Displaying #maven-dev/2017-11-11.log:

Sat Nov 11 10:21:46 2017  tibor_:Joined the channel
Sat Nov 11 11:53:01 2017  rfscholte:Joined the channel
Sat Nov 11 12:59:41 2017  hboutemy:Joined the channel
Sat Nov 11 16:03:40 2017  tibor_:Joined the channel
Sat Nov 11 19:09:09 2017  stephenc:hboutemy: rfscholte: so I am planning rounding out my video series with a retrospective on-air hangout
Sat Nov 11 19:09:49 2017  stephenc:It would be about 30min long and about a week or two after the series is “finished”
Sat Nov 11 19:10:08 2017  stephenc:I want to let the Maven project have time to put it into practice
Sat Nov 11 19:10:31 2017  stephenc:It would be great if one or both of you could take part
Sat Nov 11 19:10:39 2017  stephenc:Are with of you game?
Sat Nov 11 19:11:03 2017  stephenc:(I’m trying to recruit aheritier as moderator)
Sat Nov 11 19:11:42 2017  stephenc:Probably some time early December
Sat Nov 11 19:12:51 2017  stephenc:Also I’d like 1-2 people from infra
Sat Nov 11 19:13:22 2017  stephenc:And I’m going to try and get 1-2 from Jenkins community
Sat Nov 11 19:14:00 2017  rfscholte:I have 2x 2hours presentations I need to build up from scratch, so not sure if I can really add value to the retro.
Sat Nov 11 19:23:06 2017  stephenc:rfscholte: it would be ~30min in 2-3 weeks time
Sat Nov 11 19:24:32 2017  stephenc:I have 4 aims from the series:
Sat Nov 11 19:24:44 2017  stephenc:1. To give Maven better use of Jenkins
Sat Nov 11 19:25:07 2017  stephenc:2. To provide infra with a better git integration
Sat Nov 11 19:25:20 2017  stephenc:3. To help people write Jenkins plugins
Sat Nov 11 19:25:57 2017  stephenc:4. For me to actually produce an updated tutorial and fight my perfectionist tendencies
Sat Nov 11 19:26:22 2017  stephenc:I can cover #4
Sat Nov 11 19:26:26 2017  rfscholte::)
Sat Nov 11 19:26:41 2017  stephenc:And I’m looking for some people to help assess the other three
Sat Nov 11 19:27:53 2017  stephenc:hboutemy is an ideal person... in fact I had the idea for the series because Hervé wasn’t fully understanding my vision from a 4 line email
Sat Nov 11 19:28:44 2017  stephenc:But you could add value too as someone looking at the end product
Sat Nov 11 19:29:11 2017  rfscholte:sure, I can help with that
Sat Nov 11 19:30:55 2017  stephenc:Thanks
Sat Nov 11 19:32:03 2017  stephenc:I’ll see about arranging a time once I get some more bodies ;-)
Sat Nov 11 19:32:07 2017  stephenc:If you have contacts in infra that would be interested in taking part, please let me know
Sat Nov 11 19:32:21 2017  stephenc:S/would/might/
Sat Nov 11 19:32:34 2017  rfscholte:olamy: ^^?
Sat Nov 11 19:33:18 2017  stephenc:TZ could get tricky for olamy
Sat Nov 11 19:33:34 2017  rfscholte:that's what I was thinking... :D
Sat Nov 11 19:34:07 2017  stephenc:Perhaps humdeb...
Sat Nov 11 19:34:22 2017  stephenc:I think he’s behind the gitbox stuff
Sat Nov 11 19:35:01 2017  stephenc:Gavin would be ideal also... but same TZ issues as loamy
Sat Nov 11 19:35:14 2017  stephenc:curses autocorrect
Sat Nov 11 19:36:19 2017  stephenc:Perhaps if I pick Jenkins community people in EU, we could find a time that works for olamy/gavin...
