sub: from automake 1.3
* AcSplit/CF_CHECK_ERRNO: OS/2 EMX has sys_errlist in <stdlib.h>
* AcSplit/CF_FIONBIO: RCS_BASE
* AcSplit/CF_FIONBIO: ci-aclocal generated header
1998-02-20
* AcSplit/CF_PATH_EDITOR: make this a command-line argument
* AcSplit/CF_PATH_MAILBOX, AcSplit/CF_PATH_MAILER:
make this a configure option
1998-02-19
* AcSplit/CF_DISABLE_ECHO:
my fix for autoconf makes quotes in the help-message superfluous
1998-02-17
* AcSplit/CF_CURSES_LIBS: correct a typo (spurious '\' before 'test')
add logic to allow --enable-5lib option (needed for ded, because I need to
have bsd curses on SunOS).
add fallbacks for undefined $cf_cv_ncurses_header
* AcSplit/CF_NCURSES_LIBS, AcSplit/CF_NCURSES_VERSION:
fallback for undefined $cf_cv_ncurses_header
1998-02-15
* AcSplit/CF_CURSES_LIBS: add xcurses library for UnixWare
* AcSplit/CF_CURSES_CHTYPE: typo
* AcSplit/CF_CURSES_TYPE: RCS_BASE
* AcSplit/CF_CURSES_CHTYPE:
add a check to see if 'chtype' is a scalar (OSF1 4.0 uses a struct, so we
cannot do attribute OR'ing on it)
* AcSplit/CF_CURSES_TYPE, AcSplit/CF_TD_CURSES_LIBS:
ci-aclocal generated header
* AcSplit/CF_TD_CURSES_LIBS: split-out CF_CURSES_TYPE
1998-02-10
* AcSplit/CF_PATH_EDITOR, AcSplit/CF_PATH_MAILBOX, AcSplit/CF_PATH_MAILER:
RCS_BASE
1998-02-07
* AcSplit/CF_SYS_TIME_SELECT, AcSplit/CF_UNSIGNED_LITERALS: RCS_BASE
* AcSplit/CF_UNSIGNED_LITERALS: ci-aclocal generated header
* AcSplit/CF_SHARED_OPTS: install HP-UX libs w/o execute-privilege
add simple chaining of libraries
* AcSplit/CF_SRC_MODULES: integrated tack
* AcSplit/CF_MAN_PAGES:
run the make_sed.sh script rather than trying to do that inline here.
also, make uninstall a little more efficient by not running the renaming then.
* AcSplit/CF_TYPEOF_CHTYPE:
use unsigned literals for chtype if we've got them
* AcSplit/CF_GNAT_VERSION: Juergen stopped supporting gnats before 3.10
* AcSplit/CF_ISASCII, AcSplit/CF_LIB_PREFIX: RCS_BASE
* AcSplit/CF_BOOL_SIZE:
fixes from Fred Fish: check if bool wants to be unsigned, and fix a quote.
* AcSplit/CF_LIB_RULES:
lib-subsets (to allow building terminfo library)
fixes for Ada95 install/uninstall
* AcSplit/CF_CFG_DEFAULTS: use AC_FD_MSG rather than '6'
* AcSplit/CF_CURSES_TERMCAP: moved the details into CHECK_DECL_HDRS
* AcSplit/CF_TYPE_OUTCHAR:
restructured, using CHECK_DECL_HDRS to hide details
* AcSplit/CF_CURSES_LIBS:
oops: have to check for termcap libraries BEFORE curses libraries!
* AcSplit/CF_X_TOOLKIT:
both IRIX 6 and IRIX 5 have broken nsl, socket libraries
* AcSplit/CF_FUNC_TGETENT: rewrite by Bjorne Helgaas
1998-02-02
* mkdirs.sh: space on 1st line
1998-01-14
* AcSplit/CF_HELP_MESSAGE: ci-aclocal generated header
* AcSplit/CF_HELP_MESSAGE: use new macro AC_DIVERT_HELP
* AcSplit/CF_SLANG_LIBS: hack to get rid of some spurious blanks
* AcSplit/CF_CHECK_FUNCDECL:
make this work with compilers other than gcc, which is too tolerant
1997-12-23
* AcSplit/CF_CURSES_LIBS:
restructured so I'm checking for more combinations (prompted by bug
report that SUSE installs ncurses, not "curses", but also by noting that
the tgoto/initscr combinations weren't expanded properly when I went to
add ncurses to the list)
1997-12-17
* AcSplit/CF_CURSES_TERMCAP:
first pass of debug/test, corrected inverted test for termcap.h, and
also add logic to see if we're ok with just term.h
* AcSplit/CF_TYPE_OUTCHAR:
split-out the header file, since config.log gets hard to read.
