On Tue, May 03, 2005 at 07:34:37PM -0400, Timothy Miller wrote:
>
> > > What I said only applies to VGA modes. Anything x.org does bypasses
> > > the VGA emulation and uses the video controller directly, which has
> > > all the timing flexibility you need.
> >
> > OK. And specifying exact timings other than through xorg.conf would
> > only apply to the Fixed Frequency Monitor feature (when optional features
> > become open to consideration).
>
> Yes.
>
> And for fixed-frequency monitors, we'll pay attention to EDID. If the
> EDID is wrong or absent, there isn't a simple solution, although some
> ideas have been discussed.
I'd think a fixed-frequency monitor with EDID capability would be a
vanishingly rare case. The typical fixed-frequency monitor I've seen has a
3 or 5 BNC input, no electrical outputs of any kind, and not so much as a
NAND gate on-board. But if EDID were present, the board would have to
interrogate it and sync up with no help from the host computer, because
there's no way to make BIOS settings, install the OS, or do anything else
until you can see the default boot-time display.
Having wrestled quite a bit with fixed-frequency monitors and poorly
thought-out boards that create can't-get-there-from-here paradoxes, I've
distilled some specific ideas about the capabilities a fixed-frequency board
needs in order to cope with the real world.
What I don't want to do is divert more than minimal time and
attention to the problem right now. The spec has to stay stable while the
first-generation board is carried through to completion. The most that
might be done is to provide hooks to plug in an option board, but I'm
hesitant to suggest even that at the present stage of the project.
When it's time to open the floor again to additional features and
extra-cost options, I'll be eager to talk about requirements and some
possible design approaches. What I wouldn't be at all enthusiastic about is
half-measures. Been there, done that, got the tire tracks on my T-shirt to
show for it.
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
