Mail Archives: djgpp/2000/10/02/08:29:29
Peter Remmers <Peter DOT Remmers AT t-online DOT de> wrote:
> Hans-Bernhard Broeker <broeker AT physik DOT rwth-aachen DOT de> schrieb ...
>> Edmund Horner <ejrh AT paradise DOT net DOT nz> wrote:
>> > Can the mmap() function be implemented in the DJGPP libc?
[...]
> Wouldn't it be possible to write a filesystem extension for
> say "/dev/fb0" which sits on VBE2.0, and then have a mmap() for getting
> a pointer to the framebuffer?
Possible: yes. But it wouldn't gain us a thing, compared to direct
access to the framebuffer via the farptr functions or nearptr
mode.
We'ld have to build a whole tower of things on top of each other
(framebuffer access via farptr or nearptr, plus an FSEXT, plus a
mmap() imitation) in order to simulate portability where none really
exists, not even among different flavours of Unix.
Stuff like graphics hardware access *is*, unvariably,
system-dependent, so there almost no point at all in trying to use
'portable' features to implement it. We'ld be adding new problems,
instead of solving the existing ones.
That's why X11 is so popular on Unix boxes: you don't have to care how
that X server accesses the graphics hardware, at all. Your app talks
X11, and that's it. Let the X11 implementor (or, in the context of
DJGPP: the makers of Allegro and GRX) deal with the gory details.
--
Hans-Bernhard Broeker (broeker AT physik DOT rwth-aachen DOT de)
Even if all the snow were burnt, ashes would remain.
- Raw text -