Electrolysis/Meetings/2014-09-11

From MozillaWiki
Jump to: navigation, search


« previous week | index | next week »

jimm

  • pdf.js changes merged to the pdf.js parent repo, working through reviews. we'll be bringing the code down in a pdf.js update.
  • still trying to put out webrtc fires (bug 1060738, bug 1066242). Generally this is mostly complete, just working through some reviews.
  • M2 Bugs
    • none!
  • M3 Bugs
    • bug 863512 (thumbnails) - making progress on this, nothing to post yet. Hopefully tomorrow.

mconley

  • Might be voice-only or absent, because I can't seem to join the Vidyo room from home without getting kicked after a few seconds. :( Do we have dial-in info somewhere?
    • I can't connect with my phone either - or rather, when I do connect, everything sounds super garbled, like a busted MP3. :( Pretty bummed.
  • M2 Bug 1047603 - [e10s] Non-remote tabs in e10s windows do not handle target="_blank" or window.open links properly.
    • This has an r+'d patch, just waiting on Mossop to review some tests for it. Or if somebody else wants to take that review, go for it!
  • M3 Bug 1060093 - [e10s] Cannot debug scripts running in the content process
  • Bug 927424 - Implement |shouldFocusContent| for RemoteFinder.jsm
    • Fixed - filed follow-up bug 1064654
  • Bug 899347 - Make click-to-play work in e10s
    • Last reviews coming in. So far, all r+'s! Nice to get credit for mbrubeck's work! :)
    • Chasing down what might be a last-minute intermittent orange that my patch introduces... but it might be rare enough that I just land anyway.

allison

  • was shimming away addSystemEventListener
  • taking bug reports/needinfos
  • review for felipe
  • monday the crashland thing hit, started digging into ime m-2
    • targeting only windows, non tablet pc, with ime enabled when firefox starts
      • as you can tell there are a lot of edgecases not covered here
    • there is no good programmatic way to detect imes outside of windows.
      • theres a partial solution for mac
    • we have a test patch, I need to test a few more chinese imes
      • on windows, I can only have one type of chinese enabled at a time, and there are several
    • then it will fold into felipe's safe-mode patch
    • makato might get ime working enough that we don't need this icky patch at all

alex_tz

  • ...and here I am on my last meeting :(
  • my project depends on bug 1035454 so I can't move on to testing it right now
  • I did some benchmarking and so far it turns out that there is a ~2x speedup (the improvement is in the time spent between OnStart and OnStop in the child) when the main thread in the parent is loaded (it turns out my previous measurements don't reflect a real life scenario because downloading is not processed on the main thread, so I will soon come up with new tests)
  • met with Blake and sworkman to discuss about the second stage of the project; we agreed on the solution and I will soon file a bug for this with a full description

felipe

  • Bug 1063848 - Safe mode should disable autostart - landed
  • Bug 1063842 - In product notification about autostart - landed, notification disabled and waiting to be enabled whenever
  • Bug 1065880 - Telemetry for e10s - waiting on review from bsmedberg
  • Probably moving back to the getUserMedia bug 973001

Bug Triage

Old M2 bugs should be moved to new e10s milestones as follows:

  • P1 -> M3
  • P2 -> M4
  • P3 -> M5
  • P4, P5, or -- -> tracking-e10s=+
  • add other milestones to tracking-e10s flag
    • also ask to reorder values: ---,?,m2+,m3+,m4+,+,later,-