Re: What's the status of ncurses now?

> Over in the FreeBSD camp, we're hearing all kinds of horror stories
> of disputed copyrights and worse, so I figured I'd come over here
> and ask a group of folks a little closer to the source.

We have a new primary maintainer and a plan to resolve the copyright
problem.
 
> There has also been some concern expressed about the direction in
> which ncurses appears to be going - more and more features which
> deviate a fair bit from the original "curses mandate" to provide a
> straight-forward screen I/O library, the most complex object in
> which would be the "window" and anything else being layered on top
> (in a different library).

I don't see that we've deviated from the original "mandate" at all.

True, we support processing of mouse input events under xterm, and have
a few option entry points not present in SVr4, and have C++ and Ada
bindings. But designers are free to ignore all that and see a straight
XSI- or SVr4-compatible curses library.

> Now that forms and menus and such have appeared, it almost seems as
> if this layering has been violated somewhat, and rather than trying to
> creeping-featurize ncurses it seems (IMHO) a far better course to simply
> start developing another, entirely independant, library of CUI objects.

The forms, menus, and panels features are separate libraries which
don't creeping-featurize the base curses library at all (in fact, that
is the reason they are separate libraries).

In this as in other design decisions we hew pretty strictly to the
SVr4 API. Except we do much better cursor-movement and screen-update
optimization.

-- 
		Eric S. Raymond