13:01:08 <gsmet> #startmeeting
13:01:08 <jbott> Meeting started Tue Mar 27 13:01:08 2018 UTC.  The chair is gsmet. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:01:08 <jbott> Useful Commands: #action #agreed #help #info #idea #link #topic.
13:01:10 <jbossbot> Title: MeetBot - Debian Wiki
13:01:16 <gsmet> #topic Progress Davide
13:02:02 <DavideD> Last week was chaotic for me so I haven't been very effective
13:02:27 <DavideD> There has been several PR that I helped merging in the project
13:02:43 <DavideD> In particular we fixed a bug on id generation with mongodb
13:03:06 <DavideD> I've also followed some users on Stack Overflow
13:03:33 <DavideD> I have the idea that it's hard for people coming from MongoDB to move to JPA
13:03:57 <DavideD> It seems they expect to use OGM the same way they use MongoDB Driver
13:04:28 <gsmet> is there a query builder in the MongoDB driver?
13:04:45 <DavideD> Not really, They like our query builder
13:05:01 <yrodiere> Isn't the target of OGM people who use JPA and want to start using NoSQL datastores? i.e., the exact opposite?
13:05:02 <DavideD> but I don't think they take the time to check how the whole library works
13:05:28 <DavideD> You know maybe is just us being contacted by people that have problems
13:05:48 <DavideD> It's not really a huge amount of feedback
13:06:00 <DavideD> but I'm wondering if we whould write a blog post about it
13:06:27 <DavideD> In particular, a couple of user had problem with native queries because
13:07:10 <DavideD> they try to convert partial the result into entities without having all the field
13:07:20 <DavideD> I guess I need to update the FAQ at least
13:08:10 <gsmet> a blog post might be nice
13:08:21 <gsmet> explaining how they should use a result transformer
13:08:39 <DavideD> Supporting the result transformer would also be nice :)
13:08:45 <gsmet> not sure they would read it though but you could point them to it rather than writing the same thing over and over again
13:08:54 <DavideD> But I'm not sure that's the right approach in most situation
13:09:36 <DavideD> yes, that might help
13:10:11 <sannegrinovero> I wonder if the real reason is that people don't understand the JPA model of deferring operations.
13:10:15 <gsmet> it's a good thing we drive the future efforts with the feedback we have
13:10:52 <gsmet> it's more that they do a native query and find it handy to manipulate entities afterwards
13:11:06 <DavideD> exactly
13:11:12 <DavideD> that's what I think
13:11:17 <gsmet> I can understand how they are not very excited in using the JPA DSL
13:11:29 <gsmet> so we should provide a way to support this use case
13:11:33 <fax4ever> if the user need a projection on a native query so only raw type result is supported
13:11:36 <gsmet> as it's a pretty natural one
13:11:56 <gsmet> anyway, maybe we should have this discussion outside the meeting
13:11:59 <fax4ever> addEntity is allowed only without projection
13:12:30 <fax4ever> the same behaviour of ORM
13:12:39 <DavideD> anyway
13:12:44 <DavideD> Let's move on
13:12:49 <fax4ever> an the same behaviour of JPQL
13:12:56 <fax4ever> sorry :)
13:13:02 <DavideD> no worries
13:13:17 <DavideD> Now, I'm having a look at the ignite dialect
13:13:18 <gsmet> anything else for the past 2 weeks
13:13:19 <gsmet> ?
13:13:25 <DavideD> see if we can release it
13:14:06 <DavideD> I think that's all for the past 2 weeks
13:14:11 <gsmet> ok, thanks
13:14:18 <gsmet> #topic Next 2 weeks Davide
13:14:48 <DavideD> I think we could have a quick release of OGM 5.3.1
13:15:18 <DavideD> And we could also release 5.4 alpha1 after the holidays
13:15:38 <gsmet> maybe we should wait for the next ORM CR?
13:15:48 <DavideD> we can always release alpha2
13:15:50 <gsmet> I don't feel comfortable releasing something with CR1
13:16:38 <gsmet> OK, do we have something exciting for alpha1?
