06:36 <travis-ci> jruby/jruby (load_service_redux:a36f7c9 by Charles Oliver Nutter): The build has errored. https://travis-ci.org/jruby/jruby/builds/514504171 [202 min 47 sec]

07:09 <kares[m]> so I think I have a 'hack' to speedup BigDecimal division by 25x for some cases ...

07:09 <headius[m]> ooo neat

07:10 <kares[m]> might be relying on Java BigDecimal ctor specifics ... but if it works the same on J9 let's call it a day

07:11 <kares[m]> should work since the java.math.BigDecimal is pretty much the same .java source

14:21 <headius[m]> J9 should be using the same implementation...I will look forward to your PR

14:21 <headius[m]> I had looked into some lib...jscience I think

16:23 <lopex> there's also https://github.com/tools4j/decimal4j/wiki/Performance

16:23 <lopex> though probably not vey popular

16:26 <kares[m]> did poke around

16:26 <kares[m]> decimal4j might be a good candidate if anyone is up for a full rewrite of the backend

16:27 <kares[m]> which might be necessary if we hit any further issues

16:27 <kares[m]> as far as I tested only division was considerably slower ...

17:26 <headius[m]> someone could port MRI's impl 😁

17:26 <headius[m]> I don't know where it came from but it's complicated

17:30 <kares[m]> and slightly opinionated ... they test for some div results doing rounding on a 'hard-coded' precision

18:46 <JulesIvanicGitte> Hi everyone, Hi @headius,

18:46 <JulesIvanicGitte> Tomorrow (it’s 8pm here) I’ll contiue my work on my memory leak.

18:46 <JulesIvanicGitte> Is there a build of JRuby containing this fix: https://github.com/jruby/jruby/pull/5672 I can use tomorrow ?

18:46 <JulesIvanicGitte> (edited) ... my memory leak. ... => ... my thread leak. ...

18:47 <headius[m]> Jules Ivanic (Gitter): No build at the moment...I need to get my queue of PRs merged in

18:48 <JulesIvanicGitte> hum :/

18:48 <JulesIvanicGitte> (edited) hum :/ => Hi everyone, Hi @headius,

18:48 <JulesIvanicGitte> Tomorrow (it’s 8pm here) I’ll continue my work on my thread leak.

18:48 <JulesIvanicGitte> Is there a build of JRuby containing this fix: https://github.com/jruby/jruby/pull/5672 I can use tomorrow ?

18:49 <JulesIvanicGitte> How can I help you to move forward on this issue ?

18:55 <headius[m]> Jules Ivanic (Gitter): well I'm looking at getting it green right now so it can go into 9.2.7

18:56 <yaauie> any chance I could get some help with the Open3 issue I posted a couple weeks ago? I've added repro and threaddump, but this is a bit beyond my experience with Java process management (see: https://github.com/jruby/jruby/issues/5642 )

18:56 <headius[m]> Are you able to do a build yourself and test it out a bit? It didn't improve things in my small tests but they were spinning up enumerators like mad

18:58 <JulesIvanicGitte> > Are you able to do a build yourself

18:58 <JulesIvanicGitte> No but I can learn :)

19:00 <headius[m]> Ok, well all you need to do is clone JRuby and mvn package -Pbootstrap

19:00 <headius[m]> that will make the clone a fully functional JRuby install

19:00 <JulesIvanicGitte> ok easy

19:01 <JulesIvanicGitte> I’ll do that tomorrow morning so. Thanks a lot @headius :)

19:01 <headius[m]> no problem...hopefully I'll have PR in a mergeable state soon