12:07 <fidothe> @headius[m] dumb question. How does the gsub bench you added in `bench/core/string/bench_gsub.rb` work? Like, where is that `benchmark` method defined and how should it be run?

15:23 <headius[m]> That one depends on a benchmark harness from TruffleRuby that does some extra warmup

17:10 <fidothe> I need to add one for String#sub too, as part of https://github.com/jruby/jruby/issues/5095 (The CRuby changes affect `String#sub` and `String#gsub`). If I look at TruffleRuby will it be obvious where the harness is?

18:06 <fidothe> Yes, if I look at TruffleRuby it is obvious. 🎉

20:49 <kares[m]> the indy regression still had 2 nasty NPEs hiding ...

20:49 <kares[m]> should be fixed now + we have those cases tested

22:06 <headius[m]> oh I'll have a look

22:07 <headius[m]> fidothe: cool, thanks!

22:07 <headius[m]> it would be nice if those benchmarks were in a shared repo somewhere, since most of them were written by others anyway

22:55 <headius[m]> oh thank goodness, I bootched my thread patch

22:55 <headius[m]> fixed it, now passes specs so far

22:56 <headius[m]> I was worried I'd have to implement my own Condition.await

23:39 <headius[m]> kares: your fix should resolve the last problem reported here, yes? https://github.com/elastic/logstash/issues/11196

23:40 <headius[m]> That issue is kind of conflating multiple things, since the original ClassCastException is fixed now, but I think it should all be good now yes?

23:51 <headius[m]> oh jeez

23:51 <headius[m]> my PR seems to have fixed the test:mri:stdlib hang

23:51 <headius[m]> so that's good but also bad because I was wrong about it being concurrent-ruby

23:51 <headius[m]> 🤷‍♂️