17:25:55 #startmeeting #startmeeting Java SIG -- https://fedoraproject.org/wiki/SIGs/Java 17:25:55 Meeting started Tue Nov 2 17:25:55 2010 UTC. The chair is sochotni. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:25:55 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:25:58 eh 17:26:05 #meetingname java-sig 17:26:06 listening 17:26:08 #meetingname java-sig 17:26:08 The meeting name has been set to 'java-sig' 17:26:16 #topic roll-call 17:26:30 #info orionp cspike hannes| sochotni present 17:26:40 mbooth was here a minute ago 17:26:43 akurtakov had something this evening so he won't be joining in 17:26:47 mbooth: ping 17:26:51 pong 17:26:58 #info mbooth present as well 17:27:00 great 17:27:24 #topic Maven-3 17:27:35 so...that's me I guess 17:28:26 I have submitted maven vanilla package for review 17:28:32 #link https://bugzilla.redhat.com/show_bug.cgi?id=648945 Maven 3 review 17:28:51 I guess I'd like someone from the "old school" do the review (a.k.a. dbhole, fnasser etc) 17:29:06 but we'll see :-) 17:29:11 as for the custom resolver 17:29:21 is there a git repo yet? 17:30:24 #link http://sochotni.fedorapeople.org/0003-Use-custom-resolver.patch current version of the patch needed to use custom resolver 17:30:32 that patch is for maven 3 17:30:44 cspike: I have one local...but I think I'll have to change it... 17:30:54 because idea was to have separate independent package 17:31:02 unfortunately that's not possible 17:31:16 well...it's possible but doesn't make sense 17:31:39 so you don't plan to make it public? 17:31:46 so I'll be incorporating custom resolver into maven srpm just like mave2 is now 17:34:17 sochotni: and btw. any ETA? 17:34:36 sec.. I forgot to copy that repo out 17:34:38 so doing it now 17:36:13 #link http://sochotni.fedorapeople.org/maven-javadir-resolver custom java resolver 17:37:07 it needs maven3 parts to compile obviously and for testing I have it symlinked to /usr/share/maven/lib 17:37:10 ETA... 17:37:12 anyway... 17:37:17 I can currently build several packages 17:37:29 e.g. apache-commons-io/lang 17:37:35 atinject too 17:38:07 currently solving issue with animal-sniffer (problem with half-generated files) 17:38:16 might be bug in maven itself.. 17:38:44 #info Currently several apps compile in mvn3 "jpp" mode. 17:38:53 sochotni: eta == estimated time of arrival 17:39:13 cspike: I assumed you meant ETA of new maven jpp mode 17:39:16 was I wrong? 17:39:42 basically it would be nice to know if you're planning to push maven 3 into rawhide before f15 branch 17:39:57 and yes, with a working jpp mode, of course 17:40:03 cspike: vanilla definitely 17:40:40 cspike: we'll see about the jpp mode 17:40:55 but I assume that will be in F-15 as well 17:41:02 at least as a "tech preview" 17:41:16 so i guess all the maven plugins have to be at least touched to confirm that they work correctly with maven3 17:41:23 exactly 17:41:28 *ouch* 17:41:30 some are known not to work 17:41:36 new versions are planned 17:41:46 maven3 site plugin is being reworked completely 17:41:55 basically whole reporting in maven is new 17:42:00 so...we'll see 17:42:30 ok, next topic? 17:42:44 #topic Java packages to be monitored by the SIG 17:43:01 seems like we have all "basic" packages in place https://fedoraproject.org/wiki/SIGs/Java/Monitored_packages 17:43:32 I'd say akurtakov can go ahead and request new java sig pseudo-user with watchcommits on those packages 17:43:45 looks good to me 17:44:24 Is it easy to add new ones once we have the java-sig user? 17:45:11 I.e., can we give the user watchcommits on other packages after he's been created? 17:45:17 mbooth: you can make cvs flag '?" and request CC for that user 17:45:30 Ok, cool 17:45:31 see package change request 17:46:12 and this also depends on other topic.. 17:46:20 #topic Java Packaging guidelines 17:46:37 #undo 17:46:37 Removing item from minutes: 17:46:38 I'm here if you have questions. 17:47:07 #info Package list in good shape akurtakov can go ahead and ask for java-sig user 17:47:16 #topic Java Packaging guidelines 17:47:21 And I can set up watchchcommit access for a user if you need it; there's no other way to really request it. 17:47:22 tibbs: great 17:47:54 tibbs: isn't that same as adding him to "cc" in review request? 17:48:24 Yes, but there's no easy way to do that for existing packages. 17:48:37 So you can just give me a list and I'll do it directly in pkgdb. 17:48:44 tibbs: yes that's what is part of "create java-sig pseudo-user" 17:48:54 tibbs: https://fedoraproject.org/wiki/SIGs/Java/Monitored_packages 17:48:59 that is the list 17:49:04 but we don't have user created yet 17:50:00 last meeting we decided to have pseudo-user's email go to java-devel ML 17:51:11 until tibbs gets back to us...We had some interesting suggestions regarding new guidelines 17:51:29 a) versionless jars & javadoc dirs 17:51:50 it's true that javadir would look much nicer :-) 17:52:38 sochotni: i couldn't find any guideline that actually says that a versioned jar file has to be present. so why are they existing anyway? 