Sat Nov 11 19:36:19 2017  stephenc:Hmmm... thinking
Sat Nov 11 19:36:19 2017  stephenc:has to go out now
Sat Nov 11 19:36:25 2017  stephenc:Thanks!
Sat Nov 11 19:36:42 2017  rfscholte:np, cu
Sat Nov 11 20:50:28 2017  hboutemy:looks a good idea
Sat Nov 11 20:50:59 2017  hboutemy:I'm interested to try
Sat Nov 11 21:13:54 2017  rfscholte:hboutemy: Hi Hervé. Would be nice to see Doxia released soon, so m-site-p works with Java9. Any chance to have a look at it?
Sat Nov 11 21:15:31 2017  hboutemy:I'm trying to have some commits for MPDF plugin
Sat Nov 11 21:15:41 2017  hboutemy:but it does not go as expected...
Sat Nov 11 21:16:15 2017  hboutemy:https://issues.apache.org/jira/browse/MPDF-81
Sat Nov 11 21:16:32 2017  hboutemy:if you can add a comment, perhaps it will put sane pressur on Alex
Sat Nov 11 21:17:12 2017  hboutemy:it would be stupid not to have anything from this in Doxia 1.8
Sat Nov 11 21:34:46 2017  rfscholte:let me see what I can do
Sat Nov 11 21:57:16 2017  hboutemy:rfscholte: did you see, I published plexus-language site
Sat Nov 11 21:57:29 2017  hboutemy:(and wrote instructions on how to init such a site)
Sat Nov 11 22:16:41 2017  stephenc:hboutemy: great to hear you’re interested... given I started the series for you, it wouldn’t be a retrospective without you ;-)
Sat Nov 11 22:16:49 2017  hboutemy::)
Sat Nov 11 22:17:19 2017  stephenc:I’m thinking some time the first week of December
Sat Nov 11 22:17:19 2017  hboutemy:I'm eager to test the result
Sat Nov 11 22:17:46 2017  stephenc:I should have something we can install by ep9
Sat Nov 11 22:17:48 2017  hboutemy:then to work on plugins+shared git migration...
Sat Nov 11 22:18:00 2017  stephenc:I’ve ep7 recorded already
Sat Nov 11 22:18:37 2017  hboutemy:ep6 is not published yet, isn't it?
Sat Nov 11 22:18:41 2017  stephenc:Just waiting on our marketing dept to publish them... but I’ll record 8 & 9 this week
Sat Nov 11 22:19:00 2017  stephenc:Ep6 is mostly preparation
Sat Nov 11 22:19:21 2017  stephenc:Ep7 is me battling with html scraping and date parsing
Sat Nov 11 22:19:27 2017  hboutemy::)
Sat Nov 11 22:19:38 2017  stephenc:The two least favourite tasks in software eng
Sat Nov 11 22:20:00 2017  olamy:december?? that’s summer here so all aus guys are in the beach!!
Sat Nov 11 22:20:03 2017  olamy::-)
Sat Nov 11 22:20:16 2017  hboutemy:lucky guy: it's raining here :/
Sat Nov 11 22:20:21 2017  stephenc:I have more html scraping to do in 8 & 9 but it’s just full in the blanks
Sat Nov 11 22:20:35 2017  olamy:haha still a bit chill ATM only 28
Sat Nov 11 22:20:45 2017  hboutemy:grrrr....
Sat Nov 11 22:20:46 2017  stephenc:Who would be good to invite from infra
Sat Nov 11 22:20:56 2017  hboutemy:Gavin, for sure
Sat Nov 11 22:21:06 2017  hboutemy:he's managing Jenkins a lot
Sat Nov 11 22:21:19 2017  stephenc:Ok, that probably means I need to keep EU and not US
Sat Nov 11 22:21:20 2017  hboutemy:Olivier, are you still involved in Jenkins maintenance?
