Comment 336 for bug 351186

I hope it's not rejected. I understand the desire for the correct solution, and on the face of it you could argue that the fglrx driver should not have slow performance with the existing xorg code, however I think we should be pragmatic given the facts. Lots of people want to use the fglrx driver and want to use desktop effects. It's not nice to do driver specific hacks but looking at it another way it's a bit like having having to deal with different driver capabilities to provide useable results, something that's not unusual in the real world of software on any platform. I might be wrong but it doesn't seem that the conditional patch would have lots of negative effects on many other clients, I suppose with extensive testing the patch could show some other issues. I guess the days of manually patching your own Linux system are not over. I think I would have to get the latest source next time and apply the patch and make sure I don't make the same mistake with the package version!