RE: Re: "Drivers? We don't need no stinking..."

> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]] On Behalf Of Sven Luther
> Sent: Wednesday, March 26, 2003 7:30 AM
> To: [email protected]
> Subject: Re: [forum] Re: "Drivers? We don't need no stinking..."
> 
> 
> On Tue, Mar 25, 2003 at 10:14:27PM -0500, David Wexelblat wrote:
> > > With the Wretched Offender's 'kdrive' as an example, perhaps
> > > it is possible to prune the code.  As for chips, maybe I
> > > should replace my aging G400 and 300 MHz K6-2 with a
> > > 2 GHz CPU or two, and run dumb framebuffer mode; that was
> > > Intel's suggestion, several years ago.
> > > 
> > 
> > Vote with your dollars. If you are committed to the Open 
> > Source-only path, the don't by hardware from vendors that
> > don't release docs. This has worked in the past, rarely.
> 
> Well, but not these days anymore, what would you end buying ? 

Didn't work all that well in the past, either. It worked with Diamond, and a
couple other instances.

> Nvidia is out, matrox was giving docs in the past, but with 
> the new hardware (parhelia), i think there are not so open 
> anymore, true, who would buy a parhelia, but still. ATI does 
> release docs, but they don't release all of them too, 
> especially in the Video domain. I don't really know about the 
> others, but none release docs without you signing an NDA 
> (BTW, are all the docs you offered NDA clean ?), and anyway, 
> often you cannot even buy their hardware. 
> 

I didn't say this was easy. Suing them to have the release the docs isn't
going to go anywhere either. Incenting them to do 4.x loadable drivers is
better than having no support at all. Incenting the vendors to do their own
drivers, Open Source or binary, is still really the best approach - this
hardware is SO complicated that really good drivers almost certainly can't
be done without close consultation between hardware and software developers.
At least, not in a competitive timeframe - I recall it taking a year or more
for really, really good driver writes (like Mark Vojkovich, back before he
worked for NVIDIA) to get things tuned and up to snuff once he had docs. A
big part of that was that he was doing it part time (Thomas Roell at XiG,
for example, was able to things much more quickly, based on two factors -
one, it's his job, and two, he's always pretty much been in the upper
echelon of driver writers anyhow).

And yes, all the old docs that I had (which David Dawes now has) are
NDA-free. I had a policy of never touching NDA docs back then. Not because
of some OS religion; I simply had no interest in any issues related to doing
Open Source on NDA'd hardware back in the very early days when this was
very, very touchy stuff. The NDA stuff didn't really become an issue until
the 3rd generation of accelerators, IIRC (some companies were anal, but not
many - certainly not the way it is now).

> Friendly,
> 
> Sven Luther

-- 
David Wexelblat, Chief Architect    mailto:[email protected]
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:[email protected]