Desktop Team Meeting October 30th, 2007

Agenda

• 0.7.2: Pick up date is tomorrow! There are some l10n bugs that really need to get in there and one blocking bug (assigned to Robin). Things blocking anyone?
• 0.7.3: There are a bunch of interesting things landing in 0.7.3: month view, script recordin, leopard. News on those?
• 0.7.future: I've been working in reducing the list to what we want to do in the next 3 months (till end of 0.7.5). I've been churning lots of bugs recently. Not finished yet but should be soon.

• Misc
  • Have something to announce? Issues to bring to the group? Please take the stage, er, mic.

Minutes

0.7.2

• Stearns: DV layout won't make it (bug 10133)
• Morgen: trying to make sure that fixes for bugs 6539 and bug 6540 (both related to use of Chandler behind firewall and HTTP proxy) will make it
• BKirsch: landing a sizeable bunch of l10n/i18n stuff today and tomorrow

0.7.3

• Month view (Reid): we have a good static display, still work to do to make it usable
• Script recording (John): having problem with the layers running script recording, bear and heikki helping, still not able to run on Linux, some failures on OSX
• Leopard (Andi): ready to roll, just waiting we get the build story straight, we should choose a subset of the possible 16 builds and limit ourselves somewhat. Still, we'll have to build more than the current 8. Heikki: plan to have 2 new TBox machines: one for Leopard, one for Linux GG. Heikki and Philippe to solve the machine issue then Bear and Heikki to work on the TBoxes.

0.7.future

• Everyone like the idea to drive toward getting rid of "Preview" in our moniker. Seems to be one thing hurting us.

Misc

• FOSS Camp (Bear): Preview, as a name, is hurting. People impressed by what we have when we show the feature set. To be in Ubuntu main distro, will be hard. Getting to the Universe Ubuntu seems to be possible. That requires us to make sure that all our dependencies packages are in Ubuntu. Step 0 though is to build the Debian packages.
• FLOSS (Jeffrey): Jeffrey to go with Matthew and Katie. We'll use Oracle usability testing lab. Sprinting then on Chandler.
• OpenLaszlo (Philippe): Great meeting with Laszlo folks. They like what we have on the server side and were impressed by our client (though they won't use it for obvious reasons). See the email I sent on the strategy list.

Status

Andi

• http://people.osafoundation.org/~vajda/logs.cgi

Brian (bkirsch)

• Your status here...

Bryan (stearns)

• Working on my two remaining 0.7.2 bugs: one will make it (I hope: looking into why the fix for bug 10811 seems to cause test failures intermittently), one won't (10133: new DV layout; big & risky).

Grant

Progress

• 6540: Can't specify an HTTP proxy
• 10900: TestRecurringEvent fails around 11:00 pm
• 10901: TestEIMDiffRecordSetInMemory.py fails around midnight
• 10973: Stale Item has no attribute 'dismissItem'

• Wrote up an outline of the plan for the Chandler Architecture Pilot Project.
• PTO Oct 16.
• Met with PJE to discuss implementation for pilot project.
• Helped Robin with a Tinderbox problem.
• Ran a (very abbreviated) Chandler desktop release meeting.
• Wrote up some brief ArchitecturePilotProjectNotes.
• Investigated Bugs 11051, 10980.
• Ongoing work and discussions on the Architecture Pilot Project.

Plans

• Finish up any remaining 0.7.2 issues
• Ongoing architecture pilot project work

Jeffrey

• Demoed Chandler to lots of friends
• Lots of responses to user questions on the users list
• Lots of multiweek work
• Hassled with the dying ethernet adapter on my laptop
• Changed recurring event description text for bug 11135
• Started meeting with Oracle folks about user testing before FLOSS sprint
• Started talking to Grant and PJE and reading about Trellis for work on APP

John

• Moved to Wing 3.0. Everything is working except for windows which I still have problems running Chandler from the debugger. I checked in updated project files for the new version of Wing.
• Fixed a bug on Mac and Windows that caused recorded scripts to not be run correctly from the command line
• Tinderboxes are running recorded scripts on windows and macintosh. Bear said he'd have linux going shortly. So far only the windows tinderbox is green. All the problems I've seen so far appear to be with the tinderboxes and not related to script recording. Mac is failing, apparently because it doesn't have a monitor attached to the computer running the tinderbox.

• Narrowed down a fix for the Wing crash debugging Windows.
• Merged the recent trunk changes into the recorded scripts branch and ran into a bunch of problems:
  • An apparent bug in SVN caused problems running after the merge. I eventually recovered from the problem and also updated to the latest version of SmartSVN^? which fixes one of the problems I ran into.
  • The recorded scripts identified a bug in rt.py, which I eventually narrowed down to a bug in rt.py where it doesn't set the path correctly for localization. Bear has a fix for this bug.
• There remain various problems with the tinderboxes -- false positives on tests and false negatives and the Linux tinderbox still isn't running. I have narrowed down one easily reproducable false positive -- which I'm hoping to get some help fixing. These problems all appear unrelated to the recorded scripts themselves, but block progress on the branch and the eventual merge with the trunk.

Morgen

• Your status here...

Philippe (pbossut)

• Your status here...

Phillip (PJE)

APP stuff finished lately:

• README in APP (rearchitecture branch)
• Contextual and Plugins dev links up on CheeseShop
• AddOns package released
• Trellis time services support in eby-sarna SVN; new docs here

Reid

• figured out svn merge "problem" for multiweek
• working on multi-day lozenges for multiweek
• fixed Bug#11117 (reworking conflict dialog for l10n)
• fixed Bug#10957 (move Send button)

Robin

• Your status here...

-- PhilippeBossut - 30 Oct 2007