08:26 <GitHub136> [jruby] kares closed pull request #5251: remove MaxPermSize (master...remove_maxperm) https://git.io/fNYBC

08:26 <GitHub140> [jruby] kares pushed 1 new commit to master: https://git.io/fNY51

08:26 <GitHub140> jruby/master 52bd002 ahorek: [build] remove MaxPermSize (#5251)

11:06 <GitHub107> [jruby] ahorek opened issue #5252: can't use Date.iso8601 with a string subclass https://git.io/fNOT1

13:21 <TheBino> Hey guys

13:22 <TheBino> did anyone stumbled across this error ever using the jdbc-adapter? Bundler::GemRequireError: There was an error while trying to load the gem 'activerecord-jdbcmssql-adapter'. Gem Load Error is: undefined method `type_cast_from_database' for class `ArJdbc::MSSQL::UUIDType' Backtrace for gem load error is:

13:57 <GitHub99> [jruby] kares created string-encode (+10 new commits): https://git.io/fNO88

13:57 <GitHub99> jruby/string-encode c253b45 kares: handle String/CharSequence decoding slightly differently...

13:57 <GitHub99> jruby/string-encode 335dc2e kares: [refactor] avoid a byte[] copy on encode when possible...

13:57 <GitHub99> jruby/string-encode 1a09fe3 kares: [fix] encode UTF-16 without unwrapping Java String internals...

13:58 <GitHub189> [jruby] kares deleted string-encode at cb57a76: https://git.io/fNO8u

14:14 <GitHub27> [jruby] ahorek opened pull request #5253: prevent an infinite loop, check string subclasses (master...loop_fix) https://git.io/fNOBE

15:23 <GitHub55> [jruby] ahorek opened pull request #5254: update rubyspec (master...update_rubyspec) https://git.io/fNOre

15:35 <GitHub92> [jruby] kares pushed 5 new commits to master: https://git.io/fNOom

15:35 <GitHub92> jruby/master ec73864 kares: [refactor] use IntHashMap for Set<int> semantics - avoid boxing

15:35 <GitHub92> jruby/master 7cfe316 kares: [refactor] warning - method should be static

15:35 <GitHub92> jruby/master 3e41dbe kares: [refactor] make constant final + avoid cast on clone()

15:35 <GitHub150> [jruby] kares closed issue #5252: can't use Date.iso8601 with a string subclass https://git.io/fNOT1

15:37 <GitHub143> [jruby] kares closed pull request #5253: prevent an infinite loop, check string subclasses (master...loop_fix) https://git.io/fNOBE

17:30 <GitHub87> [jruby] ahorek closed pull request #5254: update rubyspec (master...update_rubyspec) https://git.io/fNOre

20:10 <GitHub7> [jruby.github.io] enebo pushed 1 new commit to master: https://git.io/fNOhW

20:10 <GitHub7> jruby.github.io/master 9c77019 Thomas E. Enebo: Remove MARKETO? and add engine yard thanks link at bottom of home page for hosting our S3 images

20:14 <GitHub162> [jruby.github.io] enebo pushed 1 new commit to master: https://git.io/fNOh1

20:14 <GitHub162> jruby.github.io/master 720227e Thomas E. Enebo: Center EY info to make it look a little nicer

20:17 <GitHub1> [jruby.github.io] enebo pushed 1 new commit to master: https://git.io/fNOhF

20:17 <GitHub1> jruby.github.io/master 344f8e4 Thomas E. Enebo: Center EY info to make it look a little nicer

20:37 <ChrisBr> enebo: headius: test/mri/ruby/test_hash.rb succeeds now and all missing todos are implemented. There is only one todo I'm not sure how to fix, otherwise it would be ready to go from my side

20:38 <ChrisBr> I'm also not up to date regarding the rest of the test suite, so if some other tests outside of test_hash.rb fail please let me know

20:39 <ChrisBr> this todo is what I'm talking about: https://github.com/jruby/jruby/pull/5215/files#diff-12e25a31f9cf95ea1b40547ab4b57c3bR346

20:40 <enebo> ChrisBr: I believe travis should run your branch

20:42 <ChrisBr> enebo: yo, but is master currently green? So should everything succeed?

20:43 <enebo> I have been working on oj lately :|

20:43 <enebo> to your question about TODO

20:43 <enebo> that constructor seems to be for Java proxy objects and it is an empty array

20:44 <enebo> I see no other callers

20:44 <enebo> so I am not sure if that helps you or not

20:45 <ChrisBr> hmm

20:45 <ChrisBr> so it will always be an empty array?

20:45 <ChrisBr> why is it even there as a parameter then?

20:45 <enebo> ChrisBr: within core for sure

20:45 <enebo> ChrisBr: I don't know if we realyl can expect a java native extension to be calling that

20:45 <enebo> ChrisBr: perhaps headius or kares will know more

20:46 <ChrisBr> ok

20:46 <enebo> ChrisBr: and also I see a single error in test:mri on travis involving a lambda arity error

20:46 <enebo> ChrisBr: So your branch is up to date should show that single failure

20:47 <headius> hey hey

20:47 <headius> master should be green but there's some unstable tests

20:47 <ChrisBr> my issue with the todo is that if we leave the signature, it is impossible to transform it to a valid hash as I don't have a head and don't know what is the first element in the hash

20:48 <headius> is that constructor called?

20:48 <headius> it's protected so we could ditch it any time really

20:48 <ChrisBr> (I believe it was already before 'invalid')

20:49 <ChrisBr> because just assigning the table without head would result in an invalid hash, that was the case before

20:50 <ChrisBr> according to enebo it is only called with an empty array in Java proxy objects

20:50 <headius> oh I see

20:50 <headius> I'll look too

20:50 <enebo> MapJavaProxy.RubyHashMap

20:50 <headius> ah yeah

20:50 <headius> so that proxy class extends RubyHash to wrap all Map instances from JDK

20:51 <headius> almost all methods should be overridden to hit the wrapped map

20:52 <headius> the only place it's used initializes it to new IRubyObject[0]

20:52 <headius> so basically it's just trying to initialize a totally empty shell of RubyHash

20:55 <ChrisBr> so if we replace it in RubyHashMap we can drop it?

20:59 <headius> yeah I think so

21:00 <headius> RubyHashMap should not need the impl of RubyHash to work right, I believe

21:00 <headius> it probably should not be implemented this way

21:00 <headius> I think it was just the smallest amount of plumbing possible to make it duck-type as a Hash better

21:04 <nirvdrum> headius: I just got "Optimizing Java" and started reading the foreword. I got to the part about working at the University of Minnesota and said to myself: "I wonder if this person knows Charlie or Tom. It sounds a lot like the sort of thing they told me they were doing then."

21:04 <headius> hah nice

21:04 <headius> I didn't know it was out

21:05 <nirvdrum> I saw Chris tweet about it and decided to get a copy. I think it came out in May.

21:05 <nirvdrum> They didn't give you a promotional copy?

21:16 <headius> I might have missed an email for address

22:35 <GitHub149> [jruby] enebo pushed 1 new commit to master: https://git.io/fN3m0

22:35 <GitHub149> jruby/master 785f34b Thomas E. Enebo: Lambdas should not spread non single arity calls before doing an arity check....