21:04:49 #startmeeting 21:04:49 Meeting started Wed Jun 1 21:04:49 2011 UTC. The chair is mojavelinux. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:04:49 Useful Commands: #action #agreed #help #info #idea #link #topic. 21:04:53 and the pointy hat? 21:05:07 Add me as a chair please 21:05:07 the pointy "red" hat! 21:05:12 lol 21:05:13 we hail you 21:05:19 #chair lightguard_jp 21:05:19 Current chairs: lightguard_jp mojavelinux 21:05:20 isn't that a cardinal? 21:05:30 no underwear on underneath :) 21:05:34 haha 21:05:36 yikes! 21:05:41 You've been hanging around Pete too much 21:05:47 lol 21:05:50 hehehe 21:05:55 Okay, let's get started 21:05:59 First item up 21:06:07 #topic Action Items from last week 21:06:19 bleathem: Your report 21:06:52 get on the pulpit man 21:07:05 He must be praying now 21:07:07 brother leathem 21:07:12 ARISE 21:07:26 is that what bleatham is short for? brother leathem 21:07:33 never knew you were a monk brian! 21:07:36 haha 21:07:43 sorry 21:07:47 afk for a minute 21:07:50 There he is 21:08:13 #info I spoke with Jay re: the dataexport from jsf data tables 21:08:42 cool 21:08:56 #info short answer is RichFaces 4 planned to add support for this, but not shceduled for a release version yet 21:09:10 Good enough for me 21:09:18 #info follow-up for further discussion on the RichFaces forums, to discuss any Seam tie-in 21:09:23 I know primefaces already has it, pretty sure icefaces does as well. 21:09:31 not sure about openfaces 21:09:43 right, so JBoss needs a solution to 21:09:47 too 21:09:48 funny that was one of the first jsp tags I used 21:09:50 displaytag 21:09:53 for that very reason 21:10:00 that was like 8 years ago 21:10:06 time flies eh mojavelinux? 21:10:16 reminiscing? 21:10:22 primefaces lets apache poi do the excel formatting. very simple but effective 21:10:28 just saying, nice to see that we are getting around to it :) 21:10:44 hum, shouldn´t be too hard to do that either 21:10:45 Okay, we'll leave that up to the RichFaces team to work out, and people that ask can pester them :) 21:10:55 yeah definitly a RF thing 21:10:56 it may be cool if RF provides basic functionality, but Seam integration could provide some added "sauce" 21:10:59 okay, but that would be a great feature for an example app 21:11:00 Okey 21:11:04 so let's make sure we use it as soon as it's ready 21:11:08 via Seam Reports or something 21:11:13 oh, yeah! 21:11:14 plugin 21:11:16 Yeah ! 21:11:20 Forge ! 21:11:29 well, I meant plugin like extension 21:11:30 spi 21:11:31 or something 21:11:34 ah ok 21:11:40 /done 21:11:41 That would do as well :) 21:11:46 as in, you get the data dump and you can feed it into your report mechanism 21:11:55 Next two official items from last week went to mojavelinux. 21:12:20 oh yes, cool 21:12:24 About the definitions and arquillian embedded jboss modules 21:12:26 okay, so notice on the home page we now have the defintiions 21:12:32 I updated the community integrator definition 21:12:44 http://seamframework.org/Community/CommunityRoles 21:13:15 nice 21:13:39 #link http://seamframework.org/Community/CommunityRoles 21:13:46 I checked on the embedded AS 7 and it sounds like that is already working (or close to it); basically, AS 7 can be embedded cleanly 21:13:53 That may show up twice, oh well :) 21:14:09 #info AS7 can be embedded cleanly for arquillian tests 21:14:20 is there somewhere we can get it yet? 21:14:57 I'm a bit behind on following the arquillian dev, but as far as I know, the container should be available 21:15:03 check the github repo for Beta1 21:15:04 I wish configuring arquillian be a little easier. 21:15:06 cool 21:15:49 not up to date with Beta1 yet 21:15:58 it will be simpler, that's what is currently being worked on 21:16:04 Ok 21:16:07 We have a couple of followup items that weren't official as well. 21:16:17 we are in a bit of a between period right now, because the docs aren't yet put together with the latest of how arquillian in configured 21:16:21 but it gets a lot simpler 21:16:27 ALR is the king of developer simplicity, so if he's on it, there is nothing to worry about 21:16:28 One is removing "Committer" from the module pages 21:16:31 :D 21:16:31 and I've proposed moving to a yaml baesd format for configuration to make it even simpler 21:16:37 This has been done and now says "Contributor" :) 21:16:44 Oh my, looks like I won something ! :D 21:17:05 yaml +1 ! 