Re: [AD] Allegro 4.0.3

[Eric B. <ebo...@li...>]

> Here's the list of patches that I think would break strict forward
> compatibility:
>
>
> 2002-12-17 10:05  ebotcazou
>
> 	* docs/src/allegro._tx, include/allegro/system.h,
> 	src/unix/usystem.c, tools/grabber.c: Lisa Parratt added IRIX
> 	detection
>
> 2002-11-03 11:48  elias		alleg072.diff
>
> 	* resource/keyboard/de.cfg, resource/keyboard/fr.cfg, src/font.c:
> 	added euro character to default font
>
> 2002-10-27 19:19  gradha	alleg084.diff
>
> 	* resource/language/catext.cfg: Adding Catalan translation by Oscar
> 	Giner (O_...@ho...)
>
> 2002-10-07 09:14  ebotcazou	alleg143.diff
>
> 	* allegro.cfg, resource/language/nltext.cfg: Benny Colyn added a
> 	Dutch translation
>
> 2002-10-06 23:00  ebotcazou	alleg147.diff
>
> 	* docs/src/allegro._tx, include/allegro/system.h,
> 	src/unix/usystem.c, tools/grabber.c: Evert Glebbeek added the
> 	detection of SunOS/Solaris
>
> 2002-04-12 15:45  ebotcazou	alleg475.diff
>
> 	* docs/src/ahack._tx, docs/src/allegro._tx,
> 	include/allegro/alcompat.h, include/allegro/gfx.h,
> 	include/allegro/rle.h, include/allegro/text.h,
> 	include/allegro/internal/alconfig.h,
> 	include/allegro/platform/al386gcc.h,
> 	include/allegro/platform/aldjgpp.h,
> 	include/allegro/platform/almaccfg.h,
> 	include/allegro/platform/alunix.h, src/ppc/ppcdef.inc,
> 	tests/test.c: Peter and I made it so that Allegro programs can
> 	compile with gcc -ansi -pedantic
>
> 2002-04-11 22:30  ebotcazou	alleg478.diff
>
> 	* docs/src/allegro._tx, include/allegro/platform/alqnx.h,
> 	include/allegro/platform/alunix.h, include/allegro/platform/alwin.h:
> 	Added the symbol ALLEGRO_MAGIC_MAIN

Hmm.. as I really would like to have patch #6 in the 4.0.3, I propose to take 
the following "jesuitical" stand: ok, we promised forward ABI compatibility, 
but since we didn't say anything about API compatibility, we can do whatever 
we want with the latter provided that we don't break the former (we know that 
we are commited to backward API compatibility though).

But I also realized that patch #1 and #5 might break strict API *backward* 
compatibility (in the sense that the API symbol os_type can hold an unknown 
value for a 4.0.2 program), so I propose that we keep only the 5 other 
patches.

-- 
Eric Botcazou