Electrolysis/Meetings/2014-07-17

From MozillaWiki
Jump to: navigation, search


« previous week | index | next week »

jimm

  • m1 update
    • bug 691601 (e10s support for gFormSubmitObserver) - currently blocked by bug 1015721, r?->dao
    • bug 948574 (remote nsISiteSecurityService::IsSecureHost/IsSecureURI) - blocked by reviews, sr?->ptheriault, ni?->sstamm
    • bug 942707 (pdf viewer doesn't work with e10s enabled) - blocked by reviews, r?->yury (back next week)
  • working on win tests, and I took m2 bug 903022 (Save pages to disk (Save link as, Save page as, etc.))

allison(ally)

  • 693555 - Update inline spell checker underlining for multiprocess : relanded & it stuck!
  • 1030449 - Update Spellchecker Personal & Additional Dictionaries for multiprocess started, found regression in single process
  • 1039764 - Spellchecker context menu broken in (single process) Nightly: r+, updated and waiting for trees to re-open (Bug 1040308)

cpeterson

  • we got some press about e10s
  • Add-on testing:
    • I've received a few emails from people testing add-ons on arewee10syet.com
    • I wrote a draft of blog post about add-on testing. Process your feedback!
    • juanb suggests that we schedule a QMO testday to occur our blog post, so we can reference the testday in the blog post.
      • Maybe schedule testday for August 1 or 8.

mconley

bjacob

  • m1 bug 1034321 (printing hangs) landed, done
  • working on m1 bug 586656 - plugin key events
    • Got proof-of-concept working on linux and on windows, working on OSX now
    • On OSX I had to fix focus events first before I could reproduce; have simple good patch for that now under review.
    • Problem is WidgetKeyboardEvent has non-serializable void* field for external OS-level struct
      • Initial approach on linux and windows was to use/add proper fields in WidgetKeyboardEvent instead
      • Now considering keeping that void* field but turning it into a proper sized buffer field so we could serialize it as such. Would make this require much less OS-specific changes.
        • Problem: some of these OS-specific structs contain pointers, which we should serialize as null, as the pointer value would be bad in the other process.

mrbkap

  • bug 687194 - Got r- from Mossop, looking into how much work addressing his questions is.
  • bug 952201 - Fixed orange runs on try and re-requested review (while addressing comments on the first version of the patch).