Re: [AD] Rewritting set_gfx_mode (was: GFX_WINDOWED) |
[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
- To: Allegro Conductors <conductors@xxxxxxxxxx>
- Subject: Re: [AD] Rewritting set_gfx_mode (was: GFX_WINDOWED)
- From: Eduard Bloch <edi@xxxxxxxxxx>
- Date: Mon, 9 Apr 2001 07:37:45 +0200
#include <hallo.h>
Grzegorz Adam Hankiewicz wrote on Sat Apr 07, 2001 um 11:16:18AM:
> Besides, 3) doesn't really help binary package distribution of Allegro,
> because Eduard may create a precompiled Allegro for Debian which doesn't
> use xxx, and the users having this xxx won't be able to use it because
> there are only dummy symbols for it despite the fact that the lib is
> there, forcing the mantainer to keep with the different precompiled
> versions dance or forcing the user to get the source package and build it,
> right?
Is not really true. Game developers have just to install the BIG library
and keep in mind that they should not rely on a certain driver and use
common functions instead. I imagine each Allegro package build of a) the
library and b) the dummy library containing the symbols that the real
library doesn't have, i.e. padding the set of symbols to a "compability
level". The dummy library will be preloaded before the real library. If
neccessary, Allegro should be modified to detect if the driver/function
is really available or is just a fake.
Pros: this should work even with old binaries. I have allready a game
package which is compiled with Allegro 3.9.34 and definitely won't run
with 3.9.35 because of staticaly built-in symbols. Symbols to functins,
the game doesn't use whatsoever!
Contras: The binaries still have to contain lots of symbols, not used at
all.
> I am voting for a set_gfx_mode rewrite (possibly seconded by everybody),
> and for implementing it myself (which is surely a bad idea :-).
> Suggestions?
I am not a C expert but I could help doing it this way. (Since I have
more spare time now ;).
Gr{us,eeting}s,
Eduard.
--
====================================================================
Eduard Bloch <blade@xxxxxxxxxx>; HP: http://eduard.bloch.com/edecosi
0xEDF008C5(GnuPG): E6EB 98E2 B885 8FF0 6C04 5C1D E106 481E EDF0 08C5
**
Diagnose? Erklärungen? Reproduzierbare Lösungen? Sowas wollen nur
Leute, die von EDV nichts verstehen.