21:39 <joshuacronemeye4> Hi there. Working on my first jruby PR. On the travis build I am getting test failures that seem unrelated to my changes. https://travis-ci.org/github/jruby/jruby/builds/738384442 I tried running some of the failing tests locally, but I don't see the same failures. How reliable is the build?

21:40 <headius[m]> Yeah don't worry too much about those, I think a recent update broke CI temporarily

21:40 <headius[m]> If it's okay locally and we can verify it's okay on our end it should be fine

21:40 <byteit101[m]> headius: should I cache the arguments of ctors with arguments for potential init args?

21:41 <headius[m]> Not sure what you mean, cache what arguments exactly?

21:42 <byteit101[m]> in the javaproxyclassfactory we do super(*args) for all ctors, but don't save the args for calling the ruby #initialize function, should I do that for potential tie in later, or keep it as it is right now?

21:50 <joshuacronemeye4> thanks @headius

21:56 <byteit101[m]> headius: Why is the reification code using SkinnyMethodAdapter, vs the underlying ASM types?

22:33 <byteit101[m]> How deeply should I worry about thread-safe constructors?

23:41 <byteit101[m]> headius: Hmm, my current implementation makes it hard to seperate allocate from initialize with a custon new. Should I allow them to be split for java extensions?