Confirm
313
edits
Changes
→9/5/2013
= 9/5/2013 =
* Milestone 9
** http://i.imgur.com/RnD9h.gif
* Perf regression and other landing blockers
** bug 880611 (ts_paint regression)
*** We thought we could chalk this up to TabsInTitlebar._update and CustomizableUI.initialize and put this thing to bed, but that doesn't account for all of it.
**** We're operating under the assumption that the remaining regression is related to the fact that reflows are (for some reason) slower on UX than they are on m-c. We're still trying to determine why that is the case.
**** Gijs' analysis of reflows: https://bugzilla.mozilla.org/show_bug.cgi?id=902857#c31
** bug 902024 (TART regression)
*** TART regression data: https://docs.google.com/a/mozilla.com/spreadsheet/ccc?key=0Asj8iLTl0K0UdGJzVnVYYkJTSDREbndRWWFTLWNyS1E#gid=1
*** Bug 907544 has landed, which gives wins for both m-c and UX, but doesn't give UX much of an advantage (except that it seems to have caused us to beat m-c on tpaint for Windows XP - http://bit.ly/14uhJzz)
*** We're still fishing for the cause here... we're creating some reflow profiles for layout to look at. We've got mstange on board to help us examine those.
*** vladan is looking into getting bug 900524 fixed so we can get some better data on Windows XP
** regression data: https://docs.google.com/spreadsheet/ccc?key=0Asj8iLTl0K0UdDcxemJEWVVXWTlKek1wR09iS2FsOXc#gid=0
* Bug tracking (8/29 to 9/5)
** P1: -7 (net -4)
** P2: -1
** P3: +1
** P4: +0
** P5: +2
** Firefox team coordination
*** Need an update from Dao on bug 868640 (Implement private browsing indicator when menubar is hidden)
**** Assigned for over 1 month, needinfo for 1 week
**** Might need a new assignee at this point
*** Still waiting on graphics from Stephen for bug 892076 (customize and help button in panel UI)
**** Will ping on IRC
** Cross-team coordination
*** Our Jetpack issues have been resolved, big thanks to Gijs, Zer0, and Kwierso
*** Gfx
**** Bug 907544 was fixed by BenWa
**** Still hoping we can get some attention on bugs 907546, 907426, 907542
***** We might want to reduce the priority of these (or remove blocking status entirely), since I found out yesterday that these will only really affect Windows Vista+, which we already kick butt on.
****** Sounds good to me!
*** Splitting up mochitest-bc (bug 819963) is no longer a blocker due to bug 894930 being fixed
* Roundtable
** https://bugzilla.mozilla.org/show_bug.cgi?id=882009
* Action Items (Last Week)
** (MattN) look for worst regressions on subtests across platforms
*** https://docs.google.com/a/mozilla.com/spreadsheet/ccc?key=0Asj8iLTl0K0UdGJzVnVYYkJTSDREbndRWWFTLWNyS1E&pli=1#gid=4
**** 10.6 is in the worst shape both in terms of the regression % (19.38%) and absolute times
***** Can check if lion+ specific styles are related? MattN will look at this
**** 10.7 (9%) is second worst
*** Am I correct that we should still be leaning on XP to get that cleared, since that's the "major" platform that is regressing?
**** Yes, I think so but we can investigate 10.6 in parallel
* Action Items (New)
** [mconley] Gather a TART profile on OS X 10.6 and 10.8 and determine why 10.6 is so slow
=8/29/2013=
