[hibernate-dev] IRC meeting - 6/14/2010

[Steve Ebersole]

Attached is the log of the IRC meeting for June 14, 2010.

-- 
Steve Ebersole <steve at hibernate.org>
http://hibernate.org
-------------- next part --------------
[Monday 14 June 2010] [10:01:58] <sebersole> morning (and evening to some) everyone
[Monday 14 June 2010] [10:02:04] <sebersole> lets go ahead and get started
[Monday 14 June 2010] [10:02:12] <stliu> hehe, okay
[Monday 14 June 2010] [10:02:26] <sebersole> so first order of bus is 3.5.3
[Monday 14 June 2010] [10:02:37] <sebersole> gbadner: how do you see that for this week?
[Monday 14 June 2010] [10:03:07] <gbadner> wed should be fine
[Monday 14 June 2010] [10:03:10] <sebersole> cool
[Monday 14 June 2010] [10:03:18] <gbadner> should I set up the 3.5.3 tag?
[Monday 14 June 2010] [10:03:33] <sebersole> wdym?
[Monday 14 June 2010] [10:03:39] <sebersole> we tag as we release
[Monday 14 June 2010] [10:03:39] <gbadner> in jira
[Monday 14 June 2010] [10:03:42] <sebersole> oh
[Monday 14 June 2010] [10:03:50] <sebersole> yeah thats fine
[Monday 14 June 2010] [10:03:56] <sebersole> really its a split
[Monday 14 June 2010] [10:03:57] <gbadner> okiedok
[Monday 14 June 2010] [10:04:12] <sebersole> rename 3.5.x as 3.5.3
[Monday 14 June 2010] [10:04:26] <sebersole> and set up a new 3.5.x
[Monday 14 June 2010] [10:04:38] <gbadner> oh, ok; didn't catch that last time
[Monday 14 June 2010] [10:04:56] <sebersole> well the "other" way is fine too
[Monday 14 June 2010] [10:05:00] <sebersole> i just find that easier
[Monday 14 June 2010] [10:05:33] <sebersole> otherwise you need to reopen all the closed ones to move them
[Monday 14 June 2010] [10:05:55] <gbadner> oh. right; got it
[Monday 14 June 2010] [10:06:24] <sebersole> some other loose ends from last week
[Monday 14 June 2010] [10:06:41] <sebersole> first, we got the intellij licenses squared away
[Monday 14 June 2010] [10:07:12] <epbernard> (thanks JetBrains :) )
[Monday 14 June 2010] [10:07:15] <sebersole> so if you did not hear and need one give me a shout
[Monday 14 June 2010] [10:07:19] <sebersole> so true
[Monday 14 June 2010] [10:07:23] <hardy> got the info and the new license works fine :)
[Monday 14 June 2010] [10:07:26] <sebersole> for the fast turnaround
[Monday 14 June 2010] [10:07:37] <stliu> and i finished the prod work for eap 5.1 last week(finally)
[Monday 14 June 2010] [10:08:08] <sebersole> also, i was unable to get anywhere with the JIRA->Fisheye integration
[Monday 14 June 2010] [10:08:15] <hardy> :(
[Monday 14 June 2010] [10:08:27] <sebersole> i opened a remedy ticket
[Monday 14 June 2010] [10:08:40] <sebersole> most of us know how well that works
[Monday 14 June 2010] [10:09:10] <sebersole> in the meantime i will ask contegix today to also enable the straight svn plugin
[Monday 14 June 2010] [10:09:40] <sebersole> as well as to analyze the logs to see if they see any issues in the fisheye plugin logs
[Monday 14 June 2010] [10:10:10] <sebersole> and lastly strong got gradle integrated into hudson
[Monday 14 June 2010] [10:10:22] <sebersole> unfortunatly they installed an older gradle version
[Monday 14 June 2010] [10:10:38] <sebersole> so we need to wait still for them to install the lastest
[Monday 14 June 2010] [10:11:02] <sebersole> (another remedy ticket, so wait accordingly)
[Monday 14 June 2010] [10:11:13] <stliu> they installed on a test hudson instance now
[Monday 14 June 2010] [10:11:24] <stliu> not the official hudson we're using
[Monday 14 June 2010] [10:11:32] <sebersole> oh, so there you have it
[Monday 14 June 2010] [10:11:41] <stliu> anyway, i will try that plugin when the new version be installed
[Monday 14 June 2010] [10:11:46] <sebersole> wait++ accordingly
[Monday 14 June 2010] [10:11:52] <stliu> yep :)
[Monday 14 June 2010] [10:11:56] <hardy> :)
[Monday 14 June 2010] [10:12:14] <gbadner> stliu, I fixed some more tests for HHH-5063 and HHH-5135; I resolved the issues
[Monday 14 June 2010] [10:12:15] <jbossbot> [HHH-5063] Projections that have aliases same as the property name result in invalid sql [Resolved, Major, Gail Badner] http://opensource.atlassian.com/projects/hibernate/browse/HHH-5063
[Monday 14 June 2010] [10:12:16] <sebersole> alright thats all the "old business" i had
[Monday 14 June 2010] [10:12:17] <jbossbot> [HHH-5135] "Ambiguous column" exception thrown with columns having the same name as a function registered with the dialect (e.g. to_date, floor) [Resolved, Major, Gail Badner] http://opensource.atlassian.com/projects/hibernate/browse/HHH-5135
[Monday 14 June 2010] [10:12:48] <sebersole> anyone else have old business?
