Skip to main content
Jump to: navigation, search

TPTP-PMC-20080227

Logistics

Attending: Oliver, Chris, AlexN, AlexA, Paul, Joanna

  • Harm could not attend

Finalizing 4.4.1

Oliver provides his congrats for finishing up 4.4.1. Notes

  • 4.4.1 is released.
  • Test reports >90% done
  • Platform skipped a few tests because of 4.5 testing conflicts
    • Ask if we need to go fill them in now.
    • General consensus was that it is not necessary.

4.5 Discussion

We started with IPv6 discussion which had a big mail thread earlier this week.

IPv6 Discussion

Chris had not yet had a chance to read Harm's latest message. Paul and company summarized the main points

  • Backwards compatibility requirements are the "simple" scenario that Chris mentioned in email (need to continue to support IPv4)
    • In addition to supporting old agent controllers, this is needed because few have moved to to IPv6
  • Intel to cover AC and low level network
  • IBM is finding resources to cover the UI and callback stuff.
    • Paul would think that Joanna would have someone else to handle some of the UI stuff
    • Paul noted that Duane did some stuff in this area before
    • Paul also noted from test effort a while ago that anything doing string manipulation of an IP address must be tweaked
  • Testing resources appears to be the last piece that is not fleshed out
    • Alan will be covering bits of this

Goal to get resources moving and planning all the bits so that we can know soon if there are any additional gaps.

Java6 Support

We started with discussion of the existing enhancement for updating BCI engine to post Java6 JVMs that will eventually start to require correct setting of StackMap atribute.

  • We are doing this today as part of our due diligence.

As part of this effort, Asaf noted as raised a defect against probekit BCI

  • Essentially some deprecated bytecodes are used
  • Navid is still default owner for this defect. Joanna is looking into that.
  • Paul notes that Asaf has a patch for TI profiler that might be able to be harvested
  • Chris noted that in 4.4 we ran into probekit resourcing issues right at end. Would prefer to see resource identified sooner rather than later.
    • AI Joanna to assign someone
  • Paul reminded team that there will need to be some effort required to go thru a full test pass
    • Would have been nice to have I5 test pass on Java6 as well. Too late.
    • Paul did do some touch testing back a while ago and didn't see issues.
  • AI for all project leads to create test plan for covering Java6 testing during I6
  • Paul updating reference platform BVT to test with both Java5 and Java6. This will simplify full test pass.

AlexA noted that if one wants to really test the StackMap enhancement one needs to disable the workaround that Java6 VMs provide in the presence of bad StackMap attributes.

  • Java6 works with old verifier. Testing requires not just running with Java6 but also requires a new flag in order to ensure that verifier backward compatibility is not used.
    • AI AlexA to send details to trace project mailing list and make sure it is reflected in defect on bugzilla.
    • AI AlexA to send Chris details about this flag to include in the minutes
      • Flag for Sun and JRockit is -XX:-FailOverToOldVerifier to disable use of the old verifier path

4.5 I6

General reminders about 4.5

  • Reminder to leads API freeze is at the end of I6
  • Reminder that all things that effect API or translation HAVE to be done for I6
  • Enablement and translation stuff is due as well
    • Alex has posted a note reminding folks to all the project dev mailing lists.

Oliver notes that hopefully everyone has confirmed resources/plan for I6 and has updated their weekly schedules accordingly.

Test

  • Got it (planning) done on Monday
    • Everything on weekly schedule
    • Light work this week; full work starting next week
  • Oliver asks if there is anything that needs to be brought up for PMC?
    • Not really
  • Paul did triage a bunch of defects not touching UI to move to I7
    • Joanna has provided a student to help with some of defect backlog
    • Paul confident that the enhancement work will get done by I6

Oliver has asks if there been agreement to slip any items totally out of 4.5.

