FC3: Mesa and 'nvidia' driver

Neal D. Becker ndbecker2 at verizon.net
Tue Nov 23 14:01:31 UTC 2004

Colin J Thomson wrote:

> On Tuesday 23 Nov 2004 12:49, Neal D. Becker wrote:
>> Alexander Volovics wrote:
>> > On Mon, Nov 22, 2004 at 06:09:54PM +0100, Alexander Dalloz wrote:
>> >> Am Mo, den 22.11.2004 schrieb Alexander Volovics um 16:35:
>> >> > Is it still necessary to remove the (xorg-x11) Mesa-libGL and
>> >> > Mesa-libGLU rpm's before installing the nvidia driver.
>> >>
>> >> The NVidia driver package comes with it's own GL libraries. I would
>> >> physically uninstall the Mesa-libGL RPM and reinstall it with switch
>> >> "--justdb". So RPM thinks it is installed and you are not annoyed by
>> >> future RPM actions that the missing package fails in dependency. So
>> >> did I.
>> >
>> > Thanks for the reply. I will use the neat "--justdb" trick too.
>> >
>> > Alexander
>> But if rpm thinks it's installed, then it will try to install updates. 
>> If you enable automatic nightly yum it will overwrite the NVidia version.
>> You can always re-install the NVidia manually to fix this, but it is
>> annoying.  Any better solution welcome.
> I find this interesting, I have never had to remove *any* libGL rpm to get
> nvidia's driver to run on my system rh8/fc1/2 and the driver seems to do a
> good job backing up the original libGL files..
> If there is an upgrade to xorg, I usually uninstall nvidia's driver do the
> update to xorg and resinstall nvidia's driver..
> [colinjt at localhost colinjt]$ rpm -qa|grep libGL
> xorg-x11-Mesa-libGL-6.7.0-10
> xorg-x11-Mesa-libGLU-6.7.0-10
> Interested in any views,

>From what I can tell, it's just a matter of luck.  NVidia puts lib
in /usr/lib (or /usr/lib64), while Mesa is in /usr/X11R6/lib, so they
actually can coexist, as long as you are configured to search /usr/lib
before /usr/X11R6/lib.  I'm guessing that's how things are working, but if
true, it sucks.

More information about the users mailing list