* AcSplit/CF_DEFINE_PATHNAME: ci-aclocal generated header
* AcSplit/CF_DEFINE_PATHNAME: fixes for 'NONE' vs 'NONE/'
* AcSplit/CF_SLANG_CPPFLAGS: log the headers we try (for debugging)
* AcSplit/CF_DEFINE_PATHNAME: RCS_BASE
* AcSplit/CF_SLANG_LIBS:
check if we can link slang w/o termcap (applies to recent versions)
* AcSplit/CF_CURSES_TERMCAP: RCS_BASE
* AcSplit/CF_TYPE_OUTCHAR: make this depend on CF_CURSES_TERMCAP
1997-12-16
* AcSplit/CF_IMAKE_CFLAGS:
add logic to ensure we don't allow -nostdinc and -I/usr/include (used
on SGI to redirect the standard includes in the X build tree), because
it confuses gcc.
* AcSplit/CF_SYS_SELECT_TIMEVAL: ci-aclocal generated header
* AcSplit/CF_SYS_SELECT_TIMEVAL: oops: 'no' has to be zero
* AcSplit/CF_PROG_YACC: RCS_BASE
* AcSplit/CF_TYPE_OUTCHAR:
change the ifdef for curses.h in termcap-mode so we don't use it unless
we've verified (test-not-done!) that <curses.h> does contain prototypes
for termcap (or that it's needed in some sense)
* AcSplit/CF_PROG_YACC: ci-aclocal generated header
* AcSplit/CF_IMAKE_CFLAGS:
warn, don't error-out if we cannot find the imake program
1997-12-13
* mkdirs.sh: RCS_BASE
1997-12-04
* AcSplit/CF_SYS_SELECT_TIMEVAL: better include sys/types.h too
* AcSplit/CF_SYS_SELECT_TIMEVAL: RCS_BASE
* AcSplit/CF_SIZECHANGE:
include <termio.h> if we cannot include <termios.h>
* AcSplit/CF_CHECK_ERRNO:
force the HAVE_xxx to be defined even if the DECL_xxx isn't needed
1997-12-01
* AcSplit/CF_CHECK_ERRNO: oops: more than one err
1997-11-30
* AcSplit/CF_MISSING_CHECK:
use CHECK_DECL_HDRS rather than explicit list, to make this reusable.
(vile now uses estruct.h's list)
* AcSplit/CF_CHECK_ERRNO: RCS_BASE
* AcSplit/CF_ERRNO: ci-aclocal generated header
* AcSplit/CF_ERRNO, AcSplit/CF_SYS_ERRLIST: use CF_CHECK_ERRNO
1997-11-28
* AcSplit/CF_TIOCGWINSZ: ci-aclocal generated header
* AcSplit/CF_TIOCGWINSZ: consolidated changes from CF_CHECK_1_DECL
* AcSplit/CF_CHECK_1_DECL: simplified by reusing tin.h
1997-11-25
* AcSplit/CF_FUNC_MEMMOVE: quoted a variable that needed it
1997-11-24
* AcSplit/CF_NCURSES_LIBS:
reset $LIBS if we think we've found initscr in -lgpm
1997-11-23
* AcSplit/CF_CHECK_FUNCDECL: nope: double is bad here
* AcSplit/CF_CHECK_FUNCDECL: make this more rigorous
(use a less common type)
* AcSplit/CF_CHECK_FUNCDECLS: use CF_UPPER, reverse the order of ifdefs
* AcSplit/CF_REMOVE_BROKEN:
(Philippe De Muyter) define REMOVE_BROKEN if unknown (cross-compiling)
* AcSplit/CF_CHECK_FUNCDECL, AcSplit/CF_CHECK_FUNCDECLS: RCS_BASE
* AcSplit/CF_REMOVE_BROKEN, AcSplit/CF_UNION_WAIT:
ci-aclocal generated header
* AcSplit/CF_UNION_WAIT:
make this a stronger test by compiling against 'wait()', and linking
to ensure that we'll use the macros rather than implied functions.
