15:05:04 <sebersole> #startmeeting
15:05:12 <jbott> Meeting started Mon May  2 15:05:04 2011 UTC.  The chair is sebersole. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:05:12 <jbott> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:05:57 <sebersole> so the main thing we need to discuss today is the new metamodel and whether we will drive forard with that
15:06:36 <sebersole> tbh i am not feeling confident we will make our dates
15:06:43 <sebersole> for sure not as it is going today
15:07:22 <gbadner> no, I don't think we can
15:07:36 <sebersole> trust me, i understand the new code is "better"; its the reason i have been working on it for nearly 2 years in spare time
15:07:55 <sebersole> but the dates are the dates at this point
15:09:20 <sebersole> hardy: ?
15:09:25 <hardy> i already mentioned my concerns regarding the deadlines as well
15:10:19 <sebersole> so do we just scrap it and move forward with the "revert"?
15:11:04 <hardy> what is your revert?
15:11:08 <sebersole> stliu: you have been very productive here, what are your thoughts?
15:11:41 <stliu> i'm looking gail's task list
15:11:59 <sebersole> well i envision revert as keeping the old mapping stuff
15:12:18 <sebersole> as i said last week, ideally incorporating the new domain model
15:12:33 <sebersole> sorry
15:12:38 <sebersole> the new relational model
15:13:02 <hardy> if we keep the old mapping stuff (not even taking the relational stuff) what do we have to do then?
15:13:18 <sebersole> we have to change the binding code
15:13:22 <sebersole> but
15:13:31 <sebersole> the "build persister code" is the same
15:13:47 <sebersole> well more-or-less, depending on relational model
15:14:03 <gbadner> then there wouldn't need to be an integration step?
15:14:19 <sebersole> what is "integration step"?
15:14:53 <gbadner> the mapping code is already operational
15:15:04 <sebersole> no its not
15:15:26 <sebersole> you mean the new binding/modeol code?
15:15:58 <sebersole> my point was simply that the code to build persisters from o.h.mapping objects is already in place
15:16:08 <gbadner> yes, that's what I meant
15:16:27 <hardy> sebersole: so which binding code has to change?
15:16:46 <sebersole> AnnotationBinder and HbmBinder
15:16:50 <gbadner> sebersole, I like the idea of tweaking the mapping code
15:16:55 <sebersole> gbadner: i hate it
15:16:59 <sebersole> but its what we have
15:17:10 <hardy> i hate it as well
15:17:20 <gbadner> sebersole, at least that way, we can see if it's working as we go
15:17:30 <sebersole> see if what is working?
15:17:43 <hardy> AnnotationBinder is a very fuggly piece of code to change
15:17:47 <sebersole> anyway
15:17:51 <stliu> hardy, what if we could get everyone working on this, like emmanuel, sanne?
15:18:00 <sebersole> stliu: not an option really
15:18:05 <gbadner> we can run the full unit test suite and see if our changes break anything
15:18:20 <sebersole> sanne has to get search ready for 4.0
15:18:26 <gbadner> right now, w/ the new code, there's no way to do that
15:18:48 <hardy> it would be nice to get more on board, but there is also the question on how much we can split the tasks
15:18:57 <hardy> AND there is of course Search
15:19:05 <stliu> search 4.0 has been released, hasn't it?
15:19:17 <hardy> I pretty much are off the list for Search 4.0, so it is Sanne
15:19:22 <hardy> 3.4 is released
15:19:30 <stliu> oh, yeah, sorry
15:19:38 <sebersole> gbadner: consuing mapping and consuming metamodel are still completely different / parallel tasks
15:19:47 <hardy> 4.0 is the next release where we similar to Core planning to change several APIs
15:19:51 <sebersole> so how does that help test?
15:20:27 <hardy> btw, emmanuel is the next few weeks on vacation and conferences - I don't think he can help muhc
15:20:44 <sebersole> stliu: how much time can you give to the effort over the next 5 weeks?
15:21:06 <stliu> 3-4 weeks i'd think
15:21:20 <gbadner> what I'm saying is that if we tweak the old code, we can see if it causes unit test failures
15:21:37 <stliu> there is one eap release i have done before May 11
15:22:25 <sebersole> jpav: how far along are you on the tasks hardy gave you ?
15:22:42 <jpav> I was close to done until the JIRA was updated today
15:22:55 <sebersole> no idea what that means
15:23:00 <jpav> But I still had several things I needed to run by hardy
15:23:19 <jpav> hardy added several additional global configuration components that need to be accounted for
15:23:26 <hardy> right
15:23:48 <sebersole> see this is my concern.  all i see is a lot of uncertainty :)
15:23:53 <hardy> how do you take care of them? I mean did the new metamodel cater for them already?
15:23:57 <stliu> HHH-6132?
