Skip to main content
Jump to: navigation, search



Paul Oliver Guru Chris AlexA Regrets: Joanna, AlexN, Harm


Any objections to last week?

  • No

Iteration I4 status

Oliver noted that realistically there is not time for another full test pass

  • Need to be in quite good shape by end of this week (end of TP1)

Monitoring and Platform indicate that they are on track

  • Sent email updates because they were OOP

Monitoring had one defect that popped up; fixed; will be tested in next build

Trace project/profiling

  • profiling
  • failure on XP (regression)
    • BLOCKING Defect 213201 currently assigned to Joanna
  • Trace where have resources tests are done
    • Results not yet checked in; Guru talked to Paul about this


  • 1 major defect
    • Will be fixed today or first thing tomorrow

Test pass 2 will be small

  • Be aware of holiday schedules as you plan it out

4 critical issues outstanding

213030 cri P1 Platform 4.5 0 4.5i4 plan 0.0 0.0 EMF trace model can't be updated
205821 cri P1 Trace.Ex 4.5 0 4.5i4 plan 0.0 0.0 For Java 5 and above ARM needs to use TI agent and not the PI agent
208394 cri P1 Platform 4.5 0 4.5i4 plan 0.0 0.0 Exceptions when opening the Platform documentation.
212883 cri P1 Test.Doc 4.5 0 4.5i4 plan 4.0 0.0 URL and JUnit Plug-in Test documentation missing from TOC.

AI: Chris to cc Rohit on minutes to underscore what he is covering for AlexN

Committer Cleanup

Chris mentioned the mail thread about removing defunct committers

  • Removing some key old committers (Randy) will impact build
  • AI for Kiryl and Joel to resolve build issues with changing account
    • Paul noted a similar issue on the IBM side with Hubert's account may exist

Trace and Profiling Discussion w/ Guru

Guru noted regarding Test cases for profiler in Trace project

  • Profiler test cases in Trace are somewhat redundant with profiling cases in Platform
    • Will be purging these redundant tests (especially since some Platform equivalents are automated)

Oliver a bit concerned about lack of line items in weekly milestones for Trace project

  • All blanks today
    • Guru noted that items with dependency on monitoring project are getting done
      • Other items are delayed until IBM resource comes onboard.
      • Commentary from Chris: AI to Trace lead to follow instructions in TPTP-PMC-20071128 for how this is supposed to look

Oliver would like to have a discussion about Trace/Profiling tracking w/ Guru, Alex, Chris

  • AI to Chris to help Oliver schedule a time to talk

Profiling lead user effort

There is ongoing discussion with an anonomous contributor regarding helping to document the user test case, instructions, and methodology.

  • Originally was proposed as a rework of key TPTP profiler documentation
  • Oliver is discussing it as
    • write user doc that is suitable for Harm's use case.
      • Probably want it brief (O(3pages))
  • While contributor want to be somewhat anonymous, need to know that resulting work will be EPL licensed.
    • Oliver to talk to contributor and find out if this is viable.

Alex mentions

  • Do not yet have all required background on use case (Harm has been in and out lately)
  • Use case is not simple (many potential parts)

Where are we with profiling initiative

  • Harm and Alex are communicating
    • Harm is continuing to put some details on wiki
    • Harm is away on vacation until around the 4th
    • Want to schedule 10 minutes on each meeting in new year to follow thru

Brief discussion on second use case

  • Harmony test case
    • Should this be documented in wiki (needs to flesh out Harmony_profiling_with_TPTP)
      • At some point would like it to be documented as much as tomcat use case.
    • Difficult to get Sergey to document and be official lead user
      • Harmony guys are somewhat overloaded.
      • In addition, Alex raised some concerns about having sufficient resource to actually pursue 2 use cases at a time
      • Wants to first hammer down what we expect to get done in I5


  • focus on 1 use case at this point then as it gets wrapped, move to another
    • Lets us plan when to engage Harmony team (at a time we can dedicate to help)
  • Who will be main person to evaluate success of a use case 1?
    • Harm?
    • Team as a whole?
    • Question about qualitative success versus quantitative success
  • Alex and Harm to be on phone on Jan 2 meeting
    • For that meeting:
      • A few mins to talk about I4 and delivery
      • Then really drill in and talk about use case
    • need to accept the use case; want to have it work
    • To make quantitative, need to explicitly accept/reject the line items of the use case
      • Document bits that are accepted... Drive to quantitateive success
    • Some desired features will become longer term goals
      • Oliver suggest documenting in a separate document
      • Example, pull out line items of use cases that can't be done in 4.5
      • At same time ensuring that use case is still "usable"
    • The separate longer term document helps us drive 4.6.

Test initiative

Oliver noted that documents look professional

  • Oliver will be sending out details to wider audience
  • Testing appears under control... As bugs come in test project will handle them
  • Expect to see overall resource benefits (more getting done because of reduced testing load) in near future

Joel will be documenting how he integrated into build process.

  • Oliver will distrib to community and start getting feedback/questions from community
    • Schizophrenia regarding whether or not eclipse wants community input (eclipse usable by itself versus simple vehicle for products based on eclipse)
    • board has expressed that they want OOB/usability
    • a few suggested pointing at commercial products and that eclipse just there to support commercial products.

next time for board meeting; Oliver would like collateral all ready to go

  • there are things where community has value -- e.g., standards, etc.
    • Oliver believes getting board reps on board will help with getting resources


PMC cancelled next week (its Boxing Day in Canada)....

Alex asked Paul a question about a bugzilla comment

  • why do we have two entries in bugzilla plan and priority separately.

Paul said a few words about the meaning of plan and priority

  • P1 -- commited to fixing in 4.5
  • Plan designation

Paul noted still lots of defects without P1 and without plan keyword.

  • Leads need to be aware and resolve
    • Paul (for Test) typically runs checks at end of each iteration.

These definitions cause recurring confusion

  • Paul will put on wiki the definitions and bkms for updating and attach to minutes.

Back to the top