Sat Nov 11 22:21:50 2017  stephenc:Olivier i’ll ask bobby and Daniel from the Jenkins side
Sat Nov 11 22:22:31 2017  stephenc:If you want to join, given I’d like gavin to partake, it shouldn’t be a ridiculous time for you
Sat Nov 11 22:22:58 2017  olamy:hboutemy not really
Sat Nov 11 22:23:03 2017  stephenc:hboutemy: do you know is it humdeb... who’s behind gitbox?
Sat Nov 11 22:23:07 2017  olamy:I thing Andrew is still involbed
Sat Nov 11 22:23:21 2017  hboutemy:Andrew is silent these days
Sat Nov 11 22:23:26 2017  olamy:%s/thing/think
Sat Nov 11 22:23:34 2017  hboutemy:I don't really know who worked on GitBox
Sat Nov 11 22:23:47 2017  stephenc:I can prod Andrew if necessary... but he’s Boston
Sat Nov 11 22:23:48 2017  olamy:Gav is the guy to talk with
Sat Nov 11 22:24:49 2017  stephenc:I’d like to get feedback from Gav... plus whoever is behind gitbox (as that’s what I’m scraping)
Sat Nov 11 22:26:12 2017  hboutemy:https://github.com/apache/infrastructure-puppet/commits/deployment/modules/gitbox
Sat Nov 11 22:27:33 2017  hboutemy:Gav + Humbedoo + a little bit of pono
Sat Nov 11 22:28:00 2017  stephenc:Ok, I’ll reach out to gav & hum
Sat Nov 11 22:28:13 2017  stephenc:Gav’s in Australia iirc?
Sat Nov 11 22:29:11 2017  hboutemy:I don't know
Sat Nov 11 22:30:02 2017  stephenc:I think he is, at least he was eating dinner with his family during morning time for me a week or two ago
Sat Nov 11 22:31:03 2017  olamy:Gav is in Melbourne
Sat Nov 11 22:31:09 2017  stephenc:Ack
Sat Nov 11 22:31:53 2017  stephenc:Olivier, do you want to join the party... though too many and it may be hard to keep under 35min
Sat Nov 11 22:32:31 2017  stephenc:And obviously there’s the risk that I start ragging on the evil plugin ;-)
Sat Nov 11 22:32:55 2017  hboutemy:ok, I'm not available, finally.. ;)
Sat Nov 11 22:33:17 2017  stephenc:Ha ha!
Sat Nov 11 22:33:23 2017  hboutemy:notice, while at it
Sat Nov 11 22:33:51 2017  hboutemy:it would be nice to have a look at why triggering on SNAPSHOT update is not feasible
Sat Nov 11 22:34:08 2017  stephenc:Ahhh...
Sat Nov 11 22:34:25 2017  stephenc:So I have a spec on how to do snapshot triggers
Sat Nov 11 22:34:27 2017  hboutemy:for a new series of episodes
Sat Nov 11 22:34:46 2017  stephenc:I wrote it in Feb 2015
Sat Nov 11 22:35:08 2017  stephenc:But our PMs have not priority raised it
Sat Nov 11 22:35:29 2017  hboutemy:I'm not sure it is well understood
Sat Nov 11 22:35:34 2017  stephenc:Full spec, covering not just Maven but NPM, etc
Sat Nov 11 22:36:06 2017  hboutemy:sure, the interest is not to have a Maven-specific hack
Sat Nov 11 22:36:19 2017  stephenc:The current evil plugin trigger is “ugh”
Sat Nov 11 22:36:48 2017  stephenc:To do the trigger correctly, you need to model where artifacts are published
Sat Nov 11 22:37:07 2017  hboutemy:Maven specific hack, I suppose badly hacked with low-level Maven internals
Sat Nov 11 22:37:42 2017  stephenc:Eg, if I do mvn install, that should trigger jobs that run on that agent and do not use an isolationist local repo
Sat Nov 11 22:38:08 2017  stephenc:Evil plugin doesn’t even understand what is going on
Sat Nov 11 22:38:08 2017  hboutemy:oh, you add this level of precision
Sat Nov 11 22:38:17 2017  stephenc:In my spec, yes
Sat Nov 11 22:38:47 2017  hboutemy:sure, while at it, you have to provide a real model for artifact publication/consumption
Sat Nov 11 22:38:50 2017  stephenc:Also deployment to staging repo by modelling the repo visibility
Sat Nov 11 22:39:06 2017  hboutemy:ouch, tricky
Sat Nov 11 22:39:23 2017  stephenc:Otherwise the chaining of jobs can give false results
Sat Nov 11 22:39:41 2017  hboutemy:how do you model release from staging repo to release repo?
