TPTP-PMC-20080813

Logistics

Attending: Oliver, Chris, Harm, Eugene, Joanna, Paul

Any objections to last week's summary?

• No

Does anyone need to leave early?

• No

4.5.1 Status

Joanna gave a summary. There were some last minute blocking defects to resolve regarding AC and IAC.

• The 8/12 build was given to IBM consuming products to test with
• Joanna has asked the TPTP team to finish testing by tommorrow
• Test pass results in low 20s right now. A number of results need to get checked in.
  • Joanna still need to update to remove monitoring stuff that was killed.

There are still are some defects that we know about that we'd have liked to fix.

• A few JVMTI defects assigned to Yunan. We think it will require some significant code changes
• Oliver asks if this would this require a 4.5.1.1.
  • The desire for a 4.5.1.1 for this would depend on consuming component
• Eugene noted that these are not new regressions. They were in 4.5 as well.
  • Issues related to TI memory profiling
• The Code base is pretty much frozen for 4.5.1.

Oliver asks if this acclerating 4.5.1 freeze and the associated process was worth the extra work to avoid a point release.

• If we get our delivery pretty much done early then consumers can iterate a bit with it to allow avoidance of last minute blockers.
• Joanna noted that we were not the first/only component to do this for the consumer
  • At least 4 projects are doing something similar
• This is a good indicator that the consumer should change their scheduling to better align with Eclipse schedules which are documented well in advance.

Oliver asks if for future we can align our test schedules with consumer test schedules somewhat more

• Joanna was so tied up in 4.5.1 did not have a chance to talk to them yet. She will.
• Basic query: will IBM consumers pick up intermediate milestones or will they wait till something more final

Oliver asks if Joanna needs anything from folks on call.

• Nothing right now.

Website Discussion

It is pretty hard to find the key parts of profiler stuff by going to the TPTP web site.

• The web site is somewhat out of date. Is there anyone to help fix?
• Harm notes that there are no more resources to actually go create new content
  • Harm notes that some other projects that are going to a really simplified web site
  • Less material equals less maintenance
  • Harm will look at it and propose something simpler.
• Harm notes that we are obliged to have something that describes architectural structure (it is linked by the foundation)
  • That requires some updates. It may be possible to do the update by reducing the level of detail.

Profiler Discussion

Oliver asks Eugene what the next steps are for the profiler

• Eugene plans to send wiki page to lists.
• Use wiki as a source for educational material

Oliver asks where the back/forth discussion should be?

• Newsgroup would be best place for this.

Oliver asks if we can we launch this initiative at the same time as September release.

• Oliver is looking forward to this community engagement

Oliver poses the query: Let us say we get 4 months down road and run into a key issue for profiler user.

• May not want to take 4.5.2 milestone since it has a bunch of stuff unrelated
• May want to take 4.5.1 with a patch.
• Have users using as close to what they were using with just patches they need.
• Loosely this is the equivalent of some sort of hotfix.

Harm suggestion:

• Do not logically branch.
• Users can pull from update manager updates specifically for the profiler component
• If there was a case where we needed to do a hot patch could put it into bugzilla as a zip that consumer can download.

Misc

Administrivia (committerships, etc) should be pretty much done

• permissions should go automatically when the votes finish

Siemens had an earlier mail thread about joining TPTP to work on static analysis. Eugene has not heard back from them. Oliver will probe them again.

• Harm asks if we should poll the memory profiler guys to see if they are still interested in incubating in TPTP.
  • Oliver asks if there are any new points that we wait to raise with them.
  • No new points. Just the general benefits of incubating in mature product and better match to TPTP

A thesis student in Europe doing thesis on test stuff has contacted Paul regarding willingness to provide patches and such.

• Does not yet know exact topic but it will be in the area.
• Oliver asks if he should contact him or his advisor to try to set up discussion around areas that would be mutually beneficial.
  • Paul will send details and Oliver can send mail as PMC lead.

Oliver was at a lead planning discussion yesterday. Nobody seemed to know the XML roadmaps were due on September 30.

• For TPTP side, someone needs to to own the edit
• We will farm out the actual content creation effort and have a designated editor
• We should try to identify master editor in a week or so.