[00:16:40] *** admc has joined #hudson
[00:17:58] *** alexlod has quit IRC
[00:21:18] *** alexlod has joined #hudson
[00:25:10] *** velo has quit IRC
[00:26:01] *** alexlod has quit IRC
[00:26:47] *** admc_ has quit IRC
[00:28:56] <rtyler> kohsuke: it's too bad ohloh doesn't actually tell you which files are the other licenses
[00:29:08] *** alexlod has joined #hudson
[00:38:09] *** alexlod has joined #hudson
[00:38:55] *** alexlod has quit IRC
[00:42:46] *** jieryn-w has quit IRC
[00:45:23] <mrooney> just grep for GPL and BSD?
[00:49:35] *** waz has quit IRC
[01:07:28] *** keshureddyp has quit IRC
[01:07:31] *** waz has joined #hudson
[01:26:28] *** davidstrauss has joined #hudson
[01:36:14] *** keshureddyp has joined #hudson
[01:43:28] *** abayer has quit IRC
[01:43:36] *** mindless has quit IRC
[01:45:06] *** abayer has joined #hudson
[02:04:30] <mrooney> abayer: I don't think it is the wiki excerpt?
[02:05:03] <mrooney> I am pretty sure I have my wiki page for the NCover plug-in properly set up but it doesn't have a description the Available listing
[02:06:39] <mrooney> Anyone else know how to give my plug-in a description in the Available listing?
[02:18:19] *** admc has quit IRC
[02:40:39] <abayer> mrooney - it is the wiki excerpt, but it'll only get updated when the cronjob that creates the Update Center runs.
[02:42:11] <kohsuke> mrooney: make sure your pom.xml points to it
[02:42:17] <abayer> ...that too.
[02:42:18] <kohsuke> see other plugin's pom.xml for example
[02:50:08] <keshureddyp> i see error - impossible connect to irc server when i save on configuration page . Does any one experience same ?
[02:50:57] <kohsuke> I suggest filing a bug with a stack trace
[03:11:07] *** jieryn-w has joined #hudson
[03:58:09] *** rromanchuk has quit IRC
[04:08:46] *** waz has left #hudson
[04:36:39] *** davidstrauss has quit IRC
[04:42:47] *** rromanchuk has joined #hudson
[04:54:00] *** admc has joined #hudson
[05:03:33] *** admc has quit IRC
[05:14:00] *** Lewisham has joined #hudson
[05:20:35] *** alexlod has joined #hudson
[06:17:12] *** Lewisham has quit IRC
[06:19:26] *** akostadinov has joined #hudson
[06:21:59] *** davidstrauss has joined #hudson
[06:59:08] *** jieryn-w has quit IRC
[07:11:20] *** Lewisham has joined #hudson
[07:14:52] *** rromanchuk has quit IRC
[07:17:17] *** rromanchuk has joined #hudson
[07:36:07] *** khmarbaise has quit IRC
[07:50:23] *** akostadinov has quit IRC
[07:55:42] *** davidstrauss has quit IRC
[08:27:14] *** glass-eye has joined #hudson
[08:34:34] *** akostadinov has joined #hudson
[08:36:31] *** Lewisham has quit IRC
[08:53:03] *** davidstrauss has joined #hudson
[09:12:16] *** Lewisham has joined #hudson
[09:34:14] *** davidstrauss_ has joined #hudson
[09:41:57] *** dvrzalik has joined #hudson
[09:54:23] *** davidstrauss has quit IRC
[09:58:07] *** davidstrauss has joined #hudson
[10:05:14] *** davidstrauss_ has quit IRC
[10:06:46] *** Lewisham has quit IRC
[10:09:30] *** matt-wynne has joined #hudson
[10:43:47] *** davidstrauss has quit IRC
[10:52:59] *** AgentIcarus has joined #hudson
[10:53:55] <AgentIcarus> golly a maxb
[10:53:59] <AgentIcarus> morning all
[11:34:35] *** trondn has joined #hudson
[11:34:56] <trondn> anyone online?
[11:38:46] *** trondn has quit IRC
[12:00:47] *** bobo has quit IRC
[12:01:51] *** akostadinov has quit IRC
[12:06:52] *** trondn has joined #hudson
[12:56:16] *** akostadinov has joined #hudson
[13:08:57] *** velo has joined #hudson
[13:10:28] *** bobo_ has joined #hudson
[14:40:25] *** abayer_ has joined #hudson
[14:41:52] *** jdolan has joined #hudson
[14:42:09] *** abayer_ has left #hudson
[14:57:11] *** davidstrauss has joined #hudson
[14:58:54] *** abayer has quit IRC
[15:16:47] *** kstreith has joined #hudson
[15:33:08] *** jieryn-w has joined #hudson
[15:37:14] *** mart_ has joined #hudson
[15:38:15] *** trondn has quit IRC
[16:09:37] <mart_> hi, are there any docs on how to get hudson to automatically install maven?
