Comment 15 for bug 103223

Sitsofe, yes this is an interesting (and probably not altogether uncommon) case we should test for.

I am doubtful the automatic installation will be able to detect this case properly all the time, although if it can see that a monitor is connected to one card, and nothing is connected to the other, that ought to be a sufficient clue. However in the not-so-uncommon case where the monitor can't be detected, then it would seem reasonable to default to the onboard intel port (although I suppose one could argue that if there is a card beyond the onboard chip, the user's intent is likely to be to override the motherboard, but trying to guess user intentions can sometimes make things worse...) In that case, the approach Rashid took may be the best we can hope for.

Rashid, thanks, you've give sufficient information on this bug. It will probably take a while for us to get this fixed, but at least now I think we understand it.

Can anyone confirm this bug? My test systems all either are laptops or don't have a built-on graphics chip, so I don't have one to test with at the moment.