<JRubyGithub>
[jruby] rejeep opened issue #3060: Empty intern array (%i[]) does not work http://git.io/vLWCr
JRubyGithub has left #jruby [#jruby]
enebo has joined #jruby
<bbrowning>
enebo: I think https://github.com/jruby/jruby/issues/3054 may be why jruby-master-spec-ruby is failing on the Process.spawn tests w/ heap space problems on CI
<enebo>
bbrowning: oh is this recent. last night while bbq’ing I was trying to run specs and they eventually crapped out with OOM
<enebo>
I was just going to see how I could be doing that this morning
<enebo>
hmmm well I think I still added my own OOM
<enebo>
I just stashed and I see the same issue
<bbrowning>
there's an infinite loop in spawn impl though :)
<bbrowning>
and it definitely chews up infinite memory, eventually
<enebo>
bbrowning: yeah
<enebo>
bbrowning: We can either spec or add the incr and hope for the best
<enebo>
err tag spec
mister_solo has joined #jruby
<bbrowning>
enebo: the loop actually makes no sense to me there anyway
<bbrowning>
err wait maybe it does
<bbrowning>
I dunno what this is really supposed to do
<bbrowning>
but either way, it needs either no loop, to increment the loop counter, or to break out of the loop
mister_s_ has joined #jruby
<enebo>
bbrowning: yeah I don’t know the behavior or what this should be doing entirely but it seems it should support n fd’s
mister_solo has quit [Ping timeout: 244 seconds]
<enebo>
heh…I figured out my problem last night
<enebo>
I think I was on 4.8Ghz wifi and my network was not working…there must be some specs which OOM if network is up but not working?
<enebo>
my network was not working because my patio never seems to see the 4.8 spectrum
Guest11186 has joined #jruby
mrmargolis has joined #jruby
drbobbeaty has joined #jruby
Guest11186 is now known as marr
drbobbeaty has quit [Client Quit]
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
sdogruyol has quit [Remote host closed the connection]
rsim has joined #jruby
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
sdogruyol has joined #jruby
rsim1 has quit [Ping timeout: 265 seconds]
drbobbeaty has joined #jruby
qmx is now known as qmx|z
drbobbeaty has quit [Client Quit]
drbobbeaty has joined #jruby
<brightball>
hey bbrowning, is there a good way to talk to the jRuby team about a consult on a project?
<bbrowning>
brightball: I dunno - contact them here or send them an email, I guess?
<brightball>
part of my job is maintaining a huge REE 1.8.7, Rails 2.3 legacy system with about 200,000 lines of code and I'd be really interested in getting some expert advice on switching it to jruby.
drbobbeaty has quit [Client Quit]
<brightball>
bbrowning: are you on the team? for some reason I thought you were
<bbrowning>
brightball: I contribute to jruby fairly often, but I'm not on the "jruby team" :)
<brightball>
well then, what good are you :-)
drbobbeaty has joined #jruby
qmx|z is now known as qmx
drbobbeaty has quit [Client Quit]
skade has joined #jruby
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
sdogruyol has quit [Remote host closed the connection]
kfpratt has joined #jruby
iamjarvo has joined #jruby
drbobbeaty has joined #jruby
qmx is now known as qmx|z
drbobbeaty has quit [Client Quit]
qmx has joined #jruby
mje113 has joined #jruby
sdogruyol has joined #jruby
skade has quit [Quit: Computer has gone to sleep.]
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
<bga57>
brightball: If you also want to convert your Rails system to Torquebox, bbrowning's probably the guy for you.
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] eregon pushed 1 new commit to master: http://git.io/vLWhP
<JRubyGithub>
jruby/master 1e55229 Benoit Daloze: [Truffle] Introduce BooleanCastWithDefaultNode which cast and allow a default for NotProvided....
JRubyGithub has left #jruby [#jruby]
drbobbeaty has joined #jruby
mrmargolis has quit [Remote host closed the connection]
rsim has quit [Quit: Leaving.]
drbobbeaty has quit [Client Quit]
mrmargol_ has joined #jruby
cristianrasch has quit [Quit: Leaving.]
cristianrasch has joined #jruby
cristianrasch has quit [Read error: Connection reset by peer]
cristianrasch has joined #jruby
deobalds has joined #jruby
subbu has joined #jruby
mrmargol_ has quit [Remote host closed the connection]
mrmargolis has joined #jruby
qmx has left #jruby ["Leaving"]
<brightball>
bga57: it was definitely on my radar
camlow325 has joined #jruby
iamjarvo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
camlow325 has quit [Remote host closed the connection]
GitGud has left #jruby [#jruby]
dinfuehr has joined #jruby
dinfuehr has quit [Ping timeout: 265 seconds]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] djberg96 opened pull request #3061: Add incrementor back PopenExecutor to prevent infinite loop. (master...master) http://git.io/vL8yE
JRubyGithub has left #jruby [#jruby]
deobalds has quit [Quit: Computer has gone to sleep.]
vlad___ has joined #jruby
x1337807x has joined #jruby
skade has quit [Quit: Computer has gone to sleep.]
vlad___ has quit [Quit: Page closed]
subbu is now known as subbu|lunch
x1337807x has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
wscheele has joined #jruby
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] chrisseaton pushed 2 new commits to truffle-head: http://git.io/vL8xr
<JRubyGithub>
jruby/truffle-head f156e5f Chris Seaton: Merge branch 'master' into truffle-head
<JRubyGithub>
jruby/truffle-head daf94b9 Chris Seaton: [Truffle] Update for latest Truffle snapshot.
