On Fri, Apr 29, 2011 at 3:20 PM, Max Rydahl Andersen <max.andersen@redhat.com> wrote:

> JBT 3.3.0.M1 / JBDS 5.0.0.M1:
> * Savara/Scribble missing (refactored?) feature: https://issues.jboss.org/browse/JBIDE-8817 (blocks aggregate build)

cc'ed Gary  -  I understand that as already have been fixed. Gary - maybe missing some release/push somewhere?

> * AS - single test failure: https://issues.jboss.org/browse/JBIDE-8816

cc'ed Rob - that looks like you not updating the tests after you refactored publishers ?!

> * VPE had 219 tests on Jan 11; now only 17 tests are running and of those, 6 are failing

cc'ed Yahor - doesn't this normally just happens because we run on a server without the right libraries and thus the eclipse simply dies?
Or is there something else known about this Yahor ? you mentioned looking into reducing number of tests to avoid duplication ? Sounds like
a rather big reduction though ?
Any prooflink? I found only these two: [1], [2]. No failed tests here, and their number is >>17.

> * Seam has 3 test failures: https://issues.jboss.org/browse/JBIDE-8759

Looks like Victor already commited fixes for this one (based on jira comments)

> JBT 3.2.1.M1 / JBDS 4.1.0.M1:
> * Aggregate builder is fixed (I think!), but hit a network snag trying to fetch latest bits because I disabled the "wait until all upstream are done" option in the job in order to test my latest publish.sh fixes. Respin will fix this & allow me to verify my fix worked, then I can re-enable the block-while-upstream-spinning option.
> * One last Seam test failure: https://issues.jboss.org/browse/JBIDE-8818

cc'ed Alexey - not much information available - any ideas ?

> * AS & Deltacloud jobs time out after 2hrs; job timeout increased to 3hrs. Would respin but Hudson is in shutdown mode.

Did it timeout because of failure to get all dependencies or was it actually doing testruns?

> * JSF was clean (no failures) but now has 7 new tests and 14 new failures: https://issues.jboss.org/browse/JBIDE-8819

Alexey - another case of tests running before builder completes ? Can't we query the state of the builder and fail on that so we know when that is the case ?


jbosstools-dev mailing list