Bug 273690

Version:           unspecified (using KDE 4.6.2) 
OS:                Linux

I often work with many window open on several workspace / desktops. I use no deskop effects or animations when switching windows. With few window the alt-tab list comes up fairly rapidly with only a slight but acceptable hesitation. But with many window (the alt-tab list approaching the screen height, say) the hesitation or latency after pressing alt-tab before the list appears approaches a second or more. However, regardless of how many windows are open or how slow the list appears, once the alt-tab list appears, scrolling though it is perfectly fast.
I have 4G of memory and don't seem to be using even half of it according to "system monitor" graphical display. The graphic card is radeon of some sort and using the free mesa driver which works well with other apps (e.g., smooth movement with google earth now, unlike with older free radeon drivers) so don't think this is a problem caused by video performance.

Note: The latency is more noticeable when some other activity for several minutes occurs between alt-tabs. If alt-tabs are done back-to-back timewise, the latency is not as apparent. 

Reproducible: Didn't try

Steps to Reproduce:
Open lots of windows on two or more desktop
Work in one window for a few minutes 
Do alt-tab to change to another window 
Time between pressing alt-tab and seeing the list of windows can about 1 sec.

(As noted above, if alt-tab pressed again with little delay since the last time it was pressed, the latency delay is not as large.)

Actual Results:  
Can take about a seconds for alt-tab window list to appear.

Expected Results:  
Window list should appear "immediately" when alt-tab is pressed regardless of how many windows are open and how much time has elapsed between alt-tab presses.


Previous used KDE 3.5.x and it had zero latency when popping up the alt-tab list regardless of conditions. Kde 4.x has always been sluggish in this regard compared to 3.5 on several hardware platforms I have used.