On 5/1/05, Patrick McNamara <[EMAIL PROTECTED]> wrote: > >We're not trying to meet the VGA spec. We're tying to FAKE meeting > >the VGA spec well enough that (a) programs think we meet the spec, and > >(b) you see basically what you're supposed to see. > > > > > Hrm, if programs think we meet the spec, doesn't that mean we do, at > least in some manner? :)
Well, just because the register exists doesn't mean we pay attention to it the way they expect us to. Say, for instance, there's a way to specify video timing. We are only intested in the resolution. We'll totally ignore the clock programming and other timing numbers. > I was taking the "it just works" approach. If we really aren't that > worried about VGA compatibility the we can cut corners with a lot of the > VGA control registers. They won't necessarily have a lot of meaning. > We just need to make sure they are initialized to a valid valid when we > switch modes and then we can pretty much ignore what is written to them > so long as we return it for any subsequent reads. Right. > > I haven't read the VESA specs yet so I don't know what they say about > such stuff, but how my effort do we want to put into conforming to them? Enough effort to make DOS, Windows, and Linux text console work correctly, not to mention BIOS config screens, and all that. If some other stuff works too, well, great. We want it to be usable to the user, not necessarily even always wonderful looking. Although, we may want to do pixel-doubling in 320x200 mode. > >So, if some DOS-based game doesn't sync to frames correctly... well, > >what the heck are you doing running DOS-based games on this thing > >anyhow? > > > > > > > People do some really strange things. Well, with this card, they might want to lower their expectations for a FEW things. :) > >Well, what we're doing is trading off logic cells for memory cells. > >We'll design a minimalist processor and have it execute out of a RAM > >block. > > > > > > > I never have wrapped my head around FPGA internals. I'll let you > experts handle that. I'll just keep tossing ideas out and seeing what > sticks. :) > Well, you go in a direction I like. Whether or not that's a GOOD direction is for history to judge. :) _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
