| [wtp-dev] Minutes for the WTP Status Telecon, 2005-11-03 |
1. Review of Open Action Items [1] - Arthur Ryman
| 110309 | enh | P3 | jst.jsp-inbox@xxxxxxxxxxx | ryman@xxxxxxxxxx | NEW | --- | [action] Add WTP 1.0 JST Component API Plan Items to Master Bug 110276 |
| 110308 | enh | P3 | david_williams@xxxxxxxxxx | ryman@xxxxxxxxxx | NEW | --- | [action] Add WTP 1.0 WST Component API Plan Items to Master Bug 110275 |
| 113134 | nor | P3 | david_williams@xxxxxxxxxx | jeffliu@xxxxxxxxxx | NEW | --- | [action] Assign target milestone to critical/blocking bugs |
| 113137 | nor | P3 | tbashor@xxxxxxx | jeffliu@xxxxxxxxxx | NEW | --- | [action] Document the set of facets that we'll need for J2EE projects |
| 114446 | nor | P3 | cbridgha@xxxxxxxxxx | jeffliu@xxxxxxxxxx | NEW | 1.0 M9 | [action] Open API requests for non-API usages in WTP |
| 114914 | nor | P3 | syeshin@xxxxxxxxxx | jeffliu@xxxxxxxxxx | NEW | --- | [action] Post some examples of what not to do for context help/infopop IDs |
Review bug backlog, builds,
and plans. Assign two teams to test in disconnected mode.
David - Build complete, test finished.
A couple compile errors, teams please look at the compile errors. I believe
one is in jst.j2ee and the other in jst.ws.
David - Plan is still to smoke test
and fix any blocking bugs by noon tomorrow.
Jeff - Who's going to test offline?
Craig - I will.
Phil - I will.
Lawrence - How is disconnected testing
done? We should have a case where we start off with nothing (no cache),
and make sure we don't get any errors. Another case is to have things cached
and make sure the cache is reused.
David - We have 5 blocking bugs in the
open state. Do we need them for this week's build?
Jeff - A lot of them are in data tools.
Der Ping, can u review them and send a quick update to the dev list?
Larry - They should be in today's I-build.
Is it too late?
David - No, not too late. We want to
fix any blocking problems so that developers can make progress with our
I-builds. Before EOD today is still good.
David - Reminder everyone that there
are two weeks left for M9. Code function, API changes should be done by
next week. M9 test plan done by next week. After end of next week, we will
have a week of testing and critical bug fixes.
Ted - UI changes allowed post M9?
Lawrence - what kind of UI changes?
Ted - Clean up module depedency wizard.
Lawrence - From the doc standpoint,
late UI changes are difficult to deal with. Please let Lawrence and Susan
know ahead of time.
David - Post M9 UI changes should be
kept to a minimal. Exceptions can be review by the PMC.
Lawrence - or at least a note to the
dev list.
Chuck - Going back to the jst.j2ee compile
error. This will break EJB creation, we will drop a fix.
Chris - Chuck's fix will fix jst.ws's
compile error as well.
3. WTP docs - Susan Yeshin
4. WTP 1.0 API Scans - Jeffrey Liu
Review action to finalize
component.xml
https://bugs.eclipse.org/bugs/buglist.cgi?query_format=advanced&short_desc_type=allwordssubstr&short_desc=Update+component.xml&product=Web+Tools&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&bug_status=RESOLVED&bug_status=VERIFIED&bug_status=CLOSED&cmdtype=doit&order=Reuse+same+sort+as+last+time&field0-0-0=noop&type0-0-0=noop&value0-0-0=
Jeff - component.xml are all updated.
Thanks.
Jeff - API reports needs to be run more
often, can we add it back into the normal build process? It only takes
a couple minutes to run.
Chuck - We are cleaning up our APIs,
would be nice to see these reports more often.
Jeff - Please also look at the Unit
test and Javadoc coverage. Our goal is 100% coverage.
David - Any chance these API reports
will break the build?
Jeff - I do not think so, but I'll test
it.
David - if it's 1 or 2 minutes, then
that's ok.
Chuck - Would be nice if we can generate
these reports from the workspace, then we don't even need the reports to
be generated as part of every build.
Jeff - Great idea, will work on it.
Will post instructions on how to do that.
[action]
Jeff to make the API scanner capable of running from within the Eclipse
workspace.
5. WTP 1.0 Performance - Jeffrey Liu
Performance results for 1028 I-build [1]. Three issues that I want to discuss:
1. It shows quite a few regressions, some even in the 100%+ range. However, the difference could be 10ms vs 20ms. So, are these really test data that we want to show in the summary page?
Jeff - Tim, I believe you own most of the testcases on the summary page. Do we need all of them to show up in the summary?
Tim - Probably not. You must be using a very fast machine, it took 10 - 20s on my machine.
Tim - We should investigate why we are seeing hug regressions for some of them.
Jeff - I'll work with Tim offline on this.
2. Since performance tests are not run as part of the normal build process, performance JUnit failures are not identified until after the weekly I-build is declared. As a result, no performance data for the failing testcase. I think we need to make sure when the performance testcase gets run, they will run successfully. Any suggestions on how to do that?
Conclusion is to run run performance tests more often. We don't need the data from daily builds because they may not be accurate due to things like compile errors or failing JUnit tests. But at least they will tell us which performance testcases are failing. We'll open bugs for those and get them fixed.
3. Start reviewing performance regressions
Craig - to avoid ms testcase, we can wrap it in a loop
Lawrence - or use a slow machine
Jeff - a slow machine will be useful, let's discuss this offline.
Tim - Are we doing scalability testing? For validator, editors, etc...
Tim - We can doing scalability testing based on file size, project size, etc...
[action] Jeff to start the work on scalability testing.
[1] http://download.eclipse.org/webtools/downloads/drops/I-I20051028-200510280017/perfresults/graph/performance.php
6. WTP Features and Subsystems Status [1] - David Williams
[1] http://www.eclipse.org/webtools/development/arch_and_design/subsystems/SubsystemsAndFeatures.html
David - WST done.Still issues in WS-I and WSDL, will be fixed next week.
David - I would expect JST to be more difficult. Since we only have a couple weeks left, we may not have enough time for this. I'll discuss with the PMC about the risk that's involved.
7. Flexible Project Development Status - Chuck Bridgham
Chuck - No update, we can probably take this off the agenda.
8. Project Facets Development Status -Konstantin Komissarchik
Konsta - off the agenda.
9. Other Business -
Open
Ted - Wanted to ask Naci about the EJB facet work.
Chuck - You talking about the UI?
Ted - yes
Chuck - we are working on it this week.
Tim - Doing the paper work for promoting Amy Wu to be
a committer.
Congratulation Amy!
Jeffrey Liu
IBM Rational Software - Performance Analyst
IBM Toronto Lab.
8200 Warden Ave. Markham, Ontario, L6G 1C7
Internal mail: D3/R8V/8200/MKM (D3-268)
T/L: 969 3531
Tel: (905) 413 3531
Fax: (905) 413 4920
jeffliu@xxxxxxxxxx