17:53:04 Easy way to know what version is installed? 17:53:09 I thought the recent guideline revision did away with the versioned javadoc stuff. 17:53:13 orionp: that was my idea also 17:53:18 Holdover from jpackage guidelines, I'd say 17:53:36 tibbs: if by recent you mean the one from ~3 weeks ago...noone in Java SIG knew about that 17:53:51 orionp: rpm query 17:54:17 orionp: yeah, that was also my idea...you can figure it out by rpm -qf usually 17:54:20 but... 17:54:43 I guess sometimes jar could have different version than rpm package.. 17:54:49 although that would be a bit weird.. 17:55:24 sochotni: examples? as far as i know the package naming guidelines, that should be impossible... 17:55:33 ok...how about: if no one cames with a really good reason to have versions in jars -> let's get rid of versions in jars 17:55:47 cspike: should be but people are inventive :-) 17:56:48 ok, someone has to update the packaging guideline draft and the review template... 17:56:58 there are few packages where we have multiple versions but those are usually distinguished by number in name.. (junit3/4) 17:56:58 volunteers? :) 17:57:09 cspike: I'll do it 17:57:23 I am mergin Abo's draft anyway 17:57:28 well..not a draft anymore 17:57:43 I don't mean to interrupt, but where will the java-sig user's mail go? 17:57:56 tibbs: see few lines up: java-devel ML 17:57:58 tibbs: java-devel mailing list 17:58:11 Isn't that closed to nonsubscribers? 17:58:21 #action sochotni will update Packaging draft with versionless jar filenames 17:58:31 tibbs: yes, would that be a problem? 17:58:39 Well, yeah. 17:58:49 Otherwise you won't be getting much mail. 17:59:10 :-) I meant: can you just subscribe pseudo user there? 17:59:38 frankly I don't know who list admin is 17:59:51 ah: java-devel list run by gbenson at redhat.com, green at redhat.com, mark at klomp.org 18:00:16 can we actually use the draft in a review request yet? it would be stupid to force someone to provide a versioned symlink just to remove it as soon as the draft is approved... 18:00:45 cspike: we can't be sure that it will be approved...although I'd say that if Java SIG sees no problem... 18:01:18 cspike: I don't see a problem with that... How long does the Packaging committee take to ratify things like this? 18:01:19 Generally if the SIG wants it and it's not insane, FPC will approve it. 18:01:44 If you open a ticket we'll address it at the next meeting which I'm pretty sure is tomorrow. 18:01:57 fedorahosted.org/fpc 18:02:52 tibbs: there are a few cleanups before it's ready to go 18:03:06 tibbs: you have meetings bi-weekly right? 18:03:25 Ostensibly. 18:03:39 But it really depends on everyone's schedule and how much work there is to do. 18:03:53 let's assume in a month then :-) 18:04:01 i would suggest to do it as soon as possible to avoid doubling the work 18:04:37 sochotni: what about you finish the merge, get rid of the versioned symlinks and then just open the ticket 18:04:38 cspike: I believe there are no insande changes so guidelines will be approved once we agree on them.. 18:05:09 well..why not...if there is something we'd like to change...next meeting 18:06:07 #action sochotni file FPC ticket for new draft approval once versionless things are in place 18:06:26 sochotni: oh, just found something... could you update the plugin names in the maven specfile template before you open the ticket? 18:07:02 the maven2 stuff should be obsolete 18:07:27 cspike: I see them as maven-X 18:07:48 cspike: refresh? 18:07:54 sochotni: oops, sorry, was looking at the current guidelines. wrong page 18:07:59 cspike: :-) 18:08:24 so I'll do it today just to be sure I don't forget 18:09:00 tibbs: so to add pseudo-email to java-devel I guess I should contact java-devel admins right? 18:09:32 Well, the mail won't come _from_ the pseudo-user. 18:09:35 It will go to it. 18:09:54 So subscribing java-sig@fp.org won't help you. 18:10:08 tibbs: ah, sorry...mixed up 18:10:13 I've already asked for the pseudo-user to be created. 18:10:22 tibbs: thank you 18:10:37 Once it's done I'll add it as a CC on all of those packages. 18:10:57 tibbs: you are my new hero :-) 18:11:58 BTW, I have an item when you open the floor. 18:12:22 I gues we could discuss change of time for the meeting...and then open-floor 18:12:49 hello all, has the java sig meeting started yet? 18:12:55 get on to the next topic? 18:13:03 cspike, mbooth, hannes|, orionp: is it worth starting discussion on change of time for meetings? 