Date: Mon, 13 Jul 1998 11:02:29 +0300 (IDT) From: Eli Zaretskii To: Dave Love cc: Andris Pavenis , "Gurunandan R. Bhat" , Alexey Yakovlev , djgpp-workers AT delorie DOT com, fortran AT gnu DOT org Subject: Re: Inconsistencies between g77 v0.5.23 and v0.5.19 In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Precedence: bulk On 12 Jul 1998, Dave Love wrote: > Eli> DJGPP shows that `unix' can be defined by a compiler which > Eli> targets other systems. > > Sure, and there are already special cases for Windows32 stuff, but > we're interested in a characteristic of Unix (POSIX, I guess). I admit I don't know enough about how g77 is configured. But it puzzles me why that NON_ANSI_RW_MODES needs to be defined for native Unix systems? I assume that what that does is to avoid using the "b" qualifier in an `fopen', but AFAIK using "b" does no harm on Unix, so why do you need that define at all? If the reason is that older non-ANSI Unix libraries don't support "b" in a destructive way, then just test for ANSI (__STDC__) or Posix, and if one of those is supported, always use "b". This should solve the problem for DJGPP, as well for any other standard-complying environment. Or did I miss something? > I don't think so. The behaviour for cygwin32, at least, depends on > how you mount the file system, not on the library intrinsically. True, but AFAIK using "b" won't hurt, regardless of how the file system is mounted. > Eli> Since the patch is for configure.in, the DJGPP version *must* be > Eli> built by running the configure script. > > But not necessarily natively under DJGPP. The DJGPP port of Bash is perfectly capable of running GNU configure scripts natively, even on plain DOS. I have been doing that for almost two years now (although never for g77), with no problems at all.