Message-Id: <200006051811.OAA03423@qnx.com> Subject: Re: ANSI C and stdio.h To: djgpp-workers AT delorie DOT com Date: Mon, 5 Jun 2000 14:11:36 -0400 (EDT) From: "Alain Magloire" In-Reply-To: <200006051549.SAA22788@mailgw1.netvision.net.il> from "Eli Zaretskii" at Jun 05, 2000 06:48:13 PM X-Mailer: ELM [version 2.5 PL0b1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Reply-To: djgpp-workers AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk > > > Date: Sun, 4 Jun 2000 16:12:03 -0400 > > From: DJ Delorie > > If you can demonstrate a standards-conforming program that the average > > user would be expected to produce that doesn't work with djgpp's > > headers, we'll fix it. I know we can come up with hand-crafted > > programs that demonstrate non-conformity. I'm not interested in > > those, because common sense should prevail in the rare esoteric cases. Problems like this do not appear in simple "hello world" type of programs, but rather creep in when you are doing advanced or large or medium or big projects with interdepencies etc ... when you move the enviroment around, this is where you hit the brick wall. Standards can be viewed like guidelines when coding portable code and usually they have a very flexible vocabulary to permit some common denominator. IMHO, not complying to std should be seen as a bug. Of course the platform may have good reason to do so, for example this is OK in POSIX.1 but can not be implemented cleanly here : fd = open (file, ...); unlink (file); write (fd, ...); > > > > Standards aside, we've had a lot of luck with our headers and have > > gotten almost no FAQs about them due to obsessive standards > > compliance. There are a number of headers that are this way (for > > example, sys/stat.h doesn't require sys/types.h) and I'd rather this > > continue working than cause problems for users in the name of > > "standards". > > I couldn't agree more. > In some way, if I was a DJGPP user, I could take issue with the comments above, it means that I'm not bright enough to understand a std and always need a coussin ... My 0.002 $ canadian, the point is not to harasse people but at least to give, when possible, a clean environment to work and develop safely where I would expect some reasonnable/portable ANSI code to work properly without any #ifdefing. And it should include advanced/guru users, DJGPP is not a toy meaning it seems quite mature and can/should be view as a "real/rigourous" environment of work. sigh .. again I'm on the other end of the spectrum. I suppose it depends on how far you want to carry the "DJGPP userfriendliness", for me, it's a bug, very minor. -- au revoir, alain ---- Aussi haut que l'on soit assis, on n'est toujours assis que sur son cul !!!