radeon driver heading in wrong direction :-(.

Marko Vojinovic vvmarko at gmail.com
Sun Jan 31 16:35:57 UTC 2010


On Sunday 31 January 2010 14:25:39 Gilboa Davara wrote:
> On Sun, 2010-01-31 at 20:30 +0800, Ed Greshko wrote:
> > In my case, none of the above...
> 
> Oh, come on!
> I'm using the same binary driver as you are.
> And yes, nVidia binary driver, while -far- better than ATI's driver,

Lesser of two evils, right? ;-)

> has
> had it's share of issues.
> E.g.
> - Initial F12 xorg + nVidia driver combo resulted in unbelievably slow
> performance under KDE.

But this is worked around by now. Unbelievable, given the fact that those are 
actually closed source drivers, yes? :-)

> - Legacy driver releases tend to lag the "current" driver badly. In a
> desktop, you could always switch to the latest version, but you laptop
> still carries a GF5600M, you're more or less screwed.

Umm, my initial comment that spawned this part of the thread was about advice 
on buying *new* cards, not buying *old* cards. If you want to buy an old card, 
feel free to buy a new Intel card instead, you'll get the same level of 
performance.

> - Xen kernel were never supported by nVidia.

You want 3D graphics in a virtual environment? To what purpose? Playing quake3 
on a mail/web/file-server under a virtual machine?

> - Having to compile a kernel without 4K stacks for months, until nVidia
> added support for it.

As compared to ATI not providing support for current version of X for the same 
number of months and still counting? Tricky question: what is easier --- 
recompiling a kernel, or downgrading X? :-)

>  .... Again, nVidia is doing an admirable job at keeping their drivers
> stable and current (compared to say, ATI or Intel Poulsbo), but claiming
> the using them do not come at a price, is ridicules, at best.

Fair enough. But this price is lower then in ATI and Intel case, at any rate. 
It's not perfect, but is just the best offer available.

Best, :-)
Marko




More information about the users mailing list