Displaying #maven-dev/2017-03-18.log:

Sat Mar 18 09:59:04 2017  rfscholte:Joined the channel
Sat Mar 18 10:46:01 2017  hboutemy:Joined the channel
Sat Mar 18 11:07:38 2017  rfscholte:Joined the channel
Sat Mar 18 11:46:38 2017  stephenc:rfscholte: how are you doing?
Sat Mar 18 11:47:42 2017  stephenc:paging rfscholte
Sat Mar 18 11:48:08 2017  rfscholte:stephenc: good (apart from ping on several issues ;) )
Sat Mar 18 11:48:28 2017  stephenc:is MNG-6176 good to merge once the CI passes?
Sat Mar 18 11:48:46 2017  stephenc:(actually as it shouldn't affect ITs we can probably merge early)
Sat Mar 18 11:48:47 2017  rfscholte:just pushed it, tested it with JDK7 and JDK8
Sat Mar 18 11:49:13 2017  rfscholte:I can merge it right now if you want
Sat Mar 18 11:49:32 2017  stephenc:I'm just thinking that we might be able to get the release staged now
Sat Mar 18 11:49:45 2017  stephenc:or do you know if anything else is pending
Sat Mar 18 11:50:23 2017  stephenc:rfscholte: how many javadoc errors in maven-compat?
Sat Mar 18 11:51:29 2017  stephenc:because I'd be happy to just fix those so we don't need the profile
Sat Mar 18 11:51:44 2017  rfscholte:don't know. Just don't want to spend time on it, as all those classes should be treated as removed
Sat Mar 18 11:51:58 2017  stephenc:ok well I'll fix them
Sat Mar 18 11:52:16 2017  rfscholte:did MNG-6182 got reverted?
Sat Mar 18 11:53:06 2017  stephenc:I'm not sure
Sat Mar 18 11:53:12 2017  rfscholte:and I hope hboutemy will pick up MNG-6188 ;) but not for alpha-2
Sat Mar 18 11:53:40 2017  stephenc:if it is not in alpha-2 and it is anything other than a bug fix, we'll need a beta-1
Sat Mar 18 11:53:52 2017  stephenc:in fact if we need a bug fix, we'll need a beta-1
Sat Mar 18 11:55:29 2017  hboutemy:I don't understand MNG-6188: both what happens and what can be done against it
Sat Mar 18 11:56:16 2017  rfscholte:what I did is rerun previous command, with a -X
Sat Mar 18 11:56:38 2017  rfscholte:I didn't want debug logging, so I did CTRL+C during the process
Sat Mar 18 11:56:55 2017  rfscholte:and now the default color was mint :(
Sat Mar 18 11:57:20 2017  rfscholte:I think JANSI should be unregistered in a finally block
Sat Mar 18 11:59:40 2017  hboutemy:doing this is easy: see MavenCli
Sat Mar 18 12:00:10 2017  hboutemy:but I don't really know if the finally block will be called when doing CTRL-C
Sat Mar 18 12:01:11 2017  hboutemy:is the issue easy to reproduce?
Sat Mar 18 12:01:29 2017  hboutemy:since I suppose the console does not stay much in a colorized state...