1997-11-08
* AcSplit/CF_SHARED_OPTS:
extend the rpath option to include Solaris (-R)
* AcSplit/CF_CGETENT, AcSplit/CF_STRUCT_SIGACTION,
AcSplit/CF_STRUCT_TERMIOS:
RCS_BASE
1997-11-07
* AcSplit/CF_UNION_WAIT: correct type for the first test.
add some logging to see what's going on if it fails.
1997-11-06
* AcSplit/CF_ADD_INCDIR:
don't special-case /usr/local/include, since gcc 2.6.3 doesn't include it
anyway (and apparently it's possible to misinstall 2.7.2 so that doesn't
work either).
1997-11-01
* AcSplit/CF_SHARED_OPTS:
remove an unneeded blank in '-d y' so gcc 2.7.2 works on Solaris shared libs
* AcSplit/CF_LIB_RULES: add dependency on CF_SYSTYPE
* AcSplit/CF_CFG_DEFAULTS, AcSplit/CF_LIB_SUFFIX:
changes to use config.guess/config.sub
* AcSplit/CF_SHARED_OPTS: mods to use config.guess/config.sub
add -lncurses dependencies for Linux.
* AcSplit/CF_CHECK_CACHE: show mismatching system-name, if error
* AcSplit/CF_REGEX: add a check for -lgen for regexpr.h macros
* AcSplit/CF_REGEX: ci-aclocal generated header
1997-10-22
* AcSplit/CF_NCURSES_CPPFLAGS: minor cleanup, add CF_NCURSES_VERSION
* AcSplit/CF_NCURSES_VERSION:
add an AC_TRY_RUN test, since unproto splits the preprocessor stuff I'd
concocted, making it unusable for getting the actual version.
* AcSplit/CF_FUNC_SYSTEM:
add dependency so I don't need to invoke CF_UNION_WAIT explicitly
* AcSplit/CF_FUNC_SYSTEM: ci-aclocal generated header
1997-10-21
* AcSplit/CF_FUNC_WAIT, AcSplit/CF_WAIT_HEADERS: RCS_BASE
* AcSplit/CF_FUNC_WAIT, AcSplit/CF_WAIT_HEADERS:
ci-aclocal generated header
* AcSplit/CF_NCURSES_BROKEN: fix order of if/then/message
* AcSplit/CF_UNION_WAIT: split-out/expanded tests for union-wait
1997-10-18
* AcSplit/CF_NCURSES_CPPFLAGS: nit
* AcSplit/CF_ARG_OPTION: ci-aclocal generated header
* AcSplit/CF_ARG_OPTION: trimmed blanks
* install.sh: space on 1st line
* AcSplit/CF_WITH_PATH: fix my last change (pattern was not complete)
* AcSplit/CF_NCURSES_CPPFLAGS: correct a guess/message
* AcSplit/CF_SIZECHANGE: correct quoting of NEED_PTEM_H, for SCO
* AcSplit/CF_NCURSES_BROKEN: nit in output-message
* AcSplit/CF_NCURSES_LIBS:
ncurses 4.1 971011 needs extra boost for ncurses+gpm test
1997-10-11
* AcSplit/CF_LIB_RULES: fix for uninstall in parallel directory
* AcSplit/CF_WITH_PATH: modify the workaround for exec-prefix=NONE
1997-10-04
* AcSplit/CF_LIB_RULES: uninstall rather than deinstall
(use jargon rather than plain English)
* AcSplit/CF_X_ATHENA:
integration with vile, need to define HAVE_LIB_xxx
1997-09-28
* AcSplit/CF_IMAKE_CFLAGS:
oops: we must use $(ALLDEFINES) rather than $(STD_DEFINES), since the latter
doesn't set XTFUNCPROTO!
* AcSplit/CF_LIB_RULES: add logic for de-installing libraries
* AcSplit/CF_MAN_PAGES: add logic so I can de-install man-pages
* AcSplit/CF_GNAT_VERSION: juergen added $cf_compile_generics symbol
1997-09-25
* AcSplit/CF_X_ATHENA: add --with-Xaw3d and --with-neXtaw options.
correct order of Xmu_s vs Xext libraries
1997-09-20
* AcSplit/CF_LIB_SUFFIX: comment
* AcSplit/CF_LIB_RULES: pass $target down to mk-1st.awk
* AcSplit/CF_SED_CONFIG_H:
add parameters 3 & 4, for preamble and trailer here-documents
* AcSplit/CF_GCC_WARNINGS: don't reset EXTRA_CFLAGS here!
