14:11:01 <gsmet> #startmeeting
14:11:01 <jbott> Meeting started Tue Feb 28 14:11:01 2017 UTC.  The chair is gsmet. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:11:01 <jbott> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:11:03 <jbossbot> Title: MeetBot - Debian Wiki
14:11:04 <gsmet> cough
14:11:07 <DavideD> :)
14:11:15 <gsmet> I'll put the start in the email
14:11:22 <gmorling> thx
14:11:29 <gmorling> weird you can set a topic without starting before
14:11:53 <DavideD> I'll repeat it, after the release I need to to start on the new contrib repository for the dialects
14:12:25 <gsmet> ok, nice
14:12:37 <gsmet> DavideD: ok for you?
14:12:43 <DavideD> yes, I think that's all
14:12:46 <gsmet> #topic Progress Emmanuel
14:12:50 <gsmet> emmanuel: ?
14:12:59 <emmanuel> I’ve been back to some coding tasks
14:13:16 <emmanuel> First making ORM work nicely in environments where the DB might not be present at boot time
14:13:59 <gmorling> will ogm benefit from that, too?
14:14:10 <gmorling> i remembe we discussed it long time ago in the context of mongo
14:14:11 <emmanuel> I’m happy with my first run, I’ve seen a few of you gave feedback on the PR but I’ve moved to another topic. Hopefully will be able to close and get some time on it in the next few weeks
14:14:25 <emmanuel> gmorling: no, most of the work is based on ConnectionProvider
14:14:30 <emmanuel> and we don’t use that
14:14:36 <gmorling> i see
14:14:57 <gmorling> would be nice, was on my wish list or long time (more for ogm than for orm fwiw)
14:15:11 <sannegrinovero_> we should be able to adopt the same concept & same config properties.
14:15:16 <emmanuel> we don’t have a common access pattern for the connection
14:15:23 <emmanuel> maybe except Dialect
14:15:34 <emmanuel> we could try and do what I did for ConnetionProvider to Dialect
14:15:40 <emmanuel> yet another delegator
14:15:54 <emmanuel> At least the JVM will give up any hope of inlining at this stage :)
14:16:36 <emmanuel> My second big topic is a presentation on BV 2.0 that I need to have done for mid march
14:16:52 <emmanuel> We had the idea of starting from the Hiking demo and add examples here with Gunnar
14:17:06 <emmanuel> So I’ve spent some time dusting it off
14:17:23 <emmanuel> But it’s slow and I am getting really concerned that I iwll lack the time
14:17:44 <emmanuel> Without help, I’ll give up I think and do a slide only presentation
14:17:56 <emmanuel> at least that’s my feeling atm
14:18:11 <gmorling> i see
14:18:29 <gmorling> i guess i can help out somehow, I’ll need it for my talk, too
14:18:39 <sannegrinovero_> you need help to finish the "dust off" or beyond that?
14:19:00 <sannegrinovero_> I might be able to spend 1h to update the dependencies.
14:19:16 <emmanuel> I’m at 6hrs of clock time and the demo is likely missing some complex examples, has not be migrated to WF 10 and the UI has no concept of error report
14:19:48 <emmanuel> that’s it for what I ahve done
14:19:55 <gsmet> ok
14:20:03 <gsmet> #topic Next 2 weeks Emmanuel
14:20:57 <emmanuel> I have 2 presentations to prepare for some customer meetings which will eat some of my time. There are also some energy spend on the other projects liek daikon
14:21:19 <emmanuel> but I’ll give 1 or 2 more days on the demo and fall back to pure slides if that is failing.
14:21:29 <gsmet> OK
14:21:35 <emmanuel> with that BV pres + the ES pres I’ll be easily busy for the next 2 weeks
14:21:44 <gmorling> sounds good, so let’s discuss on thursday then how things stnd
14:22:02 <gsmet> thanks!
