The nVidia card on my FC3 machine works fine, and other than
building an new module and re-creating initrd to include the
nvidia module after upgrading the kernel, I haven't had any
issues for a while.
On Mon, 2005-10-01 at 11:28 -0500, Jeff Spaleta wrote:
On Mon, 10 Jan 2005 10:37:17 -0500, Paul Iadonisi
> Take a deep breath, Jeff. ;-) Though I do think you are mostly
> right, you'll note that 'gslink(a)one.net' did, in fact, imply in his
> original post that it is nVidia that needs to provide an updated driver.
> Though I would be curious as to what the problem actually is, the
> implication didn't seem to be that the glibc update was the source of
> the problem, but rather that nVidia is lagging again.
Scapegoating nVidia whenever an individual sees a video related
problem on their system, while fun to do, isn't particularly
constructive. Other than's gslink's naked claim that the problem is
widespread and everyone running nvidia drivers on the list should
avoid the glibc updates what evidence is there that there is a
widespread problem? Did gslink cite a discusion where nvidia
developers have ackownledged a problem? Did gslink cite an on-going
discussion among nvidia hardware owners where problems have been
cross-checked and confirmed on multiple systems? There isn't much
supporting evidence to suggest this is a problem with the nvidia
libraries or kernel drivers or even a problem associated with the
glibc update. There could be a number of problems that gslink is
misintepreting as being something inside the nvidia drivers. I think
gslink's initial post is classic FUD, and i strongly challenge the
call for avoiding the glibc update that gslink made, without more
supporting evidence beyond gslink's personal experience. His comment
about rpm saying files are missing.. speaks of deeper issues wholely
unrelated to how the nvidia drivers and libraries actually operate.
More likely than not he used nvidia's installer instead of the nvidia
rpms and forced the removal of the xorg-Mesa-libGL which provides
libGL.so.1 thus breaking the rpm dependancy chain for ANY application
that requires libGL.so.1. Sadly gslink hasn't provided us with any
specifics as to the exact errors he's seeing, so I can't be sure thats
the problem. We can't be sure of anything really other thank gslink
feels that avoiding the glibc update is the correct course of action.
If avoiding the glibc update is something nvidia developers have
stated as a work-around to a widespread problem i demand the citation
to such an annoucement. Telling other users to avoid updating glibc is
pretty extreme in my opinion, and I want confirmation from multiple
sources that avoiding the glibc update is the prefered workaround...
otherwise this is at best unconstructive and at worst dangerous
advice. Luckily that glibc update wasn't marked as a security update ,
if it were, I'd actually be angry at the suggestion that people
shouldn't update it. As it stands, im just slightly annoyed.
-jef"There is a vast difference between asking for others to confirm a
problem that you are seeing and making a bold proclamation of a
solution to a widespread problem which you assume exists."spaleta
The Internet Centre