Don Simons wrote:
> I don't understand what this gcc and CYGWIN stuff means very well, but I
> don't think it's a good idea to make such a change in the official
> distributions. AFAIK the problem that started this thread is due to an
> ERROR in the g77 Win compiler; an error that doesn't seem to occur in
> ANY other compiler, Unix, Linux, or DOS/Win. I don't want any
> distributed version my code changed in a way that "has the side-effect
> that the generated .tex file has inappropriate end of line characters."
> just to compensate for the error in that one compiler. By all means,
> feel free to make any changes you want in your private version if you
> must, to make the compilation work with this one compiler, but please
> don't mess anything up for others.
>
> If using compiler directives to effect this change does not affect
> average Joe who tries to compile the source for example in Linux by
> typing "g77 source.f", then fine. But if average Joe has to use a
> makefile to do the compilation, where he otherwise would not have to, or
> do any more than he has to now, then my objection stands.
This might be a misunderstanding.
Sebastien wrote to me about his attempts to compile the unix/c
version under Win32. The gcc compiler has been ported to Win32 for
the so called Cygwin environment. To get working midi files we tried
to change the fopen() calls to use binary mode all the time. To
achieve this, we can fill in the following lines in libI77.c:
#ifdef __CYGWIN__
char *f__r_mode[2] = {"rb", "r"};
char *f__w_mode[4] = {"wb", "wb", "r+b", "r+b"};
#else
char *f__r_mode[2] = {"rb", "r"};
char *f__w_mode[4] = {"wb", "w", "r+b", "r+"};
#endif
As you can easily see, this won't affect anybody except Win32 gcc
users, because I don't see any reason why __CYGWIN__ should be
defined on any non-Cygwin system.
Regards, Andreas.
--
Andreas Kurth
Mannheim, Germany
http://sites.inka.de/wam56
_______________________________________________
TeX-music mailing list
[EMAIL PROTECTED]
http://sunsite.dk/mailman/listinfo/tex-music