14:22:08 <gsmet> #topic Progress Guillaume
14:22:29 <gsmet> so I revived an old patch of mine (from 2011-2013) for search
14:22:45 <gsmet> using the SimpleQueryParser to provide a nice FTS feature
14:22:54 <gsmet> I created a PR, feedback welcome
14:23:03 <yrodiere> gsmet: done
14:23:12 <gsmet> (note: it might give us some hints on how to clean up the Lucene related API too)
14:23:18 <gsmet> yrodiere: ok, thanks
14:23:29 <gsmet> I then switched to HV
14:23:44 <gsmet> working on introducing the ContainerElementNode and other things
14:24:05 <gsmet> sent an email to the list about that but we didn't get that many answers
14:24:27 <gsmet> I'm currently working on an issue with message interpolation that I would like to have fixed for 5.4.1
14:24:41 <gsmet> that's pretty much it for the last 2 weeks
14:24:53 <gsmet> #topic Next 2 weeks Guillaume
14:25:10 <gsmet> so finish the work started on ContainerElementNode following the discussion with Gunnar
14:25:23 <gsmet> fix this bug and find a way to backport it to 5.4
14:25:32 <gsmet> and I would like to start the work on the TCK
14:25:42 <gmorling> yes, that’d be very good
14:25:44 <gsmet> the state of the tests is not very good on the value extraction front
14:25:58 <gsmet> there are the ones inherited from the old impl, the new ones...
14:26:10 <gsmet> I would like to clean it all up and move them to the TCK
14:26:22 <gmorling> a good first step would be to take a look at the spec diff and review whether there are any “tck-testable” annotations missing
14:26:23 <gsmet> (btw, I also moved the TCK doc to asciidoctor)
14:26:43 <gsmet> yup, I plan to do that and then articulate the tests with the spec
14:26:52 <gmorling> cool
14:26:59 <gsmet> that's all from me
14:27:09 <gsmet> #topic Progress Gunnar
14:27:37 <gmorling> so i’ve been working on several smaller things for BV (clarifications and such)
14:27:54 <gmorling> the most relevant one is to spec (and implement) how value extractors are found
14:28:31 <gmorling> we now employ the service loader scheme, allowing providers of alternative collection types (e.g. guava or other jvm langs) to add a value extractor for those types, too, and it will automatically be used
14:28:55 <gmorling> i’m in touch with stephane from the ceylon team, eager to hear back how it works out for them
14:29:04 <yrodiere> gmorling: do they need to depend on the bean validation API?
14:29:10 <gmorling> yes
14:29:20 <gmorling> they’d implement a type from the BV API
14:29:54 <yrodiere> ok. I guess they can use provided dependencies... But I'm quite sure they don't want to add a transitive dependency to their packages
14:30:11 <yrodiere> *to bean validation in their packages
14:30:13 <gmorling> yes, it’d be provided/optional or something like that
14:30:34 <gmorling> but then, if bv is around and they provide such extractor, it would automatically be used
14:30:47 <yrodiere> Sounds great :)
14:30:55 <gmorling> other than that, i’ve began to update the API for programmatic constraint definition in HV to support the new container element constraints, too
14:31:11 <gmorling> i’m doing that primarily to understand what’s needed to do it in XML, too
14:31:14 <gmorling> (which is part of the spec)
14:31:41 <gmorling> there is a nasty api riddle, which i’ll share with you later
14:31:47 <gmorling> maybe one of you has a great idea
14:31:55 <gmorling> details on that separately :)
14:32:10 <gmorling> which brings me to my next steps, @whipmaster
14:32:18 <gsmet> #topic Next 2 weeks Gunnar
14:32:36 <gmorling> so i’ll finish up the prog api work hopefully tomorrow
14:32:49 <gmorling> and then i’ll also have to switch to some preso work
14:33:15 <gmorling> there is an “adopt-a-jsr” hangout tomorrow with heather from the jcp, i’ll have 15 min to tell about BV and where people can help us
14:33:19 <gmorling> so i need to prepare that
14:33:50 <gsmet> feel free to ping me if you need proofreading
14:33:57 <gmorling> and then i’ll need to work on my talk on BV for javaland (end of march); also update/condense my HS+ES talk as slots are there only 40 min and so far it lasted 60 min
14:34:04 <gmorling> yepp, thanks, will definitely do
14:34:42 <gmorling> that’s pretty much it; just some nice news: the multi-release jar post on in.relation.to got a second life, it’s on top of reddit/java right now: https://www.reddit.com/r/java/
14:34:42 <jbossbot> Status 429
14:35:09 <gmorling> someone had shared it again on twitter and there :)
14:35:25 <gmorling> that’s it unless someone has questions for me
14:36:03 <gsmet> gmorling: no question, you have some disqus comment to answer on your blog post ;)
14:36:17 <gmorling> ah, cool, didn’t see that; thanks :)
14:36:20 <gsmet> ok, let's move to Sanne
14:36:25 <gsmet> #topic Progress Sanne
14:37:04 <gsmet> sannegrinovero_: ?
