Kaiwai Gardiner wrote: > On Thu, 2008-05-01 at 09:52 -0500, James Cornell wrote: > >> Ben Taylor wrote: >> >>> James Cornell wrote: >>> >>>> I'll let others suffer for now with RadeonHD, I still don't trust >>>> ATI, especially since they are now part of AMD, the management >>>> perspective is a bit blurred and to me, untrustworthy at the moment. >>>> It's not proven, and neither is the track record under AMD >>>> management, and performance wise NVIDIA has never let me down. >>>> >>>> >>> AMD/ATI are releasing the specs for the R500/R600. >>> >>> Suse/Novell is driving RadeonHD, but it appears that RedHat is pushing >>> the radeon driver from their end. This can be confirmed by the fact >>> that my x1550 is supported by both radeonhd and radeon using >>> the current git head. To my eye, the radeon driver is better performant >>> compared to the radeonhd on my 32-bit Linux box. I've not been >>> able to compile radeon git on SXCE. >>> >>> Ben >>> >> Exactly why it doesn't and will not matter to me for the foreseeable >> future. The hardware is definitely wimpier, why else would they even >> consider pushing specs, they're dieing of course. The main problem here >> is "I've not been able to compile radeon git on SXCE" that gives me a >> lot of confidence. Sun needs to support it, not the community, they >> made a deal with NVIDIA to distribute drivers in SX:CE and SX:DE as well >> as Solaris 10, and even though the specifications are open, quality >> control wise, until we can even get some sort of module loading, how can >> we trust the module, let alone debug things that are alien to the >> system. (Linux is the main target, giving us the nasty bits due to >> retarded programming model on their part) >> > > Hmm, and the Linux retardness doesn't stop there, oh no, opensource > applications seem to be teeming with linuxisms, where the only > understanding of 'testing' means 'does it compile'? > > >> Sorry if this offends anyone, but my point is, it's nice to see >> specifications out, but it does no justice for implementation, let alone >> on OpenSolaris. The confidence is not there about it, nor is there >> proven test cases or any real packages supporting it to any degree. >> Performance is also questionable to those who have founded upon working >> solutions. If I were to upgrade my notebook, I would surely consider >> RadeonHD because performance could be theoretically better, but on the >> desktop, sorry AMD will never compete based on past, present, and future >> unknowns. I'll let you guys have the brunt of the pain, sorry. >> > > For me, it goes back a long way. ATI have been screwing over Linux users > for years, and Xorg developers for years (back to Xfree86 days) - quite > frankly, AMD did nothing within the first months of owning ATI; and > quite frankly, they would have had to do *alot* to win back any possible > support. > > Apart from a few fruity fringes, most OSS/Alt-OS people I know are > rigidly keeping away from ATI/AMD due to their past. Its up to AMD/ATI > to prove themselves rather than expecting that the act of waving around > code like some engorged bodily appendage makes up for years of screwing > over developers. > > Matthew > > I think you nailed it. This is my REAL issue... they've been screwing me over for almost half my life and I'm not even that old. I personally boycott them. I'm looking into Intel for future notebooks if the need comes up where I don't need a dedicated GPU. I'm sure this offends ATI diehards, but then again the fact remains their driver quality even for Windows users for the last decade has not been even comparable to NVIDIA, and in some cases worse than Intel's cards before they actually put effort into them.
James James
