X-Authentication-Warning: acp3bf.physik.rwth-aachen.de: broeker owned process doing -bs Date: Wed, 24 Nov 1999 16:41:16 +0100 (MET) From: Hans-Bernhard Broeker X-Sender: broeker AT acp3bf To: djgpp-workers AT delorie DOT com cc: rainer AT mathematik DOT uni-bielefeld DOT de Subject: Re: AW: ANNOUNCE: rsxntdj 1.6 BETA In-Reply-To: <199911241439.JAA25063@envy.delorie.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Reply-To: djgpp-workers AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On Wed, 24 Nov 1999, DJ Delorie wrote: > > If other versions of GCC do the right thing, it is probably a > > configuration problem. [...] > When you set up gcc for a given target, you have to configure it to > match the target. If we're supporting windows, we should reconfigure > gcc to use 16-bit wide characters. However, you'll have to rebuild > *all* of libc to support them properly (after fixing the wchar type in > the system headers). Just to get this into me own head: isn't it RSXNTDJ we're talking about here, i.e. DJGPP's own libc is not really part of the game, originally? But to continue to use RSXNTDJ, we'ld need a gcc configured for wchar_t == short int, unlike DJGPP's wchar_t == int. Which would mean either we modify DJGPP's libc and the i?86-pc-msdosdjgpp gcc settings (simultaneously!) just to allow the RSXNTDJ addon to continue working, or RSXNTDJ would have to bite the bullet and bring its own custom-configured gcc. The latter would invalidate almost all of the point there ever was in having RSXNTDJ, instead of using RSX with emx, or Cygwin. The former would mean a cartload of work, for some volunteer, and considerable coordination to avoid confusion by non-compatible versions of gcc and the libc. We might be forced to proclaim DJGPP version 3.0 just for this minor gotcha :-( Hans-Bernhard Broeker (broeker AT physik DOT rwth-aachen DOT de) Even if all the snow were burnt, ashes would remain.