Thunderbird/StatusMeetings/2008-06-03
From MozillaWiki
last week | index | next week »
Thunderbird Meeting Details
- Tuesday, June 3, 16:30 UTC (9:30am Pacific, 12:30pm Eastern)
- How to dial-in
Agenda
Action Items
Open
- AI:clarkbw continuing to drive MailNews:Address Book New Card
- Getting started, mockups soon to come up post-discussion.
- Standard8 and dmose will help with this.
- AI:davida guided form doesn't allow version unsetting
- AI:davida bugzilla should _refuse_ unspecified version
- AI:davida version field UI in bugzilla should be clarified
- AI:davida should have "file bug" menu item in nightlies/alphas
- Standard8: carry over to next week for davida response
- AI:clarkbw create design for UI around Emre's Implementation Proposal pieces
- AI:emre begin work on imap preemptive message download feature
- AI:dmose Thunderbird 3 roadmap and planning with davida.
- AI:asuth Post initial global database schema/plan on wiki and post to newsgroup for discussion.
- Lightning-related.
Closed
- AI:asuth figure out what installer work needs to be done for bug 375922
- asuth: figured out. porting from firefox installer-related patches in progress.
- AI:emre begin work on offline operation playback feature: Patch is submitted.
- AI:group write down desired IMAP offline behavior on wiki page
- done.
- AI:Standard8/clarkbw: mac addrbook UI
- Standard8: done
Thunderbird:Thunderbird_3.0a2
- Why? - Release goals TBD
- What
- Address book OS X bug 397811 (Completed)
- Bundle attachment?
- Progressmeter bug on OS X
- Auto-configuration, phase 1?
- Better Gmail/IMAP setup?
- Addressbook interface redesign
- Joshua's nsIAbCard patch (bug 413260)
- Autocomplete from XPFE to Toolkit
- Password manager from Wallet to Toolkit
- Move Default Profile Location to FF/SM equivalent
- xpcshell Unit tests passing bug 406227
- Performance for startup
- STEEL
- Things marked blocking-thunderbird3+ or its nominations
- Crashers
- move further on de-folderization of mail metaphors
- When
- Planned code freeze: July 8, 23:59 Pacific
Thunderbird v.next and branch tracking
Discussion so far in mozilla.dev.planning (link missing the first few posts).
Following discussion on irc today, several points came out from discussion about possibly moving to 1.9.1 for our main work:
- Core devs willing to help us come to a reasonable solution, possibly an intermediate one whilst working out when to move to hg fully etc (e.g. see bug 437059
- We don't want to re-invent the build system.
- Option A on SeaMonkey/1.9.1_Repository_Options may be the best solution for the time being and/or overall
- Would require a separate pull script for mozilla-central (at least whilst we're not in hg I believe).
- ted & co willing to help us get the build system working with that sort of layout (Standard8 needs to file bug)
- Depends on when we will ship.
- 1.9.0.x is already frozen.
- We need to define our release goals for Thunderbird 3.
- We don't have very many core patches that are Thunderbird-specific and have a high impact on core.
- 1.9.1.x may possibly slip into next year.
Full discussion on the talk page
QA Updates
- Previous bugday on 29 May 2008.
- Only session 2 staffed.
- See the spreadsheet of results.
- Next bugday on 05 June 2008.
- continued focus on recent bugs which have no comments/feedback
- nth10sd will be there for session 2 only.
- We may only have one staffed session for bugdays in the near future, until helpers' numbers increase.
Roundtable
- Status Updates
- dmose
- reviewed patch to drop nsISupportsArray usage from addrbook where possible (bug 410177)
- reviewed reply quoting patches (bug 384599 and bug 23394)
- drove IMAP deadlock fix into tree (bug 434642)
- filed and worked on patch for compose window crasher bug, but it turned out to be an already-patched DUP (bug 436544)
- reviewed unit test setup/teardown cleanup patch (bug 434810)
- put together draft review requirements pages (see announcement)
- started reviewing nsIAbCard refactoring (bug 413260)
- misc other small reviews
- nth10sd
- Mastering hg and gdb.
- Went on an assertion-finding spree. (12 found so far)
- Provided normal and full backtraces for each assertion trapped, using gdb.
- Did some work exploring the use of DOMi on Thunderbird trunk.