13:16:51 <gsmet> (apart from the ORM upgrade)
13:16:53 <DavideD> There is the upgrade to WildFly 12
13:17:30 <DavideD> and Infinispan 9.2
13:17:58 <gsmet> looks good enough :)
13:18:18 <DavideD> Cool, I think we can wait after the holidays and see if we can fit something else
13:18:23 <DavideD> But I wouldn't wait longer for now
13:18:54 <DavideD> next week I need to have a look at ORM and see how it integrates with OGM. I guess I will focus on that
13:19:10 <gsmet> ORM 6?
13:19:23 <sannegrinovero> [I'll need to leave in 20 minutes to start my webinar]
13:19:38 <DavideD> yes 6
13:19:50 <DavideD> I think I've finished
13:19:54 <gsmet> ok, thanks
13:20:06 <gsmet> let's move to Sanne and back to the usual order after that
13:20:10 <gsmet> #topic Progress Sanne
13:20:13 <sannegrinovero> thanks
13:20:29 <sannegrinovero> I've been focusing mostly on the ORM upgrade, still.
13:20:52 <sannegrinovero> And sharing / collecting our concerns :/
13:21:12 <sannegrinovero> On the Search side I *had* a working JGroups update
13:21:37 <sannegrinovero> but then I wanted to fix some dodgy things we do in that backend as part of it
13:21:44 <sannegrinovero> so it no longer works :)
13:22:09 <sannegrinovero> I'm usually adamant about keeping things separate but in this case
13:22:23 <sannegrinovero> the previous patterns just don't fit the new semantics of JGroups 4 well
13:22:37 <sannegrinovero> so I'll need to work more on it, or back off.
13:23:03 <sannegrinovero> On other topics, still analyzing situation of ORM backwards compatibility
13:23:19 <sannegrinovero> and fixing some weird things, some caused by our wish to move to ByteBuddy as default.
13:23:28 <sannegrinovero> (apparently it breaks Spring proxy detection logic)
13:23:47 <sannegrinovero> I'll be on holidays since today for a week
13:24:02 <sannegrinovero> so back on these same topics the week after, hopefully it will all look better.
13:24:28 <sannegrinovero> oh, and about to present OGM to the nice folks at EPAM in 15 minutes.
13:24:59 <fax4ever> good luck for presentation
13:25:04 <yrodiere> break a leg :)
13:25:08 <sannegrinovero> Really hope to resume focus on Search properly when we return..
13:25:09 <fax4ever> ahah
13:25:28 <sannegrinovero> right, see yrodiere already starts hating me :D
13:25:31 <sannegrinovero> that's all, thanks.
13:25:41 <gsmet> thanks sannegrinovero
13:25:50 <gsmet> #topic Progress Fabio
13:25:59 <fax4ever> thanks
13:26:15 <fax4ever> In the last two weeks...
13:26:44 <fax4ever> I solved the issues: OGM-1412, OGM-1413, OGM-1424, OGM-1423. With help and full review of DavideD.
13:26:46 <jbossbot> jira [OGM-1412] Upgrade Infinispan to 9.2.0.Final [Resolved (Fixed) Improvement, Major, infinispan, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1412
13:26:46 <jbossbot> jira [OGM-1413] Upgrade WildFly distribution used for integration tests to 12.0 [Resolved (Fixed) Improvement, Major, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1413
13:26:47 <jbossbot> jira [OGM-1424] Infinispan remote: avoid word "Template" in favour of "Cache Configuration" in dialect properties and annotations [Resolved (Fixed) Improvement, Minor, infinispan, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1424
13:26:47 <jbossbot> jira [OGM-1423] Infinispan remote: default Hibernate OGM cache configuration for created client side caches [Resolved (Fixed) Improvement, Major, infinispan, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1423
13:27:03 <fax4ever> The issues was relative to porting to Infinispan 9.2.0 and porting to I Wildfly 12.0.0.
13:27:23 <fax4ever> As a follow up we defined our Hibernate OGM configuration, possible in Infinispan 9.2+.
13:27:35 <fax4ever> This configuration is transactional
13:27:57 <fax4ever> so we don't see the log waring
13:28:13 <fax4ever> I'm continuing to work on Hibernate ORM & OGM Microservices demo with Andrea Boriero.