* AcSplit/CF_GCC_WARNINGS: add ncurses' tweaks to the heap
1997-09-18
* AcSplit/CF_SED_CONFIG_H: keep old behavior: remove the input-file
* AcSplit/CF_X_TOOLKIT: make dependency upon CF_CHECK_CACHE explicit
* AcSplit/CF_NCURSES_CPPFLAGS:
oops: forgot the final else-clause in the check for predefined header
* AcSplit/CF_IMAKE_CFLAGS: missing -f on 'test'
1997-09-17
* AcSplit/CF_NCURSES_BROKEN:
correct sense of cache-variable, which was causing this test to be ignored
* AcSplit/CF_NCURSES_CPPFLAGS: fix to work with 1.8.7
1997-09-16
* AcSplit/CF_CURSES_LIBS: add a test for Hcurses
* AcSplit/CF_NETLIBS:
cache the result of this macro so we'll get the same result if we rerun
w/o deleting config.cache
1997-09-15
* AcSplit/CF_TYPE_FD_SET: RCS_BASE
* AcSplit/CF_FUNC_TGETENT: comment
* AcSplit/CF_IMAKE_CFLAGS: missing quotes
* AcSplit/CF_X_TOOLKIT: use config.guess, so case-names are lower-case
* AcSplit/CF_X_ATHENA, AcSplit/CF_X_TOOLKIT:
add logic from vile's configure.in
* AcSplit/CF_IMAKE_CFLAGS: parameterize, since xterm needed it
1997-09-14
* AcSplit/CF_IMAKE_CFLAGS:
simplified this by using Kevin Buettner's logic from vile
1997-09-13
* AcSplit/CF_ALL_MAKEFILES: RCS_BASE
* AcSplit/CF_ALL_MAKEFILES: ci-aclocal generated header
* AcSplit/CF_GMTOFF: FreeBSD has a 'timezone()' function.
* AcSplit/CF_FIND_TDLIB: gcc 2.6.3 doesn't search /usr/local/include
* AcSplit/CF_TOP_SRCDIR: RCS_BASE
* AcSplit/CF_TD_SRC_MAKEFILES: ci-aclocal generated header
* AcSplit/CF_TD_SRC_MAKEFILES:
move most of the logic into CF_SRC_MAKEFILE
1997-09-12
* AcSplit/CF_SRC_MAKEFILE: RCS_BASE
* AcSplit/CF_SRC_MAKEFILE: ci-aclocal generated header
* AcSplit/CF_FIND_TDLIB: comment
* AcSplit/CF_DISABLE_ECHO: add symbol RULE_CC
* AcSplit/CF_SED_CONFIG_H:
adjusted so I don't try to sed _or_ grep wide files (SCO's grep cannot do it)
* AcSplit/CF_TD_CONFIG: ci-aclocal generated header
* AcSplit/CF_TD_CONFIG: corrected the variable to test/update
* AcSplit/CF_TYPE_OUTCHAR: fix OUTC_ARGS when it's a char
* AcSplit/CF_HELP_MESSAGE: RCS_BASE
* AcSplit/CF_WITH_PATH:
add a hack to let me use variables for the default, e.g., --libdir
* AcSplit/CF_CURSES_LIBS: add hack for HP's cur_color library
* AcSplit/CF_COLOR_CURSES: make sure we have a color defined.