[16:09:46] *** mart_ is now known as mart
[16:10:10] <bobo_> hm, i just checked the box?
[16:10:29] <mart> and then...?
[16:10:35] <bobo_> its installed on demand
[16:10:42] <mart> so just save?
[16:10:45] <bobo_> yepp
[16:10:49] <bobo_> and build your project
[16:11:20] <bobo_> atleast that worked for jdk and ant, not sure i tried maven
[16:13:03] <mart> trying again...
[16:13:23] <mart> oh, ok.  worked this time.  probably just PEBKAC.
[16:14:04] *** akostadinov has quit IRC
[16:15:17] <AgentIcarus> I'm seeing some strange behaviour in my hudson now I've upgraded to 1.316
[16:15:31] <AgentIcarus> I've got a multimodule maven project (using a top-level aggregator pom)
[16:15:46] <AgentIcarus> and the sub projects depend on each other in some sane ways
[16:16:00] <AgentIcarus> when I build now some of my projects get built 5 times+
[16:16:14] <bobo_> mart: :-)
[16:16:22] <AgentIcarus> is this expected behaviour?
[16:30:32] *** timp has joined #hudson
[16:33:19] <mart> bobo_: d'oh, so tried it on our production hudson, and the install failed with java.io.FileNotFoundException: /.timestamp (Permission denied)
[16:33:56] <mart> no idea what it's trying to do with the filesystem root. :(
[16:34:27] <bobo_> eh? yeh that sounds weird
[16:35:02] <mart> ah, there was a MAVEN_HOME box that I didn't fill in.
[16:35:10] <mart> I guess that's where it's trying to unpack to
[16:35:59] <bobo_> no i think thats if you already have one installed
[16:36:23] <bobo_> but if its a debian/ubuntu you can just apt-get it and enter it in the box
[16:36:32] *** mindless has joined #hudson
[16:40:32] <mart> can't really use apt-get - I'm looking into issues that I suspect are caused by specific version combinations. :(
[16:43:35] <bobo_> ah =(
[16:43:46] <bobo_> although you can apt-get specific versions aswell i belive?
[16:45:01] <mart> the only thing I know of is specifying stable, testing or unstable - and if hudson'll install maven for me, I'm too lazy to figure out whether those releases have the versions I want :)
[16:54:42] <mart> ok... so the actual problem I'm having is this:
[16:54:46] <mart> I have a simple build with parent/reactor P, child A and child B, where B depends on A
[16:54:58] <mart> in a multi-module build.
[16:55:29] <mart> and when maven builds B, it's using the jar for A from its local repository, not from the workspace.
[16:56:02] <mart> so, I keep getting build failures, whenever I change an interface in A.
[16:57:51] <mart> I've verified this by using -X to show maven's debug output, and I an see from the 'classpathElements' in the log that it's using the wrong classpath
[16:58:04] <AgentIcarus> is this in hudson or command line?
[16:58:37] <mart> hudson.  command line always using the local repo, which is good.
[16:59:36] <mart> but somehow hudson interferes with installing A to the local repo, which means that when maven needs it to compile B, the old version is still there
[16:59:49] <AgentIcarus> interesting
[16:59:54] <AgentIcarus> are you building modules in parallel?
[16:59:59] <mart> nope.
[17:00:10] <mart> I'm using clean install deploy site site:deploy as goals in hudson.  is that normal?
[17:00:56] <AgentIcarus> well deploy will call install
[17:01:51] <mart> true, but I added install because it wasn't putting the jar in the repo at the correct time.
[17:02:15] <AgentIcarus> weird. your versions match (dependency, project)
[17:02:58] <mart> parent defines a snapshot version, both children inherit the same
[17:03:31] <AgentIcarus> your hudson config builds the top-level pom?
[17:03:53] <mart> yep
[17:05:00] *** rromanchuk has quit IRC
[17:06:23] <AgentIcarus> If you go to the Project > Modules (in the sidebar) what do you see?
[17:06:44] <mart> I see the two child projects, which is what I expect
[17:06:53] <AgentIcarus> under a parent one?
[17:07:03] <mart> yes, sorry.
[17:07:54] <AgentIcarus> if you click "A" is "B" listed as a downstream project?
[17:08:30] <mart> no.
