[vorbis-dev] OggSplit 0.1.0

[Philip Jägenstedt]

On Wed, 6 Aug 2003 14:12:18 -0400 (EDT)
squirrel at WPI.EDU wrote:

> > On Wed, 6 Aug 2003 18:17:44 +0100
> > Ralph Giles <giles at xiph.org> wrote:
> >
> > In any case, where does OS X specify the error codes, if not error.h?
> > For debian, I include <errno.h> and that's all (although now system.h
> > does that for me). Since its OS X's libc library that will be returning
> > errors from fopen(), it also has to have some way of interpreting the
> > errors, right? (using strerror I hope).
> 
> I thought that too when reading this thread, so I checked out error.h - it
> just seems to be a fancy error-reporting utility; it contains no error
> codes (they're all in errno.h). AFAIK, error.h isn't standard - it's
> probably just a GNU extension. It only defines a couple of functions,
> error, error_at_line, and *error_print_progname. Does OggSplit use any of
> these?

Ah, thanks for that information. No, I don't use any of these
functions... It seems that error.h is included from xmalloc.c, where the
function "error" is actually used. From the beginning I had my own (very
simple) xmalloc and xrealloc functions, but thought I'd use the
autoproject xmalloc for increased portability... but now that it's
actually _reduced_ portability, I should consider if I shouldn't go back
to my own functions -- after all, I think there are very few systems
which don't handle malloc and realloc properly, so it shouldn't be a
problem.

It's quite late now, but I'll see if there's anything in system.h and
xmalloc.c that I _really_ need, and if not, I'll just take them out of
the picture again... during the night or tomorrow :)

// Philip Jägenstedt
--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.