| openstackgerrit | Daniel Watkins proposed openstack-infra/jenkins-job-builder master: Always deep format Jinja2 templates https://review.openstack.org/546835 | 01:28 |
|---|---|---|
| openstackgerrit | Daniel Watkins proposed openstack-infra/jenkins-job-builder master: Add support for {% include %} in Jinja2 templates https://review.openstack.org/493694 | 01:31 |
| openstackgerrit | Daniel Watkins proposed openstack-infra/jenkins-job-builder master: Add !j2 directive to allow inline Jinja2 templates https://review.openstack.org/546838 | 01:51 |
| *** kien-ha has joined #openstack-jjb | 01:58 | |
| *** kien-ha has quit IRC | 02:34 | |
| waynr | so this is a thing: https://github.com/ktdreyer/jenkins-job-wrecker | 02:46 |
| zxiiro | What... Lol | 02:54 |
| zxiiro | That's a pretty significant effort. Wow | 02:58 |
| zxiiro | I feel like if jjb had a reverse mode it would be easier. | 02:59 |
| waynr | yeah | 03:05 |
| openstackgerrit | Merged openstack-infra/jenkins-job-builder master: publishers: add Task Scanner Plugin support https://review.openstack.org/532797 | 03:10 |
| openstackgerrit | Merged openstack-infra/jenkins-job-builder master: Support version 2.0 and greater of the Priority Sorter Plugin https://review.openstack.org/544235 | 03:14 |
| *** hashar has joined #openstack-jjb | 07:16 | |
| *** electrofelix has joined #openstack-jjb | 08:40 | |
| openstackgerrit | Sorin Sbarnea proposed openstack-infra/jenkins-job-builder master: Add multibranch project-type https://review.openstack.org/403940 | 09:26 |
| ssbarnea | electrofelix: waynr ^^ this is probably the most important addition to JJB I contributed too, please have a look at it. I want to make it ready before tomorrow session. | 10:10 |
| electrofelix | ssbarnea: is there a reason that the existing scm definitions couldn't be re-used? | 10:17 |
| ssbarnea | electrofelix: you mean to read scm from root instead of under the "multibranch" branch? i doubt there is anything preventing that, i was considering making this change. | 10:18 |
| ssbarnea | electrofelix: that's how original author did it, so i kept it. but i am inclined to believe that I can load scm from there. | 10:19 |
| ssbarnea | i will give a try now, look for other things meanwhile. | 10:19 |
| electrofelix | mean more that instead of redefining the options for the github/git/bitbucket plugins within the multi branch function, to delegate the calls to the methons in the jenkins_jobs.modules.scm module e.g. git => https://git.openstack.org/cgit/openstack-infra/jenkins-job-builder/tree/jenkins_jobs/modules/scm.py?id=5c08d33daef04c2ee28c422f94c41783a27b6705#n54 | 10:20 |
| electrofelix | my first guess is that the plugin is using the scm plugins in jenkins to handle this and is just hooking into them | 10:20 |
| electrofelix | but that's just a guess | 10:21 |
| ssbarnea | electrofelix: in fact, i am looking at my test project definition and I see that I already did it. scm section is directly under job. | 10:21 |
| ssbarnea | electrofelix: now this is diverging from how the "pipeline-scm" is doing it. | 10:21 |
| electrofelix | yes, is it that method won't work here, or just doesn't look like it was tried? | 10:24 |
| electrofelix | looking at the code in the scm module, maybe it's because it didn't appear to line up. git function in scm.py is defining the class as hudson.plugins.git.GitSCM while the multi-branch is using jenkins.plugins.git.GitSCMSource. Looks like they are both the same plugin https://github.com/jenkinsci/git-plugin/tree/master/src/main/java/jenkins/plugins/git, so maybe the definition in the scm.py can be updated to take parts of what's in the workflow_mu | 10:29 |
