cpuguy83 has quit [Remote host closed the connection]
cpuguy83 has joined #jruby
cpuguy83 has quit [Ping timeout: 256 seconds]
_whitelogger has joined #jruby
ur5us_ has quit [Ping timeout: 244 seconds]
nirvdrum has joined #jruby
ur5us_ has joined #jruby
_whitelogger has joined #jruby
nirvdrum has quit [Remote host closed the connection]
_whitelogger has joined #jruby
nirvdrum has joined #jruby
nirvdrum has quit [Remote host closed the connection]
ur5us_ has quit [Ping timeout: 260 seconds]
ur5us_ has joined #jruby
ur5us_ has quit [Ping timeout: 265 seconds]
nirvdrum has joined #jruby
nirvdrum has quit [Remote host closed the connection]
nirvdrum has joined #jruby
xardion has quit [Remote host closed the connection]
cpuguy83 has joined #jruby
xardion has joined #jruby
knu has quit [Remote host closed the connection]
knu has joined #jruby
cpuguy83 has quit [Remote host closed the connection]
cpuguy83 has joined #jruby
cpuguy83 has quit [Remote host closed the connection]
cpuguy83 has joined #jruby
snickers has joined #jruby
snickers has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
travis-ci has joined #jruby
<travis-ci> jruby/jruby (no_sourceposition:c64c698 by Thomas E. Enebo): The build is still failing. https://travis-ci.org/jruby/jruby/builds/687177089 [146 min 8 sec]
travis-ci has left #jruby [#jruby]
<headius[m]> valphilnagel: hey we keep missing each other
<headius[m]> valphilnagel: if you have some tweaks that can get this working on both, please look at a fork + PR!
<headius[m]> plus a PR or issue will get conversation going... have you opened anything so far?
<headius[m]> I'm glad you're making progress and yes I think we should even favor OpenJFX since that's the path forward
<enebo[m]> lopex: how are you buddy?
<cpuguy83> @headius[m] So for the docker-jruby repo, I invited you as a collab so you could transfer it to the jruby org.
<cpuguy83> I did try to transfer it myself, but said I couldn't create a new repo in the jruby org.
<headius[m]> cpuguy83: yeah I don't know how to give you just that permission... I can only make you an admin
<headius[m]> but it occurred to me that it's better to just leave it there for the moment
<cpuguy83> I think one thing tianon liked is having the issue/PR history in tact.
<headius[m]> I don't seem to be able to get to the settings for the repo so I'm pretty sure I have to be admin on your org too
<headius[m]> it's not a permission that's easy to give out alone
<headius[m]> oh yeah we won't re-push in any case... it will maintain your history
<headius[m]> I have no Settings tab on cpuguy83/docker-jruby
<cpuguy83> Hmm yeah, it seems only collaborator is possible.
<headius[m]> by re-push I mean we won't take current revision and make it revision zero
<headius[m]> we could repush as non-fork but it doesn't matter... it won't be a fork once your repo goes away
<cpuguy83> That works. I guess the next step is to update https://github.com/docker-library/official-images/blob/master/library/jruby with the new repo and maintainer
<headius[m]> enebo: I nominate you
<enebo[m]> bleh
<cpuguy83> haha
<headius[m]> in the short term I suppose it can be either of us so we can start the transitioning
<enebo[m]> well I guess it needs to be done
<headius[m]> we have had some soft commitments to help but someone core to the project should still own it
travis-ci has joined #jruby
<travis-ci> jruby/jruby (no_sourceposition:71247c4 by Thomas E. Enebo): The build is still failing. https://travis-ci.org/jruby/jruby/builds/687199119 [139 min 15 sec]
travis-ci has left #jruby [#jruby]
<headius[m]> once we get some contributions from users maybe a new maintainer will rise to the top
<enebo[m]> bleh
<cpuguy83> Sorry to throw more work your way :(
<enebo[m]> cpuguy83: so I believe we run a command with the Dockerfile we have and it generates bits and signs it or something like that and then pushes to docker mother ship?
<enebo[m]> cpuguy83: and if we need to support 8 and 13 we would do this twice with a slightly different Dockerfile?
<cpuguy83> @enebo[m] The docker mother ship builds it, this is just the spec.
