justinmcp has quit [Remote host closed the connection]
justinmcp has joined #jruby
Puffball has quit [Remote host closed the connection]
Puffball has joined #jruby
Puffball has quit [Remote host closed the connection]
Puffball has joined #jruby
Puffball has quit [Quit: No Ping reply in 180 seconds.]
Puffball has joined #jruby
Puffball has quit [Quit: No Ping reply in 180 seconds.]
Puffball has joined #jruby
claudiuinberlin has joined #jruby
amitchellbullard has quit []
amitchellbullard has joined #jruby
phlebas has quit []
phlebas has joined #jruby
zph has quit []
zph has joined #jruby
drbobbeaty has quit [Ping timeout: 260 seconds]
flori has quit [Quit: leaving]
flori has joined #jruby
flori has quit [Client Quit]
flori has joined #jruby
flori has quit [Client Quit]
flori has joined #jruby
flori has joined #jruby
flori has quit [Quit: leaving]
Puffball has quit [Remote host closed the connection]
headius has quit [Ping timeout: 256 seconds]
claudiuinberlin has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
headius has joined #jruby
drbobbeaty has joined #jruby
claudiuinberlin has joined #jruby
Puffball has joined #jruby
Puffball has quit [Read error: Connection reset by peer]
bbrowning_away is now known as bbrowning
Puffball has joined #jruby
fidothe has quit []
fidothe has joined #jruby
Scorchin has quit []
Scorchin has joined #jruby
Puffball has quit [Remote host closed the connection]
Puffball has joined #jruby
Puffball has quit [Remote host closed the connection]
deathy has quit []
deathy has joined #jruby
aemadrid has quit []
aemadrid has joined #jruby
knowtheory has quit []
knowtheory has joined #jruby
flavorjones has quit []
flavorjones has joined #jruby
Puffball has joined #jruby
<GitHub195> [jruby] headius pushed 1 new commit to master: https://git.io/vAmzy
<GitHub195> jruby/master 7b0fbf0 Charles Oliver Nutter: Support JDK 8+, open-ended. This boots on 10 and 11 without warn.
<GitHub185> [jruby] headius pushed 1 new commit to master: https://git.io/vAmgG
<GitHub185> jruby/master acd9a53 Charles Oliver Nutter: Don't reference AnnotationBinder at runtime....
<GitHub162> [jruby] kares pushed 1 new commit to master: https://git.io/vAmg2
<GitHub162> jruby/master af82b10 kares: work-around TreeSet internals (used to track SortedSet's order)...
<GitHub15> [jruby] kares closed issue #5035: JRuby's native `SortedSet` lies about adding an object https://git.io/vAJ24
<headius> good morning!
<enebo> morning
Remo has joined #jruby
akp has joined #jruby
Remo has quit [Client Quit]
xardion has quit [Remote host closed the connection]
xardion has joined #jruby
<GitHub95> [jruby] headius pushed 1 new commit to jruby-9.1: https://git.io/vAmiS
<GitHub95> jruby/jruby-9.1 7cc201a Charles Oliver Nutter: Support Java 7+.
18VADAEEN is now known as adam12
snowp has quit []
snowp has joined #jruby
<GitHub156> [jruby] headius pushed 1 new commit to jruby-9.1: https://git.io/vAmFk
<GitHub156> jruby/jruby-9.1 02dbb67 Charles Oliver Nutter: Finish forcing method name in DynamicMethod constructors....
<headius> enebo, kares: Ok, that should be the last major commit for method names
<headius> it uses the presence of DynamicMethod.Version annotation to know whether to use the new constructor that takes name
<headius> if it is not present, it uses the old path but uses a thread local string to pass the name up another way
<enebo> sweetness
<headius> jossl generates both invokers AND populators right now, so it's still always passing null (which becomes method name "null" right now), but it doesn't error
<headius> there's still a compat issue in that if we regenerate those invokers they won't have a "name" super constructor on older JRuby
<headius> so that jossl would not work on 9.1.15 and earlier
<headius> in any case at the moment this will call old or new constructor safely so existing generated older constructors won't break
<enebo> yeah this is really nice for future work
<headius> have a look...I just went with an int instead of an enum or something
<headius> I doubt it will rev much after this
<GitHub196> jruby/jruby-9.1 7bb9781 Charles Oliver Nutter: Filter out new-style Java version numbers from ASM binding.