18:13:14 ggraz: ~50 minutes ago :-) 18:13:19 ggraz: one hour late. but you're not the only one :) 18:13:27 sochotni yeah 18:13:28 #topic Change of meeting time 18:13:44 5pm is kinda inconvenient in the UK 18:13:45 sochotni: you set up a voting somewhere, didn't you? 18:13:51 d*mn daylight saving time 18:14:11 cspike: I did but it closed already (only 2 of us voted :-/) 18:14:21 ggraz, mbooth correct 18:14:24 sochotni: i guess you and me :) 18:14:40 There was a vote? 18:14:44 see http://lists.fedoraproject.org/pipermail/java-devel/2010-October/003965.html 18:15:05 I noticed almost everyone missed it... I should have posted another thread I guess 18:15:18 let's try again...I'll do the vote from the scratch 18:15:28 Oops 18:15:41 sochotni: nah, i can't remember what i voted for 18:15:55 cspike: :-) (me neither) 18:16:26 #action sochotni will re-create voting for change of meeting time and post separate thread to java-devel 18:16:42 #action orionp akurtakov_ mbooth hannes| cspike will vote :-) 18:17:01 let's give tibbs a chance ... 18:17:06 #topic open floor 18:17:11 Yes, I am normally commuting at 5 ;-) I will vote for earlier or later 18:17:25 Sure, I just wanted to point out that there are several pretty old java package reviews open. 18:17:37 tibbs: we are slowly chipping away at them 18:17:44 but you're right.. 18:17:50 even quite a few merge reviews 18:17:54 i need to subscribe to the list first 18:17:57 OK, I didn't know if anyone was looking at them. 18:18:27 * cspike took 3 merge reviews last week 18:18:28 The oldest one is from last October. 18:18:29 tibbs: I believe most of us did a few merge reviews lately 18:18:33 akurtakov_: keeps nagging :) 18:18:41 cspike: :-0 18:18:56 jilter, I think. Plus there's jexcelapi, jrexx, the aduna stack, and others. 18:18:57 tibbs: yes, I have a few open where original reported doesn't respon... 18:19:03 I'll have to close them I guess 18:19:22 tibbs: for aduna...meforster wasn't sure if it's still neede AFAIK 18:19:28 Yes, close them out after a month. Don't keep them around; it just wastes reviewer time. 18:19:39 I was wondering if meforster was still active. 18:19:49 tibbs: sort of 18:20:03 Your pseudo-user is set up now, by the way. 18:20:08 Re: the aduna stack, Mary dropped off the radar for while, one of us should ping some of those reviews 18:20:17 I believe life intervened but she helped with maven 2.2.1 update AFAIK (before I came) 18:21:02 tibbs: thank you a lot, now just get java-devel to accept mail from all weird addresses 18:21:17 commits are from our fas accounts if I remember correctly 18:21:46 Can the java-sig user's email address just forward to the list? Is that possible? 18:22:17 mbooth: java-sig user's email IS the list now I believe 18:22:23 tibbs: right? 18:22:37 Oh, I have to request that that address accept mail. 18:22:47 on a completely unrelated topic. http://lists.fedoraproject.org/pipermail/java-devel/2010-October/003970.html 18:22:48 #action all: look at merge reviews 18:23:06 does any one know how to contact devrim? 18:23:06 cspike: Indeed 18:23:20 cspike: Did fnasser release his? 18:23:28 mbooth: i did them 18:23:38 cspike: cool :-) 18:23:55 mbooth: i mean, yes he released his and i orphaned and stuff 18:24:31 devrim actually just has to relese ownership, i can finish the rest... 18:24:46 is there a way to force this if he doesn't answer in say a week or two? 18:24:57 OK, the java-sig pseudo user will now accept email (once everything pushes out). 18:25:01 cspike: I believe we'd have to do unresponsive packager thing.. 18:25:32 ok, i'll write another pm and we'll talk about it during the next meeting 18:25:36 tibbs: so what do we need to do? Get java-devel to open-up? 18:25:54 tibbs: or do you see another way to make it work by any chance? 18:26:12 cspike: yeah. I pm'ed him few months back at the start of rename process 18:26:14 no response 18:26:44 Yes, that would be good. I am not a list admin so I don't know what options may be available besides making the list more open. 18:27:01 tibbs: ok, thank you 18:27:47 any volunteer for contacting java-devel list admins? 18:27:56 so that I don't write all the emails :-) 18:28:19 mbooth: btw, you don't have to file the releng ticket, there are some packages missing so i'll file one single ticket when all packages are dead -> https://fedoraproject.org/wiki/JakartaCommonsRename#Progress_details 18:29:26 cspike: yeah, I forgot about the jakarta rename...sorry 18:29:39 as in #topic stuff 18:30:08 np, seems everythin really important is said 18:30:26 ok, since noone volunteered I guess I'll send those emails 18:30:47 #action sochotni contact java-devel list admin and figure out a way for our pseudouser get mail 18:31:10 let's coll it a night then? :-) 18:31:15 +1 18:31:21 #endmeeting