Skip to main content

Notice: This Wiki is now read only and edits are no longer possible. Please see: https://gitlab.eclipse.org/eclipsefdn/helpdesk/-/wikis/Wiki-shutdown-plan for the plan.

Jump to: navigation, search

Difference between revisions of "TPTP-PMC-20080409"

(New page: assorted fixed from Joanna 225672 225396 will not be resolved in this pass. * workaround exists can patch in I6 and provide for consuming products * will give a build to drop today To...)
 
(On WTP EE Bundling)
 
(One intermediate revision by the same user not shown)
Line 1: Line 1:
 +
== Logistics ==
 +
* Attending: Oliver, Paul, MikhailV, Joanna, Chris
 +
** AlexA on vacation...  back next week.
  
assorted fixed from Joanna
+
== Discussion around I6 delivery ==
225672
+
Note: Chris came in partway into this discussion.
  
225396 will not be resolved in this pass.
+
There are a number of ultracritical fixes from the teams and we want to update our I6 delivery (primarily because I6 is picked up for testing by some IBM teams).  It turns out that we are quite lucky because even though the I6 deadline is passed and we already dropped, higher level Eclipse pull of I6 deliveries hasn't happened so we can squeak an update in if we are quick.
* workaround exists
+
  
can patch in I6 and provide for consuming products
+
There are assorted fixed from Joanna
* will give a build to drop today
+
* On example is [http://bugs.eclipse.org/225672 here]
 +
* Some won't be resolved this quickly
 +
** [http://bugs.eclipse.org/225396 Example] that will not be resolved in this pass (but a workaround exists
 +
* We can patch in tweaksI6 and provide result for consuming products
 +
* Final updated drop today
  
To get variances from build proper need contact with Bjorn
+
Oliver asked if there is always this leeway to redrop within a few days after the deadline.  Joanna responded that to get variances from build need contact with Bjorn
* Joanna not sure if that is true or not.  In this case lucky because build has not yet been grabbed so its not officially a refresh
+
* Joanna not sure if that is true or not that there is always this flexibility.   
 +
* In this case lucky because build has not yet been grabbed so its not officially a refresh
  
How can we avoid in future
+
Oliver asks how we can avoid this issue in the future
* For TEST, had two defects opened last wednesday
+
* For test project, Paul mentioned that a few of these ultracritical defects weren't opened until last Wednesday
** For Jerome defects he was working quite hard but is only part time.  Has been quite flexible
+
** For at least one of these, the engineer is only part time and he's been quite flexibile to get it fixed as quickly as possible
** Paul thinks that a bit more night work may have been efforted on other defect
+
** We are lucky  it didn't conflict with his other deliverables.
  
there will be people who go above and beyond and we want to make sure to recognize that.
+
Oliver hosted assorted discussion about productivity and wanted to be sure that we recogize and work to resolve less than productive folks and also to recognize and reward more than productive folks.
  
would this have been fixed w/ 2 test passes
+
Chris asked the question as to whether this was caused partly because of the fact that we let ourselves abbreviate our test passes this time.
* Got hit with unusually large number of defects in I6
+
* Joanna noted that we did hit with unusually large number of defects in I6 and that
* Joanna indicates that if we had had 2 test passes we should have realized because the volume of our modifications.
+
* Joanna indicates that if we had had 2 test passes it would have helped.
 +
* In retrospect, we should have realized because the volume of our modifications that two passes were needed.
 +
* Joanna did note some fear about I7 because there are a bunch of stuff going into I7 too...
 +
* General rule of thumb: In future we should bear in mind volume of changes as we decide whether to abbreviate test pass and perhaps be a bit more conservative.
  
Fear about I7 because there are a bunch of stuff going into I7
+
== Discussion around I7 ==
 
+
In future should bear in mind volume of change as we decide whether to abbreviate test pass.
+
  
 
I7 is in process.
 
I7 is in process.
* AlexN has sent some comments on 2 defects
 
** 225862 context for log information dialogue -- change in provisional log viewer api
 
*** May want to discuss merits on AG call.
 
*** Joanna has what she needs from this meeting
 
*** Have they sized it -- ~4hrs of work
 
** 219486 slipped past M6
 
  
Joel about WTP EPP stuff
+
AlexN has sent some comments on 2 defects
* Spoke to david williams
+
* [http://bugs.eclipse.org/225862 Here] he gives some context for log information dialogue changes
* will running JEE stuff.
+
* There have been requests for changes in provisional log viewer api
* details are known
+
** We may want to discuss merits of this change on AG call.
 +
** PMC asked if defect has been sized.  It has been sized at ~4hrs of work
 +
** Since it is small, it may be preferable to discuss over mail instead.
 +
* [http://bugs.eclipse.org/219486 This] one lipped past M6 but is pretty much done
 +
 
 +
== On WTP EE Bundling ==
 +
Joel spoke to his AI about evaluating WTP EPP stuff
 +
* Joel spoke to David Williams
 +
* Got specific details about how to get integrated into builds
 
** Eugene and Joel are working on it.
 
** Eugene and Joel are working on it.
** Should be in the package starting this week.
+
** Should be in the package starting this week!
  
