TPTP-PMC-20080109

Logistics

Attending: Oliver, Harm, Chris, Paul, AlexN, Joanna, AlexA Absent: Guru

AlexA agreed to be called Sasha to avoid confusion with AlexN.

Cleanup of committer lists

Paul has now seen official note from leads for all projects except for platform.

• Joanna agreed will send it out
  • She noted some confusion as she thought Harm had done it on her behalf previously.

Paul asked if anything needs to be done about tools project

• Harm indicated this project does not exist anymore
  • An old Hyades thing that went elsewhere some time ago

Ganymede I4 delivery

We talked about M4/I4 test status completion

• Oliver noted that in the test reports project completion was a ~50%... Whats up.
  • Platform completed what they planned to do but some may be sitting in her inbox not checked in yet
    • Paul indicates that list of tests itself may still need cleanup
    • Joanna will look at list and clean up as needed
  • Monitoring completed what they intended and have checked it in

Ganymede I5 startup

A note arose regarding making sure that anything that will involve translation needs to happen in soonest possible iteration

• Action item for leads to take this into account

Oliver posed two questions for each project: For everything you are doing for I5

• Is it logged in weekly schedule and in bugzillas
• Are test pass contents updated to be accurate for I5

Test

• yes, yes

Monitoring

• Not all in weekly schedule
  • Need to sort out because of some I4 overflow, some I5s are moving to I6
  • This sorting will be done by next week
• Test suite is updated for existing stuff.
  • May be some updates for tests for enhancements in plan
  • Sorting by next week

Platform

• Regarding weekly schedule and bugzilla -- Joanna triaged 150 bugs I5/I6/I7
  • Joanna noted that we are talking about a large number ~100 defect numbers for I5
  • Does not think it worthwhile to transcribe all of them into the table
  • Is doing out weekly update indicating the defects on tap for the next week
  • Can create a bugzilla query that one can use to automatically identify them
  • Plans to put in table high priority ones
    • Joanna has a few committers that she still needs to coordinate with regarding I5 plan
• By next week test plan will be submitted

Trace

• Bugzilla/Weekly scheudule -- Not there yet but will have it there by next week
• Test pass updates will be there next week as well
• Oliver asked if Trace/profile need any support from PMC level.
  • Alex will work with Guru to decide

Oliver asks if all work (e.g., .NET) appears in weekly table and in bugzilla

• If he does a 4.5 I5 bugzilla check, not all are in weekly table

Regarding 4.4.1 Release

Joanna raised topic of 4.4.1 release

There was a historical EMO request regarding OpenSSL dependences for 4.4.1

• We decided to defer that for 4.5 but the question remains (for 4.5 now)
  • JSSE is the 4.4.1 solution
• New questioncoming in on OpenSSL
  • Why 3rd party download, "why not distribute with TPTP"
  • Chris noted that we did this to make it easier on them to approve
    • There is still some questionabout OpenSSL build flags
  • Joanna will continue to follow up

NOTE: 4.4.1 test pass overlaps with 4.5 test pass for I5

• 1/28 is start of test pass for 4.4.1 (full 2 week test pass)
• Most stuff that is changing/has changed in in Platform
• Joanna would like to move test pass in but
  • JSSE discussion/finalization is still in process between Igor/Jonathan

11 open defects targeted to 4.4.1

• several security items
• broken links
• Attach failure for AC
• Release note issues to Alex and Joanna
• Probekit one to be moved -- (201287) probe instrument driver does not instrument directories that end in certain string
  • Harm finds the deferral a bit odd
  • Joanna went and looked and noted she checked in a patch already.
    • She will review it and possibly leave to 4.4.1

Harm notes that since the open defects are primarily platform, she is in drivers seat for deciding if to start test pass early.

AlexN will be handling key interface to translation team

AI all leads to come back with recommendation/plan of how they will cover 4.4.1 test pass

Test Automation Initiative

Paul is still working on it...

• Reports are automated
• Still working on running of automated tests

Harm has question

• sounds like it is oging the way of some of the code coverage effort where the full benefits seem to be continually on future
• Paul indicates the problem is an overload on his own time.
  • Either need to pull in more folks or accept the deferral
  • Paul overloaded between project lead, AG lead, plus automation effort

Note: This is effectively an incubator item so we need to be careful about progress expectations in weekly call

• Follow up for a few more weeks and if overload continues will treat as separate discussion item

AlexN notes that the savings he entered in the table (estimating value) is speculative assuming automated build tests

• Monitoring hasn't saved too much from automation of test themselves (because they'd already done this)
• Oliver asks if may be some justification to help a bit with Paul's bottleneck
  • Paul notes this may work well in theory but there is a learning curve to helping with these bits

POG Initiative

Ideally Harm's scenario will work in upcoming iteration. A meeting about POG follows this PMC meeting. See TPTP-POG-20080109 for a summary.

MISC

Chris noted that last year we discussed having a meeting to discuss component assignment to projects

• Oliver says to be careful about just rearranging to make it look pretty rather than to extract true efficiency
• AI to Chris to pull reasoning for this discussion from last years minutes to refresh our motivation

Ganymede release train

• Oliver unsure if rapid development/iterations is as valuable for long lead development.
• Definitately valuable for rapidly evolving
• No real resolution made...
  • We seem to plan to stay on Ganymede and discuss future simultaneous releases.