I'm running win2000 comm build 2001042504. Steps to reproduce: 1) Load http://www.activestate.com/ in the browser. 2) Launch mirc. 3) Move mirc ontop of the browser window. 4) Let go of mirc. 5) Slowly drag mirc off of the browser window *notice the trail of gray, etc* 6) scroll around in the browser to force a refresh, or quickly move mirc over the browser without dropping it on the browser to refresh. will attach a screenshot. This bug may be a dupe, or related to bug 77634

this may be due to hyatt's new performance checkin.

No, this is not related to hyatt's checkin, since I (and you jelwell) can see it in the builds from this morning, prior to hyatt's checkin (e.g., 2001-04-25-09-trunk). However, I can't reproduce this in the build of 2001-04-21-10-trunk), so this is (apparently) recent.

(I should have said 'reproduce on win2k'. Haven't looked at linux/mac).

Is this bug a dupe of bug 74097? They sould very much the same? I will let someone with more info add a 2nd openion before marking it.

I don't think 74097 has much to do with this one. That one is about loadig a specific page and scrolling down, then back up and seeing remnants from that scroll. This one is about apps other than Mozilla leaving remnants of their windows on Mozilla's layout. Mozilla doesn't seem to do this to itself...only other apps on top of Mozilla. One bug that is almost certainly a dup is bug 78183. Take a look at the image attached to that bug. It shows the problem even more than the image attached to this bug. I've been seeing this for a while and certainly before Hyatt's checkin. Happens on win2k for me at work and at home which have different video cards, so it dosnt' seem to be the fault of the video card. This one needs to be fixed fast. This is really ugly! jake

*** Bug 78183 has been marked as a duplicate of this bug. ***

*** Bug 78402 has been marked as a duplicate of this bug. ***

Is this a regression from bug 76495? Can someone narrow down the builds?

jrgm at 2001-04-25 19:23 said: > No, this is not related to hyatt's checkin, since I (and you jelwell) can > see it in the builds from this morning, prior to hyatt's checkin (e.g., > 2001-04-25-09-trunk). However, I can't reproduce this in the build of > 2001-04-21-10-trunk, so this is (apparently) recent.

*** Bug 79115 has been marked as a duplicate of this bug. ***

Additionally regarding the previous attachment. The image corruption is from scrolling up and down in the browser. The images are repainted if I cover them with another window or move the mouse pointer over them. Other images also have painting problems. There is a GIF banner at top of the page that 'sometimes' doesn't appear when the page is first loaded. Also there are some 'bullet' gif images that have 'missing lines'. At first I thought this was a problem related to interlaced gifs. But the problem with the jpeg car images disproves that. Platform: RedHat Linux 7.1, XFree86 4.0.3, ATI video driver on ATI Rage II (Mach64). Build: Moz/5 0.9 (2001050716) This didn't happen in Moz/5 0.8.1. I'll attach more example images.

Build: 2001050515 (Moz 0.9) on Win98 Notice that dragging another window over the Mozilla browser window leaves a trail of unrepainted ripples, over the whole UA (menu/toolbars, etc), not just the content area - only the title bar is painted properly. This problem was not occurring in Build: 2001032319 (Moz 0.8.1)

*** Bug 80699 has been marked as a duplicate of this bug. ***

this appears to WFM on build #2001051604 on win2k Jake

Same here on WinMe. The fix for this, if there was one checked in recently, doesn't seem to be very CPU-efficient, though. By dragging my Wintop window over top of Mozilla, I can see that Mozilla is using 100% of available CPU time. I'm on a PIII-800, and I can "feel" it sucking up the cpu. (slightly jerky, unresponsive)

Hmmmm I saw it spike at 42%, but average between 20 - 30% CPU time when I moved an IE window eradically back and forth over a Mozilla window. Doing the same thing with Mozilla over an IE window, IE spiked at about 30% and averaged between 15 - 27%. I have a PII 266 with 128 meg RAM, a Matrox Millenium II video card with 4 meg RAM, and running win2k SP2 Jake

The test case I'm using here is http://www.mozilla.org I am dragging around a Wintop window pretty quickly and watching the mozilla process. The Wintop window is about 1/4 the size of the Mozilla window. Mozilla uses about 85% CPU IE uses about 25% CPU NS4 uses about 18% CPU

Recent builds don't have the original problem. Marking as WFM. The problem of CPU usage should be filed as another bug.