Sat Nov 11 22:39:49 2017  hboutemy:or modre precisely: detect ?
Sat Nov 11 22:40:01 2017  stephenc:We have customer tickets up the wazzoo where evil plugin is just not working
Sat Nov 11 22:40:17 2017  stephenc:I haven’t re-read my spec in 2 years
Sat Nov 11 22:40:23 2017  hboutemy:by not working, you mean triggerring too much, or something worse ?
Sat Nov 11 22:40:53 2017  stephenc:Giving the wrong chains and then incorrect statuses on the commit
Sat Nov 11 22:41:27 2017  hboutemy:I don't get what a status on a commit is
Sat Nov 11 22:41:30 2017  stephenc:Also triggering too little, too much, wrong jobs
Sat Nov 11 22:41:52 2017  hboutemy:but sure, being hackish, you get what you paid for
Sat Nov 11 22:41:53 2017  stephenc:GitHub / Bitbucket/ gitea / gitlab
Sat Nov 11 22:42:14 2017  stephenc:All show a check mark in commits that have successfully built
Sat Nov 11 22:42:46 2017  hboutemy:you don't expect to change the status based on dependant builds, no?
Sat Nov 11 22:42:55 2017  stephenc:Some people do
Sat Nov 11 22:43:03 2017  hboutemy:oh
Sat Nov 11 22:43:10 2017  hboutemy:ideally, why not
Sat Nov 11 22:43:33 2017  hboutemy:but sure, if you do that, you need to have high quality on such dependent builds management
Sat Nov 11 22:43:47 2017  stephenc:The api is only successful if the integration tests of the impl pass
Sat Nov 11 22:43:57 2017  hboutemy:or, yes, false positives
Sat Nov 11 22:44:18 2017  hboutemy:I see in such api/impl builds separation, that can be useful
Sat Nov 11 22:44:30 2017  stephenc:Ok, 3 min... need keyboard
Sat Nov 11 22:45:54 2017  stephenc:right
Sat Nov 11 22:46:29 2017  stephenc:So what you need to understand is that there are a lot of features in Jenkins that people misunderstand and do not use
Sat Nov 11 22:46:42 2017  stephenc:the biggest misuse is fingerprinting
Sat Nov 11 22:47:01 2017  stephenc:Jenkins can fingerprint (i.e. just track MD5 checksums) files, docker images, etc
Sat Nov 11 22:47:19 2017  stephenc:and then it uses the database of fingerprints to build a job dependency graph
Sat Nov 11 22:47:45 2017  hboutemy:that is to be expected, yes
Sat Nov 11 22:47:49 2017  stephenc:in classic jobs, and in pipeline, the fingerprints are basically the primary mechanism for job dependency detection
Sat Nov 11 22:48:04 2017  stephenc:the secondary mechanism is from explicit job trigger calls
Sat Nov 11 22:48:32 2017  stephenc:assuming you only fingerprint output from each job (i.e. **/target/** for Maven)
Sat Nov 11 22:48:38 2017  stephenc:then fingerprints are great
Sat Nov 11 22:49:01 2017  stephenc:but if you get lazy and just fingerprint everything (**) then you get a lot of false flags on the fingerprints
Sat Nov 11 22:49:12 2017  hboutemy:sure
Sat Nov 11 22:49:32 2017  stephenc:because I can have an empty .gitignore in each "retained" empty folder in multiple projects
Sat Nov 11 22:50:11 2017  stephenc:the maven triggering relies on parsing the pom
Sat Nov 11 22:50:13 2017  hboutemy:or copy/pasted
Sat Nov 11 22:50:31 2017  stephenc:it currently is somehwat dumb... GAV -> directed graph
Sat Nov 11 22:51:08 2017  stephenc:so the first problem you hit is when people change versions... you trigger the downstream jobs that have not updated their source code to the new version
Sat Nov 11 22:51:24 2017  stephenc:that is the excessive trigger problem
Sat Nov 11 22:51:56 2017  hboutemy:the graph does not take the version into account?