1997-09-11
* AcSplit/CF_FIND_TDLIB: RCS_BASE
* AcSplit/CF_FIND_LIBRARY:
comment, set $cf_libdir to known value so it can be tested in CF_FIND_TDLIB
* AcSplit/CF_TD_CONFIG: RCS_BASE
* AcSplit/CF_SED_CONFIG_H:
use td_config if available to workaround problems with very long @DEFS@
* AcSplit/CF_MAKE_AR_RULES, AcSplit/CF_TD_TEST_MAKEFILES,
AcSplit/CF_TD_TEST_MODULES:
RCS_BASE
* AcSplit/CF_DISABLE_ECHO: SHOW_CC is only for .c.o rules
* AcSplit/CF_TD_SRC_MODULES, AcSplit/CF_TD_TEST_MAKEFILES,
AcSplit/CF_TD_TEST_MODULES:
ci-aclocal generated header
* AcSplit/CF_TD_SRC_MAKEFILES: made a workable version
* AcSplit/CF_TD_SRC_MODULES: make a workable version
1997-09-08
* AcSplit/CF_TD_SRC_MAKEFILES, AcSplit/CF_TD_SRC_MODULES: RCS_BASE
* AcSplit/CF_MAKE_INCLUDE: make this work w/o lynx's directory-tree
1997-09-07
* AcSplit/CF_CHECK_CACHE: add check for AC_CANONICAL host
* AcSplit/CF_DISABLE_ECHO: fix quoting of $@
* AcSplit/CF_DISABLE_ECHO: incorrect order for ECHO_LD
* AcSplit/CF_SED_CONFIG_H: cleanup
* AcSplit/CF_DISABLE_ECHO:
oops: vile 7.2b broke because I left out ECHO_LD
* AcSplit/CF_OUTPUT_IF_CHANGED: RCS_BASE
* AcSplit/CF_OUTPUT_IF_CHANGED: ci-aclocal generated header
* AcSplit/CF_SED_CONFIG_H: use CF_OUTPUT_IF_CHANGED
* AcSplit/CF_SED_CONFIG_H: RCS_BASE
* AcSplit/CF_MSG_LOG: ci-aclocal generated header
* AcSplit/CF_MSG_LOG: use symbol rather than number for file-descriptor
* AcSplit/CF_CURSES_FUNCS: use CF_UPPER rather than CF_UPPERCASE
* AcSplit/CF_CURSES_STYLE: ci-aclocal generated header
* AcSplit/CF_CURSES_STYLE, AcSplit/CF_HAVE_FUNCS:
use CF_UPPER rather than CF_UPPERCASE
* AcSplit/CF_ADD_LIBDIR: RCS_BASE
* AcSplit/CF_ADD_LIBDIR: ci-aclocal generated header
* AcSplit/CF_TD_SIG_ARGS, AcSplit/CF_CHECK_REGEX: RCS_BASE
* AcSplit/CF_CHECK_REGEX: ci-aclocal generated header
* AcSplit/CF_FCNTL_VS_IOCTL, AcSplit/CF_POSIX_VDISABLE: RCS_BASE
* AcSplit/CF_FCNTL_VS_IOCTL: ci-aclocal generated header
* AcSplit/CF_TD_CURSES_LIBS: RCS_BASE
* AcSplit/CF_SYS_ERRLIST: merge td_lib (which needs both DECL and HAVE)
* AcSplit/CF_FUNC_LSTAT: RCS_BASE
* AcSplit/CF_FUNC_LSTAT: ci-aclocal generated header
* AcSplit/CF_ANSI_CPP, AcSplit/CF_CURSES_CHTYPE, AcSplit/CF_CURSES_DATA,
AcSplit/CF_CURSES_FUNCS, AcSplit/CF_CURSES_MOUSE,
AcSplit/CF_CURSES_STYLE, AcSplit/CF_GMTOFF, AcSplit/CF_HAVE_FUNCS,
AcSplit/CF_INCLUDE_PATH, AcSplit/CF_PROGRAM_FULLPATH,
AcSplit/CF_PROGRAM_PREFIX, AcSplit/CF_RCS_SCCS,
AcSplit/CF_REGCMP_FUNCS, AcSplit/CF_REGCMP_LIBS,
AcSplit/CF_REGEXPR_H_FUNCS, AcSplit/CF_REGEX_H_FUNCS,