[17:08:40] <mart> hmm. that's not good
[17:09:45] <AgentIcarus> did you set parents in the child projects?
[17:09:49] <AgentIcarus> (in the poms)
[17:10:04] <mart> yep.  the dependency is listed in B as having <version>${pom.version}</version>
[17:10:26] <mart> which is being inherited from the parent
[17:10:43] <mart> and neither A nor B specify a version explicitly
[17:10:51] <mart> (they just inherit it)
[17:10:55] <AgentIcarus> yeah
[17:10:58] *** akostadinov has joined #hudson
[17:11:08] <AgentIcarus> and they have <parent>s?
[17:12:23] <mart> sure, and it's the correct GAV
[17:12:54] <mart> that is, the <parent> GAV is the same GAV specified in the parent pom.
[17:13:04] <AgentIcarus> hmmm
[17:15:30] <mart> the parent pom itself declares a parent, but that can't be unusual.
[17:16:15] <AgentIcarus> nope
[17:16:48] *** rromanchuk has joined #hudson
[17:19:06] <mart> I wonder what happens if I remove the jars from the local repo, so it can't compile against the old one :)
[17:19:46] <mart> oh, I suppose it could still download them from nexus.
[17:23:51] <mart> which it duly does. :(
[17:27:50] <AgentIcarus> :(
[17:28:00] <AgentIcarus> I'm still getting stupidly large numbers of maven module builds
[17:29:29] *** akostadinov has quit IRC
[17:29:49] <mart> AgentIcarus: so, you see upstream/downstream links between the individual modules, as well as projects?
[17:29:54] <AgentIcarus> yup
[17:29:56] <AgentIcarus> well
[17:30:02] <AgentIcarus> I don't see them on projects per se
[17:30:11] <AgentIcarus> but all my modules have upstreams and downstreams
[17:30:22] <AgentIcarus> and after each one builds it seems to schedule all its downstreams
[17:30:26] <AgentIcarus> which is a pain
[17:30:34] <AgentIcarus> I may just have it misconfigured thought
[17:32:24] <AgentIcarus> as it means my tests fail as it builds stuff out of order
[17:32:25] <AgentIcarus> rargh
[17:33:47] <mart> did you try toggling the "Block build when dependency building" option?  that's the only thing I can see that looks relevant?
[17:33:58] <AgentIcarus> yeah, I've got that turned on
[17:34:48] <AgentIcarus> that's only the project itself
[17:34:50] <AgentIcarus> not the submodules
[17:35:04] <AgentIcarus> (I think)
[17:40:57] <AgentIcarus> and it can work out _some_ upstream projects (they block in the build queue going upstream project foo is building)
[18:01:29] <W_work> ok, I know I am being stupid, but this is just so weird I have to ask: Does Hudson modify ant on a machine it is installed on in some way that caches source files? I have a WIERD problem where a build has stopped working, apparently because one of the source files is being compiled with an old version, despite the workspace only containing the new version
[18:02:06] <W_work> I even tried checking out the files in a different directory on the machine and just calling "ant" and it gives the same error. Did the same thing on /my/ machine and it worked fine
[18:02:39] <W_work> what can I be missing?
[18:04:19] <W_work> (I am getting a "cannot find symbol" error from javac, on a symbol that IS defined in the relevant source file used. I tried javap on the jar file where that class is generated, and indeed the class file does not contain that symbol, despite the source file used to compile it *does*)
[18:05:55] <bobo_> W_work: so it only works on your machine?
[18:06:20] <W_work> no, it works on all machines except the hudson master
[18:06:43] <W_work> the original developer can check out and build fine as well
[18:07:37] <bobo_> even if you build from the commandline?
[18:07:37] <W_work> I'm gonna try to use my machine as a hudson slave and see if it builds in hudson using that
[18:07:41] <W_work> yes
[18:08:07] <W_work> as I said, really spooky problem, and of course I know it is outlandish and silly to think Hudson is to blame, but I just have to ask
[18:08:19] <bobo_> yeh i doubt its hudson
[18:08:26] <bobo_> i would guess path or something
[18:08:37] <bobo_> or something in ant/lib ?
[18:09:10] <W_work> heh I'll check - it sounds more likely that some retarded developer put the jar artifacts in ant/lib
[18:10:12] <bobo_> developers (including myself!) are always retarted =)
[18:10:52] <W_work> quite
[18:11:33] <mart> I'm not.  I'm amazing, me.
