[forum] "Drivers? We don't need no stinking..."

[David Wexelblat]

> -----Original Message-----
> From: forum-admin@XFree86.Org 
> [mailto:forum-admin@XFree86.Org] On Behalf Of Alan Cox
> Sent: Monday, March 24, 2003 7:26 PM
> To: forum@XFree86.Org
> Subject: Re: [forum] "Drivers? We don't need no stinking..."
> 
> 
> > How the @#$& can people write/fix drivers when they cannot 
> > obtain the required register-level documentation for
> > the graphics chips?
> 
> The are documented drivers not ported from 3.3.6. There are 
> documented drivers with bugs. There are unaccelerated 
> documented chipsets. There is full documentation for old sis 
> 3d, for trident 3d and there is broken code for sis3xx 3D.
> 
> It isnt just about documentation, although that is a 
> significant factor.

FWIW, this is probably a place where the distribution vendors could help out
- by providing a financial incentive to the people skilled enough to do the
port/reimplementation. The people who have the skillset are, in general, not
particularly interested in (and often don't have access to) the old
hardware. If the distribution vendors were to contract out the work, that
might entice someone who would otherwise spend their free time playing with
new toys.

Back when I was playing with these drivers (in the dark ages when we walked
to school in our bare feet in the snow), going one generation back was
boring, two was insanely boring, and three wasn't even worth considering. I
can't imagine things have gotten any more interesting for the volunteers.
Back then, if someone had paid me a thousand bucks, I would have done the
back-rev work.

Then again, someone just posted a Hercules monochrome driver for 4.x on the
devel list, so what do I know about what motivates people?

FWIW2 - I sent my entire archive of old chip documentation to David Dawes. I
had docs on every VGA, SVGA, first- and second-generation PC graphics
accelerators, and most of the next couple of generations. Plus every bloody
RAMDAC and clock chip. It was a HUGE collection (about 10 linear feet of
databooks). Back in "the day", when I was doing SuperProbe, I collected
everything I could get my grubby mitts on from every vendor out there. The
point being that the docs exist for a LOT of the stuff that hasn't made it
to 4.x.

FWIW3 - If you want to learn how to do this sort of stuff, porting an old
driver to XAA, using the 3.x code and the old docs to learn about the
chipsets, then looking at docs for a chipset that has an existing 4.x
driver, will be a really educational experience. I got started with all this
by studying Thomas Roell's ET4000 code and a databook for a few days...

-- 
David Wexelblat, Chief Architect    mailto:DavidWexelblat@aol.com
America Online, Inc                 http://www.aol.com/
44900 Prentice Drive - 24B:P08      (703) 265-1158 (voice)
Dulles, VA 20166                    (703) 265-1301 (fax)
	
Please send private email to: mailto:dwex@xfree86.org