<GitHub196> [jruby] headius pushed 1 new commit to jruby-9.1: https://git.io/vAmAR
<enebo> yeah we probably maybe will once at best but a number seems very reasonable
<enebo> an enum would just be a number more or less anyways
<GitHub13> [jruby] headius pushed 1 new commit to jruby-9.1: https://git.io/vAmxZ
<GitHub13> jruby/jruby-9.1 9022774 Charles Oliver Nutter: Java 7 does not like this type mismatch for some reason.
<headius> yeah
<headius> kares: weird errors with those tests you added...it didn't like List<Object> = Arrays.asList(1, 2, 3) because the LHS wasn't List<Integer>
<headius> after all these years generics still confuse me sometimes
<headius> I guess it's a contravariant issue
<headius> a List<Object> may not be compatible with a List<Integer>
<headius> but List with no type param is ok
<nirvdrum> lopex: Were you able to reproduce the joni performance regression I mentioned last week?
<headius> 🤷‍♂️
<headius> ok, lunch break, bbiab
<headius> hopefully 9.1 should be mostly green this round
<lopex> nirvdrum: will look at it this week
<nirvdrum> lopex: No problem. I'll be curious if you can reproduce.
<lopex> nirvdrum: I hope so
<lopex> I mean I believe you there is, why is the question :P
<nirvdrum> Heh. I like to be thorough.
claudiuinberlin has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<enebo> lopex: can you add those missing ranges for EUC-JP?
<lopex> enebo: tomorrow
<lopex> enebo: not sure it would be better to extract them from binaries
<lopex> same for sjis
<enebo> lopex: yeah I guess I have no idea. We may need to extract data again in the future so that probably is a factor?
<lopex> yes
<GitHub114> [jruby] nomadium opened pull request #5039: Add FrozenError class (ruby-2.5...add-frozen-error-class) https://git.io/vAmjx
<headius> enebo, lopex: looks like only thing failing is those chr specs
<headius> if we can decide how to address those we can proceed with 9.1.16 this week
<enebo> still have 8 open .16 issues
<enebo> one of which I will punt to .17
<enebo> right now
<lopex> headius: I havent been able to reproduce that here
<headius> yeah
<headius> me neither
<headius> enebo: you are on linux, can you try to repro that?
<headius> I tried in a VM and was unsuccessful so seems something specific to travis?
<lopex> headius: tried on ubuntu
<enebo> will see
<headius> I have tags ready to go if we want to punt and investigate outside of 9.1.16
<enebo> yeah both those fail on FC27
<enebo> and Java 8
<headius> oh really?
<enebo> how is that even possible
<headius> they don't fail for me on ubuntu jdk8
<headius> I'm going to double check because that still seems strange
<enebo> openjdk version "1.8.0_151"
<enebo> I may be older 1.8 but still
<headius> something with default external encoding?
<headius> what's LANG?
<headius> en_US.UTF-8 here
<headius> Ubuntu 16.04, Java 8u131
<enebo> jruby -e '0x0100.chr("ISO-8859-9")'
<enebo> This is the repro for second error on travis
<enebo> This is all library code how could it change between OSes
<enebo> It is literally a long going into jcodings
<enebo> lopex: do some encodings get conditionally enabled depending on their environment?
