17:02:00 #startmeeting Java SIG -- https://fedoraproject.org/wiki/SIGs/Java 17:02:00 Meeting started Tue Nov 16 17:02:00 2010 UTC. The chair is sochotni. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:02:00 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:02:00 sochotni: Error: Can't start another meeting, one is in progress. 17:02:06 ehm :-) 17:02:16 #meetingname java-sig 17:02:16 The meeting name has been set to 'java-sig' 17:02:34 #topic roll-call 17:02:39 * cspike is here 17:03:19 hmm, ok...let's continue like this maybe someone else will join in later on 17:03:26 #info cspike sochotni present 17:03:55 #topic Maven 3 update 17:04:14 oh cool, how the custom resolver 17:04:21 so...current status issss... 17:04:23 +verb 17:04:42 I resolved one big issue and can currently compile most of things... 17:04:46 there is one problem though 17:04:53 * abo2 is here... 17:05:04 abo2: welcome 17:05:09 nice to see you here 17:05:12 thanks :) 17:05:48 well..the problem is that most packages that instal maven pom files and depmaps use non-working paths 17:06:02 so maven 3 can't find pom files in jpp mode.. 17:06:19 I am guessing mvn2 didnt have problem because it mostly used pom files directly from jar files 17:06:22 sochotni: let me guess: mixing jpp. jpp- and so on 17:06:26 (but that's just a guess now) 17:06:38 cspike: mostly mixed up "." and "-" 17:06:53 once I fix those filenames I always get further.. 17:07:06 I get something like this: 17:07:07 [WARNING] The POM for net.sf.kxml:kxml2:jar:2.2.2 is invalid, transitive dependencies (if any) will not be available: 1 problem was encountered while building the effective model 17:07:11 [FATAL] Non-readable POM /usr/share/maven2/repository/JPP/maven2/default_poms/JPP-kxml.pom: /usr/share/maven2/repository/JPP/maven2/default_poms/JPP-kxml.pom (No such file or directory) 17:07:32 that causes compilation failures obviously since depmaps from pom.xml are not added to project 17:08:10 ok, so we either need a provenpackager to touch all these broken packages, or you change the custom resolver and use a crazy regex 17:08:10 so far...I can't think of a way to make this work other than fixing all the packages (or doing some crazy stuff...) 17:08:24 lol 17:08:36 cspike: yeah, regext might work but might introduce bugs that would be hard to identify... 17:08:53 right now mvn will tell you exactly where is the problem 17:09:33 I am hoping to become provenpackager soon-ish so I'll be able to fix all those things 17:10:08 you could add a system.err.print (or whatever logging you use) to spit out something like "whoops, i couldn't match the expected path, trying with some other exp on blabla" 17:10:22 so you don't have to rely on mvn to tell you whats wrong 17:10:28 and where the compilation fails 17:11:02 sochotni, ;) 17:11:39 sochotni: fixing all broken packages will take quite a bit of time. don't think you will be able to get this into f15 17:11:39 cspike: yeah, I'll know what happens but I am thinking about other packagers.. 17:12:08 well there would be a lot of FTBFSs 17:12:26 ok...let's try middle ground 17:13:03 if the file doesn't exist I'll try different places.. 17:13:19 i mean it's your code, i won't tell you what to do. i just want to see maven3 as soon as possible. 17:13:31 cspike: patches welcome :-) 17:13:47 sochotni: i had a feeling that you might say that :) 17:13:59 ok, let's move on.. 17:14:20 cspike: you added the comps groups? 17:14:28 "Discuss Java and Java development comps groups content " 17:14:53 I have nothing to add to it at the moment actually... 17:15:17 sochotni: no, actually that was alex. i asked him if i should add certain commons to comps and he said he'd like to discuss it here 17:15:18 akurtakov added this I assume but he had a surprise meeting 17:15:43 #topic Discuss Java and Java development comps groups content 17:16:05 let's talk about this when akurtakov has some time to join us 17:16:17 yes, I think that would be better.. 17:16:29 #info deferred until next meeting with more people :-) 17:16:39 #topic Change of meeting time 17:17:34 soo...accroding to vote: http://www.makeavote.net/mavsta13768.html 17:17:48 most people preferred Wednesdays 13:00 UTC 17:18:26 second was Tuesday 13:00 UTC 17:19:41 well ok, i don't like it, but that's what people decided 17:20:18 well..the thing is that there were just 6 votes...hard to tell :-) 17:20:29 akurtakov didn't vote for that time as well 17:21:03 and wed 13UTC just had 3 second places, so actually no one likes it :) 17:21:19 honestly I didn't vote because all alternatives were pretty early through the day. I could understand that it should be during office hours, or= 17:21:56 hannes|: didn't have to be...but I guess for most people it works out for the best during the day 17:22:13 sochotni: what was the problem with the current meeting time? 17:22:15 sure 17:22:27 cspike: too late I guess? 