280.97 KB, image/png
Created attachment 627342 [details] Nightly (20120524) - Screenshot (Galaxy Nexus, 4.0.4) Not sure if there's a bug already on this ... Recent regression on trunk (Nightly 20120525) is revealing black artifacts shown amidst the low-resolution tiles during fast panning on an assortment of pages. To illustrate exactly what I am seeing, see attached screenshot. Steps to Reproduce :: http://www.arstechnica.com, http://www.neowin.net, conduct some fast pans -- Tested via: Samsung Galaxy Nexus (Android 4.0.4) Mozilla/5.0 (Android; Mobile; rv:15.0) Gecko/15.0 Firefox/15.0a1
I think I saw something similar on the HTC Desire testing bug 757920.
The screenshot makes it look like java screenshot layer is confused about where things should be. I've seen something like this as well.
Following the steps from the duplicate bug (bug 759048): good build: May 24th bad build: May 25th Possible regression-range: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=f43e8d300f21&tochange=1dd0c5c6d9fd
My guess is this is caused by bug 748384 but it would be nice to have the regressionwindow narrowed down further.
Inbound builds: May 23 1337822875 – good May 23 1337825996 – bad Possible inbound regression range: http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=a2a6aee1aaca&tochange=19dbf7300434
I believe this was fixed with the patch from bug 759762. The build at http://stage.mozilla.org/pub/mozilla.org/mobile/tinderbox-builds/mozilla-central-android/1338466586/ has this patch and I cannot reproduce the problem on that build (at least on arstechnica.com). Could somebody else confirm just to make sure? Then we can dupe this to bug 759762.
This is fixed; I am unable to reproduce anymore either. I'm testing on an inbound build from this morning -- 05/31 using my original STR and also gave planet.mozilla and a Tumblr blog a pan through.