<headius> still passes for me
<headius> 9.1 HEAD
<lopex> enebo: not that I know of
<headius> oh wait
<headius> I was testing String#chr specs
<headius> it's Integer#chr and it does fail for me
<enebo> yeah it would be crazy if it didn't at this point
<enebo> for all platforms
<enebo> lopex: are you sure my command-line above passes
<headius> oh good and it's failing on macos too
<headius> lopex: retry
<enebo> I really don't see how this could be anything other than bad data for 8859-9
<enebo> enc.codeToMbcLength(255) -> 1 instead of an exception in jcodings
<headius> yeah this may not be what we saw before with it failing for US-ASCII and things
<headius> we determined that was something fixed later in 2.4 and backported I think
<headius> enebo: oh that could be it
<headius> we have continued to eliminate exception throwing for errors because sometimes the error is expected and there's a fallback
<enebo> but it must be 100% failing on all platforms of jcodings is considering env in some way
<headius> I do see we also still have logic that allows 0x01 to 0xFF to use pregenerated bytelists
<enebo> by env I just mean the system it is running on
<enebo> for some reason rvm tells me I don't have 2.5 installed....mystery!
<enebo> headius: run that on 2.5 please
<lopex> it raises
<enebo> headius: other possibility is this should return 1
<enebo> and new data fixed the issue
<enebo> e.g. bad test
<headius> ah maybe
<headius> passes on 2.5
<enebo> ok so bad test I suspect
<lopex> headius: yeah, I checked against 2.5
<headius> enebo: how do you figure?
<enebo> passes on 2.5 so either 2.5 regressed 8859-1 data or MRI was wrong in this case
<enebo> however I also notice if I comment that out there is at least one other
<headius> 2.3 also passes
<headius> so I'm still not following
<headius> my 2.3 is newer than our baseline for 9.1 though
<enebo> I think I misunderstood the passes bit
<headius> passes as in doesn't fail? :-P
<enebo> you mean 2.5 still throws rangeerror or not?
<lopex> yes
<enebo> 2.3 does raise a rangeerror
<enebo> for me
<headius> it does not fail the test, so yes, it raises where we do not
<enebo> 9.1 does not
<headius> only 9.1 is wrong so far
<headius> only 9.1 does not pass the test
<enebo> ok I asked you to run my commandline and you replied with the spec test result
<enebo> I was just confused
<headius> eh?
<headius> oh, I was just running the spec
<enebo> yeah sorry I was talking about the command-line once I had it made so 'that' for us was different things
<headius> hence "passes"
<headius> ok
<enebo> passes for me was not raising
<enebo> in the context of running a CLI
<headius> heh ok let's never use that definition :-)
<lopex> same here
<headius> everyone seems to raise except us
<headius> and we used to raise before jcodings update
<lopex> headius: er, head matches mri 25. for me
<enebo> ok well that is better if you ask me
<lopex> 2.5
<enebo> it tells us jcodings just has a bug now
<lopex> both raise
<headius> lopex: 9.1 head raises?
<headius> or master head?
<enebo> vs something more nebulous like data changed
jeremyevans has quit [Ping timeout: 248 seconds]
<lopex> er, master
<lopex> jruby 9.2.0.0-SNAPSHOT (2.4.1) 2018-02-12 af82b10 Java HotSpot(TM) 64-Bit Server VM 25.131-b11 on 1.8.0_131-b11 +jit [linux-x86_64]
<lopex> RangeError: 256 out of char range
<headius> ok
<headius> so perhaps it's a stale chr port
<headius> master has a fresh chr because of Integer unification
<lopex> oh
<headius> that port likely uses return values properly for error checking while 9.1 doesn't
<headius> just a guess
<lopex> but sound reasonable
<lopex> sounds
<headius> right
<enebo> hahah
<enebo> if (0xff < i) {
<lopex> yes
<enebo> I see how it would range error unless defaultInternal exists
<enebo> am I the only person who finds the constant before < confusing
<headius> woah what the
<headius> ok = StringSupport.preciseLength(
<headius> ok is a boolean, preciseLength return an int
<enebo> no doubt defaultInternal gives the ability to transcode to something else
<headius> am I reading this right in RubyInteger?
