Electrolysis/Meetings/2015-11-12

From MozillaWiki
Jump to: navigation, search

« previous week | index | next week »

roundtable

jimm

  • bug 1190364 (m8, plugin info sometimes fails with e10s) - patch landed to clean up some potential failure in chrome apis that gets ignored by content process. Waiting to see if it helps.
  • bug 1096804 (m8, taskbar previews support) - wip
  • bug 1222662 (mouse cursor bug with plugins) - investigated, have a fix in mind.

bobowen

  • bug 1156742 - print to xps with low integrity sandbox - fonts are really hard - fixed various problems found in testing:
    • memory corruption from the way I was pulling out names from TrueType files
    • timeout due to misunderstanding the way nsPagePrintTimer works
    • added a load of null checking to the Moz2D event playback code, as it currently just expects everything to be fine
    • added a new patch to only send the font data once and cache on the parent, scaled font then reference that
    • just found issue with TrueType font collections for GDI fonts, working on fixing that now

tracy

  • pre-triage of e10s-perf bug list
  • bug 1221159 re-enable a few working tests for e10s
    • sorted out previous backout. Now waiting on ok to re-enable more tests to land new patch
  • bug 933103 - a few test cases need to be reworked to use a new version of synthesizeMouse in BrowserTestUtils.jsm
    • basically the test cases need to be rewritten to make use of modern test utils.
  • if you hadn't noticed, crash stats had major miscalculations recently; ie. reports shows DevEd at 2000 crashes per 100 active daily install.
    • Tracked in bug 1223788 - Wrong ADI data for Nov 10, 2015

mconley

  • M8 bug 1209689 - Crashed tab indicates all tabs have crashed and every tab loads the crashed tab page
    • Waiting on last reviews.
    • This will make it so that if the content process crashes, the background tabs will be restore on demand, even if the user has restore on demand disabled. This is to avoid a crash loop in the event that the crash was caused by a background tab.
  • M8 bug 1171708 - [e10s] Stop using CPOWs for window closing
    • Patches up. Writing a test for an edge-case that billm noticed yesterday.
  • M8 bug 1177310 - [e10s] Stop using CPOWs on application shutdown
    • Starting on this. This is my last M8 (for now).
  • bug 1193838 - Make it possible to gather profiles from processes that have exited, and bug 1103094 - Start profiling subprocesses if the parent process is already profiling
    • Haven't started on this yet - hoping too once I get my last M8 out the door.
  • The usual reviews
  • Gave JS / DOM onboarding. Also been doing lots of debiased interviewing duty.

Giorgio

  • Released first prototype of the Tails Download & Verify Extension ("dave"), learned lot of stuff about what's missing to make download managers and encryption-oriented add-on work in the WebExtension era
  • Removing minor e10s bugs in XUL-based NoScript and FlashGot
  • bug 1213483 Complete the implementation of chrome.webRequest

felipe

  • doing e10s experiments work with performance team
  • in talks about figuring out how to test in Beta
  • bug 1199841 and bug 1210617 - Downloads in PB window go to non-PB window
    • Patches waiting for review
  • bug 1109146 - link target no longer saved when browser.altClickSave is set to True
    • patches reviewed, need to look at a test failure


elan

  • Weekly status draft ready to share:https://wiki.mozilla.org/E10s/Status
  • Nov 13 for criteria lock is at risk but it's ok, I can move it out a week, thanks Jimm for your help with the perf section
  • a/b telemetry experiment for beta 43.0 is our next big thing

Gabor

  • bug 1159827 (m8, - showing spinner when closing a tab) - fixed review comments, but have a lot of test failures that I don't think I can fix easily
  • bug 1217876 (+, - nsIXMLHttpRequest clients in a frame script don't get any visible AuthPrompt when needed) - I tried out the patch that was posted in the other bug, but it does not seem to fix it. I debugged and added some stack trace and extra info to the bug. Waiting for the netwerk team to provide some help.
  • started going through the SDK bugs, made a lot of progress

gw280

  • bug 1221913 - m8 - newly added to the m8 list as of the last triage. asked kats what the likelihood of APZC shipping in 45 is, and it's not looking likely. This probably will need to stay in M8 as a result.
  • bug 1109714 - m8 - actively working on this. was my last m8 before the one above got added :) currently moving large chunks of code over to the parent process. also working on addressing the excellent feedback from mconley.
  • bug 1098131 - m8 - landed. took a bit longer than anticipated to fix because it turns out the exact attribute I was playing with was being cached elsewhere for just e10s, which is why the testcase was failing (it was hitting a cached value rather than the actual one to determine the current state).

mrbkap

  • bug 1171215 - (m8, websockets cookies and third party stuff) - dealing with review comments now.
  • bug 1116478 - (m8, ctr+click on mailto link opens an extra empty tab) - investigated, have idea of how to fix.
  • Responding to questions about testing.
Retrieved from "https://wiki.mozilla.org/index.php?title=Electrolysis/Meetings/2015-11-12&oldid=1105457"