21:17:26 I'm going to suggest that we have arquillian.yaml if arquillian.xml is absent 21:17:31 gotta file a jira for that 21:17:32 We're also in the process of removing the GWT module from the site and replacing it with Errai info 21:17:39 cool 21:17:59 yes, that's for driving that jason 21:18:06 so just to be clear, 21:18:10 #info there is no GWT Module 21:18:21 #info rather, we have Errai CDI 21:18:38 #info Errai CDI brings the CDI programming model to GWT, both on the client and server side 21:18:57 #info Errai is the bus that handles the event roundtripping 21:19:06 it's fast a hell 21:19:13 Look for that change to be finalized in the next day or two 21:19:27 * rruss knew it was good to send Dan to IO with Mike ... 21:19:30 #action lightguard_jp finish the GWT module -> Errai CDI 21:19:40 it's like Ajax Push on steroids 21:19:47 haha cool 21:20:00 Even better if GWT was not used 21:20:26 it's interesting that you say that george 21:21:07 Maybe Seam Remote could have something like that ? 21:21:09 because it wouldn't be too far out of the realm of possibility to reuse the infrastructure to adapt it to other frameworks 21:21:14 right, exactly 21:21:19 cbrock: will you be able to use Errai outside of GWT? 21:21:20 perhaps Seam remoting could be changed to use Errai 21:21:36 yeah, that would be awesome 21:21:37 Seam remoting currently is a little messy 21:21:49 because a lot of the logic inside of Errai is about the mechanism of handling a streaming json respose 21:21:50 and about 21:22:04 task scheduling, work pools 21:22:05 etc 21:22:08 Reusing Errai would be a good move, if it can be done 21:22:18 you could think of it as Node.js written in Java 21:22:24 that would be cool 21:22:24 yeah 21:22:29 with CDI 21:23:10 #idea use Errai to power Seam Remoting (or other Ajax/push technology stacks) 21:23:18 Who´s in charge of Seam Remoting ? 21:23:22 if you want to know why Errai and Node.js matter 21:23:25 Shane ? 21:23:34 think so yes 21:23:34 I strongly urge you to spend a few minutes and read the Node.js refcard 21:23:37 you will totally get it 21:23:49 serial request-response is dead 21:23:54 long live async requests 21:24:11 the refcard is more about the paradigm than the impl 21:24:15 Richfaces could benefit from that too ! 21:24:31 i'd have to take a look 21:24:37 http://refcardz.dzone.com/refcardz/nodejs-building-scalability 21:24:46 shane, this idea fits well with Seam Remoting 21:24:46 (morning!) 21:24:51 there is no push in remoting currently 21:25:05 it was in seam 2, but i took it out for seam 3 because i wanted to rewrite it 21:25:07 how much you can use from Errai is something I can't tell you, but certainly the technique can be used 21:25:27 but I trust Mike with this feature if we do put it back 21:25:37 because he has spent a lot of time unwrapping it 21:25:42 it has a few requirements, including 1) being able to subscribe to a jms topic/queue 21:25:46 git [12core] push 10master7 e31ac24.. 6Lincoln Baxter, III SEAMFORGE-200 21:25:47 jira [3SEAMFORGE-200] Forge should default to built-in 'mvn' command line if Maven is not detected in the underlying system [10Open (Unresolved) Feature Request,7 Major,6 Lincoln Baxter III] https://issues.jboss.org/browse/SEAMFORGE-200 21:25:47 git [12core] push 10master URL: http://github.com/forge/core/compare/555f3b4...e31ac24 21:25:50 2) receive CDI event notifications 21:25:51 There is SEAMREMOTING-6 and SEAMREMOTING-7 21:25:52 jira [3SEAMREMOTING-6] Comet requests for Seam Remoting [10Open (Unresolved) Feature Request,7 Major,6 Shane Bryzak] https://issues.jboss.org/browse/SEAMREMOTING-6 21:25:53 jira [3SEAMREMOTING-7] Seam event bus integration for remoting [10Open (Unresolved) Feature Request,7 Major,6 Shane Bryzak] https://issues.jboss.org/browse/SEAMREMOTING-7 21:26:15 both of those are being worked on in Errai 21:26:27 git [12core] push 10master7 abcc0f2.. 6Lincoln Baxter, III SEAMFORGE-200, typo 21:26:28 jira [3SEAMFORGE-200] Forge should default to built-in 'mvn' command line if Maven is not detected in the underlying system [10Closed (Done) Feature Request,7 Major,6 Lincoln Baxter III] https://issues.jboss.org/browse/SEAMFORGE-200 21:26:28 git [12core] push 10master URL: http://github.com/forge/core/compare/e31ac24...abcc0f2 21:26:31 so it will be interesting to see if we can reuse the bus w/o gwt...I'm guessing we can, but don't want to jump to conclusions 21:26:39 does it only use json though? 