TPTP-PMC-20090527

Logistics

Atending: Eugene, Paul, Chris, Oliver, Kathy

Any issues with last weeks minutes?

• no issues

anyone need to leave early?

• No

4.5.2 Discussion

• None

4.6 Discussion

Schedule; Paul generated new test report before call

• We are 52-53% thru test pass and it needs to be done this week
• 1 test suite is not appearing in the report for Test yet
• Platform project a bit behind because were waiting for a good driver using the new build process that caused some delays
  • Test project would not have been effected because of agent controller
  • JVMPI issues did show up yesterday that seem to be related to build
  • Some {Platform testing may overrun to next Monday

Paul notes that Documentation technical review test needs to be done and included in test pass

• UI changes match documentation. typically do once per release.

Kathy notes that we dropped one driver to RC2

• We will be doing a few respins based on PMC approved defect changes
• There are only 3-4 more defects in the queue
• Paul has all approvals that he needed in place already for Test
  • There is a new defect with patch that he will propose and would like Eugene to review.

Update from Kathy on the remaining items for Platform

• Currently the log file gets a message that needs to be cleaned up around interaction with the IAC preferences page
• Updating "getting started" html page
  • Question has come up whether these changes need PMC approval?
  • It is packaged with a plugin so it should go ahead and be managed (PMC approval)
  • If its our web pages (i.e., not packaged with plugins in our CVS) then we don't need PMC approaval
• After updated GCC in latest build environments, we find that a few libraries are not longer needed
• Some Heap profiling testing and documentation updates remain
  • Hope to get some of that soon.

Deferral of P1s

• JVM crash (new JVM had already resolved part of it).
  • Another new JVM patch seems to have resolved all issues under touch testing. Asking consuming product to test with it under their load
  • Might not have to defer this
• For other three on Kathy's list we discussed going ahead and deferring
  • Unanimous approval of deferrals

Release Review

paul did mega awesome job for last review or two but is too busy this time

• Does it fall to Kathy now? If start with 4.5 material should be easier
• Paul notes that the hardest part is probably to summarize the defect landscape
• Kathy will look at it and says that she might need a bit of help from Paul
  • Paul can consult

IP log track

Kathy tried using previously and ran into issues with how the tool works

• She is still trying to decide if using the tool or manul IP log is easier
  • Paul only has 1 defect from external contribution so manual or tool is really irrelevant.

MacOS

There has been a lot of renewed interest on mailing list. Eugene has been working to collect vounteers who will test early engineering builds and maybe collaborate on patches/builds.

Oliver asks if anyone have concrete suggestion about how to support the community

• Lots of people waiting for a build to try
• Suggest that we start posting some periodic drivers
• Some are reporting having issues setting up overall build infrastructure; some updated documentation about setting up infrastructure might help.
• Community needs someone that is investing some limited cycles monitoring
  • We do not plan to do anything here prior to galileo release.
  • There are a few patches that have come in for this that Kathy has targeted to 4.6.1
• Oliver asks how we want to announce to community.
  • Generally, a message that the PMC is watching the interest in MacOS and will investigate more resources after Galileo.
  • Eugene will send out update on direction to interest list and to PMC.

Kathy asks about how functional the current AC is on macos.

• Eugene notes that at least one user reports that they have a partially functional local build.

Build Transition Update

x86, em64t transfer is done (for 4.6) and builds are pretty much up to date.

• Ben had found something about a packaging of a plugin (model hierarchy stuff) with the agent controller
• Found it in earlier candidate driver but in this mornings driver it was resolved.
  • Assuming for now it was related to build system flux; Paul opened and closed a defect but wants to know when we will call the builds completely stabilized

IPF side of build is still at Intel while we finalize equipment transfer issues.

Builds remain at Intel for 4.5.1 milestone (at least until after galileo)

Post 4.6 discussion

We had a discussion about post Galielo resourcing

• • Intel expects to be flat thru spring maintenance
  • Paul is officially 0.5 time and will start enforcing that after 4.6
  • IBM's platform team will be flat

Misc

Minutes from 2 weeks ago are needed.

Chris reminded PMC that Intel China folks are on holiday for Dragon Boat festival

We discussed a bit IBMs AC porting efforts to more varied architectures

• AIX code merge is desired prior to 4.6
• IBM is doing a trial merge now
  • If at that time we feel risk is low they will bring up to promote in 4.6
  • Should know end of this week or early this week (after "final" patch is ready)

We discussed some version info naming details that planning council had some heartburn with

• the "-" in our name (eclipse.org-tptp).
• We plan to keep it this way until get feedback about what breaks from using the dash.

Chris will be in China 2 weeks in June

• Kathy mentions that we'll need to talk about knowledge transfer to IBM team of components currently primarily maintained by Intel.