No objections
+
Oliver asked if there are any objections to our inclusion.
 +
* A question narose question about what components will be put in.
 +
* So far it only includs profiling
 +
** Eugene states that it has Platform, Trace, WTP integration only
  
Question about what components will be put in.
+
Question arose asking if we can tweak list later
* so far only included profiling
+
* Joanna thinks adding will be easiser than removing
  
Eugene
+
Question arose regarding how testing needs to work with the bundle
* Platform, Trace, WTP integration only
+
* Where and when do we need to put test results
 +
* What exactly needs to be put
  
Can we tweak list later
+
Question arose about potentially putting whole distribution in.
* Joanna thinks adding will be more possible than removing
+
* May allow getting rid of all in one (it would be WTP bundle)?
 +
* May reduce some test overhead?
 +
* Joanna is unsure how that will work w/ consuming products but will look into it.
  
On testing:
+
Quick intro email... AI to someone... (Joanna?) to David williams
* where and when do we need to put test results
+
* Say we are working on patch to let us build cleanly in bundle
Quick intro email:
+
* We are trying to get test ducks lined up and need more formal details about where and what results need to be checked in.
working on patch -- trying to get test ducks lined up and need more formal details.
+
** copy oliver
* copy oliver
+
** Need some sanity test due diligence to make sure we're okay with shipping subsets
* need sanity test to make sure we're okay with shipping subsets
+
** first release this way one may need more testing that subsequent ones.
* first one may need more testing that subsequent ones.
+
* (Joanna?) will follow up, copy joel and oliver
  
why not put whole distribution in?
+
== Followup on component owners ==
* May allow getting rid of all in one (it would be WTP bundle)
+
Joanna raised question regarding movement platform pieces to trace project.  Apparantly this wasn't quite as settled during AG call as we thought.  Background: Mikhail had sent note a while back about moving JVMPI/JVMTI over to trace.  Three of the 4 components Joanna is good with moving.  Joanna has concerns about one item.
* Joanna unsure how that will work w/ consuming products
+
 
+
Joanna will follow up, copy joel and oliver
+
 
+
Packaging of platform trace pieces
+
* Mikhail had sent note about moving JVMPI/JVMTI over to trace
+
 
* Profiling perspective org.hyades.trace plugin
 
* Profiling perspective org.hyades.trace plugin
 
* Logging pieces make use of this component as well.  May need to remain in platform
 
* Logging pieces make use of this component as well.  May need to remain in platform
** Mikhail?  it was discussed.  Discussed principle before and ends up making platform unwieldyDiscussion stated that if something is primarily used in one project to move it.
+
* Mikhail noted that this was discussed.   
** Joanna wants common components to be in platform
+
** We discussed the principle of having all common things in platform and how to avoid an unwieldy platform
** question about whether this will impact packaging and dependences from monitoring.
+
** Discussion seemed to indicate that if something is primarily used in one project it would be okay to move it.
** Who will do smoke tests on the profiling UI -- two different places?
+
** Joanna wants common components to stay in platform
 +
** Joanna noted that moving it may end up making some resources (IBM) attend more meetings which may not be desirable.
 +
** There was a question about whether this will impact packaging and dependences from monitoring.
 +
*** Actual packaging was not supposed to change
 +
** Joanna asked how this will impact (smoke and other) tests on the profiling UI
 +
*** Would they be in two different places (trace and monitoring) where there is consumption or just one?
 +
* Joanna aked what the goal is...  Is it just to get bugzillas to move w/ default committer?
 +
 
 +
== POG lead ==
  
What is goal? Just to get bugzillas to move w/ default committer?
+
The team discussed the POG lead. We used [http://dev.eclipse.org/mhonarc/lists/tptp-pmc/msg04719.html some material] posted to PMC list from Intel to frame part of discussion.
* will this double meetings for IBM folks?
+
  
---
+
Oliver thinks some of these pieces (actual soliciting of new users and driving commit from companies for resources and related reprioritization is actually his role.  He is looking for an advocate to summarize the strategy and to work with the technical lead.
  
need for an advocate
+
More discussion around this will happen next week at AG.
question about how much alex could do
+
  
Board presentation is coming together
+
== MISC ==
* evangelize to try to get new users
+
Oliver notes that the presentation for the board is coming together
 +
* Oliver hopes to evangelize via the board that we are ready for more users.

Latest revision as of 16:24, 11 April 2008

Logistics

  • Attending: Oliver, Paul, MikhailV, Joanna, Chris
    • AlexA on vacation... back next week.

Discussion around I6 delivery

Note: Chris came in partway into this discussion.