14:37:42 <yrodiere> London seems like a great place to use internet :D
14:38:02 <gmorling> lol
14:38:04 <sannegrinovero> [I'm back..]
14:38:16 <gsmet> sannegrinovero_: it's your turn!
14:38:25 <sannegrinovero> ah, ok
14:38:32 <sannegrinovero> I spent most of last trying to figure if Search 5.7 was addressing the most important needs from Infinispan;
14:38:38 <sannegrinovero> this tuned out to be complex..
14:38:47 <sannegrinovero> as their testsuite is - even normally - very slow, but was now particularly unstable.
14:38:56 <sannegrinovero> Yoann and me had finished our JIRAs, so we decided to release it w/o the Infinispan team to try these out.
14:39:05 <sannegrinovero> In case of need, we'll do some maintenance releases..
14:39:23 <sannegrinovero> My next two weeks fill be a mix of:
14:39:30 <sannegrinovero> s/fill/will/
14:39:36 <sannegrinovero> - Search PR reviews, Yoann is full steam on ES5
14:39:48 <gsmet> #topic Next 2 weeks Sanne
14:39:50 <sannegrinovero> - trying to get Steve to work on Java9 :-/
14:39:58 <sannegrinovero> - take the dust off from free-form refactoring I had started in the Search backends
14:40:04 <sannegrinovero> - Work on an HACEP demo for Red Hat Summit
14:40:14 <sannegrinovero> - preparing an OGM / Hot Rod short demo for a talk here in London
14:40:18 <sannegrinovero> (in 3 weeks)
14:41:03 <gsmet> ok, looks good (and busy)
14:41:06 <gsmet> :)
14:41:11 <gsmet> sannegrinovero: anything else?
14:41:25 <yrodiere> sannegrinovero: about the free-form, did you notice the thread on the mailing list?
14:41:56 <yrodiere> (emmanuel started it)
14:42:03 <emmanuel> yes there were big questions on the strategy not adressed
14:42:05 <DavideD> sannegrinovero, when is it the talk?
14:42:09 <DavideD> in London
14:43:00 <yrodiere> I think he brexited again
14:43:31 <gmorling> gee
14:44:39 <gsmet> hmmm, let's switch to Yoann and then get back to Sanne at the end
14:44:44 <gsmet> ah
14:45:08 <gsmet> #topic Progress Yoann
14:45:08 <yrodiere> Hmm ok
14:45:25 <yrodiere> So, we wrapped up the last remaining JIRAs on 5.7 with Sanne
14:45:39 <yrodiere> Mostly bugfixes, plus authentication
14:45:46 <yrodiere> (ES authentication)
14:46:21 <yrodiere> After Sanne released 5.7, I worked on getting Mincong's JSR-352 work back to our codebase
14:46:42 <yrodiere> It's currently living in a branch on my copy of the repository, for reasons I explained on the mailing list
14:47:05 <gmorling> thanks a lot for taking up this one!
14:47:36 <yrodiere> gmorling: well, np, but wait for the results, maybe you won't like it :p
14:47:50 <gmorling> we’ll see :)
14:48:06 <sannegrinovero> I'm back??