<enebo[m]> or possibly some ENV when running the dockerfile
<headius[m]> cpuguy83: I'm hoping this isn't a lot of work once we have things configured
<headius[m]> and we would just include it as another release artifact
<enebo[m]> cpuguy83: ah so we push the spec somewhere and it does the rest
<cpuguy83> Yep.
<headius[m]> and that's what's in your repo
<enebo[m]> ok and how do we verify that it successfully completed?
<headius[m]> cpuguy83: apologies for losing your description of the process, we are new at this
<enebo[m]> the tangential question is also how long does it tend to take
<headius[m]> but it's time to address it
<cpuguy83> So really it's just keeping up with jruby and jvm releases. From there the community may want changes from time to time... actually there's some changes from the MRI ruby image that might need to get made for jruby just for consistency if nothing else.
<enebo[m]> yeah we really don't know much about docker. I have used it locally for multiple db instances but not much more than configing as a consumer
<headius[m]> right we keep getting more requests for an image with updated JDK
<headius[m]> trying to figure out the best way to do that without permuting all JRuby and JDK versions
<headius[m]> SBT seems to be doing something different here? https://hub.docker.com/r/hseeberger/scala-sbt/
<cpuguy83> @enebo[m] When you PR the spec changes there's some sanity checks that are made. The team that reviews it is also quite thorough in checking things. I'm not sure in what timing the images are actually updated.
<enebo[m]> So our release requirements are our version and any u-updates on JVM
<headius[m]> I guess this is just a way to build on the fly with a different "from" for your needs?
<enebo[m]> cpuguy83: ok so far from real time was one answer :P
<cpuguy83> And there's also automatic image updates that happens due to bugs or cve's in the base images.... nothing to do for this, it just happens.
<enebo[m]> cpuguy83: ah that is super nice
<headius[m]> so once we have a recipe for building JRuby's image that's mostly out of our hands
<cpuguy83> Yes.
<headius[m]> and we just need to update for version changes or other tweaks that make sense for JRuby
<enebo[m]> cpuguy83: that was my main worry about needing to be concerned about updates of the JVM piece
<enebo[m]> and we already have the recipe
<headius[m]> enebo: presumably by depending on something generic like openjdk8-jdk we get latest automatically then
<cpuguy83> So the images I have pinned to a somewhat generic version number so it should automatically pick up patch releases.
<headius[m]> right
<enebo[m]> so we just need some security aspect to push the spec and once per release we probably do it unless there is non-CVE base image updates maybe
<headius[m]> here's an example of what maybe we want to avoid: https://hub.docker.com/_/clojure
<enebo[m]> oh ok very cool
<headius[m]> I think we want to avoid anyway
<cpuguy83> There's also all these "adoptopenjdk" images... which I just don't understand the purpose of. https://hub.docker.com/_/adoptopenjdk?tab=tags
<cpuguy83> Like, the relationship between "openjdk" and "adoptopenjdk"
<headius[m]> it's just another build of OpenJDK
<cpuguy83> And then adoptopenjdk has openj9 and hotspot variants.
<headius[m]> so more independent of Oracle but there's still a company behind it (which I believe is either Microsoft or Amazon)
<headius[m]> yup
<headius[m]> we are figuring supported images will be only Java 8 and Java 11 (LTS release) on somebody's non-Oracle build (maybe we can find a Red Hat build)
<headius[m]> anything else people can assemble themselves
<cpuguy83> That makes sense, I did add 14 and 15, not sure if you want to keep those.
<enebo[m]> dinner prep
<headius[m]> probably not
<cpuguy83> The dockerfiles in the repo are the spec, and then there is https://github.com/docker-library/official-images/blob/master/library/jruby which tells the builders where to build from.
<enebo[m]> cpuguy83: thanks for this update on how this works. It was enlightening
<enebo[m]> cpuguy83: also not too lockstep with releases in that it may take time for it to arrive...so we will release and then at some nearisgh point we will get our image(s) out.
<cpuguy83> Also, changing the Dockerfiles does not affect the builds until you update that manifest.
<olleolleolle[m]> PS: This is all awesome, with improvement to official images being planned.
<headius[m]> yeah we'll get it done soon
cpuguy83 has quit []