There are a number of ultracritical fixes from the teams and we want to update our I6 delivery (primarily because I6 is picked up for testing by some IBM teams). It turns out that we are quite lucky because even though the I6 deadline is passed and we already dropped, higher level Eclipse pull of I6 deliveries hasn't happened so we can squeak an update in if we are quick.

There are assorted fixed from Joanna

  • On example is here
  • Some won't be resolved this quickly
    • Example that will not be resolved in this pass (but a workaround exists
  • We can patch in tweaksI6 and provide result for consuming products
  • Final updated drop today

Oliver asked if there is always this leeway to redrop within a few days after the deadline. Joanna responded that to get variances from build need contact with Bjorn

  • Joanna not sure if that is true or not that there is always this flexibility.
  • In this case lucky because build has not yet been grabbed so its not officially a refresh

Oliver asks how we can avoid this issue in the future

  • For test project, Paul mentioned that a few of these ultracritical defects weren't opened until last Wednesday
    • For at least one of these, the engineer is only part time and he's been quite flexibile to get it fixed as quickly as possible
    • We are lucky it didn't conflict with his other deliverables.

Oliver hosted assorted discussion about productivity and wanted to be sure that we recogize and work to resolve less than productive folks and also to recognize and reward more than productive folks.

Chris asked the question as to whether this was caused partly because of the fact that we let ourselves abbreviate our test passes this time.

  • Joanna noted that we did hit with unusually large number of defects in I6 and that
  • Joanna indicates that if we had had 2 test passes it would have helped.
  • In retrospect, we should have realized because the volume of our modifications that two passes were needed.
  • Joanna did note some fear about I7 because there are a bunch of stuff going into I7 too...
  • General rule of thumb: In future we should bear in mind volume of changes as we decide whether to abbreviate test pass and perhaps be a bit more conservative.

Discussion around I7

I7 is in process.

AlexN has sent some comments on 2 defects

  • Here he gives some context for log information dialogue changes
  • There have been requests for changes in provisional log viewer api
    • We may want to discuss merits of this change on AG call.
    • PMC asked if defect has been sized. It has been sized at ~4hrs of work
    • Since it is small, it may be preferable to discuss over mail instead.
  • This one lipped past M6 but is pretty much done

On WTP EE Bundling

Joel spoke to his AI about evaluating WTP EPP stuff

  • Joel spoke to David Williams
  • Got specific details about how to get integrated into builds
    • Eugene and Joel are working on it.
    • Should be in the package starting this week!

Oliver asked if there are any objections to our inclusion.

  • A question narose question about what components will be put in.
  • So far it only includs profiling
    • Eugene states that it has Platform, Trace, WTP integration only

Question arose asking if we can tweak list later

  • Joanna thinks adding will be easiser than removing

Question arose regarding how testing needs to work with the bundle

  • Where and when do we need to put test results
  • What exactly needs to be put

Question arose about potentially putting whole distribution in.

  • May allow getting rid of all in one (it would be WTP bundle)?
  • May reduce some test overhead?
  • Joanna is unsure how that will work w/ consuming products but will look into it.

Quick intro email... AI to someone... (Joanna?) to David williams

  • Say we are working on patch to let us build cleanly in bundle
  • We are trying to get test ducks lined up and need more formal details about where and what results need to be checked in.
    • copy oliver
    • Need some sanity test due diligence to make sure we're okay with shipping subsets
    • first release this way one may need more testing that subsequent ones.
  • (Joanna?) will follow up, copy joel and oliver

Followup on component owners

Joanna raised question regarding movement platform pieces to trace project. Apparantly this wasn't quite as settled during AG call as we thought. Background: Mikhail had sent note a while back about moving JVMPI/JVMTI over to trace. Three of the 4 components Joanna is good with moving. Joanna has concerns about one item.

  • Profiling perspective org.hyades.trace plugin
  • Logging pieces make use of this component as well. May need to remain in platform
  • Mikhail noted that this was discussed.
    • We discussed the principle of having all common things in platform and how to avoid an unwieldy platform
    • Discussion seemed to indicate that if something is primarily used in one project it would be okay to move it.
    • Joanna wants common components to stay in platform
    • Joanna noted that moving it may end up making some resources (IBM) attend more meetings which may not be desirable.
    • There was a question about whether this will impact packaging and dependences from monitoring.
      • Actual packaging was not supposed to change
    • Joanna asked how this will impact (smoke and other) tests on the profiling UI.
      • Would they be in two different places (trace and monitoring) where there is consumption or just one?
  • Joanna aked what the goal is... Is it just to get bugzillas to move w/ default committer?

POG lead

The team discussed the POG lead. We used some material posted to PMC list from Intel to frame part of discussion.

Oliver thinks some of these pieces (actual soliciting of new users and driving commit from companies for resources and related reprioritization is actually his role. He is looking for an advocate to summarize the strategy and to work with the technical lead.

More discussion around this will happen next week at AG.

MISC

Oliver notes that the presentation for the board is coming together

  • Oliver hopes to evangelize via the board that we are ready for more users.

Back to the top