[Monday 14 June 2010] [10:13:00] <stliu> gbadner, great thanks
[Monday 14 June 2010] [10:13:07] <gbadner> there are 2 things that seem to lock up postgresql:
[Monday 14 June 2010] [10:13:23] <hardy> fyi, I finished the first version of the collection mapping chapter update and committed the changes. I will need to go over it once more in the next couple of days though
[Monday 14 June 2010] [10:13:32] <sebersole> hardy: awesome!
[Monday 14 June 2010] [10:13:35] <epbernard> year!
[Monday 14 June 2010] [10:13:39] <sebersole> oh yeah
[Monday 14 June 2010] [10:13:56] <sebersole> epbernard finsihed the basic mapping chapter rewrite!
[Monday 14 June 2010] [10:14:28] <sebersole> so we have a good headway on these doc migrations
[Monday 14 June 2010] [10:14:37] <epbernard> Let's finish up the pure mapping conversions which we are well into
[Monday 14 June 2010] [10:14:49] <epbernard> then we can go for the configuration change
[Monday 14 June 2010] [10:15:00] <epbernard> and later the example migration over to annotations
[Monday 14 June 2010] [10:15:17] <sebersole> that was the plan because i can see synergies (sorry for using that word) during the merge
[Monday 14 June 2010] [10:15:40] <sebersole> so that Configuration and AnnotationConfiguration become one
[Monday 14 June 2010] [10:15:44] <epbernard> yep
[Monday 14 June 2010] [10:16:02] <sebersole> lord of the rings references aside
[Monday 14 June 2010] [10:16:19] <epbernard> We can stop at any phase (outlined) that ill be good on the doc side.
[Monday 14 June 2010] [10:16:26] <epbernard> ah got to go.
[Monday 14 June 2010] [10:16:57] <sebersole> then you lose gradle vote rights :)
[Monday 14 June 2010] [10:17:09] <epbernard> I'm voting with the majority
[Monday 14 June 2010] [10:17:11] <epbernard> :)
[Monday 14 June 2010] [10:17:13] <sebersole> haha
[Monday 14 June 2010] [10:17:24] <sebersole> l8r
[Monday 14 June 2010] [10:17:30] <epbernard> cheers
[Monday 14 June 2010] [10:17:34] Quit epbernard has left this server (Quit: epbernard).
[Monday 14 June 2010] [10:18:06] <sebersole> alright, so then new business
[Monday 14 June 2010] [10:18:13] <gbadner> please be aware that on postgresql:
[Monday 14 June 2010] [10:18:13] <gbadner> 1) any DB operations (even select) w/o a transaction will lock up the test
[Monday 14 June 2010] [10:18:13] <gbadner> 2) an unexpected exception will also lock up the test
[Monday 14 June 2010] [10:18:46] <gbadner> ^^^, I just spent a whole lot of time yesterday on this
[Monday 14 June 2010] [10:18:50] <sebersole> gbadner: most likely for (2) because we do not generally cleanup transations after exception
[Monday 14 June 2010] [10:19:04] <sebersole> which seems to feed into your (1)
[Monday 14 June 2010] [10:19:22] <sebersole> have you tried #postgresql ?
