Hi Wesley, Are today's CPU's not quality then? The FDIV bug would have happened with or without peer review, simply because nobody outside the company has the tools to test the hardware before it ships.
For graphics hardware that problem is no different. And even when the hardware has reached end-users, only a tiny fraction of them will take the effort to do more than just reporting the bug. A couple weeks ago I bumped into an issue in my software renderer that took me several days to solve. It was seemingly simple to solve with a patch, but it actually required a significant redesign to solve it for good. No expert would have done the correct thing without spending a great deal of time getting to know my code, and taking the effort to do more than was strictly required to fix the issue. In fact I never ever got a good patch when it was still open-source, despite sufficient interest. Anyway, what I'm trying to say is that claiming higher hardware quality because of open specifications is flawed. Companies pay QA teams big money to perform tedious jobs, and you can't rely on the community to make any investments and be that rigorous. Driver quality for open-source operating systems will be higher though, because it's more accessible. Best regards, Nicolas Capens -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Wesley Parish Sent: maandag 9 januari 2006 10:06 To: [email protected] Subject: Re: [Open-graphics] Re: Vendors won't release driver source because offlaws in silicon? You only need to remember the infamous Pentium division bug of the mid-nineties. Peer review isn't optional - at least if one wants quality. Wesley Parish On Mon, 09 Jan 2006 12:40, Timothy Miller wrote: > An anonymous reader pointed out that it was inappropriate of me to say > that the S3 rep was "full of it". What that gentleman said is > correct. It would be better to say that the whole INDUSTRY is full of > it. If they're all afraid to release specs, that'll only make it > easier for this project to succeed and catch up with them. > > On 1/8/06, Timothy Miller <[EMAIL PROTECTED]> wrote: > > On 1/8/06, Lourens Veen <[EMAIL PROTECTED]> wrote: > > > There's an article over at Linux Today > > > (http://www.linuxtoday.com/infrastructure/2006010701126NWEV) with the > > > following interesting tidbit: > > > > > > "I talked to one of [S3's] reps (who gave the impression of being > > > fairly senior in the company), and he cast some interesting insight > > > into the > > > > whole > > > > > Linux graphics driver issue. According to him, it's not an issue of > > > > revealing > > > > > cool high end features to the competition that is keeping them from > > > > releasing > > > > > the high-performance 3D drivers into the kernel. It's that the drivers > > > > will > > > > > reveal flaws in the hardware that could be used against them by their > > > competitors." > > > > > > Could be interesting marketing for Traversal. "Our hardware is so good, > > > > we > > > > > actually dare tell you about its flaws" :-) > > > > I think he's full of it. Every chip ever made has flaws, and there > > are often elaborate things you have to do to work around them in > > software. Not to say that any hardware vendor is above being a > > hypocrite, but that's exactly what it would be. Ever looked at the > > ALSA drivers in Linux? The reason it took so long to get sound out of > > a good number of different chipsets implementing AC97 is that they all > > had different bugs to work around. Every Intel processor has bugs; > > they just have the ability to upload microcode patches to work around > > them. > > > > This is all very silly. We're going to have bugs. Even after > > extensive testing in FPGA, something is going to require at least a > > software workaround. And that's just life. > > _______________________________________________ > Open-graphics mailing list > [email protected] > http://lists.duskglow.com/mailman/listinfo/open-graphics > List service provided by Duskglow Consulting, LLC (www.duskglow.com) -- Clinersterton beademung, with all of love - RIP James Blish ----- Mau e ki, he aha te mea nui? You ask, what is the most important thing? Maku e ki, he tangata, he tangata, he tangata. I reply, it is people, it is people, it is people. _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com) _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