- Will work on some form of blogpost or wiki page describing this.
- Usual bugday, triage etc.
- Standard8
- Various reviews
- Investigated and pulled mozilla-central to allow me to commit some patches against core.
- Mac OS X Address Book
- Committed bug 398711 UI
- Fixed bug 433264 to allow users of special build 2.0.0.6 + Mac OS X AB to switch to trunk without "loosing" all their address books (they would disappear from the display).
- Automated Testing
- Finished reworking how we do tests as a result of bug 434810
- MailNews:Xpcshell_tests wiki page updated to keep in sync.
- Confirmed list of failing unit tests is still the same (i.e. blockers for bug 406227
- Tidied up patch and submitted for review a SMTP fake server and basic test bug 436847
- Autocomplete
- We now have autocomplete for my domain working on toolkit interfaces (thanks to Neil)
- Main address book autocomplete in progress (bug 370306, basic search functionality in place with unit test.
- Still need to think about LDAP.
- Misc
- bug 436086 Finished changing nsMsgDBFolder::mSubFolders from an nsISupportsArray to an nsCOMArray
- bug 414432 Finished improvements to nsIMsgHeaderParser::makeFullAddress(String)
- bug 376877 Fixed duplicate packaging of all-l10n.js in installer builds.
- asuth
- Lightning
- Tried very hard to duplicate bug 367456, but it appears to have been mooted on the trunk. Another commenter says TB 2.0.0.14 w/lightning also no longer has the problem. There was a fair amount of short-term memory bloat and poor performance with the torture-test files though, which leads to...
- some pecobro enhancements to support with investigating leaks/bloat
- Made pecobro understand object allocation.
- Refactored pecobro to support caching all underlying data types, and no longer fuse/integrate the ASTs with the semantic cache (as suggested by dmose/dascher). This greatly improves the performance and makes it possible for people with more reasonable amounts of memory to run things.
- Ran into a glitch where the DTrace probes/underlying javascript engine is lying about what file is being executed at a given point of time, which made the trace parser un-happy. (Why I haven't posted a processed trace with the new mods yet). I believe this would affect venkman too.
- Will try and complete these pecobro changes/fixes and use to help track down some lightning performance issues soon.
- Global Database
- Lot of thinking/planning. Have looked at myk's snowl schema, posterity's original schema, my mods to posterity's schema, and clarkbw's examples from the newsgroup and previous proposed extension.
- justdave's WindowServer slow-down bug, bug 436897
- Spent a bunch of time trying to help remotely debug this and duplicate this locally (before the bug was logged).
- misc
- Fixed a glitch in my mail-folder-bindings fix on bug 296453; thanks to sid0.
- bug 435228 fix went into tree; uncovered a multiple-listener scenario which is logged on bug 436366 which I need to look into more deeply/propose a fix for.
- Still pending
- Installer work bug 375922 (hopefully MSDN OS license comes through soon)
- Unit test for bug 296453.
- Lightning
- bienvenu
- Many code reviews - reduced backlog of old reviews, and keeping up with new review requests
- Helped Emre with Better Faster IMAP issues - added unit test in msgdb for one of them
- Investigated adding support for IMAP RFC 4551 bug 436151
- Worked on speeding up loading saved searches bug 436960
- Emre
- Mainly worked on bug 435153, a patch is submitted for review
- Blogged about the thinking and development process at sicakborek.wordpress.com
- Started to work on bug 436615
- Started to work on bug 413874 to adapt it to the branch
- Dale from penelope project has commented on imap plans. Good news, he is available for help. Planning to talk to him this week.
- Gozer
- Mozilla Messaging NOC is up and running
- Will post details on downtime there.
- All 4 Sun X4150s are up and running, waiting for something to do.
- System Configuration managment is still lacking, but not blocking
- Apple hardware still hasn't arrived...
- Need some work on Bugzilla components for MoMo IT.
- Will post to mozilla.dev.mozilla.org as suggested by dmose.
- Build processes are the most important, and other infrastructure will have to be migrated over too.
- Mozilla Messaging NOC is up and running
- clarkbw
- Worked on an interactive status bar in the beginning of the week
- De-Folderification meeting mid-week
- Started discussion on experimental view
- dmose
Attendees
Standard8, emre, clarkbw, nth10sd, dmose, gozer, asuth, KaiRo