[18:11:38] <mart> ;)
[18:11:40] <bobo_> :-p
[18:12:29] <W_work> if developers were perfect, there'd be no need for integration in the first place, much less continuous integration :)
[18:16:18] *** akostadinov has joined #hudson
[18:28:33] *** khmarbaise has joined #hudson
[18:29:17] <W_work> ah, found the problem. developer had committed build directory
[18:29:32] *** akostadinov has quit IRC
[18:29:46] <W_work> svn checkout of course marked the class file with the timestamp of the checkout, and ant build then didn't recompile the class
[18:29:54] <W_work> I have *no* idea why it worked on my machine
[18:30:24] <W_work> perhaps something about the order things are checked out; whether it takes the source or build directory first (and thus sees the class as up to date or not)
[18:31:05] <W_work> gah half a day finding such a silly problem
[18:34:52] *** abayer_ has joined #hudson
[18:35:00] *** abayer_ is now known as abayer
[18:44:36] <abayer> heh. hudson/main can't actually build using the m2 project type's "build in parallel" option.
[18:44:51] <abayer> 'cos of an unspecified dependency on maven-plugin in war.
[18:56:04] <abayer> AgentIcarus - you can pick up Hudson 1.317-SNAPSHOT, which includes the fix for your bug report, at http://hudson.glassfish.org/job/hudson/13486/artifact/hudson/main/war/target/hudson.war
[19:03:50] *** mart_ has joined #hudson
[19:20:28] <jbriguet> ok guys, i installed hudson as a service, with no tomcat or thing like that, and it's working prefectly
[19:20:38] *** mart has quit IRC
[19:21:04] <jbriguet> (except the ssh slave thing that never achieved to update itself, had to do it by hand)
[19:21:18] <jbriguet> very nice job kohsuke :)
[19:21:41] <kohsuke> thanks. Glad to hear it worked.
[19:22:13] <jbriguet> now, i have to configure like 20 jobs :p
[19:22:22] *** mart has joined #hudson
[19:24:24] <W_work> yeah I just disabled the ssh slave plugin, I got annoyed at the update message
[19:25:00] <jbriguet> i took the hpi, extracted it in the right place (plugins), replaced the old hpi, and voila
[19:25:38] <W_work> JNLP slaves works fine for out purposes anyway
[19:31:54] *** jbriguet has quit IRC
[19:33:32] *** mart has quit IRC
[19:33:32] *** davidstrauss has quit IRC
[19:33:32] *** glass-eye has quit IRC
[19:33:32] *** mrooney has quit IRC
[19:34:13] *** mart has joined #hudson
[19:34:23] *** davidstrauss has joined #hudson
[19:34:23] *** glass-eye has joined #hudson
[19:34:23] *** mrooney has joined #hudson
[19:34:23] *** taril42 has joined #hudson
[19:34:27] *** jbriguet has joined #hudson
[19:37:48] <kohsuke> W_work: what's the update message?
[19:38:31] <W_work> kohsuke: oh, I guess I wasn't clear. The "some plugins have updates" on the management screen.
[19:38:37] <taril42> Is there an easy way to change the size (# lines or max kb) of Build Failed messages?  We're getting upwards of 400K mailouts due to the length of the command lines being echoed to console by make; I can't seem to find docs/option for this (my search-fu is weak today).
[19:38:38] <kohsuke> ah
[19:38:51] <kohsuke> W_work: I need to disable that for bundled plugins
[19:39:33] *** mart_ has quit IRC
[19:43:24] *** jbriguet has quit IRC
[19:43:24] *** velo has quit IRC
[19:44:46] *** jbriguet has joined #hudson
[19:45:32] *** velo has joined #hudson
[19:50:32] <W_work> what does Hudson actually do when you click the "restart Hudson" button after updating plugins? This is the second time now that it's shut down Tomcat. No message in the logs or anything, the process is just not there any more.
[19:50:48] <abayer> Odd. It's done great for me with tomcat5.
[19:51:37] <W_work> tomcat 5.5 here
[19:51:46] <abayer> Same here. What OS?
[19:52:32] <W_work> any idea how to find out what actually happened?
[19:52:39] <W_work> Ubunty Hardy in this case
[19:52:53] <abayer> Hmm again. I'm poking around the source to see what exactly it does now.
[19:52:56] <W_work> which is wierd as we've kind of standardized on CentOS
[19:53:13] <W_work> but that's neither here nor there
[19:56:13] *** kohsuke has quit IRC
[19:59:35] <abayer> Hrm. Looks like it supposedly only does anything with the restart if it's running as the executable war - which doesn't seem quite accurate to me.
[19:59:49] *** kohsuke has joined #hudson
[20:00:36] <mrooney> Wow are these suggestions real
[20:00:44] <abayer> ?
