To view parent comment, click here.
To read all comments associated with this story, please click here.
I don't need to look into the code and do a thorough reductionist analysis when the behavior is crystal clear: everything but Qt4 is slow on my system. If *everything* was slow, sure, blame the drivers. However, I find it very hard to believe it is just the drivers' fault if only one toolkit ever seems to have a problem. And if it is the drivers' fault, it's because the toolkit is making use of edge cases which the other toolkits can clearly get by without using and Qt4 should do the same. Again, don't need to look at code to know the correct answer here.
It should not take noticeable amounts of time to redraw a section of a double-buffered window during an exposure event. I don't need to look at code to know that that is bad behavior on the part of the toolkit.
Some of the responses you are receiving are indeed bizarre.
Operator: 911. May I help you?
Caller: My table saw just went haywire and cut off my arm!!! I'm losing blood fast!!! Help!!!
Operator: Have you gotten an anatomy book and studied it carefully to see exactly what the problem might be, and why owners of other table saws might not have had their arms cut off?
Edited 2009-08-05 14:49 UTC
But at least you need to listen to people explaining what the problem is instead of dismissing them offhand, and blaming the toolkit. Like it is known that kwin exposes more features of the graphic card than other compositing window managers. This is how bugs are found in the drivers, and actually NVidia acknowledged that the problem is on their side some time ago and they issued a fix - but don't let that bother you. Just keep harping on how KDE sucks for YOU ignoring other's comments or requests for more information. You're at 10% now (of the total number of comments) basically repeating yourself over and over again. Keep up the good work, and you may reach 20%, maybe then you'll accept that we HEARD YOU THE FIRST TIME, and you can stop wailing...
It should not take noticeable amounts of time to redraw a section of a double-buffered window during an exposure event. I don't need to look at code to know that that is bad behavior on the part of the toolkit.
everything thing but Qt4 is fast on my system.
You can try another variable - the drivers and the graphics chipset - as has been explained quite a few times in this thread and as I'd explained in another post. If it's KDE 4's fault then if you vary the graphics driver and the chipset then the issues should remain. They don't. ATI's Linux drivers are junk right now if you expect everything to work well.
Either you can't face up to the fact that ATI's Linux driver support is pretty useless or you're labouring on this for reasons that are best known to you.







Member since:
2006-03-05
It's time you jumping to conclusions you are not qualified to make. Do some research -- dive into some code, from kwin down to the X11 driver and then come back with a thorough analysis. You then will have helped a lot of people get informed and will have gained some respect. But right now, you're just spouting nonsense.