[Monday 14 June 2010] [10:19:44] <sebersole> though in my experience they really tend to be uninterested there in jdbc stuff
[Monday 14 June 2010] [10:19:45] <gbadner> no, but I found info by googling
[Monday 14 June 2010] [10:19:49] <sebersole> oh ok
[Monday 14 June 2010] [10:20:20] <hardy> there are some fishy tests in em and annotations and I think in Search we had similar cases
[Monday 14 June 2010] [10:20:44] <gbadner> is a transaction automatically started by postgresqll, if there isn't one already?
[Monday 14 June 2010] [10:20:47] <sebersole> well the testsuite generally do not handle exceptions
[Monday 14 June 2010] [10:20:57] <hardy> there are some jira issues related to this. Juca opened them on a case by case basis
[Monday 14 June 2010] [10:21:11] <sebersole> gbadner: it depends on autocommit (or should anyway)
[Monday 14 June 2010] [10:21:32] <gbadner> oh, ok
[Monday 14 June 2010] [10:21:37] <sebersole> oh, *open* a transation
[Monday 14 June 2010] [10:21:54] <sebersole> yes thats the general pattern for jdbc drivers
[Monday 14 June 2010] [10:22:20] <sebersole> when not in auto commit, a transaction is started on first DML
[Monday 14 June 2010] [10:22:34] <sebersole> jdbc defines no explicit "start transaction"
[Monday 14 June 2010] [10:23:11] <gbadner> can't remember for sure, but I think missing a session.close() can also lock up postgresql tests
[Monday 14 June 2010] [10:23:38] <sebersole> not unless its a non-primary session
[Monday 14 June 2010] [10:23:51] <sebersole> primary sessions for tests are closed automatically
[Monday 14 June 2010] [10:24:11] <sebersole> aka, sessions obtained from openSession() on the test class
[Monday 14 June 2010] [10:24:26] <sebersole> of course annotations, em, etc are different here
[Monday 14 June 2010] [10:24:33] <gbadner> ok, maybe it was an unclosed transaction then
[Monday 14 June 2010] [10:24:38] <gbadner> stliu, I tried to run the full test case on qa01 using postgresql, but got OOM
[Monday 14 June 2010] [10:25:10] <gbadner> meant the full test suite
[Monday 14 June 2010] [10:25:14] <stliu> gbadner, i guess you missed this
[Monday 14 June 2010] [10:25:14] <stliu> export MAVEN_OPTS=-Xmx1024m
[Monday 14 June 2010] [10:25:56] <gbadner> hmmm, I must have set that up on my machine at some point
[Monday 14 June 2010] [10:26:10] <sebersole> look in .profile
[Monday 14 June 2010] [10:26:22] <sebersole> thats where i set all this kind of stuff up
[Monday 14 June 2010] [10:26:28] <sebersole> ~/.profile
[Monday 14 June 2010] [10:26:38] <stliu> mine is ~/.bash_login on qa01
[Monday 14 June 2010] [10:26:41] <gbadner> ok, I'll do that and try again
[Monday 14 June 2010] [10:26:48] <sebersole> yeah depends on shell
[Monday 14 June 2010] [10:27:31] <sebersole> gbadner: i can work with you on this today if you need
[Monday 14 June 2010] [10:27:52] <sebersole> another testsuite issue i need to work through is for h2
[Monday 14 June 2010] [10:28:06] <sebersole> and this is why it takes so long for the testsuite to run there stliu
[Monday 14 June 2010] [10:28:16] <sebersole> having to do with temp table support
[Monday 14 June 2010] [10:28:37] <sebersole> which leads into another bit of old business i forgot earlier
[Monday 14 June 2010] [10:28:41] <stliu> btw, we do need take some time on the test failures on the other DBs except h2/hsqldb
[Monday 14 June 2010] [10:29:00] <sebersole> i assume noone still has looked at the temp table plan
[Monday 14 June 2010] [10:29:10] <stliu> i will try to get the failures summary for 3.5
[Monday 14 June 2010] [10:29:47] <sebersole> stliu: i will look at trunk after i get the h2/temp table stuff done
[Monday 14 June 2010] [10:29:57] <stliu> gbadner, you changed something on trunk for that postgresql lock issue?