AcSplit/CF_RE_COMP_FUNCS, AcSplit/CF_SIZE_T,
AcSplit/CF_STAT_ST_BLOCKS, AcSplit/CF_STRUCT_SCREEN,
AcSplit/CF_TCAP_CURSOR, AcSplit/CF_VARARGS, AcSplit/CF_WAIT:
RCS_BASE
* AcSplit/CF_ANSI_CPP, AcSplit/CF_CURSES_MOUSE, AcSplit/CF_REGCMP_FUNCS,
AcSplit/CF_REGCMP_LIBS, AcSplit/CF_RE_COMP_FUNCS,
AcSplit/CF_STRUCT_SCREEN, AcSplit/CF_TCAP_CURSOR, AcSplit/CF_WAIT:
ci-aclocal generated header
1997-09-06
* AcSplit/CF_SRC_MODULES, AcSplit/CF_LIB_RULES:
use CF_UPPER rather than CF_UPPERCASE
* AcSplit/CF_LIB_RULES, AcSplit/CF_LIB_SUFFIX, AcSplit/CF_SHARED_OPTS,
AcSplit/CF_SUBST_NCURSES_VERSION, AcSplit/CF_ADA_INCLUDE_DIRS,
AcSplit/CF_BOOL_DECL, AcSplit/CF_BOOL_SIZE, AcSplit/CF_CFG_DEFAULTS,
AcSplit/CF_CXX_LIBRARY, AcSplit/CF_DIRS_TO_MAKE,
AcSplit/CF_GNAT_VERSION, AcSplit/CF_INCLUDE_DIRS, AcSplit/CF_LIB_TYPE,
AcSplit/CF_LINK_DATAONLY, AcSplit/CF_MAN_PAGES, AcSplit/CF_OBJ_SUBDIR,
AcSplit/CF_PROG_INSTALL, AcSplit/CF_SPEED_TYPE,
AcSplit/CF_SRC_MODULES, AcSplit/CF_SUBST, AcSplit/CF_SYSRELV,
AcSplit/CF_SYSTYPE, AcSplit/CF_TYPEOF_CHTYPE, AcSplit/CF_WIDEC_SHIFT:
RCS_BASE
* AcSplit/CF_CXX_LIBRARY, AcSplit/CF_SUBST, AcSplit/CF_WIDEC_SHIFT,
AcSplit/CF_MISSING_EXTERN:
ci-aclocal generated header
* AcSplit/CF_MISSING_EXTERN: use CF_UPPER rather than CF_UPPERCASE
* AcSplit/CF_TYPE_OUTCHAR: merge with tin's version
* AcSplit/CF_GCC_WARNINGS: -Wwrite-strings too noisy except for xterm
set the AC_SUBSTR for EXTRA_CFLAGS here
* AcSplit/CF_ERRNO: comment
* AcSplit/CF_KILLPG, AcSplit/CF_RESTARTABLE_PIPEREAD: RCS_BASE
* AcSplit/CF_ANSI_CC_REQD: ci-aclocal generated header
* AcSplit/CF_ANSI_CC_REQD: tested with vile 7.2
* AcSplit/CF_AC_PREREQ, AcSplit/CF_ANSI_QSORT, AcSplit/CF_CURSES_TERM_H,
AcSplit/CF_MISSING_CHECK, AcSplit/CF_MISSING_EXTERN,
AcSplit/CF_PREREQ_COMPARE:
RCS_BASE
* AcSplit/CF_AC_PREREQ, AcSplit/CF_PREREQ_COMPARE:
ci-aclocal generated header
* AcSplit/CF_GCC_ATTRIBUTES:
CF_UPPER is reverse order from CF_UPPERCASE
* AcSplit/CF_ANSI_CC_REQD: RCS_BASE
* AcSplit/CF_ANSI_CC_CHECK: renamed from CF_ANSI_CC
* AcSplit/CF_ANSI_CC_CHECK: RCS_BASE
1997-09-05
* AcSplit/CF_GCC_WARNINGS: nits for X-defines
* AcSplit/CF_GCC_ATTRIBUTES:
make this standalone by checking if we're using gcc
* AcSplit/CF_GCC_ATTRIBUTES: use CF_UPPER rather than CF_UPPERCASE
* AcSplit/CF_CHECK_CACHE, AcSplit/CF_DISABLE_ECHO,
AcSplit/CF_FUNC_MEMMOVE, AcSplit/CF_FUNC_TGETENT,
AcSplit/CF_GCC_ATTRIBUTES, AcSplit/CF_IMAKE_CFLAGS,
AcSplit/CF_UPPERCASE, AcSplit/CF_VERBOSE, AcSplit/CF_X_ATHENA,
AcSplit/CF_X_TOOLKIT:
RCS_BASE
* AcSplit/CF_VERBOSE: ci-aclocal generated header