[20:01:16] <mrooney> all plug-ins should move to github? a closed-source, paid, out-of-our-control hosting solution sound great!
[20:01:20] <abayer> Oh, that.
[20:01:33] <abayer> Yeah, I started a long reply, then scuttled it.
[20:01:42] <mrooney> haha I did the same thing
[20:01:52] <mrooney> not sure what to say, really, without being flame-y
[20:02:04] <abayer> No point in arguing about something that won't actually happen. =)
[20:02:25] <abayer> People get...passionate about things like git.
[20:02:28] <abayer> Excessively so.
[20:04:06] *** davidstrauss has quit IRC
[20:06:03] <W_work> abayer: so it's something wrong in Hudson and not my Tomcat setup or anything?
[20:06:11] <abayer> kohsuke - about the "restart hudson now" button on plugin updates - UnixLifecycle seems like it only applies to executable-war instances of Hudson, but I'm running Hudson in tomcat5 and get that option, everything restarts, etc.
[20:06:21] <abayer> (W_work - I r confused. =) )
[20:06:35] <W_work> alright :)
[20:07:04] *** khmarbaise has quit IRC
[20:07:13] <kohsuke> It restarts the current proecss by execing to itself
[20:07:27] <kohsuke> So if the current process happens to be Tomcat with other webapps, they get restarted, too.
[20:07:55] <W_work> you missed out on my error; I press the button, Tomcat disappears, and does not restart
[20:08:30] <kohsuke> If you look at the Tomcat output file (wherever that is), that might have some clue
[20:08:43] <W_work> nope, not a line after the plugin update
[20:09:36] <jieryn-w> this is funny: http://dpaste.com/69602/
[20:10:22] <jieryn-w> my maven build was successful, but hudson build was not -- no indication why
[20:10:29] <kohsuke> if you run "java -jar hudson.war" from command line temporarily and try to restart, what would you see in the console?
[20:10:38] <kohsuke> (you can restart by going to http://server/hudson/restart)
[20:12:49] <W_work> well, it's all running behind an Apache proxy and stuff, but give me a sec and I'll figure it out
[20:13:32] <W_work> heh I get a 404 with a Hudson logo and winstone footer if I run the war file
[20:13:48] <W_work> probably hudson home directory not set or something
[20:15:54] <W_work> there we go
[20:15:57] <W_work> that works just fine
[20:16:25] <kohsuke> If you run Tomcat as "catalina.sh run" in the foreground and do restart, what happens?
[20:17:14] *** glass-eye has quit IRC
[20:17:14] *** mrooney has quit IRC
[20:18:25] *** glass-eye has joined #hudson
[20:18:27] *** ppawel has joined #hudson
[20:19:40] <W_work> that works fine as well
[20:20:06] <kohsuke> And now if you do "catalina.sh start" and do restart?
[20:20:10] <W_work> Jul 21, 2009 8:23:15 PM org.apache.catalina.core.AprLifecycleListener lifecycleEvent
[20:20:10] <W_work> INFO: The Apache Tomcat Native library which allows optimal performance in production environments was not found on the java.library.path: /usr/lib/jvm/java-6-sun-1.6.0.07/jre/lib/i386/server:/usr/lib/jvm/java-6-sun-1.6.0.07/jre/lib/i386:/usr/lib/jvm/java-6-sun-1.6.0.07/jre/../lib/i386:/usr/java/packages/lib/i386:/lib:/usr/lib
[20:20:10] <W_work> Jul 21, 2009 8:23:15 PM org.apache.coyote.http11.Http11BaseProtocol init
[20:20:10] <W_work> INFO: Initializing Coyote HTTP/1.1 on http-8080
[20:20:16] <W_work> oops that was a bit much
[20:21:27] <W_work> but yeah, start also works
[20:21:40] <W_work> nothing special output, but it does restart successfully
[20:21:49] <kohsuke> OK. So what is it that doesn't work?
[20:22:43] <W_work> tomcat started with /etc/init.d/tomcat55 start
[20:22:57] <W_work> if I *then* do the hudson restart, the tomcat dies without a message
[20:23:31] <kohsuke> Does that run Tomcat in a different user?
[20:23:34] <kohsuke> Maybe that's related
[20:23:34] <W_work> might be something to do with user levels? started as root but lowered or something?
[20:23:46] <kohsuke> Tomcat being Java, I don't think it does setuid
[20:23:48] <W_work> let me look at the init.d script for a sec
[20:24:22] <abayer> That could make sense - I run Tomcat as a non-root user all the way.