17:22:51 several people had different things.. 17:23:16 sochotni: so the question is, will we get some more people here, when we do it on 13utc 17:23:24 ok, honestly I don't think wednesday at 13:00 would work for more people...just me and guido placed it "on the pedestal" 17:24:01 hannes|, abo2: does wed 13utc work any better that tue 17utc for you? 17:24:32 I think hannes| actually prefers later hours just as it is now, right? 17:24:33 I don't want to decide this. 17:24:41 hannes|: :-) 17:24:57 you would not be deciding it...just voicing your preference :-) 17:25:05 sochotni: do you know, if akurtakov has any problem with tue 17utc? 17:25:16 rl meeting or something? 17:25:20 right now...yes 17:25:23 yeah of course but I don't think my preference should matter on this decision ;) 17:25:33 since the summer time changed he has different meeting right now 17:25:47 sochotni: ok, if wed 13utc works for you and akurtakov, so be it 17:26:04 but btw what if people would like to join from north america? 17:26:07 cspike: he didn't vote for wed, but I think it worked for him 17:26:30 hannes|: yes, that's true...but so far noone from US got in touch :-) 17:26:38 hannes|: they don't have java in the us, they do .net :) 17:27:07 ok, meeting time stays as it is for now...I'll talk to akurtakov and we'll see 17:27:31 so once again, democracy doesn't work :) 17:27:32 if he can't move his other meeting, or prefers that wednesday one...we'll move it 17:27:38 cspike: seems like it.. 17:27:40 cspike: I,m here because I,m ircing on my phone on the way home, but honestly I don't think I should decide... 17:28:06 abo2: so generally you travel at this time... 17:28:21 maybe you'll stay longer @ work once every 2 weeks or so :-) 17:28:39 sochotni: just talk to alex and pick a new date if the current one does interfere with his/your duties. 17:28:46 yup �) I guess it'd be typical for a european 17:28:50 ok, I guess that's it for today's agenda.. 17:28:56 fwiw I won't be able to attend any meeting at 17:00 UTC on tuesdays 17:29:07 akurtakov: yeah, thought so... 17:29:14 wednesdays 13:00 UTC work? 17:29:22 yes 17:29:31 ok, wednesdays 13:00 UTC it is... 17:29:53 #agreed (sort of) SIG meeting time changed to 13:00 UTC every other wednesday at fedora-meeting 17:29:53 that was a toughy 17:30:48 now that I think about it the vote should have been...pick all time that work for you and we should have picked time that had most "x"-es 17:31:00 bah, I am bad at politics 17:31:04 :) 17:31:11 #topic open-floor 17:31:30 abo2: so, welcome to your fist SIG meeting :-) 17:31:58 abo2: oh yeah, hope you brought food for everyone... 17:32:00 if you can please check I incorporated your previous packaging guidelines into our new ones OK 17:32:13 heh, thansk :) 17:32:14 oh...almost forgot... 17:32:28 We have new packaging guidelines (FPC approved our changes) 17:32:42 so..in short...no more -%{version} in jars/javadocs 17:32:49 and other tweaks 17:33:03 sochotni: did the fpc vote on this yet? 17:33:06 cspike: yes 17:33:09 last week 17:33:27 akurtakov and me had to explain a few things, but generally it went smooth 17:33:35 Yeah, I diffed them yesterday, it looks good. I do think my sample delete-jar code is better though... 17:34:24 or warn-abo-jars rather 17:34:29 abo2: I'd like to actually create/update set of macros for java packaging one we get major stuff in good shape 17:34:47 https://fedoraproject.org/wiki/Packaging:Guidelines is up2date? 17:34:57 Yes, macros would be very neat 17:34:57 wrong like 17:35:44 http://fedoraproject.org/wiki/Packaging:Java 17:36:18 It's current as of the guideline revision passed six days ago. 17:36:26 tibbs: thanks 17:36:31 I don't think there's anything additional pending except for perhaps some typo fixing. 17:36:53 should be 17:37:00 Feel free to file tickets for any additional changes you feel need doing. 17:37:04 cspike: ^ 17:37:15 Or if it's simple stuff like fixing typos, just ask me and I'll make the changes. 17:37:32 ok, thanks 17:38:46 creating a set of helper functions/macros for java spec files would be nice.. 17:38:53 add_to_maven_depmap could be improved a lot.. 17:39:00 but that's for a later time.. 17:39:13 unless someone finds time to do it now.. 17:39:51 so...anyone has anything to add? 17:40:15 abo2: btw...you might look @ the SIG page...maybe you'll find something you can help us with? 17:40:22 can & want that is :-) 17:41:07 oh...almost forgot.. 