New Curve for Radeon Bug

[Zarathustra]

Ok, here's a new curve - I tend to leave my machine in sleep mode most of the time rather then shutting down. It hasn't fully shut down in over a month and in that time I never experienced the Radeon bug.
Recently, I finally shut the thing down and today I got the bug twice, plus once yesterday.
Could it be that there's something misread at startup? That would answer why sometimes it happens and sometimes not and would explain why I had a bug-free June. I'm no programmer, but that's my layman's guess.

[mlinde]

Interesting thought. Try this next time you restart:

Hold down cmd-V (apple-V) until the screen starts up, it should show all the unix loading commands, look for something that isn't loading correctly.

[Zarathustra]

Uh.....give me an example - what would it say or look like?

[mlinde]

Well, I know that on my G4 I have an error that comes up (I've never actually tried to fix it) that reads like this:

Couldn't alloc class "NVDANV10HAL"

and on my old machine (8600 w/G3 upgrade) I get errors like

Can't find superclass for _20PatcheIOSCSICDDrive : 13IOSCSIDDrive

So a string that has "Couldn't" or "Can't" or even the obvious word "Error" might lead somewhere. If you do it, the easiest way is to just boot into open firmware (cmd-S) and then when you get to the command line, copy down the error (if there was one), and type "reboot" to restart normally.

[toby]

hmmm... I always get the 'Couldn't alloc class "NVDANV10HAL"' too. is it nVidia?