21:26:57 XML should not be used, not even in Seam remoting (IMO) 21:27:05 agreed 21:27:06 what's the alternative? 21:27:09 json or yaml 21:27:13 yaml being the best choice 21:27:17 yaml is the choice 21:27:23 json doesn't support recursive object graphs 21:27:26 though I think currently Errai uses JSON, but it's totally custom inside of Errai 21:27:33 so it could use anything 21:27:34 i don't know about yaml 21:27:38 no reusing buses 21:28:05 yaml does from what I understand 21:28:17 i'm looking now 21:28:17 but again, i'm a bit on the fringe of my understanding here, still have stuff to read 21:28:26 john? 21:28:33 no reusing buses? 21:28:48 sbryzak: That could be an action for next meeting, what do you think ? 21:28:59 i don't think it will work. i was looking at it for JMS & Errai, won't work. 21:29:23 Maybe Errai should be refactored then 21:29:29 ah, we'll, I was thinking along the lines of the fact that we could make changes to the project, if there were motivation to do so 21:29:39 geoge beat me to it 21:29:51 even factoring in a refactor. 21:29:53 yaml doesn't look like it supports it either 21:29:55 conceptually it won't work. 21:30:21 johnament: you mean specifically the JMS tie-in? 21:30:25 oh, wait, we are talking about jms and errai, I mean whether errai can be used for other push technologies like seam remoting 21:30:34 oh 21:30:37 right, that's seperate and likely I agree with you there that's not going to work w/o a bridge 21:30:46 * johnament shouldn't jump into the middle of an on going conversation 21:30:50 :) 21:30:57 it's okay, got us thinking :) 21:31:04 why not use the richfaces features? 21:31:17 Seam JMS Beta2 just added specific support to support RichFaces push 21:31:18 basically, we need an alignment 21:31:20 :-) 21:31:38 because the actually mechanism of push in errai is highly optimized 21:31:38 mojavelinux: Do you want the action item, or should we give it to someone else? 21:32:17 Richfaces push is going to add support for CDI 21:32:31 bleathem: event model? 21:32:33 I think the action item is that we need a plan for the seam remoting story with respect to errai...are we moving forward with seam remoting, trying to join them or just pushing forward w/ errai 21:32:39 looks like there is a lot of duplication of effort going on 21:32:44 so for shane 21:32:51 yes 21:33:10 and I guess richfaces has a part of that story too 21:33:12 johnament: I don't know the details (yet) 21:33:13 consolidation certainly required 21:33:18 #action sbryzak will talk with cbrock about errai and flesh out the seam remoting / errai idea 21:33:36 remoting would only use errai for push functionality 21:33:40 #info also need to get Jay involved in this for RichFaces push 21:33:51 errai is no replacement for the other features 21:33:57 #agreed 21:33:58 yeah, I know 21:34:09 sounds like we need a separate "push" feature that is view independent. 21:34:25 my hope is that's errai...because it's not just about push 21:34:32 I bet Seam Remoting is the place for that 21:34:34 it's also about async requests and task pooling 21:34:45 because the requests will build up otherwise 21:34:54 sounds like errai needs to separate GWT bits so it can be re-used elsewhere 21:35:00 errai is very optimized to handle massive # of incoming requests and not getting backlogged 21:35:04 kenfinnigan: agreed 21:35:13 once you start to queue 21:35:14 +1 kenfinnigan, summed it up well 21:35:17 your async goes to shit 21:35:53 it's all fun and games until the pipe clogs 21:36:00 very true 21:36:07 Any other action items from last week we need to discuss? 21:36:14 Code Review ? 21:36:19 not that i recall 21:36:19 gastaldi: It's coming 21:36:24 ok 21:36:36 ticket bastard 21:36:38 What about Forge + Seam modules ? 21:36:56 If we have time. 21:37:04 ok, let´s proceed then 21:37:10 jason and I are getting organized to revive ticket bastard, we will be making an announcement when we get it sorted 21:37:20 cool 21:37:23 ticket bastard? 