[Monday 14 June 2010] [10:30:13] <gbadner> stliu, yeah, trunk and Branch_3_5
[Monday 14 June 2010] [10:30:21] <stliu> http://hudson.jboss.org/hudson/view/hibernate/job/hibernate-core-testsuite/
[Monday 14 June 2010] [10:30:29] <stliu> it works whatever you changed :D
[Monday 14 June 2010] [10:30:47] <stliu> now the test suite is not locked on postgresql
[Monday 14 June 2010] [10:31:00] <sebersole> yellow is test failures?
[Monday 14 June 2010] [10:31:09] <stliu> yep
[Monday 14 June 2010] [10:31:10] <gbadner> I checked it in while the tests were running
[Monday 14 June 2010] [10:31:21] <stliu> http://hudson.jboss.org/hudson/view/hibernate/job/hibernate-core-testsuite/database=postgresql82,jdk=java15_default,label=hibernate/30/
[Monday 14 June 2010] [10:31:26] <sebersole> would be nice if they had a legend on these screens
[Monday 14 June 2010] [10:32:06] <stliu> now we have 10 failures on postgresql :)
[Monday 14 June 2010] [10:32:35] <sebersole> so this is another task for this week:
[Monday 14 June 2010] [10:32:50] <sebersole> start going through the test failures from hudson and analyzing
[Monday 14 June 2010] [10:33:03] <sebersole> and divying out tasks
[Monday 14 June 2010] [10:33:21] <stliu> here is the last test result report for trunk
[Monday 14 June 2010] [10:33:23] <stliu> http://hudson.jboss.org/hudson/view/hibernate/job/hibernate-core-testsuite/30/testReport/
[Monday 14 June 2010] [10:33:42] <sebersole> yes i know :)
[Monday 14 June 2010] [10:34:15] <sebersole> thats cumulative?
[Monday 14 June 2010] [10:34:17] <sebersole> i take it
[Monday 14 June 2010] [10:34:21] <stliu> sebersole, i'd suggest we should start from 3.5 branch for going through these failures
[Monday 14 June 2010] [10:35:14] <stliu> you know, most failures are both in 3.5 and the trunk
[Monday 14 June 2010] [10:35:21] <sebersole> i was noot aware they did test number aggregations
[Monday 14 June 2010] [10:37:21] <sebersole> stliu: url to the 3.5 job?
[Monday 14 June 2010] [10:37:48] <sebersole> http://hudson.jboss.org/hudson/view/hibernate/job/hibernate-core-branch35-testsuite/
[Monday 14 June 2010] [10:37:48] <sebersole> ?
[Monday 14 June 2010] [10:37:59] <stliu> http://hudson.jboss.org/hudson/view/hibernate/job/hibernate-core-branch35-testsuite/
[Monday 14 June 2010] [10:38:19] <stliu> yes, but now there is some problem with this job which i don't know why
[Monday 14 June 2010] [10:38:26] <sebersole> ok was going to ask
[Monday 14 June 2010] [10:38:27] <sebersole> :)
[Monday 14 June 2010] [10:38:35] <sebersole> hard to analyze as is
[Monday 14 June 2010] [10:38:53] <hardy> i can check ann and em tests tomorrow.
[Monday 14 June 2010] [10:39:11] <stliu> i will fix it asap
[Monday 14 June 2010] [10:39:12] <sebersole> we should start on 3.5 for all of them
[Monday 14 June 2010] [10:39:26] <sebersole> most will carry over i have a feeling
[Monday 14 June 2010] [10:39:32] <sebersole> anyway
[Monday 14 June 2010] [10:39:47] <hardy> so I should check out the 3.5 branch?
[Monday 14 June 2010] [10:40:01] <stliu> yea, hardy that's better
[Monday 14 June 2010] [10:40:07] <sebersole> you mean `svn co` ?
[Monday 14 June 2010] [10:40:12] <hardy> if I find a problem do I have to apply the change on trunk as well?
[Monday 14 June 2010] [10:40:24] <sebersole> hardy: absolutely!