Sat Nov 11 22:52:18 2017  stephenc:that should only trigger once, because it will then detect no connection (because of different V) and only the SCM commit to update the version downstream will trigger a new build and re-link
Sat Nov 11 22:53:06 2017  stephenc:The trigger does not take into account the sources that a project may pull from, consequently a job may be falsely identified as being upstream because it builds groupid:artifactid:version but does not publish it to a source that can be consumed by the “downstream” job. This can result in excessive build triggering.
Sat Nov 11 22:53:15 2017  stephenc:Similarly, the trigger does not take into account the sources that a project may push to, or provide for external triggering from those sources, e.g. Sonatype Nexus Professional’s staging feature should be able to trigger a build of the project once the missing dependency has been promoted out of its staging repository (either by polling a Nexus’s RSS feed or by a Nexus plugin pushing the
Sat Nov 11 22:53:22 2017  stephenc:promotion details)
Sat Nov 11 22:53:22 2017  stephenc:It is exclusively a feature of the Maven job type. There are customers who would like to move to Freestyle/Matrix jobs and the new Workflow job type has no means to similarly interact with these triggers.
Sat Nov 11 22:53:29 2017  stephenc:There are other dependency systems that have similar needs, e.g. NPM, NuGet, etc.
Sat Nov 11 22:53:55 2017  stephenc:The current implementation does not provide any easy hooks to allow for distribution of triggering events across a Jenkins cluster as it relies on the dependency graph system to trigger jobs and the dependency graph system does not scale across a Jenkins cluster.
Sat Nov 11 22:54:09 2017  stephenc:those are the issues from my spec
Sat Nov 11 22:54:17 2017  hboutemy:ok
Sat Nov 11 22:54:33 2017  hboutemy:sure, issues come at cluster level
Sat Nov 11 22:54:53 2017  hboutemy:not so hard when only local
Sat Nov 11 22:55:03 2017  stephenc:and scaling Jenkins past a certain point basically requires a cluster
Sat Nov 11 22:56:52 2017  stephenc:The biggest set of issues from a support perspective are where people want to block jobs until downstream jobs finish building
Sat Nov 11 22:57:17 2017  stephenc:typically you can have A,B -> C -> D -> E
Sat Nov 11 22:57:39 2017  stephenc:and then A and B have commits...
