Date: Wed, 8 Feb 1995 03:27:23 +0900 From: Stephen Turnbull To: DJGPP AT sun DOT soe DOT clarkson DOT edu Subject: problems with contrib sources.. should this be happening? I would guess that the problem with the contrib sources is that certain things change from system to system and from version to version of GCC. Even with the typos, probably somebody looked at quickly, saw it wasn't gonna work anymore, and changed it without testing. If you want to see what it takes to make sure that the same code compiles on a wide variety of systems, look at what Imake does for Xwindows. Another example is that some people use GNU Make, some people use NDMake, some people use Borland or MS makes, and some people use QMake (mostly DV & DVX users). Unfortunately, the DOS makes (at least QMake, and I think Borland and MS as well) do not require tabs before commands in makefiles. Guess what happens when you try to use such makefiles under GNU Make? I'm not sure exactly which programs you're talking about. What you could do, of course, is take the original programs and your edited version, run them through "diff -C 2" and send that somewhere. (If you think it's worth sending them to DJ, make sure you read the readme on bug fixes and the like; he requires very specific formats for very good reasons.) In the meantime (ie, until DJ says what he intends to do/wishes someone else would do), if someone else wants to coordinate the project, I would be willing to supply appropriate access and a few MB on the Yaseppochi-gumi server. But this is only worth it if you're willing to make reasonably sure that the versions you produce are going to work on any reasonable DJGPP system, including but not limited to: (1) no dependency on where the top of the DJGPP subtree is (c:\djgpp and h:\NFS\usr\local\compilers\djgpp should both work) (2) works with GNU make, and preferably tested with other common makes (3) document exactly the system you used to compile, including the versions of make, C, C++, and binutils (4) make sure that the includes are compatible with current and future (ie ANSI and Posix) configurations (5) you actually, really, truly successfully compiled the uploaded version :-) The goal of course is eventually to reintegrate with the DJGPP distribution, but in the meantime it's probably better to keep them out of DJ's hair (at least until we have V2!!---when half of them will have to change again to comply with ANSI and Posix :-( ) Certainly, RMS, DJ, CWS, LD, EB, MW, EZ, and a few other sets of initials I don't remember offhand have done far more than their share and more than most could. But that doesn't mean you can't contribute to keeping our environment clean! -- Stephen Turnbull / Yaseppochi-gumi / http://turnbull.sk.tsukuba.ac.jp/ anon FTP: turnbull.sk.tsukuba.ac.jp Check out Kansai-WWW, too ------------> http://pclsp2.kuicr.kyoto-u.ac.jp/