| electrofelix | btw, I'm not sure this is your most important contribution, your changes around the logging, avoiding spamming jenkins logs, print job names as urls, are all more important in my view because they help every user of JJB, this change is big and is adding support for other plugins but only a few are likely to benefit | 10:34 |
| electrofelix | small changes can have bigger impacts than large ones ;-) | 10:35 |
| ssbarnea | electrofelix: true. regarding scm support: i have confirmation that it works with "scm" section specified at root level, just like freestyle. | 10:41 |
| ssbarnea | electrofelix: ok for you? I (personally) prefer having scm at root level. | 10:42 |
| *** electrofelix has quit IRC | 13:14 | |
| openstackgerrit | manut proposed openstack-infra/jenkins-job-builder master: use AuthorizationMatrixProperty from cloudbees class https://review.openstack.org/546979 | 13:22 |
| *** zxiiro has quit IRC | 13:56 | |
| *** zxiiro has joined #openstack-jjb | 13:58 | |
| *** caphrim007_ has quit IRC | 14:39 | |
| *** caphrim007 has joined #openstack-jjb | 14:40 | |
| *** caphrim007 has quit IRC | 14:44 | |
| waynr | ssbarnea: looks good to me, thanks for working on that! | 15:02 |
| Odd_Bloke | Hey folks, I'd really appreciate reviews on my Jinja2 changes: https://review.openstack.org/#/c/546835/ https://review.openstack.org/#/c/493694/ https://review.openstack.org/#/c/546838/ | 15:03 |
| Odd_Bloke | (That first one fixes a bug that someone pointed out in this channel.) | 15:03 |
| ssbarnea | waynr: in fact is not working yet, just made some late discoveries. | 15:03 |
| ssbarnea | waynr: i am working on it but i struggle to find out where do I get some extra unexpected <source> section in the xml file. | 15:03 |
| ssbarnea | waynr: something that is not defined anywhere in the module. | 15:04 |
| Odd_Bloke | ssbarnea: Ah, looking back through history, it was your bug that it fixes. :) | 15:04 |
| zxiiro | morning folks | 15:20 |
| zxiiro | I'll try to get some reviews in today. Work's been busy the past couple of weeks so I've been falling behind | 15:21 |
| *** caphrim007 has joined #openstack-jjb | 16:06 | |
| Odd_Bloke | zxiiro: Thanks for those reviews. :) | 16:11 |
| *** openstackgerrit has quit IRC | 18:33 | |
| ssbarnea | zxiiro: hi! still around here? | 19:50 |
| zxiiro | ssbarnea: yep | 19:57 |
| zxiiro | what's up? | 19:57 |
| ssbarnea | zxiiro: i was wondering if there is a trick for 79 width limit, on jjb I seem to struggle to wrap lines | 19:58 |
| ssbarnea | especially those with very long java strings with full class names | 19:59 |
| ssbarnea | i have the impression that "# noqa" is going to become the new <CR>. | 19:59 |
| zxiiro | ssbarnea: use parenthesis () | 20:02 |
| zxiiro | ssbarnea: for example | 20:03 |
| zxiiro | some_string = ('a very long string that needs to continue ' | 20:03 |
| zxiiro | 'continuation of that long string') | 20:03 |
| ssbarnea | zxiiro: well, for doc lines is ok, but for class names is recipe for making hard to search inside source code. | 20:04 |
| zxiiro | you don't need to add + or % or anything to that. Python automatically continues a string on the next line when it's in parenthesis like that. | 20:04 |
| zxiiro | ah yeah if you want to search the whole string it does make it hard i dunno how else to do it. | 20:04 |
| zxiiro | although the original openstack folks are no longer working on jjb | 20:04 |
| zxiiro | I can forsee us allowing longer max lines if current committers agree to it | 20:05 |
| zxiiro | Personally I use max of 120 in my own projects. | 20:05 |
