Revision as of 14:56, 20 May 2014

E10S updates from zombie (read only this week)

- not very much progress to report this week
- got slowed down by content dump() debugging problems
- fixing sdk/tab event, leaving tab.attach() for later

- but: landed 903039 native options.. \o/ ♫
- needed a fix-up plus a follow-up 1012361 to un-orange (still r?)
- learned not to push big changes on a weekend!

@@ Line 1: / Line 1: @@
-= E10S updates from zombie =
+= E10S updates from zombie (read only this week) =
-* questions from last week
+** not very much progress to report this week
-** E10S stuff should not break Fennec (modulo my mistakes)
+** got slowed down by content dump() debugging problems
-** Frame Scripts run with chrome Principal, but we have Components/Sandbox
+** fixing sdk/tab event, leaving tab.attach() for later
-* first e10s (smoke) test running on CI (and passing ;)
-** [https://bugzilla.mozilla.org/show_bug.cgi?id=1004833 resurect cfx --e10s and add a package.json flag]
+** but: landed [https://bugzilla.mozilla.org/show_bug.cgi?id=903039 903039] native options.. ''' \o/ ♫ '''
-* the more i learn about E10S, the less i'm sure in my first proposed approach
+** needed a fix-up plus a [https://bugzilla.mozilla.org/show_bug.cgi?id=1012361 follow-up 1012361 to un-orange (still r?)]
-** (was: attach small Frame Script that forwards messages across boundaries)
+** learned not to push big changes on a weekend!
-** afaict, the original assumption when designing jetpack APIs was that a DOM window was going to be a barrier between Parent and Child processes (not true)
-** different code paths for xul:iframes not inside tabs? (they are not remote)
 = previous weeks =