TPTP-PMC-20080827

Logistics

Attending: Oliver, Joanna, Chris, Eugene, Paul

4.5.1 Discussion

Joanna noted that Lei is reporting build issues

• Chris noted that his committer account was deleted because of the committer switch from test to platform
• New committer status is in process
  • AI Joanna to set up repositories so process can complete
  • AI Chris to follow up with webmaster on getting ssh access back so builds can proceed

Test pass status

• Paul finished his testing
• Joanna scrambling to get stuff in
• Managed to get local build done despite profiler error in code
  • Eugene sent Asaf's information off to VM team who sent it back saying that they believe the issue is in the agent
  • There is also another issue that Eugene ran into is being analyzed as well
    • Appears to be some issue with pipes not being closed/cleaned completely
    • Eugene asks what the last day is... Joanna notes it is today or tommorrow
    • Eugene notes that one of the sightings is considered blocker by the consuming product at IBM

Oliver asks when a postmortem on 4.5.1 is planned

• Joanna notes it is not yet planned
• Harm would also like to do a scrub/review to identify
  • Stuff to do next
  • Code that may be brittle (particularly in the TI profiling area)
  • Example: Review Java6 support code and a few other components.

Joanna notes that it is not yet clear if we have slipped so far that a 4.5.1.1 will be strongly requested

• We have bit into part of one of the consuming products time buffer

Oliver asks about the September fall release: Are already frozen. What will happen next?

• Joanna notes that the underlying dependences may change but many of them are frozen already
  • Eclipse platform is being pushed to freeze in next day or two
  • We have a test pass on schedule around Sep 10. Need to decide if we will still do this.

In the discussion, the general belief was that we may need a linux (x86, EM64T) test pass because build was cobbled together at IBM rather than using the build infrastructure used for prior releases (not as worried about IPF at moment)

• Oliver asks how much effort it will take to do the test pass
• Paul believes smoke tests would probably be sufficient given that the smoke tests are pretty extensive
  • Paul notes that the test project will have limited resources for a bit because they are in between a few student co-ops.
  • Joanna thinks that having the new co-ops manual test is good learning exercise.

By next PMC should be done with some core testing and ready to talk about 4.5.2

• Need to get XML roadmap work done
• Joanna was going to identify a single editor for document

Does Eugene think will be ready to go with POG discussion at 4.5.1 release

• Eugene thinks that the linux patch (discussed above) may be needed
• Oliver thinks we might be able to get away with starting with windows initially
• Oliver underscores how mega-awesome Eugene's work has been here.

European Test Student

Gunter (student in Europe) joined to talk briefly about synergy to TPTP test and his research projects. Before Gunter joined, Oliver asks if TPTP Test (Paul) feel it has resources that can expend to support a bit?

• Paul believes that contribution from more sources is inherently a good thing
  • Paul notes that most of the parts of what Gunter has talked about related to parts of the test project that have NOT been deprecated (i.e., they correspond to good live code)
  • If Gunter is committed to providing some code back, it would be a net win if we can invest some light support and get some enhancement back.
  • Clearly, with our resources, how much we can do is limited.
  • If Gunter submits code back IP review will be required on patches/etc (normal Eclipse processes)

Gunter then came online and Oliver talked a bit

• TPTP guys have read his collateral.
• TPTP would be mega excited to get contributions.
• It sounds like there is some good potential alignment between TPTP and Gunter's research topics
• TPTP would like to help along a bit and get some contribution back in return
• Oliver asked Gunter to make sure that he is careful about how he creates his code so that he maintains the ability to donate back to TPTP
  • Action item for Paul and Gunter to talk offline about the legal constraints
  • Paul and Gunter will keep in touch and move the overall work along

MISC

Chris asked about status of syncing build OSes between Intel and IBM

• Joanna thought that Lei was unable to install Redhat AS 2.0
• Chris asked Joanna to send specific OS/compiler details that would be ideal to him
• Chris will talk to PRC team about what would prevent a syncup for 4.5.2 timeframe.

Got long response about heap analysis tool incubation

• They put some material up on wiki but didn't really get comments back. They wonder if there is really opportunity for synnergy
• Oliver asks if we have a story about possible synergies at this point.
  • We had some discussion about core memory profiling in TPTP but heap analysis team not really interested in maintaining bits there.