13:28:26 <fax4ever> We've arrived at a good point
13:28:35 <fax4ever> The demo will be presenented internally on April 5th to Hibernate and Infinispan guys.
13:28:56 <fax4ever> that's all about my last 2 weeks
13:28:58 <fax4ever> thanks
13:29:06 <gsmet> #topic Next 2 weeks Fabio
13:29:12 <fax4ever> ok
13:29:27 <fax4ever> Infinispan Remote it is
13:29:30 <fax4ever> in my opinion
13:29:37 <fax4ever> the most stategical dialect
13:29:50 <fax4ever> n the next two weeks I will handle with issues: OGM-1426, OGM-1428.
13:29:51 <jbossbot> jira [OGM-1426] Infinispan remote: enable support to JPQL [Open (Unresolved) Improvement, Major, infinispan, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1426
13:29:51 <jbossbot> jira [OGM-1428] Feature pack distribution for Hibernate OGM WildFly modules [In Progress (Unresolved) Improvement, Major, build, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1428
13:30:05 <fax4ever> I could finish OGM-1428 (Wildfly feature pack distribution for Hibernate OGM modules).
13:30:21 <fax4ever> OGM-1428 is about enabling support to JPQL for Infinispan remote, writing the entity query subsystem.
13:30:22 <jbossbot> jira [OGM-1428] Feature pack distribution for Hibernate OGM WildFly modules [In Progress (Unresolved) Improvement, Major, build, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1428
13:30:36 <fax4ever> I don't know if I could finish the OGM-1426, in the next two weeks, because is a very huge issue.
13:30:37 <jbossbot> jira [OGM-1426] Infinispan remote: enable support to JPQL [Open (Unresolved) Improvement, Major, infinispan, Fabio Massimo Ercoli] https://hibernate.atlassian.net/browse/OGM-1426
13:30:51 <fax4ever> Solving the issue OGM-1426 is a preriquire to enable full server side indexing (that is a sort of epic).
13:31:13 <fax4ever> moreover
13:31:18 <gsmet> wondering if it's worth trying something about the JPQL support when the ORM 6 SQM will change everything?
13:31:26 <gsmet> DavideD: what's your take on this?
13:31:38 <gsmet> maybe we could explore things for the future though
13:31:46 <gsmet> just wondering
13:32:00 <DavideD> To be honest I'm not sure what we need to do for Infinispan Remote
13:32:20 <DavideD> If it's not too hard it might be worth to start to add it
13:32:35 <DavideD> At least,  he can start to have a look and we can have a better idea
13:32:38 <gsmet> OK, so fax4ever, keep in mind that the JPQL parser will be totally different in ORM 6
13:32:51 <fax4ever> Ok thank you
13:32:51 <gsmet> so be careful not to waste too much time on parser related issues
13:32:55 <DavideD> Anyway, we talked and Fabio will start with the feature packs
13:33:13 <fax4ever> from now I'm copying what is done on MongoDB and Neo4j
13:33:36 <gsmet> OK, as long as the effort is on the ISPN side, it sounds good
13:33:38 <fax4ever> yeah Feature pack will take the precedence
13:33:56 <fax4ever> ok moreover...
13:34:05 <fax4ever> I will follow the progess of the issue OGM-1427 (performed by EPAM) and giving support to them.
13:34:06 <jbossbot> jira [OGM-1427] Infinispan remote stored procedure (Remote Task) support [Open (Unresolved) Improvement, Major, infinispan, Sergey Chernolyas] https://hibernate.atlassian.net/browse/OGM-1427
13:34:21 <fax4ever> Enabling server task execution via Hibernate OGM could be a very nice and strategial feature.
13:34:28 <DavideD> Ah, right, forgot about that one
13:34:59 <fax4ever> Davide and me have decided to give this issue to EPAM buys
13:35:00 <DavideD> have they already started?
