Ed Wilts wrote:
On Tue, Apr 08, 2003 at 06:45:52PM -0500, Thomas Dodd wrote:Perhaps if you told it an obvious wrong chip, it would fallback correrctly? What about using --set-driver=vesa ? Does it hang or does it use the vesa drive like you said to?
If I explicitely tell it to use the vesa driver, it works properly. If I give it an obviously wrong one, startx fails with a Fatal server error: no screens found. This was with i810, s3 and trident. So, at least the savage driver thinks it should be able to work.
Why is startx in here? I was talking about rh-cfg-xf. That's where I saw the fallback to vesa. Does the XF86Config file created have the vesa driver in it?
I didn't have this system available during the beta. I successfully installed AS2.1, 7.3, and 8.0 and wouldn't even thought of having issues with something like the graphical install of phoebe or 9. I do have to wonder why anaconda tries to use the native savage driver when the vesa driver is more likely to be successful. Sure, use the savage driver later (which in my case still fails), but at least let the user do the install properly. Which leads to another question - how do I tell anaconda to use the vesa driver instead of the savage driver?
I think there's an option to tell anaconda to use vesa instead of native XF86 drivers. There's always the text mode install.
I try to watch that bug, and see what happens.
-Thomas