| ssbarnea | zxiiro: yep, a huge difference from 79 to 120 :D | 20:06 |
| *** openstackgerrit has joined #openstack-jjb | 20:07 | |
| openstackgerrit | Sorin Sbarnea proposed openstack-infra/jenkins-job-builder master: Add multibranch project-type https://review.openstack.org/403940 | 20:07 |
| zxiiro | ssbarnea: Submit a patch and we can ask folks to vote on it. | 20:07 |
| ssbarnea | zxiiro: back to serious issues, i did something very ugly on ^^ and I want your opinion on it. I added hack to disable use of SCM module on multibranch pipelines. | 20:08 |
| ssbarnea | somehow the SCM module added unwanted garbage when using the the multibranch project type, which does the scm part correctly itself. | 20:09 |
| ssbarnea | zxiiro: this is the ugliness: https://review.openstack.org/#/c/403940/13/jenkins_jobs/xml_config.py | 20:10 |
| zxiiro | huh | 20:10 |
| zxiiro | yeah that does seem ugly. I'd get electrofelix's opinion on that. | 20:11 |
| ssbarnea | i guess we all agree on that, the question is how to do it nice, or at least better. | 20:13 |
| ssbarnea | look at the rest and add comments, if possible i would like to merge it tomorrow. I already tested it with gerrit scm but now I am also testing the other setups. | 20:16 |
| zxiiro | ssbarnea: I don't get the gerrit scm part. I don't think there is a gerrit scm? it's just git? | 20:17 |
| zxiiro | why do we need an alias for git? | 20:17 |
| ssbarnea | maybe we should mark implementation as experimental and not guarantee api compatibility. | 20:17 |
| ssbarnea | zxiiro: there is! but is a very new thing. | 20:17 |
| zxiiro | huh really? | 20:18 |
| zxiiro | what plugin provides it? | 20:18 |
| ssbarnea | zxiiro: https://github.com/jenkinsci/gerrit-plugin | 20:18 |
| ssbarnea | zxiiro: but there is a catch | 20:18 |
| Odd_Bloke | zxiiro: ssbarnea: Not to be that guy, but PEP-8 does allow for up to 100 characters, so I'd suggest that as a first step. | 20:18 |
| ssbarnea | zxiiro: this adds Gerrit SCM support but only for multibranch-pipelines. | 20:19 |
| ssbarnea | zxiiro: in fact this is the only place where is needed | 20:19 |
| ssbarnea | in other places the git one is perfect. | 20:19 |
| ssbarnea | zxiiro: but for multibranch-pipelines, there is special code to detect branches+CRs. | 20:20 |
| ssbarnea | have a look at https://j.sbarnea.com/job/test-jjb-multibranch-gerrit-http-ano/ | 20:20 |
| zxiiro | huh ok. I don't use multibranch-pipelines so not too familiar with it. | 20:20 |
| zxiiro | ssbarnea: connection timeout for me | 20:21 |
| ssbarnea | zxiiro: me neither because lack of gerrit support was preventing me from using them. | 20:21 |
| ssbarnea | no IPv6 support? | 20:21 |
| zxiiro | ah yeah my ISP doesn't do IPv6 yet | 20:21 |
| zxiiro | oh wait they did roll it out | 20:21 |
| zxiiro | maybe i have to do something on my end to turn it on | 20:21 |
| zxiiro | "IPv6 is not available for our Cable Internet services at this time." | 20:22 |
| zxiiro | ah nevermind I'm on cable | 20:22 |
| zxiiro | if i was on DSL i'd have it | 20:22 |
| ssbarnea | zxiiro: ahh, let me check the it. i think I should be able to do something about it. | 20:22 |
| zxiiro | heh | 20:22 |
| ssbarnea | i updated the DNS but it will take up to one hour to propagate. it was using a private address. | 20:26 |
| ssbarnea | (for IPv4, IPv6 was public) | 20:26 |
| *** hashar has quit IRC | 21:07 | |
| *** abelur has quit IRC | 23:45 | |
| *** abelur_ is now known as abelur | 23:45 | |
| *** caphrim007_ has joined #openstack-jjb | 23:54 | |
| *** caphrim007 has quit IRC | 23:57 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!