[20:25:08] <W_work> it uses jsvc
[20:25:23] <W_work> to run as the user "tomcat55" yes
[20:28:39] <W_work> guess I'll just stay away from that button :-}
[20:30:53] *** mrooney has joined #hudson
[20:34:06] <W_work> kohsuke: loved your Lego spheres by the way, and have gotten a new appreciation for how well-designed Lego really is. Did you know if you put a regular "1x1x1" Lego brick on the front of a "light-brick" (the ones with a stud on one side) it is the same width as 5 "flat" pieces? So I've made a diameter 7 (measured in "low height" units) lego "sphere" by basing it off 6 of the "light-bricks" :D
[20:35:02] <kohsuke> Yes, that's one of the important lego builder knowledge
[20:35:48] <W_work> I never knew these things as a kid, I just made clunky Lego rubber band guns
[20:41:32] *** ppawel has quit IRC
[20:52:26] *** alexlod has quit IRC
[20:52:39] *** alexlod has joined #hudson
[20:53:54] *** alexlod has quit IRC
[20:55:23] *** alexlod has joined #hudson
[21:01:32] *** matt-wynne has quit IRC
[21:02:45] *** statlor has joined #hudson
[21:02:45] *** jieryn has joined #hudson
[21:03:24] <statlor> wow lots of idlers today :D
[21:03:33] *** alexlod has quit IRC
[21:07:00] *** mobhawk has joined #hudson
[21:13:36] *** jieryn-w has quit IRC
[21:30:17] *** matt-wynne has joined #hudson
[21:33:36] *** matt-wynne has joined #hudson
[21:46:33] <statlor> http://community.java.net/projects/top.csp
[21:46:58] <statlor> we need some more members so we can beat out sip-communicator
[21:47:05] <statlor> I mean, sip-communicator? come on
[21:47:12] <statlor> we should totally be beating them
[21:47:17] <abayer> Or we could just...do something...about sip-communicator members.
[21:47:22] <statlor> haha
[21:47:35] <statlor> you've been hanging around rtyler too much
[21:47:49] <abayer> Nah, I was an evil, evil bastard long before I knew him. =)
[21:47:59] <bobo_> i like sip-communicator =(
[21:48:16] <statlor> I have nothing against it, it just needs to be crushed
[21:48:33] *** matt-wynne has quit IRC
[21:48:35] <statlor> also we need to pass glassfish in one of these categories
[21:48:37] <statlor> now that would be something
[21:49:16] <bobo_> why on earth is glassfish-theme at the top on hits?
[21:51:22] <statlor> I was just about to ask the same thing
[21:51:25] <statlor> the page looks worthless
[21:51:39] <bobo_> the project looks pretty worthless aswell
[21:51:41] <statlor> what, kohsuke is on there too
[21:51:45] <statlor> he's everywhere
[21:52:13] <statlor> of course I wanted to see screenshots of said themes
[21:53:24] <statlor> but perhaps I misunderstand what it is
[21:53:42] <statlor> I wonder if all the hits are of people thinking, "ooh themes!" and going there and finding nothing :)
[21:56:55] *** alexlod has joined #hudson
[22:00:16] <bobo_> :-)
[22:15:51] <kohsuke> glassfish-theme is a theme engine used for GlassFish community websites on java.net
[22:16:03] <kohsuke> If you see them, you'll notice that they all look the same
[22:16:30] *** kfu_ has joined #hudson
[22:16:31] <kohsuke> That's why the project gets a lot of hits --- every access to GlassFish projects on java.net loads some JavaScript and CSS from glassfish-theme
[22:21:00] <kfu_> hey anyone know why my OSX slave isnt working? I started the slave with the "javaws ....jnlp" line given in the 'manage nodes' section; i made a multi-conf job building on the OSX node, and gave a "Execute shell" build step of "/Applications/Chess.app/Contents/MacOS/Chess"
[22:21:48] <kfu_> but chess doesnt start on the slave when i build the job...
[22:25:54] *** jieryn-w has joined #hudson
[22:28:16] <kfu_> nothing in the main sysout
[22:29:25] <kfu_> tried running the slave command on the osx box using 'sudo
[22:30:06] *** tom7 has joined #hudson
[22:30:46] <tom7> kohsuke: you there ?
[22:30:51] <kohsuke> yes
[22:31:10] <tom7> have you used @LocalData much ?
[22:31:25] *** dvrzalik has quit IRC
[22:31:53] <tom7> i've got some test errors using it from time to time, a RejectedExecutionException during load
[22:32:17] <tom7> perhaps because Hudson.threadPoolForLoad is static, and is shutdown after the test ?
