Mail Archives: djgpp/1999/11/25/13:05:32
Chaos writes:
> So it apears that even covering most of VESA errors I won't
> be 100% sure if my aplication succeds due to still new video
> cards appearing?
The problem is not so much newer cards, as the number of old ones that
are broken in different ways. Some bugs can be worked around, but it
isn't always possible to do that, at least without breaking things for
other drivers.
For example many BIOS implementations have problems with the 0x4F06
VESA call. Some will just return an error code when you call this, in
which case you can try to do the same thing using a direct VGA register
write instead, but that register write doesn't work on all cards
either, and other VESA drivers will return a success code from 0x4F06
but not actually make the right adjustment to the hardware, which is
obviously impossible for you to detect or work around.
> Well, I think even Allegro will be weak at this field
Of course. Some drivers are just broken, and there is no fix other
than just to tell people not to use them. This sucks, but I don't
know what else you can do about it.
> Do I really have to implement VESA 1.2 Bank Changing?
> I' ve searched within my friends someone with video card old
> enough not having VESA 2.0.
> I've only found 1 guy besides using UniVbe. So, do I must stick
> with slow (int 0x10 yuck!) bank changing?
That's really impossible to answer: it depends whether you care about
supporting people who only have banked mode graphics drivers.
Personally I think this is worth having as an option, as I know several
people who use Allegro in banked mode, and would be totally unable
to run it if this wasn't supported. But banked graphics are certainly
far harder to implement, so if you feel that all your target users
will have at least a VBE 2.0 driver, you might not want to bother with
this.
Shawn Hargreaves.
- Raw text -