1997-08-28
* AcSplit/CF_ALT_CHAR_SET, AcSplit/CF_BOOL_DEFS,
AcSplit/CF_COLOR_CURSES, AcSplit/CF_CURSES_LIBS,
AcSplit/CF_CURS_PERFORMANCE, AcSplit/CF_DEFINE_PROG,
AcSplit/CF_FANCY_CURSES, AcSplit/CF_LOCALE, AcSplit/CF_MAKE_INCLUDE,
AcSplit/CF_NCURSES_BROKEN, AcSplit/CF_NGROUPS, AcSplit/CF_PATH_PROG,
AcSplit/CF_REMOVE_BROKEN, AcSplit/CF_SIZECHANGE,
AcSplit/CF_SLANG_CPPFLAGS, AcSplit/CF_SLANG_LIBS,
AcSplit/CF_STRIP_G_OPT, AcSplit/CF_STRIP_O_OPT,
AcSplit/CF_SYSTEM_MAIL_FLAGS, AcSplit/CF_TERMCAP_LIBS,
AcSplit/CF_TERMIO_AND_TERMIOS, AcSplit/CF_TTYTYPE, AcSplit/CF_UTMP:
RCS_BASE
* AcSplit/CF_NCURSES_LIBS:
add AC's hack to workaround broken FreeBSD linker
* AcSplit/CF_TERMIO_AND_TERMIOS: ci-aclocal generated header
* AcSplit/CF_UPPER: comment
* AcSplit/CF_FIND_LIBRARY: omit spurious attempt at 'unset'
* AcSplit/CF_ERRNO, AcSplit/CF_SYS_ERRLIST: whitespace
* AcSplit/CF_ADD_INCDIR, AcSplit/CF_ARG_DISABLE, AcSplit/CF_ARG_ENABLE,
AcSplit/CF_ARG_OPTION, AcSplit/CF_CHECK_1_DECL, AcSplit/CF_CHECK_DECL,
AcSplit/CF_CHECK_NESTED_PARAMS, AcSplit/CF_COMPTYPE,
AcSplit/CF_COREFILE, AcSplit/CF_CPP_CONCATS, AcSplit/CF_CPP_EXPANDS,
AcSplit/CF_ERRNO, AcSplit/CF_FIND_LIBRARY, AcSplit/CF_FUNC_FORK,
AcSplit/CF_FUNC_SYSTEM, AcSplit/CF_GCC_WARNINGS,
AcSplit/CF_HEADER_PATH, AcSplit/CF_LIBRARY_PATH, AcSplit/CF_MAKEFLAGS,
AcSplit/CF_MSG_LOG, AcSplit/CF_NCURSES_CPPFLAGS,
AcSplit/CF_NCURSES_LIBS, AcSplit/CF_NCURSES_VERSION,
AcSplit/CF_NETLIBS, AcSplit/CF_PROG_SUM_R, AcSplit/CF_RECHECK_FUNC,
AcSplit/CF_REGEX, AcSplit/CF_SET_GID_UID, AcSplit/CF_SIG_ARGS,
AcSplit/CF_SIG_CONST, AcSplit/CF_SOCKS5, AcSplit/CF_SYS_ERRLIST,
AcSplit/CF_SYS_NAME, AcSplit/CF_TERMIOS, AcSplit/CF_TIOCGWINSZ,
AcSplit/CF_TM_GMTOFF, AcSplit/CF_TYPE_OUTCHAR,
AcSplit/CF_TYPE_SIGACTION, AcSplit/CF_UNION_WAIT, AcSplit/CF_UPPER,
AcSplit/CF_WITH_DFTENV, AcSplit/CF_WITH_PATH, AcSplit/CF_WITH_PROGRAM,
AcSplit/CF_WITH_VALUE:
RCS_BASE
* AcSplit/CF_CHECK_DECL, AcSplit/CF_CHECK_NESTED_PARAMS,
AcSplit/CF_COMPTYPE, AcSplit/CF_COREFILE, AcSplit/CF_CPP_CONCATS,
AcSplit/CF_CPP_EXPANDS, AcSplit/CF_FUNC_FORK, AcSplit/CF_PROG_SUM_R,
AcSplit/CF_SET_GID_UID, AcSplit/CF_SIG_ARGS, AcSplit/CF_SIG_CONST,
AcSplit/CF_SYS_NAME, AcSplit/CF_TERMIOS, AcSplit/CF_TM_GMTOFF:
ci-aclocal generated header
1997-05-16
* config.guess: add case for CLIX
1997-03-11
* install.sh: RCS_BASE
1996-11-26
* config.guess: RCS_BASE
1996-11-22
* config.sub: RCS_BASE
1995-08-23
* AcSplit/CF_POPEN_TEST, AcSplit/CF_YACC_ERROR: RCS_BASE
* AcSplit/CF_POPEN_TEST, AcSplit/CF_YACC_ERROR:
ci-aclocal generated header