[22:36:13] <statlor> ahh makes sense now (glassfish-theme)
[22:36:31] *** jieryn has quit IRC
[22:39:20] <statlor> couple of questions for committers.. when I commit with an issue should I do [HUDSON-xxxx] and [FIXED HUDSON-xxxx] ? I think I just did FIXED HUDSON-xxxx (no brackets) and the comment showed up but it did not mark the issue as fixed
[22:39:32] <statlor> or is it supposed to mark the issue as fixed automatically?
[22:40:53] <kohsuke> statlor: it should be a bracket
[22:41:00] <kohsuke> tom7: I don't remember what is @LocalData
[22:41:04] <statlor> ok
[22:41:29] <tom7> the test recipe to have a configured hudson home when your test starts
[22:41:36] <statlor> kohsuke is 4045 going into the next release? I think it should :) it was a huge problem for me
[22:41:55] <statlor> and it was hitting everyone with accurev + slaves
[22:42:07] <tom7> kohsuke: i'm using this to test loading existing jobs
[22:42:16] <kfu_> hey, is there anyway to put the slave in verbose mode
[22:42:17] <statlor> 4045 is the memory leak with ExportTable
[22:42:46] <kohsuke> Yes, 4045 will be in 1.317. I should have noted that in the issue.
[22:43:13] <statlor> I'm running on a snapshot I built and it's working great now
[22:43:21] <kohsuke> tom7: could be due to thread shutdown. Do you have a stack trace?
[22:43:27] <statlor> my master wouldn't last 12 hours without it
[22:44:08] <statlor> at some point I'm going to add a "# jobs" indicator at the bottom of each view :)
[22:44:15] <statlor> should have paper cutted that one
[22:44:17] <tom7> kohsuke: http://pastebin.com/m45ca67af
[22:44:43] <statlor> I have too many jobs to count now
[22:46:13] <kohsuke> tom7: yeah, I do wonder how this works
[22:46:23] <kohsuke> I mean, Hudson.cleanUp is shutting down this pool but no one is recreating it
[22:46:30] <tom7> kohsuke: indeed
[22:46:43] <tom7> kohsuke: it probably works in fork mode if you do only one test per class
[22:47:32] <tom7> now, I tried to just remove the static from that field
[22:47:55] <tom7> but now I get "second instance" errors
[22:47:59] <tom7> still debugging that
[22:48:01] <kohsuke> forkMode is once, so Hudson does reuse the same JVM to run multiple tests
[22:48:26] <tom7> that depends on which forkmode
[22:48:34] <tom7> yes, I usually changed it to perTest
[22:49:49] *** davidstrauss has joined #hudson
[22:51:14] <tom7> probably need to make it transient...
[22:54:21] *** matt-wynne has joined #hudson
[22:57:28] *** matt-wynne has quit IRC
[23:15:38] *** gits has joined #hudson
[23:16:06] <rtyler> kohsuke: git-svn, there be monsters
[23:16:22] <kohsuke> you think it's too clunky?
[23:16:35] <rtyler> I *know* it's too clunky
[23:16:42] * rtyler points to link in email he just replied
[23:17:57] <kohsuke> In my experiments, I was able to have a lattice on git side as long as I only dcommit from one git branch.
[23:18:16] <rtyler> that doesn't work for more than one person
[23:18:20] <rtyler> or if you start to create branches
[23:18:29] <rtyler> unless you squash commit your merge when you merge into your lcoal master
[23:19:37] <kohsuke> It's very possible that I missed something, but people can create branches all they want --- they just have to go through a tracking branch to commit changes back to Subversion.
[23:19:55] <rtyler> right, and they have to squash
[23:20:03] <rtyler> git-svn and svn can't cope with non-linear history
[23:20:53] <kohsuke> That's the part I don' get --- git-svn only records one change when a big branch gets merged to the tracking branch, and I think we are fine with that.
[23:21:13] <kohsuke> OK, so you must have been trying to preserve all the commits in all the branches.
[23:21:22] <kohsuke> ... which is necessary for it to be true bi-di.
[23:22:15] *** matt-wynne has joined #hudson
[23:23:17] *** lynggaardDK has joined #hudson
[23:24:11] <lynggaardDK> I have got a weird problem with my hudson install, I cannot configure jobs as the page loading stops "mid-page"
[23:24:29] <kohsuke> any error reported by server on its console output?
[23:24:36] <lynggaardDK> how to debug ? or has anyone seen this before?
[23:24:47] *** matt-wynne has quit IRC
[23:24:52] <mobhawk> Getting "It seems that ruby gems is not installed" trying to configure Rake plugin.