Sat Nov 11 22:57:48 2017  hboutemy:yes, I suppose this one can become tricky
Sat Nov 11 22:58:01 2017  stephenc:one of the bugs that you can get is just the right frequency of commits to A can prevent B from ever building
Sat Nov 11 22:58:21 2017  stephenc:then there's situations like
Sat Nov 11 22:58:30 2017  stephenc:A -> B -> C,D -> E
Sat Nov 11 22:59:25 2017  stephenc:the diamond fork join causes issues for that block until downstream jobs finish... or block while upstream jobs are building
Sat Nov 11 22:59:43 2017  stephenc:the simple case is usually ok
Sat Nov 11 23:00:03 2017  stephenc:but to get -SNAPSHOT triggers to actually work, you need to have the CI server deploy -SNAPSHOTs
Sat Nov 11 23:00:23 2017  stephenc:and that usually screws over the developers... unless you give CI a dedicated repository to deploy to
Sat Nov 11 23:01:00 2017  stephenc:the evil job type then ends up training users to expect that there should be a simple solution
Sat Nov 11 23:01:13 2017  stephenc:but TBH this is actually a hard problem
Sat Nov 11 23:01:33 2017  stephenc:you need to know where artifacts are being deployed to
Sat Nov 11 23:01:53 2017  stephenc:or else, you need to be able to short-cut that
Sat Nov 11 23:02:22 2017  stephenc:so another feature I would like to offer is that Jenkins provides a maven repo of the upstream jobs
Sat Nov 11 23:02:47 2017  stephenc:how I would see that working in org-folder world is where things get really interesting
Sat Nov 11 23:03:00 2017  stephenc:basically the same branch name would be used to link
Sat Nov 11 23:03:14 2017  stephenc:so master -> master -> master
Sat Nov 11 23:03:37 2017  stephenc:now if I create a PR or a branch named MNG-9987
Sat Nov 11 23:04:01 2017  stephenc:as I need to branch other jobs and create the same branch name in those jobs
Sat Nov 11 23:04:26 2017  stephenc:the maven repo that gets exposed to the downstream jobs would get the -SNAPSHOTs from the same branch of upstream
Sat Nov 11 23:04:37 2017  hboutemy:ok, create some magic and start to have people understand it
Sat Nov 11 23:05:01 2017  stephenc:but it kind of needs an org-level kind of Jenkinsfile
Sat Nov 11 23:05:08 2017  hboutemy:but at least, the magic shields you from global mess
Sat Nov 11 23:05:15 2017  stephenc:yes
Sat Nov 11 23:05:30 2017  stephenc:if I have 5 branches all deploying snapshots
Sat Nov 11 23:05:36 2017  hboutemy:as I said: this is the next series of videos :)
Sat Nov 11 23:05:52 2017  stephenc:then currently I have 3.5.3-SNAPSHOT of core being deployed from each of them
Sat Nov 11 23:05:55 2017  hboutemy:perhaps not now, but in a few monthes
Sat Nov 11 23:06:21 2017  stephenc:I think that -SNAPSHOT triggers is too big for a single series
Sat Nov 11 23:06:24 2017  stephenc:;-)
Sat Nov 11 23:06:28 2017  hboutemy::)
Sat Nov 11 23:06:39 2017  hboutemy:a series of series?
Sat Nov 11 23:06:56 2017  stephenc:my spec is 1465 lines long
Sat Nov 11 23:07:25 2017  hboutemy:I don't get if it's longer than usual or not in your way of doing spec
Sat Nov 11 23:07:45 2017  stephenc:I have 68 specific requirements
Sat Nov 11 23:07:57 2017  stephenc:and there are still 4 open issues with the spec
Sat Nov 11 23:07:59 2017  hboutemy:oh, that one is huge
Sat Nov 11 23:08:26 2017  stephenc:and there are still some gaps... it was just complete enough for review... and everyone ran away from it
Sat Nov 11 23:08:33 2017  hboutemy::)
Sat Nov 11 23:08:59 2017  stephenc:anyway, I still have to finish this current series
Sat Nov 11 23:09:01 2017  hboutemy:as usual, perhaps the perfect model is too much for a single step
Sat Nov 11 23:09:08 2017  stephenc:maybe
Sat Nov 11 23:09:24 2017  stephenc:we will see how Cyrille gets on with his withMaven step
Sat Nov 11 23:09:27 2017  hboutemy:need to find little improvements to implement before the next step
Sat Nov 11 23:09:36 2017  hboutemy:Cyrille?