Sat Mar 18 12:01:47 2017  rfscholte:let me try
Sat Mar 18 12:01:53 2017  hboutemy:then you didn't have much luck when you did CTRL-C
Sat Mar 18 12:04:03 2017  hboutemy:stephenc: honestly, if we could call the release beta-1 instead of alpha-2, I'm sure we would have more testers
Sat Mar 18 12:04:39 2017  hboutemy:and even beta-1 will make people test less than if it was about milestone-1 or rc-1
Sat Mar 18 12:04:57 2017  stephenc:hboutemy: ok we'll call this beta-1
Sat Mar 18 12:04:58 2017  hboutemy:just a question of advertising
Sat Mar 18 12:05:25 2017  stephenc:rfscholte: I think I have the javadoc errors fixed now
Sat Mar 18 12:05:35 2017  stephenc:I'll just push on top of your branch if that's OK
Sat Mar 18 12:05:43 2017  rfscholte:sure, go ahead
Sat Mar 18 12:05:57 2017  stephenc:just checking that it works on both java7 and java 8
Sat Mar 18 12:06:37 2017  stephenc:java 8 fine
Sat Mar 18 12:06:43 2017  rfscholte:be aware to run site -Preporting, so both javadoc and test-javadoc are called
Sat Mar 18 12:07:01 2017  stephenc:yep
Sat Mar 18 12:07:08 2017  stephenc:just getting the main code first
Sat Mar 18 12:08:29 2017  stephenc:rfscholte: though I just do mvn clean install javadoc:jar javadoc:test-jar -DskipTests
Sat Mar 18 12:08:44 2017  hboutemy:rfscholte: adding the finally block is safe: you can do it, hoping it will improve the situation
Sat Mar 18 12:08:52 2017  hboutemy:at least, we're sure it won't get worse ;)
Sat Mar 18 12:08:58 2017  rfscholte:hboutemy: I agre
Sat Mar 18 12:08:59 2017  rfscholte:e
Sat Mar 18 12:09:49 2017  hboutemy:http://stackoverflow.com/questions/11435533/how-does-ctrl-c-work-with-java-program
Sat Mar 18 12:11:23 2017  hboutemy:http://stackoverflow.com/questions/4911745/if-i-type-ctrl-c-on-the-command-line-will-the-finally-block-in-java-still-execu
Sat Mar 18 12:11:40 2017  hboutemy:shutdown hooks seem to be the solution: finally is not sufficient
Sat Mar 18 12:12:50 2017  hboutemy:notice that I don't know if uninstalling Jansi does reset console color
Sat Mar 18 12:15:20 2017  rfscholte:it's not that easy to reproduce, it is a matter of luck+timing. But anything to help prevent this would be good
Sat Mar 18 12:15:51 2017  stephenc:ok joy!!!
Sat Mar 18 12:16:19 2017  stephenc:[INFO] Generating "FindBugs" report --- findbugs-maven-plugin:2.5.5:findbugs
Sat Mar 18 12:16:19 2017  stephenc:[INFO] Locale is en
Sat Mar 18 12:16:19 2017  stephenc:[INFO] Fork Value is true
Sat Mar 18 12:16:19 2017  stephenc: [java] The following errors occurred during analysis:
Sat Mar 18 12:16:20 2017  stephenc:so cannot build the site with java 8
Sat Mar 18 12:16:21 2017  stephenc: [java] Unable to get XClass for java/lang/StringBuilder
Sat Mar 18 12:16:41 2017  hboutemy:notice shutdown hooks are deprecated
Sat Mar 18 12:16:41 2017  rfscholte::) I added -Dfindbugs.skip locally
Sat Mar 18 12:18:24 2017  stephenc:rfscholte, I may see if I can upgrade past
Sat Mar 18 12:21:14 2017  stephenc:we are going to have to drop support for java 7 soon IMHO
Sat Mar 18 12:21:14 2017  stephenc:ok so findbugs just upgrade to 3.0.4
Sat Mar 18 12:21:14 2017  stephenc:hboutemy: where does the version for the site come from now... pluginManagement?
Sat Mar 18 12:21:15 2017  stephenc:crosses fingers
Sat Mar 18 12:23:05 2017  hboutemy:stephenc: ?
Sat Mar 18 12:23:28 2017  stephenc:if I override the version by adding a pluginManagement will that be picked up by the sire
Sat Mar 18 12:23:31 2017  stephenc:s/sire/site
Sat Mar 18 12:23:34 2017  hboutemy:you mean the m-site-p in the pom.xml?
Sat Mar 18 12:23:40 2017  stephenc:yep
Sat Mar 18 12:23:48 2017  rfscholte:stephenc: drop java7 only if usage of toolchains improves. I still see too many using Maven Runtime as JDK to compile
Sat Mar 18 12:23:50 2017  hboutemy:looking
Sat Mar 18 12:23:59 2017  stephenc:I want to force findbugs to 3.0.4
Sat Mar 18 12:24:34 2017  stephenc:obviously we should bump in maven-parent, but cannot yet as lots of plugins are still java 6
Sat Mar 18 12:24:37 2017  hboutemy:as any other plugin, its version should eb defined in pluginManagement
Sat Mar 18 12:25:16 2017  stephenc:ok, I'll add the override then
Sat Mar 18 12:26:27 2017  hboutemy:found: it's set in apache paretn in pluginManagement
Sat Mar 18 12:26:40 2017  hboutemy:I don't undestand why you want to change the m-site-p version
Sat Mar 18 12:27:00 2017  hboutemy:how is it related to java version for a plugin?