<enebo> line 352
<headius> oh I see
<headius> == n
<headius> this all looks fishy
<enebo> geezus I was looking at the no-arg
<enebo> why are those two so different
<headius> reasons
<headius> I think I'm just going to copy chr impls from master sans unification logic (if any)
<headius> gah, ok that didn't work well
<headius> ok I see the right commit on master
<headius> reworked chr and cleaned up some EncodingUtils
<headius> ok those are working and now something else fails :-P
enebo has quit [Quit: Leaving.]
enebo has joined #jruby
jeremyevans has joined #jruby
<headius> got it
<GitHub96> [jruby] headius pushed 2 new commits to jruby-9.1: https://git.io/vAYI8
<GitHub96> jruby/jruby-9.1 cd56bd7 Charles Oliver Nutter: Match master single-byte cache.
<GitHub96> jruby/jruby-9.1 2d7b193 Charles Oliver Nutter: Squashed commit of the following:...
<GitHub155> [jruby] enebo closed pull request #5039: Add FrozenError class (ruby-2.5...add-frozen-error-class) https://git.io/vAmjx
<headius> I'm going to merge to master while this mess is fresh in my head
<GitHub161> [jruby] enebo pushed 3 new commits to ruby-2.5: https://git.io/vAYIh
<GitHub161> jruby/ruby-2.5 9b5753f Thomas E Enebo: Merge pull request #5039 from nomadium/add-frozen-error-class...
<GitHub161> jruby/ruby-2.5 83334b3 Miguel Landaeta: Add FrozenError as a subclass of RuntimeError...
<GitHub161> jruby/ruby-2.5 60993dd Miguel Landaeta: Update MRI tests where FrozenError replace RuntimeError
<headius> looks good though
<GitHub164> [jruby] headius pushed 1 new commit to master: https://git.io/vAYLG
<GitHub164> jruby/master 2bc3c87 Charles Oliver Nutter: Merge branch 'jruby-9.1'
<headius> enebo: kinda piecemeal cherry-picking and copying there but chr logic should match master and pass everything now
<enebo> headius: yeah hopefully we see green
<headius> ugh, open3
<enebo> chr is basic enough where I expect quite a bit of coverage
<headius> I will see if I can come up with a short term fix
<headius> I'm not going to get the windows process logic ported for 9.1.16 at this point
<GitHub39> [jruby] headius closed issue #4890: jruby uses absolute pathnames in batch files generated by rubygems https://git.io/vbRwJ
<headius> enebo: marked that RubyGems windows batch file thing as invalid
<headius> once they fix it in RubyGems we can upgrade, but it's not our bug
<enebo> yeah
<GitHub48> [jruby] headius closed issue #4896: Jruby + Rspec scoping issues on Classes with Prepended Modules https://git.io/vbrMI
<headius> enebo: 4 left
<headius> one waiting on mkristian to decide about relocating some packages in jruby-readline, one that kares self-assigned four days ago, the open3 issue, and the end_of_day thing
<headius> I confirmed end_of_day thing still broken on 9.1 head, have not checked master
<headius> I have to run for a bit
<enebo> ok
<enebo> we may want to examine if we dropped something not targetted too
drbobbeaty has quit [Read error: Connection reset by peer]
drbobbeaty has joined #jruby
<headius> yeah
<headius> I need to catch up my triage status
<headius> will try to do that this evening
akp has quit []
akp has joined #jruby
akp has quit [Remote host closed the connection]
akp has joined #jruby
akp has quit [Ping timeout: 240 seconds]
drbobbeaty has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
bbrowning is now known as bbrowning_away
rrutkowski has joined #jruby
rrutkowski has quit [Client Quit]
rrutkowski has joined #jruby
akp has joined #jruby
akp has quit [Ping timeout: 248 seconds]
rrutkowski has quit [Quit: rrutkowski]