13:35:02 <fax4ever> guys
13:35:38 <fax4ever> I don't know really
13:35:55 <DavideD> ok, I was just curious, no problem
13:36:00 <fax4ever> but now main issue and the 3 sub-issues are assigned
13:36:09 <fax4ever> it is a not trivial issue
13:36:25 <fax4ever> so let's see ;)
13:36:44 <fax4ever> it's all about my next 2 weeks
13:36:51 <fax4ever> thank you
13:36:51 <gsmet> ok, thanks
13:36:57 <gsmet> #topic Progress Guillaume
13:37:23 <gsmet> so, first, we got the ORM 5.3 regression that's been bothering me fixed
13:37:39 <gsmet> I also worked on the blog to improve the list of posts
13:37:55 <gsmet> we had a couple of TCK appeals and bug reports for HV
13:38:02 <gsmet> so I worked on getting all them fixed
13:38:25 <gsmet> HV 6.0.9.Final will be released today
13:38:43 <gsmet> I finally opened the Lucene support PR for the Search 6 POC
13:39:09 <gsmet> and opened a couple more with JQAssistant support and a move to AssertJ
13:39:25 <gsmet> speaking to AssertJ, I also moved Search 5 to AssertJ
13:39:55 <gsmet> that's pretty much it for the last 2 weeks
13:40:01 <gsmet> #topic Next 2 weeks Guillaume
13:40:14 <gsmet> so, the Apache BVal folks are still making progress on BV 2.0
13:40:22 <gsmet> so I expect a couple more TCK appeals
13:40:26 <gsmet> hopefully nothing big
13:40:39 <gsmet> they are basically reimplementing BVal from scratch
13:40:51 <gsmet> so they even open TCK appeals on very old 1.1 tests
13:41:28 <gsmet> once we have everything fixed, I will release TCK 2.0.3.Final and a new HV conforming to it
13:41:45 <gsmet> that will hopefully be the last 6.0.x release as we are moving to 6.1 work
13:41:59 <gsmet> Marko opened a couple of PRs for 6.1 that I still have to review
13:42:31 <gsmet> and I plan on making progress on the Lucene support for Search 6 once Yoann will merge it
13:43:00 <gsmet> #topic Progress Yoann
13:43:34 <yrodiere> So I worked on the Search 6 POC, mainly
13:43:54 <yrodiere> fixed a problem related to annotation discovery
13:44:06 <yrodiere> Added some logging that should help us debug
13:44:17 <yrodiere> I also polished the showcase and sent some emails to try to get some feedback on Hibernate Search
13:44:21 <yrodiere> 6
13:44:43 <yrodiere> But, well, as I suspected now is not a good time since everyone's busy :)
13:45:03 <yrodiere> Apart from that I mainly worked on the support for automatic reindexing
13:45:23 <yrodiere> i.e. when an entity is updated, reindex all the entities that include it in their indexed document
13:45:44 <yrodiere> I find it really hard, but I finally managed to get some results
13:46:26 <yrodiere> Basically I managed to get it working as long as there are no associations in embeddables and with limited (all-or-nothing) dirty checking
13:46:49 <yrodiere> Oh and today I reviewed Guillaume's PR :)
13:47:07 <gsmet> yeah saw that, lots of fun coming :)
13:47:46 <yrodiere> Also I finally filed a bug to the ES team because of their Lucene dependency in the high-level rest client
13:47:57 <yrodiere> Since Guillaume was so worried about it :)
13:48:18 <yrodiere> turns out they won't be able to do anything anytime soon, the problem runs deep in their architecture
13:48:25 <yrodiere> I think that's all
13:48:33 <gsmet> #topic Next 2 weeks Yoann
13:49:13 <yrodiere> Well, next two week we may have to do some more work on the Lucene side, but I expect that will be on Guillaume's side, mostly
13:49:52 <yrodiere> I really hope I can land the automatic reindexing feature, at least with all-or-nothing dirty checking, and hopefully with fine-grained dirty checking
13:50:14 <yrodiere> then (but that'll proabably be after the two weeks) I'll work on partial document update
13:50:31 <yrodiere> well, at least I'll investigate if it's possible :)
13:50:36 <yrodiere> that's all
13:50:45 <gsmet> ok, thanks
13:50:56 <gsmet> anything else guys?
13:51:16 <yrodiere> nope
13:51:28 <fax4ever> it's ok from me
13:51:45 <gsmet> OK, let's close the meeting then, thanks everyone
13:51:48 <gsmet> #endmeeting