21:37:26 que es? 21:37:31 ticket monster -> ticket bastard 21:37:35 Ticket Monster 21:37:39 lol 21:37:47 still don't get it. 21:37:50 is that a demo app? 21:37:55 mojavelinux: read thru the design doc, looks like some good thinking has already been done 21:38:01 yes, this is the seam booking predecessor 21:38:05 ticket monster is on hold for a while 21:38:06 ah 21:38:07 https://github.com/seam/ticket-monster 21:38:20 pete needs to get the infinispan/cdi integration into an infinispan release 21:38:27 he said that it would happen in early july 21:38:35 can we take out infinispan for now? 21:38:38 just do JPA? 21:38:50 what is it using in infinispan? 21:38:54 i guess so.. but that was one of the big features for the demo 21:38:59 +1 to that, could then introduce infinispan and hibernate ogm together 21:39:10 that's fine, our concern isn't for a demo at this point, it's having a new example to work on 21:39:35 i fixed most of the build the other night 21:39:40 and we actually have design documents for this thing, so it makes it reasonable to keep it going 21:39:50 fantastic 21:40:01 just have to pull out the seam-clouds stuff 21:40:03 can we make those design documents public? 21:40:09 cause they are useful 21:40:17 i don't see why not 21:40:24 me either 21:40:30 okay, we need an action item then 21:40:37 agreed 21:40:40 just copy paste it to the wiki, clean up some formatting 21:41:03 #action Shane will copy paste the ticket monster (bastard) design documents to the Seam wiki so more people can play 21:41:14 #action or, put them in a design.md file in the source tree 21:41:28 either way 21:41:41 feel free to reassign 21:42:17 Okay, moving on to the next item :) 21:42:24 I think you're all interested in this one. 21:42:32 as I understood this app is for module leads playing with other modules right? 21:42:58 jose_freitas: Not necessarily. 21:43:15 #topic Results of the Polls 21:43:49 #info Community integrator was a pretty tight race 21:44:03 #info George Gastaldi will be our next Community Integrator 21:44:14 congrats gastaldi! 21:44:18 Yay !! Finally I won an election ! :D 21:44:20 congratulations george 21:44:25 Thanks ! 21:44:38 to the next you can try for senator 21:44:39 ticket monster doc is shared: https://docs.google.com/document/pub?id=1JIVsO90PsLIYI66mEwOWC5qZ2GxHBRWXT3oYnjBTOyQ 21:44:40 congrats gastaldi ! 21:44:48 it won't be easy to move it to seam wiki 21:44:55 we'd lose all the comments, which i think are important 21:44:59 plus it's full of tables/images 21:45:30 oh, okay, didn't think of making it public in google docs 21:45:34 congrats george 21:45:39 let's just add a link to it, perhaps from the Examples Ideas page 21:46:00 Thanks everyone 21:46:03 wow, it has a lot of structure already 21:46:10 congrats george! 21:46:23 #info Forum Jedi support overwhelmingly went in favor of Jason Porter 21:46:28 and also, congrats to george and antoine for becoming officially listed on CDI 1.1, and for their participation 21:46:44 bah, the published doc lost the comments anyway 21:46:47 Congratulations Jason ! 21:46:56 yeah, the forums finally have someone at the wheel again :) 21:46:59 congrats jason, best man got the job 21:47:13 yeah 21:47:22 now the lightguard has another reason to be a guardian 21:47:23 := 21:47:25 I realized that I didn't set terms on them, for revote...how about we say 6 months? 21:47:28 congrats 21:47:30 haha nice one jose_freitas 21:47:39 6 month term sounds good 21:47:46 6 months sounds like a good term 21:47:46 yeah 21:47:58 Hope I don´t get hit by a bus :) 21:48:02 agreed 21:48:07 #agreed revote for these positions again in 6 months 21:48:07 #action set event in calendar for next vote on community roles 21:48:17 barring a special election :) 21:48:20 gastaldi: now you jinxed it! watch out next time you cross the street! 