00:20 <travis-ci> jruby/jruby (travis_sucks:cf91bc9 by Charles Oliver Nutter): The build was canceled. https://travis-ci.org/jruby/jruby/builds/739129283 [9 min 49 sec]

01:14 <headius[m]> makes no sense

01:39 <travis-ci> MariuszCwikla/jruby-openssl (issue_6409_ssl_min_version:4ddcdb0 by Mariusz Ćwikła): The build is still failing. (https://travis-ci.org/MariuszCwikla/jruby-openssl/builds/739046330)

02:25 <headius[m]> kares: I'm stumped as to why this is failing on CI and not locally: https://travis-ci.org/github/jruby/jruby/jobs/739129302

02:25 <headius[m]> I just had it fail on GH actions the same way

02:25 <headius[m]> I don't see what recent change would have done this because it was passing just a few builds back on master

02:26 <headius[m]> it does not fail locally

02:27 <headius[m]> I addd some debugging messages here that seem to show the HashMap inserted into the array is getting converted to a normal Ruby Hash somewhere: https://github.com/jruby/jruby/runs/1312328723?check_suite_focus=true#step:9:186

02:30 <headius[m]> signing off for today, will continue investigating tomorrow

04:38 <byteit101[m]> mariuszcwikla: tried solargraph eclipse, and I only get null pointer exceptions

07:53 <mariuszcwikla[m]> byteit101: it's a pity; works fine here.

18:56 <headius[m]> howdy all

18:59 <headius[m]> I just learned that travis-ci.org is being EOL on Dec 31 and the slow build times we've been seeing lately are likely due to them moving infrastructure there. I'm going to look at migrating the JRuby build to .com so we'll have reliable builds at least until we migrate to something else

19:00 <headius[m]> kares enebo just FYI in case you start looking for build results on .org

19:07 <headius[m]> I am migrating the following repos to travis-ci.com today: arjdbc, arjdbc sqlserver, jcodings, joni, jruby, jruby-launcher, and jruby-openssl (cc mariuszcwikla )

19:12 <headius[m]> so it looks like it's working 👍

19:12 <headius[m]> https://travis-ci.com/github/jruby/jruby/builds/193680445

19:13 <headius[m]> An immediate down side is that we're back to only 5 jobs at once, but I'll see if I can work with Travis folks to get the same treatment we had on .org

19:59 <travis-ci> jruby/jruby (master:3a3d2c6 by Benoit Daloze): The build is still failing. https://travis-ci.com/jruby/jruby/builds/193680445 [214 min 6 sec]

20:01 <headius[m]> so with just 5 jobs we're looking at about 50 minutes for the full build on .com

20:01 <headius[m]> I sent a request for more jobs, we'll see how that goes

20:01 <headius[m]> now back to this stupid failure

20:23 <headius[m]> noice, I have it failing locally

20:44 <travis-ci> jruby/jruby (master:49370de by Charles Oliver Nutter): The build is still failing. https://travis-ci.com/jruby/jruby/builds/193705980 [201 min 53 sec]

20:45 <headius[m]> yes yes

21:56 <headius[m]> gods, finally found it

21:57 <headius[m]> rspec-expectations 3.9.3 added logic for passing kwargs through to "be" matchers using **kwargs, which appears to trigger conversion logic on the normal positional args

21:57 <headius[m]> yay kwargs bug

22:02 <headius[m]> it gets converted to a Hash... probably due to **kwargs handling being too aggressively converting

22:18 <headius[m]> waste of a day's work but at least we have an answer

22:32 <headius[m]> FWIW the travis .com jobs seem to be running faster

22:32 <headius[m]> and this is some great news: https://github.com/jnr/jffi/pull/90

22:33 <headius[m]> automated build/test of jffi native backend under a bunch of different platforms (Linux only but hey)

23:07 <travis-ci> jruby/jruby (master:f1d8b25 by Charles Oliver Nutter): The build was fixed. https://travis-ci.com/jruby/jruby/builds/193876783 [202 min 27 sec]

23:11 <headius[m]> yahoo

23:12 <headius[m]> ttfn