[23:26:55] *** matt-wynne has joined #hudson
[23:27:07] <lynggaardDK> kohsuke: not that I can see, have looked both in glassfish domain/logs/server.log and in the nexus logs (nexus is installed in same glassfish)
[23:28:40] *** velo has quit IRC
[23:28:46] <lynggaardDK> last thing I see in the view source is: "<td class="setting-description">            Specify a regular expression using the syntax supported by            the Java <a href='http://java.sun.com/j2se/1.5.0/docs/api/java/util/regex/Pattern.html'>Pattern<a> class.            </td></tr><tr><td class="setting-leftspace">" which might  be from task finder plugin
[23:29:50] <kohsuke> I really think the server must be reporting an error somewhere
[23:29:59] <kohsuke> Otherwise it won't just stop the processing in the middle
[23:31:02] <kohsuke> maybe you can try "java -jar hudson.war" for a while which is known to produce output to console.
[23:31:02] <statlor> maven nexus you mean?
[23:31:14] <kohsuke> Just so that we can get GlassFish out of the picture for the sake of troubleshooting
[23:31:22] <lynggaardDK> I think I have found something... will restart and see if the entry comes back after clearing the logs:
[23:31:30] <lynggaardDK> 2009-07-21 23:30:10 StandardWrapperValve[Stapler]: PWC1406: Servlet.service() for servlet Stapler threw exception
[23:31:30] <lynggaardDK> org.apache.commons.jelly.JellyTagException: file:/opt/glassfish-2.1/domains/domain1/generated/jsp/j2ee-modules/hudson_1.316/loader/lib/form/hetero-list.jelly:66:76: <st:include> No page found 'config.jelly' for class hudson.plugins.claim
[23:31:30] <lynggaardDK> .ClaimPublisher$DescriptorImpl
[23:31:30] <statlor> er sonatype nexus
[23:31:52] *** trondn has joined #hudson
[23:31:55] <lynggaardDK> statlor: yes sonatype nexus 1.3.4
[23:32:04] <statlor> I love the nexus :)
[23:32:37] <trondn> Hi.. anyone familiar with the sloccount plugin??? It would be nice if I could get a trend graph on the "front page" of my project... anyone knows if that is a lot of work to implement?
[23:32:56] <lynggaardDK> I has a habit of stealing log output, and you have to hack it on glassfish. something about recursive calls in the logging frameworks and consoleappender
[23:34:36] <lynggaardDK> statlor: nexus is very good, but still have some annoying defects  I hope gets fixed soon
[23:35:51] <statlor> you will like hudson even more :) assuming you can get it to work of course
[23:38:13] <lynggaardDK> kohsuke: while I wait for my "server" to restart, I have a plugin question, how easy would it be to develop a plugin or functionality which could "obsolete" a job... it should forget/delete all arfifacts except for the last successfull build, but keep fingerprints/build logs etc
[23:39:06] <kohsuke> that feature is already available
[23:39:22] *** gits has quit IRC
[23:39:35] <lynggaardDK> statlor: I do love it, and we use it happily at my work. It is just not as updated as my home machine
[23:41:18] <lynggaardDK> kohsuke: where is that functionality ? all I can find it "discard old builds"?
[23:41:35] <kohsuke> look under the artifact archiver
[23:42:36] <lynggaardDK> kohsuke: ok will do... I now have a log statement from the server, for the half loading configure page: http://hudson.pastebin.com/d42928dd0
[23:42:41] *** gits has joined #hudson
[23:42:42] *** matt-wynne has quit IRC
[23:43:12] *** admc has joined #hudson
[23:43:35] <kohsuke> Either the claim plugin is broken or you had some file system related issue.
[23:43:49] <statlor> hey, I've got an issue open to expose several accurev-specific pieces of info as environment variables to build scripts.. what would be a good place for me to look to see how this is done currently?
[23:44:45] <kohsuke> SCM.buildEnvVars
[23:45:08] <statlor> thanks! headed home bbl
[23:45:23] *** statlor has quit IRC
[23:47:00] *** lynggaardDK1 has joined #hudson
[23:47:34] <lynggaardDK1> (crashed, gotta love the new OSS ati drivers ;-))
[23:47:50] *** lynggaardDK has quit IRC
[23:48:33] *** trondn has left #hudson
[23:49:13] <kfu_> i gonna run the master on OSX i think
[23:50:06] <kfu_> lets see if the jobs will start
[23:55:22] *** jieryn-w has quit IRC
[23:55:37] *** kfu_ has quit IRC
[23:57:58] *** mindless has quit IRC

top