21:48:23 that also means that bleathem and lightguard_jp owe time 21:49:27 :) 21:49:46 Next item, unless others have questions 21:50:10 jason, you can add that to your sig on the forums 21:50:22 mojavelinux: Good point 21:50:33 #topic Code Reviews 21:50:46 I'll let gastaldi talk about this, he brought it up 21:51:27 Ok, it all happened when Antoine joined the room and wanting a code review for Seam Social 21:51:47 I said that I´ll do it if he do it for Seam Reports also 21:52:05 It happens that we found better design ideas 21:52:35 So, I thought that this practice would apply to other modules as well 21:53:10 and in fact it has, but perhaps accidental progress, or by chance 21:53:21 yeah 21:53:34 and I know that cody has piped up at a number of meetings asking for a code review, so I think in the interest of seeing seam be successful and consistent 21:53:58 we can have sort of a call for review and if you have some cycles, try to pitch in on reviewing those apis 21:54:04 We could have Crucible running for code reviews 21:54:15 jboss is looking at getting that setup 21:54:15 that might be overkill 21:54:24 bleathem: how so ? 21:54:26 For now, GitHub works just as well 21:54:30 crucible is cool, but wouldn't that add a lot of admin overhead? 21:54:40 You just have to be looking at a commit to make comments 21:54:50 the only shortcoming of github is that you can only comment on commits, not just on a file in a branch 21:54:51 I don't think we need code review integrated into our work flow 21:54:58 ah, ok. Github may do the job as well 21:54:58 You can do line by line commenting as well in GitHub 21:55:03 right, I think if we use the develop -> master flow 21:55:04 doing it as "one offs" is better suited 21:55:13 then we can comment on the proposed merge 21:55:23 that also helps people know what to comment on 21:55:36 lightguard_jp this comes back to the git workflow 21:55:48 so perhaps, we need to doc how the review process fits in there 21:55:50 then this will work 21:55:54 I've been looking more closely at the git workflow - I like what I see 21:55:54 do we need an action item to move all modules to git flow? 21:56:07 yes, I think we just need to do it at this point 21:56:12 what means to move them to git flow? 21:56:22 Seam Faces will adopt it (as soon as I find the time) 21:56:23 does it mean* 21:56:44 #action take steps to shift modules to using git flow; first step is to clarify what it is on the mailinglist 21:56:56 jose_freitas: see http://nvie.com/posts/a-successful-git-branching-model/ 21:57:05 bleathem: Install git flow, run git flow init you're done :) 21:57:06 there have been some follow-up posts as well 21:57:12 was looking for that, it's got great explanation on the process 21:57:34 lightguard_jp: is it transaparent for other committers? 21:57:48 It's essentially a wrapper around branching 21:57:52 thanks bleathem, you´ve pointed this link before, but I didn´t read yet. shame on me 21:57:54 lightguard_jp: in that they just fork patch and pull request? 21:57:54 It's completely transparent 21:58:02 yep 21:58:05 some extra data in .git/config 21:58:10 but they have to fork develop, not master right? 21:58:11 i think it actually solves some confusion with git 21:58:14 yes 21:58:30 I agree, makes it easier to manage in my view 21:58:35 this is where I wish github would fix their forking 21:58:39 so that you can redirect your fork 21:58:48 Set up develop as the main branch in GitHub admin for your module and it'll work just fine. 21:58:49 follows a more natural dev process 21:58:52 it's just a matter of changing the remotees on their end 21:58:57 so I was wondering if master should be develop 21:59:02 Might want to make mention of your using the workflow in the readme or description though 21:59:04 rather than release 21:59:18 lightguard_jp: good call 21:59:20 ahhhh 21:59:26 You could do that too 21:59:26 yeah 21:59:32 master release instead of develop master 21:59:46 master is just a name 21:59:47 this conversation is too multi-threaded, I'm losing track of who is responding to who :P 22:00:08 me too 22:00:10 bleathem: We can talk more later about it if you have questions 22:00:10 lightguard_jp: but master is what contributors fork by default right? 22:00:14 ok 22:00:22 * bleathem shelved 22:00:28 k, we can get that started on the list 22:00:31 agree with bleathem 22:00:31 bleathem: Just because that's the default branch GitHub gives to them, it's configurable. 22:00:39 yep, its' configurable 22:00:43 oh, didn't know that 22:00:44 that's how you can make develop the main branch 22:00:47 it's in the admin page 22:00:50 bleathem: with git flow it becomes develop, which as lightguard_jp mentioned you can change in the admin of github 22:00:56 in other words, RTFM !! 22:00:59