  • Paul not yet slipping out of 4.5 (yet) just out of 4.5 I6
  • Paul believes some will slip out of 4.5 but will finalize that list at end of I6

Trace

  • AlexA notes there is still only 1 developer on trace
    • Working on tutorial right now
    • Working on one other defect
  • AlexA notes that there is still no response from Joe Toomey regarding other Trace defect (see last week's summary)
    • TI support in execution framework
      • Joanna did follow up and asked about more details about implementation
      • Yesterday Ashish Patel provided some updated details
      • AI to Joanna to follow up with Joe again today
      • Oliver asked if AlexA could be added to distribution list of discussion
    • Essentially, we need a commit and a date from Joe Toomey

Oliver asks whether Trace all planned for I6.

  • AlexA responds yes.

Monitoring

  • Started defect work this week
  • Still have a few defects for this week that have not been fixed yes
    • Two will be fixed today
    • Remainder will slip to next week.
  • Team does have a few defects that they are looking into taking off of the plan.
    • Is talking to consumers to find out if they really need to be P1 or not
  • AlexN is working to decide which test cases to run with JDK 1.6
  • After doing worksheet/schedule planning, it looks like most planned defects can covered

Platform: Oliver paid extra attention to this section because the platform project is so central to stuff. :-)

Joanna notes that she is using the weekly report to list what has been checked in/done during the week.

  • Some progress on backlog of bugs
  • Critical stuff this coming week are making progress on IPv6 and Java6
    • Joanna has asked someone to size probekit fix for deprecated Java6 bytecodes
  • Joanna also working on test plan for I6
    • Working to resolve Trace and platform test case duplication
    • Some failure platform.ui, probekit.editor and one other place (automated tests) that are failing but not updating correctly in reports
      • Oliver notes that automated tests do need some maintenance
      • Paul notes that this will get better when BVTs run these because it will raise attention early.

MISC

Discussion regarding lead rights for AlexA

  • Guru had not yet gotten portal fixed to allow him to move stuff over to AlexA.
  • Oliver would take care of it
    • AI Chris/Guru/AlexA: Send precise list of what rights need to exist for AlexA to Oliver.
      • (Update) Guru's portal rights became resolved shortly after the meeting so this is AR is defunct.

Oliver was invited to give a TPTP presentation to the board on June 18

  • Wants to show that we are exciting
  • Wants to see if can pull in more resources from somewhere
  • Will be interesting to talk to board because there are members that do not really want OOB w/ Eclipse. They want consuming products to provide usability. Other members want OOB.

POG update

  • Alex notes that there are no new actions regarding it
  • Team has been concentrating on enhancements for last few weeks
  • No new lead user reports lately
    • Waiting for update from Harm (who sent suggestions last week)
      • Alex has written some bugzillas about views/profiler.
    • Next thing on Harm's lead user case is for Harm to look at it and find next pass of suggestions
  • Alex notes that we would like a wider range of scenarios/users (beyond profile on server)
    • Chris noted that team is discussing with Intel folks doing Harmony performance analysis to see if they can commit to being lead users for their use cases.

Oliver asks when he should send out a wide announce about POG?

  • Depending on lead user status may wait till after I6 and some of more defects are resolved
  • A few known lead users are more containable and able to be prioritized than widespread community reporting

Oliver sending out results from initial netbeans comparison

  • Initial view is there
  • Decide on next steps for analysis
  • Oliver would like a profiler specific call soon.


On EclipseCon

Oliver noted that details are being finalized for meeting at Eclipsecon

  • Trying to set up internet connection
  • Room will be reserved between 9 and 4
  • Projector might come in handy
    • AI to Chris to see if can bring a projector
      • (Update) Chris can borrow a projector from Intel but no screen
  • Oliver was going to lobby to see if Joanna could come
    • What would be required to get Joanna to F2F
      • AI Joanna to look into it and get back to Oliver

Paul asked whether anyone from WTP is coming to discuss EPP

  • They have not yet confirmed

Paul asked whether Andreas has confirmed coming to discuss

  • yes he has but Oliver needs to give him a time to come.
    • Oliver plans to give Andreas about 30 minutes.

Paul noted that the TPTP BOF is scheduled at 8:45p on Wed

  • AI all involved parties: Show up

Back to the top