15:24:02 <jbossbot> jira [HHH-6132] Process and bind global configuration annotations [Open (Unresolved) Sub-task, Major, John Verhaeg] http://opensource.atlassian.com/projects/hibernate/browse/HHH-6132
15:24:15 <sebersole> not much we can "hang our hat on"
15:24:37 <gbadner> sebersole, I agree
15:25:23 <sebersole> gbadner: did you finish up that jira i asked you to look at?
15:25:51 <gbadner> oy, which one?
15:26:09 <hardy> stliu: yes, that's the issue
15:26:47 <sebersole> personally i think we ought to power through this
15:26:56 <sebersole> but i cannot do it on my own
15:27:24 <gbadner> you'd hardly be doing it on your own...
15:27:28 <sebersole> its the last chance we will get to MakeThisRight for quite some time
15:28:17 <sebersole> well i think for sure we need to reorg the worjk quite a bit
15:28:57 <sebersole> if stliu is on board for a good chunk of work i feel better
15:29:29 <stliu> i will :)
15:29:33 <sebersole> sweet!
15:29:38 <hardy> right, maybe stliu can continue helping me on the annotation front ;-)
15:29:44 <sebersole> negative
15:29:48 <sebersole> ;)
15:29:52 <hardy> ahhhhhhhhhhh
15:29:57 <sebersole> nah, actually thats probably best
15:30:17 <sebersole> smarlow_: have we figured out the best way to pass on the jandex stuff yet?
15:30:29 <sebersole> when i looked it did not seem possible tbh
15:30:36 <stliu> well, if gail want to mock hbm elements to annotation, i'd like continue working on that
15:30:51 <sebersole> stliu: thats not possible
15:30:57 <hardy> if this is the path we want to go
15:31:03 <hardy> it makes things more complicated
15:31:41 <hardy> even though it sounds nice in theory I think it would be faster to continue to have jpa and hbm separately
15:31:58 <sebersole> gbadner: HHH-2049
15:32:02 <jbossbot> jira [HHH-2049] LEFT OUTER JOIN subcriteria filters children [In Progress (Unresolved) Bug, Major, Gail Badner] http://opensource.atlassian.com/projects/hibernate/browse/HHH-2049
15:32:06 <smarlow_> sebersole:  I'm currently, sort of passing it to myself, for the annotation scanner that I pass via "hibernate.ejb.resource_scanner"
15:32:11 <hardy> if time wouldn't be a factor I would definitely give it a go, but in the current situation ...
15:32:28 <sebersole> hardy: these kinds of things are impl/internal details
15:32:33 <sebersole> we can easily try later
15:32:39 <hardy> exactly
15:32:43 <sebersole> as it does not effect the apis
15:32:48 <gbadner> sebersole, ok, thanks
15:32:52 <hardy> +1
15:33:05 <sebersole> we really need to be focused here on stuff that affects apis
15:33:10 <sebersole> of rht next 5 weeks that is
15:33:27 <sebersole> smarlow_: cool
15:33:50 <sebersole> so from there it really comes down to the last big question i see for 4.0
15:34:06 <sebersole> how to manage EJB3Configuration/building-EMF
15:34:41 <sebersole> do we want to support users directly building EMF?
15:34:45 <sebersole> or
15:34:54 <sebersole> do we only support them using the JPA defined mechnisms
15:35:18 <sebersole> hardy: thoguhts?
15:35:45 <sebersole> i have not really been involved in that enough to make an educated decision
15:35:49 <hardy> not many. I don't have the big picture here
15:35:54 <hardy> same here
15:36:10 <hardy> just JPA mechanisms sounds good at the first thought
15:36:22 <sebersole> emmanuel indicated that JPA bootrapping was the preferred mechnism
15:36:33 <sebersole> so maybe we can work on that assumption for the time being
15:36:40 <hardy> definitely the preferred way
15:36:44 <hardy> +1
15:37:07 <sebersole> alright, we definitely need that in place for alpha3
15:37:11 <hardy> not sure what the usecases are for the direct building of the EMF
15:37:16 <hardy> i think emmanuel is the one to ask here
15:37:33 <sebersole> well it would corelate to whenever EJB3Configuration is used today
15:37:43 <sebersole> used directly i mean
15:38:05 <hardy> well, the only usage I am aware of is for testing purposes
15:39:06 <hardy> what I mean is that our EM tests are using the EJB3Configuration under the hood. at least I think so
15:40:08 <sebersole> hardy: yeah i think so
15:40:14 <hardy> there is for example a BaseEntityManagerFunctionalTestCase which uses it
15:40:28 <hardy> and a whole bunch of tests derive of it
15:41:08 <hardy> that does not say much of course about the usage of EJB3Configuration "in the wild"
15:41:51 <hardy> but the sensible thing to do has always been to use the persistence API
15:41:55 <sebersole> well its not worse than Configuration itself :)
15:42:33 <sebersole> anyway, i cannot find the email discussion anymore about building emf
15:43:50 <sebersole> between emails, and wikis and gists getting hard to track info
15:45:06 <sebersole> so how much metamodel work is not yet pushed?
