TPTP-PMC-20080123

Logistics

Attending: Paul, AlexA, AlexN, Chris, Harm, Oliver, Bing (on behalf of Joanna)

Project Structure and Methodologies

Committer Cleanup

Paul launched discussion regarding Committer Cleanup bugzilla comment 12. Asking for confirmation from projects.

• Trace
  • still some folks who are inactive but leads have not yet polled the members.
  • AI to AlexA to folloup with Guru -- to see how far Guru had gotten
• Monitoring had one name that had questions (Balen)
  • This name was resolved during the call and will not be removed during 4.5 timeframe
    • During planning for 4.5, he was listed as participating at 10%
• Platform
  • AI for Bing to folloup with Joanna -- 1 name remaining

Harm suggested that if this is going to lag, we should go ahead and cleanup what we know about since it will already be a lot better than it was.

New Trace Lead

Oliver welcomeed to AlexA as new trace project lead and gave him a "welcome to leadership" speech.

• You will drive things forward for your projet
• Make sure you attend the calls
• Feel free to ask for help

How We Work

Oliver gave a basic reminder to project leadership on how we should be working in the project.

• All work should be driven by bugzillas.
  • No work should be done without bugzilla
• If bugzilla has any significant impact should include AG review
  • At the very least a discussion with Paul to decide if full meet is needed.
• If you touch other components besides your own
  • Need to involve members from other components
  • For example if UI will change in response to something need to involve them earliest possible time
• Feel free to open a new bugzilla for additional work that needs to be done.

4.4.1 Discussion

Oliver asks each project how things are going for 4.4.1

Test

• No real engineering work. Just some copyright fixups
• Have committer who has been doing testing on candidate. No reported issues

Monitoring

• Have to finish defect related to plugin versioning and PII

Platform

• Bing (reading off of Joanna's mail)
  • a defect that she wants AlexN to resolve
  • Still open issue with OpenSSL for sizing
    • Team noted this is item actually moved to 4.5 (deciding on how to package it)
  • another defect related to BTM (for 4.5)
    • Wanted to switch MX4J to new version (This is a 4.5 item... see below)
  • Profiler issue (patch for attach/detatch on IBM VM) still remains and is being actively worked.

Trace

• There are no open defects
• There are some remaining questions about BTM .NET test
• AlexA will talk with the team regarding ensuring that the tests are run in test pass

Closing on 4.4.1 (entering test) this weekend

A bit more on IBM JVM

A bit more discussion on the IBM VM issue. The guys (primarily Asaf) are also working with Eugene on Java 1.5 IBM issues

• Do we predict success in 4.4.1 for IBM profiler
• Eugene reports most scenarios are resolved with hacked module that Asaf provided
• Asaf has sent a report that there are some more complicated scenarios that remain to be seen
• Additional investigation underway
  • Oliver to follow up with Eugene after the call
  • Alex to make sure that Asaf and Eugene close

Harm raised the question that since things are a lot better, is it possible to go ahead and get it promote what we do have in parallel with Java2D demo tests.

• [Note, Discussion on getting this patch in place proceeded via mail threads after the call]

A bit more on SSL in 4.4.1

Question was raised regarding JSSE solution was planned to be in 4.4.1

• Try to get the code checked in
• Harm noted this seems to still be tagged as P3 and is open with patches attached
• Jonathan said still working on it as of last week and has a bit more code to go.

4.5 Discussion

Oliver assumes everyone is done planning out iteration. Oliver going thru weekly schedules and asking each project how they fare

Test project

• A few smaller things have slipped.
• Enhancements are progressing. a bit behind. (BVT and Emma are slipping)
• Nothing too alarming given are in last development week for 4.5. Some discussion around some participats commitment to task.
  • Oliver can't help with bugfixes but if someone who is supposed to be committed is NOT doing their task
  • Oliver can try to motivate them by sending puttputt boats or talking to their management
  • Leads should spend as much time on technical lead tasks. Don't get TOO tied with nontechnical people management before involving Oliver