[Monday 14 June 2010] [10:40:56] <sebersole> you must apply it to both until we stop 3.5 maintenance
[Monday 14 June 2010] [10:41:05] <sebersole> after 3.6 goes Final
[Monday 14 June 2010] [10:41:10] <hardy> ok
[Monday 14 June 2010] [10:41:56] <hardy> btw, I have to leave soon again (as every second Monday). So if there is anything you need from me let me know
[Monday 14 June 2010] [10:42:10] <sebersole> stliu: btw did you rebase on that gradle jdocbook plugin this time before coding and committing?
[Monday 14 June 2010] [10:42:39] <stliu> oh, i forgot that again, sorry :(
[Monday 14 June 2010] [10:42:43] <sebersole> hardy: gradle... i just need to know if you see showstoppers there?
[Monday 14 June 2010] [10:42:48] <hardy> regarding gradle. +1 from my side.
[Monday 14 June 2010] [10:42:54] <hardy> no, no showstoppers
[Monday 14 June 2010] [10:42:59] <sebersole> hardy: cool
[Monday 14 June 2010] [10:43:11] <hardy> as said, I see how it overcomes maven limitations
[Monday 14 June 2010] [10:43:13] <sebersole> stliu: you need to, i did the fixing last time
[Monday 14 June 2010] [10:43:34] <stliu> yeah, i will do it next time, promise
[Monday 14 June 2010] [10:43:49] <sebersole> stliu: i mean you need to do it this time
[Monday 14 June 2010] [10:43:52] <hardy> it'sa little sad though that we could not stay with maven
[Monday 14 June 2010] [10:44:29] <sebersole> stliu: i will check first to see if it will be a clean apply
[Monday 14 June 2010] [10:44:30] <hardy> finally all projects are using maven and we had the mixed build tool approach behind us
[Monday 14 June 2010] [10:44:36] <sebersole> if not, you will need to fix it
[Monday 14 June 2010] [10:45:00] <sebersole> hardy: well i feel that too
[Monday 14 June 2010] [10:45:01] <stliu> sebersole, the change i made is very simple, you can just copy the diff to apply to your repo
[Monday 14 June 2010] [10:45:16] <sebersole> shoot i spent a lot of time, effort, blood, sweat and tears on maven
[Monday 14 June 2010] [10:45:30] <sebersole> stliu: its a matter of practice ;)
[Monday 14 June 2010] [10:45:45] <hardy> sebersole: same here. A shame that some major issues never got addressed from the maven side
[Monday 14 June 2010] [10:45:58] <sebersole> i'd aegue that a small change like this is better to drive that home
[Monday 14 June 2010] [10:45:59] <stliu> yeah
[Monday 14 June 2010] [10:46:08] <stliu> btw, guys, there is a great git doc, http://progit.org/
[Monday 14 June 2010] [10:46:14] <sebersole> rather than waiting till you have a 3000 line commit
[Monday 14 June 2010] [10:46:21] <sebersole> yep
[Monday 14 June 2010] [10:46:22] <hardy> i guess at some stage we have to open the discussion on whether we want Search and Validator migrated as well
[Monday 14 June 2010] [10:46:52] <sebersole> hardy: i'll leave that to y'all
[Monday 14 June 2010] [10:47:05] <sebersole> there is really no need if you dont feel the same pains there
[Monday 14 June 2010] [10:47:08] <hardy> Search would be very easy and it would be beneficial since there we also want to release the testing jar
[Monday 14 June 2010] [10:47:12] <hardy> very similar problem
[Monday 14 June 2010] [10:48:13] <hardy> Validator might be harder since the setup is a little bit special for the module which runs the TCK. This was synchronized with Weld. So there we might have to wait/investiagte
[Monday 14 June 2010] [10:49:04] <sebersole> hardy: like i said, there is no push imo if the tool currently in use works
[Monday 14 June 2010] [10:49:27] <gbadner> sebersole, hibernate-envers 3.5.2-Final didn't make it onto r.j.o/maven2/...
