Re: Generate string.h and time.h if nonexistent; kill POSIX

["Zack Weinberg" <zackw at stanford dot edu>]

On Fri, May 11, 2001 at 06:09:20PM -0700, Mark Mitchell wrote:
> >>>>> "Gerald" == Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at> writes:
> 
>     Gerald> Even the GNU coding standards do no longer require support
>     Gerald> for them, by the way, and explicitly allow requiring newer
>     Gerald> dialects of C.
> 
> In that case, I am definitely in favor of converting GCC to ANSI/ISO
> C89.
> 
> However, we need an SC decision before anyone can start on that.

Just for the record, as someone who's been digging through the
old-platform support cruft a lot lately:

I don't oppose dropping support for pre-C89 hosts or targets.
However, I think there is not a great deal to be gained by either.
Supporting pre-C89 hosts is fairly easy with our collection of wrapper
macros and replacement library modules.  We can write code that looks
almost like C89, enough to get most of the benefit in fact.  It's not
much of a maintenance burden.

Supporting pre-C89 targets does take more effort, and I'm definitely
in favor of dropping targets which are provably dead and buried.
There is also more to gain in continuing to support pre-C89 targets;
someone with an old machine to maintain is more likely to need target
support than host - they can build a cross-compiler from a modern box
and it may well be easier to work with that way.

If we're going to raise bars, I think we'd get a larger, more
immediate benefit from requiring GNU Make.

-- 
zw   You can tell [the lunatic] by the liberties he takes with common sense,
     his flashes of inspiration, and by the fact that sooner or later he
     brings up the Templars.
     	-- Umberto Eco, _Foucault's Pendulum_