14:48:19 <gsmet> sannegrinovero: we have questions for you, we switched to Yoann but we'll get back at you
14:48:21 <yrodiere> Then I worked on making Hibernate Search able to target multiple versions of ES. This involved in particular adding abstraction layers over ES requests and their creation
14:48:28 <sannegrinovero> sorry all, router seems to reboot constantly
14:49:06 <sannegrinovero> ok. go on yrodiere :)
14:49:07 <yrodiere> I also polished my branch that targeted ES5 exclusively
14:49:13 <yrodiere> (not ES2)
14:49:51 <yrodiere> and finally I worked on getting a branch working with both ES2 and ES5, with automatic detection of the version running in the cluster
14:50:20 <yrodiere> That's 10,000 additions, 3,000 deletions, and I'm sorry for reviewers :p
14:50:51 <gmorling> uff
14:51:02 <yrodiere> Ah, I didn't say it, but in the process I replaced Jest with the official ES REST client
14:51:30 <yrodiere> We could revert that, but I'm quite sure that's the most future-proof solution
14:51:57 <yrodiere> I think that's all
14:51:58 <gsmet> yeah +1, sounds like the best move
14:52:03 <gsmet> ok
14:52:09 <gsmet> #topic Next 2 weeks Yoann
14:52:32 <yrodiere> So, there's the work on JSR-352. I will probably spent 2 days on that
14:52:58 <yrodiere> Also, I talked a bit with the local JUG and may do a presentation, so I should work on this
14:53:06 <yrodiere> (HSearch + ES)
14:54:14 <yrodiere> Time shouldn't be an issue (presenters routinely spend more than was scheduled), but I'd like to present useful examples, so I may work a bit on this
14:54:54 <yrodiere> I don't know if you guys have similar goals for your presentations? A focus on practical examples?
14:55:09 <yrodiere> (I mean, for ES + HSearch)
14:55:50 <gmorling> i like to show some code / do some coding
14:55:58 <gmorling> if that qualifies as practical example
14:56:26 <emmanuel> yrodiere: we ahve a 30 mins coding seed with the work I did for Devoxx
14:56:34 <emmanuel> you can start with that and expand
14:56:44 <emmanuel> it’s on the internet :)
14:56:45 <yrodiere> Ok, is it on hibernate-demos?
14:56:50 <emmanuel> no
14:57:12 <emmanuel> https://github.com/dadoonet/hsearch-es-demo
14:57:22 <emmanuel> and the video is on Devoxx’s youtube channel
14:57:56 <yrodiere> gmorling: sure. I meant I'd like to build a small app that solves the issues we usually bring forward ("LIKE %" is evil and such)
14:58:21 <yrodiere> emmanuel: thanks, I'll have a look after the meeting
14:58:54 <yrodiere> Anyway... I'll also work on HS 5.8 with the time left. I expect to work mostly on fixes for my pending PRs, but I may start to work on other issues too. Though I think we may have to have a meeting with Sanne to prioritize and make the roadmap clearer
14:59:00 <yrodiere> And that's all
14:59:26 <sannegrinovero> I missed most of it :( I'll read the logs later.
14:59:43 <gsmet> #topic Questions for Sanne
14:59:47 <sannegrinovero> could someone tell me what was the last you've received from my status update?
14:59:52 <gsmet> guys, you had a few questions up there?
14:59:57 <sannegrinovero> Apparently I was still typing a lot when I was offline
15:00:02 <gsmet> 15:40 < sannegrinovero>  - preparing an OGM / Hot Rod short demo for a talk here in London
15:00:02 <emmanuel> gtg
15:00:05 <gsmet> 15:40 < sannegrinovero> (in 3 weeks)
15:00:11 <yrodiere> (15:54:36) sannegrinovero left the room (quit: Ping timeout: 260 seconds).
15:00:27 <gsmet> yrodiere: useful stuff :)
15:00:36 <sannegrinovero> ok that's half a page missing :(
15:00:40 <sannegrinovero> I'll paste some stuff:
15:00:51 <sannegrinovero> among plans for next few weeks:
15:00:53 <sannegrinovero> - reboot the OGM / Infinispan 9 update quest :)
15:01:01 <sannegrinovero> which I'm afraid will imply contributing several failing tests to Infinispan ..
15:01:08 <sannegrinovero> finally, there are a couple of internal refactorings I'd like to get off my plate:
15:02:12 <gsmet> mmmh
15:03:40 <gsmet> ok, let's wrap it up, I don't think we'll get a discussion with Sanne today
15:03:46 <gsmet> #endmeeting