15:45:20 <sebersole> jpav's "global" stuff i guess
15:45:31 <sebersole> hardy: ?
15:45:35 <sebersole> gbadner: ?
15:45:39 <jpav> afaik
15:45:44 <stliu> mine is not pushed yet
15:45:48 <hardy> i have some work from today
15:45:59 <sebersole> stliu: all of it?
15:46:03 <hardy> I tend to pull/push pretty miuch daily
15:46:07 <stliu> maybe Wed or Thu
15:46:18 <sebersole> stliu: thats all of orm.xml ?
15:46:22 <gbadner> sebersole, my stuff is checked in
15:46:30 <sebersole> meaning none of it is pushed?
15:46:40 <stliu> sebersole, yes, and i have to ask jason to get jandex-2 resolved first
15:46:45 <sebersole> ok
15:46:51 <sebersole> just checking
15:47:06 <sebersole> can you make it in time for release?
15:47:14 <sebersole> alpha3
15:47:24 <sebersole> oh, i guess not
15:48:16 <stliu> me?
15:48:28 <hardy> for stliu's work we really need the jandex issue resolved
15:48:41 <stliu> i think i can, alpha 3 is May 11, right?
15:48:50 <stliu> i just send a mail to jason
15:48:55 <hardy> nope, this wed
15:48:59 <sebersole> stliu: no, alpha3 is in 2 days
15:49:08 <hardy> jason is probably at JBossWorld.
15:49:19 <hardy> then again, it is such a small change
15:49:36 <hardy> stliu: have you created a push request for Jason
15:49:44 <stliu> well, our jira shows alpha 3 is May 11 ..
15:49:59 <hardy> maybe we gained a week ;-) !?
15:50:04 <stliu> hardy, no, it is a very simple fix
15:50:13 <hardy> i know, one method
15:50:18 <stliu> http://opensource.atlassian.com/projects/hibernate/browse/HHH#selectedTab=com.atlassian.jira.plugin.system.project%3Aroadmap-panel
15:50:30 <stliu> and beta 1 is Jun 11
15:51:05 <sebersole> 11 is the year ;)
15:51:48 <sebersole> jira shows 4/May/11
15:51:52 <hardy> he he
15:51:56 <stliu> damn ...., you know, in china, we use yyyy-mm-dd
15:52:01 <sebersole> yeah i do too
15:52:07 <sebersole> less confusing
15:52:12 <hardy> and the swedes use mm-dd-yyyy - the worst of all!
15:52:24 <sebersole> thats what most of US uses
15:52:33 <hardy> i have missed more than one appointment because of it
15:52:38 <gbadner> yup, terrible for sorting
15:52:42 <hardy> makes no sense to me
15:53:31 <sebersole> so for alpha3... i'll work on finishing up the api work and looking over the metamodel code
15:53:49 <hardy> ok
15:53:53 <sebersole> after alpha3 i'll start directing the metamodel work a bit more and lending a hand
15:53:59 <hardy> so we keep battling on
15:54:03 <stliu> so, my work probably can't go into alpha 3
15:54:12 <gbadner> sebersole, that would be great
15:54:17 <sebersole> hardy: right keep on keeping on for now
15:54:21 <hardy> ok
15:54:28 <sebersole> gbadner: well only way its gonna get done
15:54:44 <gbadner> you know the most about it
15:54:52 <jpav> hardy: Are you going to be online for a few more hours today?
15:55:06 <sebersole> most about code i did not write or conceptualize?
15:55:11 <sebersole> ok
15:55:16 <gbadner> yup
15:55:28 <hardy> I have a meeting now, but can check in later again
15:55:32 <sebersole> k, hardy
15:55:40 <jpav> k
15:56:30 <stliu> jpav, fyi, i will mock all jpa global configurations to index
15:56:50 <stliu> depends which entity be found first
15:57:14 <jpav> The code I've been working on is starting with the indexes
15:57:40 <stliu> that's good
15:57:58 <stliu> gbadner, btw, we need HHH-2049 resolved asap
15:58:03 <jbossbot> jira [HHH-2049] LEFT OUTER JOIN subcriteria filters children [In Progress (Unresolved) Bug, Major, Gail Badner] http://opensource.atlassian.com/projects/hibernate/browse/HHH-2049
15:58:28 <sebersole> smarlow_: btw, not sure if you saw but i pushed those changes for jboss jta platform
16:01:35 <smarlow_> sebersole:  yup, I synced up friday with the changes and pushed updates to the as7 hibernate40_2 branch :)
16:01:52 <sebersole> #endmeeting