20:14 <headius[m]> g'day

20:14 <headius[m]> ndbroadbent: cool, glad you are back!

20:15 <headius[m]> ndbroadbent: Adding support for sorbet calls to Java would be interesting, but the main problem I see with type checking is that the types end up creeping into all code otherwise you can't check something

20:16 <headius[m]> so in this case you'd not only need to know the types of what you're passing to a method, you would need to know the class the method comes from statically

20:16 <headius[m]> we can do that within the Java integration subsystem easily enough (obviously we have to because we have to look up all the methods) but I don't know how to propagate that information to Sorbet. I assume there's a way because there's core classes written in C etc...same thing

20:26 <headius[m]> kares: I'm going to be removing these tests because I don't believe they should pass; see https://github.com/jruby/jruby/pull/5984

20:27 <headius[m]> https://github.com/jruby/jruby/blob/d565996f5d3b2b98ae921b801853cf2583358732/spec/java_integration/object/marshal_spec.rb#L25-L50

20:28 <headius[m]> You ended up committing them as pending so most of the test wasn't running anyway, but the attempt to call `capsMethod` when it doesn't exist on a public class or interface fails with the corrected method binding in that PR

20:28 <headius[m]> I'm just removing those two that are marked pending, I will leave it to you if there's still something worth testing that we should restore