17:41:20 we have a FE-JAVASIG tracking bug (thanks to akurtakov) 17:41:39 anything that needs our attention can block it.. 17:41:47 #link https://bugzilla.redhat.com/show_bug.cgi?id=652183 FE-JAVASIG tracking bug 17:41:54 I know more about shell scripting than java packaging, I guess that's where I can help. I can write Java code though so I'm not totally helpless. :) 17:42:32 abo2: I am not much of a Java developer myself... 17:42:51 abo2: so maybe you could see if you could create additional macros for java packaging? 17:43:18 for example add_to_maven_depmap would automatically install pom/jar file with correct name 17:43:26 (that'a problem right now) 17:43:58 ^ that would be cool, you could skip the jpp.- stuff 17:44:17 sochotni: I would like to see add_to_maven_depmap enhanced once we get maven 3 and a way to move depmap from /etc 17:44:18 I should be able to but I guess I need to grok maven first 17:45:45 abo2: ah, no need to rush...just an idea 17:46:09 akurtakov: yeah, it just came to my mind with those problems with maven3 actually.. 17:47:12 so I'm available now 17:47:23 do you want to speak about comps? 17:47:40 sure, why not... 17:48:11 #topic Discuss Java and Java development comps groups content 17:48:13 so we have 2 Java groups 17:48:39 those being? 17:48:59 Java and Java Development 17:49:16 let's talk about Java 17:49:46 it's just gcj and openjdk... 17:50:04 the gcj is mandatory.. 17:50:08 what do we want it to be - yeah, I would propose dropping the gcj and making it only jre 17:50:16 I mean openjdk 17:50:25 Is it likely to cause any problems? 17:50:26 and plugin 17:50:33 I guess not as long as we keep libgcj 17:51:29 sochotni: why do we need to keep libgcj? 17:52:00 wouldn't it stop gcj-compiled packages from working? 17:52:25 wait... 17:52:36 that would be pulled in through requires I guess 17:52:39 so not a problem 17:52:42 sochotni: yes 17:53:04 sochotni: so let it be openjdk + plugin + webstart ? 17:53:09 whatever the current rpms 17:53:11 are 17:53:34 anyone? 17:53:37 yeah I'd say that's reasonable for "Java" group... 17:53:44 at least from my POV... 17:54:02 If I was a user and installed "Java" I'd want to have webstart and firefox plugin as well 17:54:22 not much else that a user can expect... 17:54:36 cspike: ? 17:54:41 abo2: ? 17:54:44 sounds good to me 17:55:20 so we do agree to change it 17:55:28 now move to Java Development 17:55:46 #agreed Java group will contain only openjdk (sub) packages, gcj will be removed 17:56:07 I'd say we should split it into different groups? 17:56:21 yeah, javadocs 17:56:26 I thought the same too 17:56:35 either drop them or make a separate group 17:57:03 that too, plus maybe separate webservice development and application development? 17:57:18 sochotni: ?? 17:57:42 nah, just read through the comps...thought it's more mixed than that 17:57:57 I was expecting to find even tomcat and eclipse in there 17:58:33 but those javadocs should be separated I'd say.. 17:58:41 most people read docs on the web anyway I'd say 17:58:56 my proposal is no javadocs, no gcj 17:59:24 just ant+maven+essentian ant tasks + essential maven plugins 17:59:24 drop them. a person knowing about javadocs will find out how to install them 17:59:28 I'd say gcj is in the same place as with "java" group 17:59:47 akurtakov: right on 17:59:51 let's do it 18:00:17 it doesn't make sense to create separate javadoc group 18:00:21 #agreed "Java Development" will be stripped of javadocs and gcj parts. Basic maven/ant plugins and essential libs stay 18:00:37 at least until we manage to crosslink all javadocs we ship 18:00:55 +1, they are not *that* useful at the moment anyway 18:01:10 ok, I'm glad we are on the same page 18:01:31 yeah, see no problem with removing them from comps.. 18:01:47 #agreed Javadocs group will not be created 18:02:13 there is also Eclipse group which we can add to the devel group if you think so 18:03:02 hmm...not yet. I'd say people don't have to use Eclipse for java development :-) 18:03:14 althought Netbeans are going away it seems 18:03:37 I totally agree - IDEs have no place in basic devel groups 18:03:39 and people do use eclipse for a lot of different things 18:03:54 that may have nothing to do with java at all 18:04:02 cspike: also true 18:04:06 cdt... 18:04:16 or whatitsname 18:04:50 sochotni: looking for eclipse-dltk-ruby ? :) 18:05:02 no comment 18:05:08 :-) 18:05:20 ok, I gotta run or my dinner will get cold... 18:05:28 should I close it up or you want to continue? 18:05:32 I'll give you chair.. 18:05:41 just finish 18:05:44 +1 18:05:48 #endmeeting