Platform project

• Bing again reading off of Joanna's mail
• BTM build is broken. Can AlexN work with MattMings?
• MX4J versioning update
  • AlexN understands that this has dropped off and will not be changed. It was considered too complex
  • Initially had defect targeted for I5. After further scoping decided too complex
• Profiler defect -- Patch solved attach/deteach but start/stop monitoring aspect remains
  • Oliver asks if this is a cause for large concern?
  • Not at this time

Monitoring project

• Will finish feature work today
• Code ready; needs review and work with Joel to update build script
• 3 defects have slipped this week
  • AlexN is not concerned with this but thinks they can catch up
• Has updated schedule with defects they will work on next week
  • Oliver compliments AlexN on the weekly schedule for Monitoring
• AlexN has commited to do two defects on platform project related to model cleanup and model generation
  • Will talk to Harm about it and hope to have them done by end of day tommorrow

Trace Project (separate from profiling activities managed in Platform)

• AlexA Currently working with 1 committer
  • Team reports blockage on one enhancement related to TI support for BTM (replacing old PI)

An aside related to BTM

• AlexA is still working to get communication channels opened with bugzilla filers and asignees
  • Oliver asks if he need help.
  • Alex said that a summary of who will participate on Trace from IBM would help
    • Harm, Eugene, Richard have sent mail.
    • Joe is engineer at IBM who would be involved. He just had a baby and has been in and out
    • AI: Bing to involve Joanna to make sure that someone on Joanna's team is covering for Joe
    • IBM reports that Joe will be back next week
  • AlexN talked a bit about that bit of the bug that Richard with investigating
    • IBM to figure out how to fix the problem that they have committed to.
• In any event, it does not look like enhancement will be complete in I5
  • Some other related BTM PI-TI bug issues may be impacted

AlexA saw that 4 guys are involved in BTM trace bugs

• Alex still trying to set up dialogue with them
  • The two IBM developers Alex should talk to the most would be Richard and Joe Toomy

AlexA sees 5 bugzillas filtered under BTM and asked for clarification regarding which he needs to drive under Trace.

• Example is this trace or is it monitoring?
• This is a monitoring defect
• BTM is theme that spans monitoring, trace, platform
• AlexA does not need to worry about all things related to BTM just that subset in Trace based on the component assignments.
• AI to AlexA and AlexN to follow up offline regarding the contributions and mappings of some of the items to projects.

In general, the only item that relates to Trace for BTM is ARM.

Finishing up 4.5 discussion

Anything else for 4.5?

• No response

AlexN asks if we still have a week of development next week (since we adjusted schedules last week)

• Response: next week is intended to do full 4.4.1 test pass because we didn't have time for both 4.4.1 test pass and 4.5 in same week
• If there is a bit of of time left after the test pass, it did not appear that there was objection to 4.5 work.

AI to Paul to finish updating master plan page to indicate what we are doing with test for 4.5

Test Initiative

• Made forward progress and got to a point of trying to test
• Machine running test had crash that requires OS reinstallation
  • Optimistic based on behavior before the machine crashed.
• If can get limited functionality running for each build (e.g., some of the junits) it is better than holding it up until all issues are resolved.

Harm asks if what works now lets us run black box junit tests as part of build

• If JUnits run, why don't we go ahead and get them running
• This is pretty much what Paul was suggesting and he will be talking to team about getting it done
  • This does not need to be gated by having JUnit plugin tests working.

Paul will try to get something more tangibile for next week discussion

POG initiative

AlexA gave brief update.

• Already opened 20 bugzillas
  • 3 targeted I5
  • 2 of those resolved already
  • 1 should be done soon
  • 6 are even targeted to 4.4.1
    • (those related to problems with IBM JVM)

Oliver asks if we have already identified which defects will be fixed for Jan 25 drop

• Yes

Oliver if this these make good progress to usability (to Harm as lead user for this use case)

• Yes

Harm is working with each build as it comes thru

• Will pull new candidate drop as t comes thru

MISC

Oliver creating Report to TPTP sponsors (IBM/Intel)

• test automation
• focus on use cases
• Hopes that this will be a precursor for discussion on resources for 4.6
  • Avoid classic problem with doing a good job but right folks at customer don't understand amount done.

Want more clear discussion on needs from users (IBM, Intel, community (possibly)) to hopefully make more resources fall out for 4.6.