[Monday 14 June 2010] [10:49:27] <gbadner> any idea why? I can see that it was built and deployed to my local repo
[Monday 14 June 2010] [10:49:39] <sebersole> regardless of what *other* people want to say, maven is falling on its face in the hibernate build
[Monday 14 June 2010] [10:49:50] <sebersole> they dont use it here day to day (if at all)
[Monday 14 June 2010] [10:50:01] <sebersole> so i take their opinions with a grain of salt in that regard
[Monday 14 June 2010] [10:50:44] <sebersole> gbadner: no idea, you are the first to have done this lovely nexus deploy stuff
[Monday 14 June 2010] [10:51:02] <hans__> hardy: A high priority feature on our road map is a deep import of Maven builds. That way you can have an integrated build and yet don't need to migrate everything.
[Monday 14 June 2010] [10:51:04] <gbadner> ok, I'll look into it; very strange
[Monday 14 June 2010] [10:52:06] <hardy> hans__: thanks for the info
[Monday 14 June 2010] [10:52:43] <hardy> btw, with gradle we still have to keep the pom.xml for the project description part, right?
[Monday 14 June 2010] [10:53:04] <hardy> or is there anything similar in gradle. asking more out of interest.
[Monday 14 June 2010] [10:53:19] <hans__> hardy: We have something similar.
[Monday 14 June 2010] [10:53:20] <stliu> nope
[Monday 14 June 2010] [10:53:21] <sebersole> hardy: no
[Monday 14 June 2010] [10:53:31] <sebersole> gradle generates the pom during "deploy"
[Monday 14 June 2010] [10:53:47] <sebersole> this is a critical piece
[Monday 14 June 2010] [10:54:27] <gbadner> stliu, any idea how hard it would be to also summarize test failures like http://pastebin.com/wxV5eCDh
[Monday 14 June 2010] [10:55:02] <gbadner> to make it easier to see all the configs where each test fails?
[Monday 14 June 2010] [10:55:08] <sebersole> so i have nothing else at the moment
[Monday 14 June 2010] [10:55:08] <hardy> sebersole: were is this part in the gradle branch?
[Monday 14 June 2010] [10:55:17] <hardy> I was looking for it
[Monday 14 June 2010] [10:55:19] <sebersole> hardy what part?
[Monday 14 June 2010] [10:55:28] <hardy> the "project description part"
[Monday 14 June 2010] [10:55:30] <sebersole> its just automatically done
[Monday 14 June 2010] [10:55:36] <hans__> sebersole: You know about the 0.9 features to fully customize the pom and generate it to wherever you want?
[Monday 14 June 2010] [10:55:46] <sebersole> yeah, i did hans
[Monday 14 June 2010] [10:55:51] <stliu> gbadner, seems hudson does not support that way
[Monday 14 June 2010] [10:55:56] <sebersole> in fact i think i do that in one
[Monday 14 June 2010] [10:56:01] <hardy> but you still have to for example list the jira instance somewhere
[Monday 14 June 2010] [10:56:41] <hardy> hans__: btw, nice gradle docs. finally got some time to go through it
[Monday 14 June 2010] [10:56:57] <hans__> hardy: You would do that in the customizing section for the pom.
[Monday 14 June 2010] [10:57:05] <sebersole> hardy: yes that kind of info we need to supply
[Monday 14 June 2010] [10:57:14] <hardy> ok
[Monday 14 June 2010] [10:57:28] <hardy> anyways, gtg.
[Monday 14 June 2010] [10:57:35] <hardy> talk to you tomorrow
[Monday 14 June 2010] [10:57:42] <sebersole> k, hardy later
[Monday 14 June 2010] [10:57:50] Quit hardy has left this server (Quit: hardy).
[Monday 14 June 2010] [10:59:18] <sebersole> stliu: a set of related tasks is to get the jdocbook stuff published this week
[Monday 14 June 2010] [10:59:35] <sebersole> core as well as the maven and gradle bits
[Monday 14 June 2010] [10:59:49] <stliu> okay
[Monday 14 June 2010] [10:59:58] <stliu> that's cool
[Monday 14 June 2010] [10:59:59] <sebersole> i can work on that though
[Monday 14 June 2010] [11:00:32] <sebersole> after that we can start "merging" gradle into trunk
[Monday 14 June 2010] [11:01:05] <sebersole> hibernate trunk that is
[Monday 14 June 2010] [11:01:22] <sebersole> the last piece there is testing and testsuite
[Monday 14 June 2010] [11:01:30] <stliu> yep
[Monday 14 June 2010] [11:02:10] <sebersole> hans__ , lightguard_jp : ok to bug you this week in ragrds to that stuff?
