00:22 <rwilliams[m]> Is there an easy way to get rbenv to point to the jruby that's build when building the source?

00:22 <rwilliams[m]> * Is there an easy way to get rbenv to point to the jruby that's built when building the source?

00:29 <rwilliams[m]> I just linked the jruby source directory to ~/.rbenv/versions/jruby-dev

00:29 <headius[m]> Yeah I don't know rbenv well but that seems fine to me

00:29 <headius[m]> rvm has mount but I don't know if there's an equivalent for the other switchers

00:29 <rwilliams[m]> Yeah i kept having issues with rvm recently so i switched

00:32 <rwilliams[m]> daniel_jruby_question_asker: your username is great

00:32 <daniel_jruby_que> Looking at JavaMethod, it looks like there's code to set accessible for public methods on classes that aren't public, but that doesn't seem to catch this here?

00:32 <daniel_jruby_que> Haha, thanks

00:32 <daniel_jruby_que> It is descriptive

00:34 <daniel_jruby_que> Anyway I'm encountering this issue with existing code that works with Java 8 on jruby 9.2.9.0 but fails in this way on trying to upgrade from there to Java 11

00:41 <headius[m]> ok there is one known bug in 9.2.9 regarding such methods, but in general I expect you are encountering a package that isn't open

00:41 <headius[m]> on Java 11 we don't attempt to circumvent Java module visibility so methods that we might have forced to be accessible before will not be accessible now

00:42 <headius[m]> yeah I see your issue above now

00:42 <headius[m]> that's definitely the bug

00:43 <headius[m]> daniel_jruby_question_asker: I expect it's this: https://github.com/jruby/jruby/issues/5969

00:43 <headius[m]> which is fixed by https://github.com/jruby/jruby/pull/5984

00:44 <headius[m]> enebo: probably need a 9.2.10 sooner than later

00:44 <headius[m]> this one is apparently affecting lots of folks immediately after updating

00:45 <daniel_jruby_que> Awesome, thanks so much!

00:45 <headius[m]> there are some workarounds to help get you past that...unfortunately a JVM --add-opens flag is the best workaround we have right now

15:05 <satyanash> so, we just noticed a bug where arjdbc for postgresql fails to return the correct PK when insert_returning is true, but the primary key is not the first column. Have raised a draft PR with a failing test here: https://github.com/jruby/activerecord-jdbc-adapter/pull/1045/files

15:06 <satyanash> Looks like a regression as it was discovered when trying to upgrade from Rails 4 to 5.

17:16 <headius[m]> Nice thanks