Sat Nov 11 23:09:52 2017  stephenc:I think the org folder and template jenkinsfile is a big step forward for maven project
Sat Nov 11 23:10:15 2017  hboutemy:yes: we'll have full Git + more efficient builds
Sat Nov 11 23:10:15 2017  stephenc:that should at least let us move CI maintenance out of our immediate concern
Sat Nov 11 23:10:45 2017  hboutemy:the current jobs for plugins are a mess
Sat Nov 11 23:10:55 2017  stephenc:with our current manual -SNAPSHOT deployment policy, and the current small set of committers, the -SNAPSHOT issue is not a big problem right now
Sat Nov 11 23:11:07 2017  stephenc:and you can always reference them as timestampped
Sat Nov 11 23:11:25 2017  stephenc:Cyrille LeClerc
Sat Nov 11 23:11:37 2017  stephenc:he has written the pipeline withMaven step
Sat Nov 11 23:11:45 2017  hboutemy:ah ok, didn't know
Sat Nov 11 23:12:03 2017  stephenc:it's the current best wrapper around maven for pipeline
Sat Nov 11 23:12:40 2017  hboutemy:so, now it's late here
Sat Nov 11 23:12:44 2017  stephenc:withMaven(maven:'3.5.2', jdk:'java 8', mavenLocalRepository:'.repository') { sh 'mvn verify' }
Sat Nov 11 23:12:59 2017  hboutemy:nice
Sat Nov 11 23:13:08 2017  stephenc:the withMaven step will fingerprint all the artifacts, record the test results, findubgs, checkstyle, etc
Sat Nov 11 23:13:16 2017  stephenc:and turns on ansi colour
Sat Nov 11 23:13:29 2017  stephenc:it injects a *listen only* extension into the maven build
Sat Nov 11 23:13:36 2017  stephenc:which is better than evil plugin...
Sat Nov 11 23:13:40 2017  hboutemy:perfect
Sat Nov 11 23:13:50 2017  stephenc:evil plugin injects a read-write extension into the maven build
Sat Nov 11 23:14:03 2017  stephenc:and then evil plugin actually does some write
Sat Nov 11 23:14:25 2017  stephenc:the biggest evil that evil plugin does is -Dmaven.test.failure.ignore=true
Sat Nov 11 23:14:33 2017  stephenc:but it doesn't stop there
Sat Nov 11 23:14:35 2017  hboutemy:notice that just keeping Maven 3 compatibility (I suppose) is something that helps
Sat Nov 11 23:14:43 2017  stephenc:yep
Sat Nov 11 23:15:02 2017  stephenc:next week I want to try and tackle sorting out maven 4.0.0 epics in jira
Sat Nov 11 23:15:05 2017  hboutemy:it's always easy to analyze 10 years later
Sat Nov 11 23:15:17 2017  stephenc:define the scope
Sat Nov 11 23:15:24 2017  hboutemy:yes, I didn't have time to work on it yet
Sat Nov 11 23:15:35 2017  stephenc:and then I'll commit the version bump to master and say... we are ready
Sat Nov 11 23:15:54 2017  stephenc:or do you want that we get a updated clirr replacement first
Sat Nov 11 23:16:03 2017  stephenc:before we say "we are ready"
Sat Nov 11 23:16:23 2017  hboutemy:did you see the guy who provided some reports?
Sat Nov 11 23:16:29 2017  hboutemy:in a Jira issue
Sat Nov 11 23:16:36 2017  stephenc:nope
Sat Nov 11 23:16:53 2017  stephenc:I'll take a look during the week anyway
Sat Nov 11 23:17:05 2017  stephenc:my wife is back from dropping the baby-sitter home
Sat Nov 11 23:17:11 2017  stephenc:so I need to stop on the computer
Sat Nov 11 23:17:16 2017  hboutemy:https://issues.apache.org/jira/browse/MNG-6301
Sat Nov 11 23:17:19 2017  stephenc:and Saturday nights
Sat Nov 11 23:17:22 2017  hboutemy:bye
Sat Nov 11 23:17:51 2017  stephenc:oh MNG-6301 perfect

Comments