[Monday 14 June 2010] [11:02:57] <hans__> sebersole: You are welcome :)
[Monday 14 June 2010] [11:03:00] <sebersole> specifically we need to scope out what we want done here
[Monday 14 June 2010] [11:03:07] <sebersole> hans__: thanks! :)
[Monday 14 June 2010] [11:03:12] <hans__> sebersole: And we are almost in the same timezone (I'm in Denver until Saturday).
[Monday 14 June 2010] [11:03:46] <sebersole> oh! nice! (you know you can almost get from denver to austin on the same road! ;)
[Monday 14 June 2010] [11:04:02] <sebersole> of course its like a 20 hour drive :D
[Monday 14 June 2010] [11:04:03] <hans__> sebersole: I would actually love to do that if I had more time :)
[Monday 14 June 2010] [11:04:35] <sebersole> so the test reqs I see:
[Monday 14 June 2010] [11:05:11] <sebersole> (1) one module -> 2 artifacts (and src/test/java)
[Monday 14 June 2010] [11:05:55] <hans__> sebersole: We could schedule a session to get this started?
[Monday 14 June 2010] [11:06:16] <sebersole> (2) easy config for developer locally (similar to old local/properties but accounting for jdbc deps through gradle/ivy)
[Monday 14 June 2010] [11:06:26] <sebersole> hans__: sure
[Monday 14 June 2010] [11:06:38] <sebersole> and i got around to installing skype again
[Monday 14 June 2010] [11:06:45] <hans__> sebersole: (2) Right, we discussed that a little bit already.
[Monday 14 June 2010] [11:06:56] <hans__> sebersole: Excellent :)
[Monday 14 June 2010] [11:07:07] <hans__> sebersole: Although phone is fine for me as well. I can use Skype out.
[Monday 14 June 2010] [11:07:29] <sebersole> (3) the "aggregated" test execution for hudson (looping)
[Monday 14 June 2010] [11:07:48] <hans__> sebersole: What is that about?
[Monday 14 June 2010] [11:07:59] <sebersole> the stuff we discussed last time
[Monday 14 June 2010] [11:08:07] <sebersole> at this meeting last week
[Monday 14 June 2010] [11:08:34] <hans__> sebersole: I remember. The multiple db stuff, right?
[Monday 14 June 2010] [11:08:46] <sebersole> about giving hudson job the ability to checkout and run a given task (i assume) that loops over all the databases
[Monday 14 June 2010] [11:08:49] <sebersole> yes
[Monday 14 June 2010] [11:08:55] <stliu> run testsuite parallel?
[Monday 14 June 2010] [11:09:12] <sebersole> well parallel is fine for the each db
[Monday 14 June 2010] [11:09:21] <sebersole> meaning thread per db
[Monday 14 June 2010] [11:09:48] <sebersole> but i dont think we will be able to do threads within tests against a given db
[Monday 14 June 2010] [11:10:11] <hans__> sebersole: Except possibly for h2?
[Monday 14 June 2010] [11:10:32] <sebersole> ideally
[Monday 14 June 2010] [11:10:56] <sebersole> ideally if h2 (the default) works we still want to publish the nightly snapshot
[Monday 14 June 2010] [11:11:11] <sebersole> regardless of what happens in the other test runs
[Monday 14 June 2010] [11:11:25] <hans__> sebersole: right. That should be easy to express.
[Monday 14 June 2010] [11:11:29] <sebersole> cool
[Monday 14 June 2010] [11:12:11] <sebersole> not sure how to express (4) concisely
[Monday 14 June 2010] [11:12:23] <hans__> Hudson could use an init script.
[Monday 14 June 2010] [11:12:46] <hans__> tasks.withType(Test).each { task -> task.ignoreFailures = false }
[Monday 14 June 2010] [11:13:06] <hans__> I mean = true
[Monday 14 June 2010] [11:13:09] <hans__> testH2.ignoreFailures = false
[Monday 14 June 2010] [11:13:18] <sebersole> ah
[Monday 14 June 2010] [11:14:37] <hans__> We could schedule something for today or tomorrow?
[Monday 14 June 2010] [11:14:55] <sebersole> hans__: sounds good