Permalink
Commits on Jan 4, 2016
  1. @tschneidereit
Commits on Dec 18, 2015
  1. @rpl

    Bug 1146606 - forward and cache remote readyState updates and remove …

    …CPOWs usage. r=mossop
    rpl committed Dec 18, 2015
Commits on Dec 10, 2015
  1. @mikeconley

    Bug 1195295 - Follow-up: remove other race-y SDK test that I failed t…

    …o remove in cc674d838615. r=Mossop
    
    Mossop r+'d this over IRC.
    mikeconley committed Dec 10, 2015
  2. @mikeconley

    Bug 1195295 - Remove race-y SDK test that relied on slow content-sess…

    …ionStore.js load. r=Mossop
    mikeconley committed Dec 10, 2015
Commits on Dec 1, 2015
  1. @gijsk

    Bug 1222490 - part 6: remove miscellaneous other bits and bobs referr…

    …ing to panorama/tabview/tab groups, r=ttaubert
    gijsk committed Dec 1, 2015
Commits on Nov 24, 2015
  1. @KWierso

    Merge inbound to m-c a=merge

    KWierso committed Nov 24, 2015
Commits on Nov 23, 2015
  1. @Mossop

    Bug 1226868: Fix test metadata to correctly check things and verify t…

    …hat it has done so afterwards. r=bz
    Mossop committed Nov 23, 2015
Commits on Nov 19, 2015
  1. @Mossop

    Bug 1146926: Allow attaching a worker to a CPOW window without invoki…

    …ng CPOW operations. r=gabor
    
    By sending the CPOW to the child processes the one that owns it will create the
    child worker and then send back the url of the window to set up the parent
    side of the worker.
    
    There are two breaking changes here. Workers invoked in this way no longer
    attach synchronously. We no longer pass the window through the attach event.
    Mossop committed Nov 19, 2015
Commits on Nov 18, 2015
  1. @Mossop

    Bug 1146926: Allow sending CPOWs from parent to child processes. r=gabor

    Defines an emitCPOW(event, arguments, cpows) to send a dictionary of CPOW
    objects to a child process or frame. The only process that will receive the
    event is the one that owns all of the CPOWs sent. The CPOW dictionary will be
    appended to the arguments emitted from the port.
    
    Also gets rid of some of the redundancy in message handling so we can more
    easily change the protocol in the future.
    Mossop committed Nov 18, 2015
  2. @Mossop

    Bug 1146926: Fix SDK remote tests to pass with e10s enabled. r=gabor

    By not waiting for the tab to finish loading we end up accidentally killing
    the child process somehow and later tests that expect a child process to be
    present fail.
    Mossop committed Nov 18, 2015
Commits on Nov 17, 2015
  1. @Mossop

    Bug 1224577: Fix removing frames from the map so chrome windows aren'…

    …t leaked from every add-on. r=bz
    Mossop committed Nov 17, 2015
Commits on Nov 10, 2015
  1. @jaragunde
Commits on Nov 5, 2015
  1. @BavarianTomcat
Commits on Nov 4, 2015
  1. @jryans
Commits on Nov 3, 2015
Commits on Oct 30, 2015
  1. @KWierso
Commits on Oct 29, 2015
  1. @arai-a
  2. @nt1m

    Bug 1169679 - Update add-on manager entry design to match in-content …

    …prefs styling. r=dao
    nt1m committed Oct 29, 2015
Commits on Oct 28, 2015
  1. @miketaylr
  2. @jsantell
  3. @akrawchyk
  4. @syg

    Bug 932517 - Treat let as a contextual keyword in sloppy mode and mak…

    …e it versionless. (r=jorendorff)
    syg committed Oct 28, 2015
Commits on Oct 26, 2015
  1. @BavarianTomcat
Commits on Oct 23, 2015
  1. @KWierso
  2. @rvandermeulen

    Backed out changesets 4c98cfb04c9e and 4c98cfb04c9e (bug 1169679) for…

    … being the likely cause of OSX 10.10 mochitest-bc failures.
    rvandermeulen committed Oct 23, 2015
Commits on Oct 22, 2015
  1. @KWierso
  2. @BavarianTomcat
  3. @nt1m

    Bug 1169679 - Update add-on manager entry design to match in-content …

    …prefs styling. r=dao
    nt1m committed Oct 22, 2015
Commits on Oct 21, 2015
  1. @jlongster
  2. @nt1m

    Bug 1169679 - Update add-on manager entry design to match in-content …

    …prefs styling. r=dao
    nt1m committed Oct 21, 2015
  3. @jryans
Commits on Oct 18, 2015
  1. @arai-a
Commits on Oct 16, 2015
  1. @Mossop

    Bug 1129662: sdk/page-worker should use a remote page. r=krizsa

    This makes page-worker load its pages in the remote process. It does so by
    creating a single frame in the hidden window used to ensure we have a remote
    process when necessary and then a module in the remote process is used to
    create windowless browsers to load the pages.
    
    This does break one API, getActiveView, but I don't think we should be
    maintaining that and it has been unstable since its inception anyway.
    
    Once downside, the l10n module now has to use the observer service to detect
    documents rather than the DOM event, this might be causing more CPOW traffic
    since that observer notification is shimmed so we may need to use the shim
    waiver there.
    Mossop committed Oct 16, 2015
Commits on Oct 14, 2015
  1. @gijsk

    Bug 1211470 - r=zer0

    gijsk committed Oct 14, 2015