Re: [AD] pack_ferror() - differences between platforms

[Ben Davis <bnd25@xxxxxxxxxx>]

> > Apparently, pack_feof() returns nonzero as soon as you reach the end of
> > the file. It doesn't want for you to try to read BEYOND the end of the
> > file.
>
> Old news. See http://glub.ehu.es/mlist/allegro_conductors/2001/3095.html

Yep, I remember that thread. It didn't occur to me, of course, until I'd 
rediscovered this behaviour.

> > Disregard the three patches I posted last time; if there are no
> > objections, I shall begin work on a patch that will arrange for
> > pack_ferror() to return nonzero after you try to read beyond the end of a
> > file. I'll include the extra documentation for pack_feof() in the patch.
>
> Already available at
> http://perso.libertysurf.fr/ebotcazou/dev/packfeof_ansi.zip
>
> We ruled it out (well, Peter did but I concurred) though. See the thread
> "pack_ungetc" starting July 09.

No no no no no, you misunderstood me. Reread my patch suggestion above, and 
note where I say pack_ferror() instead of pack_feof(). So the patch will 
really have two separate parts: a change in pack_ferror()'s functionality, 
and a change in pack_feof()'s docs. Sorry for not making this clearer.

So, any objections? :)

Ben