Electrolysis/Meetings/2014-07-24

From MozillaWiki
Jump to: navigation, search


« previous week | index | next week »

ally

  • finished off the contextmenu issue in single process fx
  • investigating 1042687, I can't reproduce, but that doesn't mean it's not real
  • switched from the spellchecker dictionaries bug to the context menu one, bug 1030451
    • blake will be taking the dictionaries bug
    • wrestling with suggest() turns out to be embedded in the checkword() call
  • taking over the shims work from billm
    • still generally getting up to speed
    • first bug: 1041720 create an e10s content shim for bugzilla tweaks

jimm

  • m1 update
    • bug 948574 (remote nsISiteSecurityService::IsSecureHost/IsSecureURI) - landed
    • bug 691601 (e10s support for gFormSubmitObserver) - blocked by review in bug 1015721, r?->dao
    • bug 942707 (pdf viewer doesn't work with e10s enabled) - blocked by review, r?->yury (back this week I think?)
  • Windows mochitest-plain, two bugs remaining:
    • bug 1041695 - some sort of gfx leak when media classes destroy D3D objects on a non-main thread. These objects send ipc events in their finalize methods that don't make it across to the parent. asked davidp for help on this.
    • bug 1043597 - interesting race condition with dom events that modify content state and cpows which check it. Currently working on an async content query api in SimpleTest as an experimental fix.
  • bug 903022 (Save link as/P1) - wip

mconley

  • Send me your badge nominees!
    • If you're sending me their email address, please also include a short message about why they're receiving the badge - this goes in the award email. I could leave it blank, but that's kinda anticlimactic. I can fully accept generic message templates and fill in the blanks.
    • I can also sent you claim codes if you want to just hand those out.
  • (M1) Bug 1009628 - Need mozAfterRemotePaint event for remote iframes
    • Reviews from Enn are done, first round of review from dao finished
    • Nits addressed, up for a second round of review from dao.
  • (M2) Bug 1013744 - [e10s] Can't open Web Console using CMD+OPT+K and CMD+OPT+I keyboard shortcuts in e10s
    • Landed and stuck! evilpie grabbed the helm after my first patch and drove it home, since our event serialization stuff can be pretty tricky to untangle. Thanks evilpie!
  • Bug 961362 - [e10s] HTML5 video's fullscreen feature zooms video to fill tab, not screen
    • Initial investigations done, and there are a number of issues. Basically, there's code in the implementation that goes up the document tree until it finds the chrome window, and then triggers fullscreen. The problem is that remote documents never get to chrome windows, and so it just fails.
    • Luckily, b2g solved this problem in bug 684620. There's infrastructure in there I can use, and I've got entering and exiting OS fullscreen on OS X working.
    • It's almost there - the last big problem is that the MozEnteredDomFullscreen event is being dispatched in such a way that the event handler set in browser.js that does the hiding of the UI and warning/permission dialog don't come up. This is because the XULDocument dispatches the event to the Chrome event handler, which bypasses the browser.js event handlers. smaug is helping me untangle this.

cpeterson

 date        #reqs   #pages
 Jul/20/14   651     200 +++++
 Jul/21/14   513     156 ++++
 Jul/22/14   600     178 +++++
 Jul/23/14   4521    1448+++++++++++++++++++++++++++++++++++++
  • e10s bugs with unset tracking-e10s flag:
  • ACTION: Create M2 burndown chart
  • ACTION: hide addon bugs from triage query

bjacob

  • m1 bug 586656 - e10s plugin events:
    • Landed patch that fixes the issue at hand (pluginEvent's were not serialized over IPC)
    • Confirmed to work now on Linux and Windows
    • On Mac, another bug prevents from even reproducing the bug: focus events don't work. Stuck investigating that. Bug 586656 comment 50.
  • Non-e10s: bug 978966 - many WebGL test failures on WS2012 VMs (future Win test slaves)
    • Debugged: the problem is that anti-aliasing is broken in the Microsoft Basic Render Driver
    • Patch under review to disable anti-aliasing there. With that, these slaves can run the full WebGL 1.0.1 test suite without any failure.

mrbkap

  • Got bug 687194 reviewed, working on writing a test now
  • Started looking at content prefs (mostly to support remembering spellchecking dictionaries) bug 1027898
    • Should be pretty easy, current plan is to ignore nsIContentPrefService (which is deprecated anyway).
  • Bug 952201 (non-AMO addons) checked in.

alex_tz

  • sent a first patch for bug 1015466 (which covers the project that I will be working on this summer)
  • got review from felipe on bug 1017914 and I will look into that in the following week
  • my patch for bug 966395 still fails tests; I will take care about that soon
Retrieved from "https://wiki.mozilla.org/index.php?title=Electrolysis/Meetings/2014-07-24&oldid=1008203"