Sat Mar 18 12:27:00 2017  stephenc:I want to force findbugs to upgrade from 2.5.5 to 3.0.4
Sat Mar 18 12:27:17 2017  stephenc:that's coming from maven-parent:27
Sat Mar 18 12:27:23 2017  hboutemy:not related to m-site-p version
Sat Mar 18 12:27:30 2017  stephenc:no
Sat Mar 18 12:27:45 2017  stephenc:I was asking when the site plugin runs, will it pick up the reporting plugin version from plugin management
Sat Mar 18 12:27:57 2017  stephenc:because for the old site plugin the answer was no
Sat Mar 18 12:28:27 2017  hboutemy:currently, m-site-p is 3.4
Sat Mar 18 12:28:42 2017  stephenc:and that picks up reporting plugin versions from pluginMgmt
Sat Mar 18 12:28:46 2017  hboutemy:I don't remember if pluginManagement is used for reporting plugins in this version
Sat Mar 18 12:28:48 2017  stephenc:ok, it's good
Sat Mar 18 12:40:21 2017  stephenc:rfscholte/hboutemy: so https://builds.apache.org/view/Maven/job/maven-3.x-jenkinsfile/job/MNG-6176/ build #3 is onto the ITs
Sat Mar 18 12:40:39 2017  stephenc:I'm going to merge to master as javadoc comments are not going to affect ITs
Sat Mar 18 12:40:45 2017  stephenc:any objections?
Sat Mar 18 12:40:53 2017  stephenc:waits 2 min
Sat Mar 18 12:41:39 2017  hboutemy:seems reasonable
Sat Mar 18 12:42:16 2017  stephenc:finished waiting
Sat Mar 18 12:46:50 2017  rfscholte:go ahead
Sat Mar 18 12:49:11 2017  stephenc:rfscholte / hboutemy from you PoV is there anything else left to land in master
Sat Mar 18 12:49:52 2017  stephenc:or will I wait until monday
Sat Mar 18 12:50:08 2017  rfscholte:MNG-6182 is still there, so it seems
Sat Mar 18 12:50:48 2017  rfscholte:that's the only one left for me
Sat Mar 18 12:54:03 2017  stephenc:I do wonder about that one
Sat Mar 18 12:56:08 2017  stephenc:So in https://git-wip-us.apache.org/repos/asf?p=maven.git;a=commitdiff;h=ab800b0cfae4e3ca9453304e3b9727ba4a4b712b it seems he is adding at least "claimed working" impls
Sat Mar 18 12:58:29 2017  stephenc:that is at least *less* concerning to me
Sat Mar 18 12:58:57 2017  stephenc:but I am not seeing any unit test
Sat Mar 18 13:03:20 2017  rfscholte:what triggered me was the unclear comment: [MNG-6182] ModelResolver interface enhancements.
Sat Mar 18 13:04:23 2017  stephenc:I've responded on the M/L
Sat Mar 18 13:04:36 2017  stephenc:I'm ok merging "dead" code if it has a test
Sat Mar 18 13:04:51 2017  stephenc:so that ide integrators can start using it ahead of 3.5.1
Sat Mar 18 13:05:15 2017  stephenc:I initially thought it was merging code with the signature changes and just a "throw new Unimplemented()"
Sat Mar 18 13:05:39 2017  stephenc:the change here is to make the API available, but we are not using it internally
Sat Mar 18 13:05:55 2017  stephenc:which seems somewhat reasonable... if risky
Sat Mar 18 13:06:15 2017  stephenc:add a unit test - no matter how crap - and we at least have coverage
Sat Mar 18 13:06:31 2017  stephenc:OK, I'll give until monday for the MNG-6182 situation to be resolved
Sat Mar 18 13:06:37 2017  stephenc:after that the train is leaving the station
Sat Mar 18 13:06:44 2017  rfscholte:ok
Sat Mar 18 13:08:53 2017  stephenc:ok, family time now
Sat Mar 18 13:09:02 2017  stephenc:waves bye for now
Sat Mar 18 13:21:55 2017  hboutemy:stephenc: rfscholte: what hurts me his always claims about "fixing", "enhancing" = judgements, but no facts = "adding xxx" or "changing yyy to zzz"
Sat Mar 18 13:23:01 2017  hboutemy:but I'm like you: looks here like adding dead code (that is expected to be used in future releases)
Sat Mar 18 17:22:03 2017  tibor_:Joined the channel
Sat Mar 18 18:00:42 2017  hboutemy:Joined the channel
Sat Mar 18 18:06:48 2017  jasonvanzyl:Joined the channel
Sat Mar 18 22:46:01 2017  olamy:Joined the channel

Comments