JRubyGithub has left #jruby [#jruby]
<wscheele>
anyone familiar with status of ssl cipher support issues in jruby/jruby-openssl? or know of any potential workarounds? i am attempting to send some data to hipchat with logstash, but api.hipchat.com refuses the jruby ssl connection with a handshake_failure, because (from what i can see) they can't agree on a cipher
<chrisseaton>
kares might be able to help with that
<asarih>
chrisseaton: I am, but I need to get on the bus soon-ish.
<chrisseaton>
sorry to ask you a dumb question, but I've RTFMd and I can't figure out - can I run only some tests on a particular branch in Travis?
<chrisseaton>
So I have a separate branch, and I don't want to run all tests as most of them are redundant with the master branch
<asarih>
chrisseaton: you'll have to edit `.travis.yml` on that branch. merging would be a little troublesome, though.
<chrisseaton>
Ah that's a good idea - so it reads masters, sees what branches, and then reads again on the branch
<chrisseaton>
Merging is fine - this branch is a pain to merge anyway
<wscheele>
kares: could you perhaps s[ph]are some thoughts?
camlow325 has joined #jruby
<chrisseaton>
he may not be around - I'm sure he'll follow up on it when he can
<wscheele>
chrisseaton: thx, i hope so. tried some upgrades and looked at some code, but would prefer to have a workaround with released versions instead of having to hack jruby-openssl/jruby/logstash to make it all work...
camlow325 has quit [Remote host closed the connection]
camlow325 has joined #jruby
mrmargol_ has quit [Read error: Connection reset by peer]
_djbkd has quit [Remote host closed the connection]
_djbkd has joined #jruby
skade has joined #jruby
<wscheele>
chrisseaton / kares: turns out the ssl handshake errors were not related to needed cipher fixes. it seems the ftw library used by logstash builds an incompatible SSLContext; I attempted to run logstash with the 0.9.7 jruby-openssl fixes, and had the same issues. only after removing the configured sslcontext from the SSLSocket did the connection succeed. this also worked against the older jruby version.
x1337807x has joined #jruby
_djbkd has quit [Remote host closed the connection]
_djbkd has joined #jruby
bbrowning is now known as bbrowning_away
mrmargolis has joined #jruby
lanceball is now known as lance|afk
mrmargol_ has quit [Ping timeout: 244 seconds]
dinfuehr has joined #jruby
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] r6p opened issue #3062: Unable to java_import a class on jruby-9.0.0.0.rc1 http://git.io/vL4pR
JRubyGithub has left #jruby [#jruby]
temporal_ has quit [Ping timeout: 264 seconds]
Aethenelle has quit [Quit: Aethenelle]
rsim1 has quit [Quit: Leaving.]
skade has quit [Quit: Computer has gone to sleep.]
imperator has quit [Quit: Leaving]
subbu|busy is now known as subbu
dinfuehr has quit [Remote host closed the connection]
tcrawley is now known as tcrawley-away
dinfuehr has joined #jruby
mrmargol_ has joined #jruby
dinfuehr has quit [Ping timeout: 246 seconds]
temporalfox has joined #jruby
mrmargolis has quit [Ping timeout: 264 seconds]
mrmargol_ has quit [Remote host closed the connection]
_djbkd has quit [Remote host closed the connection]
_djbkd has joined #jruby
skade has joined #jruby
_djbkd has quit [Remote host closed the connection]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian pushed 3 new commits to jruby-1_7: http://git.io/vLBmY
<JRubyGithub>
jruby/jruby-1_7 78502ad Christian Meier: [build] no more jruby.home usage in build
<JRubyGithub>
jruby/jruby-1_7 85d972c Christian Meier: convert File.dirname("file:/my.jar!/a") to "file:/my.jar!" to avoid !// in uri...
<JRubyGithub>
jruby/jruby-1_7 0046bed Christian Meier: detect uri pattern and not windows drives on File.dirname
JRubyGithub has left #jruby [#jruby]
_djbkd has joined #jruby
mkristian has quit [Quit: Ex-Chat]
enebo has quit [Quit: enebo]
_djbkd has quit [Read error: Connection reset by peer]
_djbkd has joined #jruby
temporalfox has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
<Antiarc>
So I'm poking at startup times, testing with just require 'active_support/all'
<Antiarc>
Lots of exceptions. Those are costly in JRuby - I'm wondering if there might be lots of swallowed exceptions during gem requirement that are inflating boot times.
<Antiarc>
Small sample size, but that feels like a substantial improvement to me
<nirvdrum>
Yeah, that looks pretty good. I'm a bit confused as to what this was doing before, just by inspection.
<Antiarc>
What do you mean?
dinfuehr has joined #jruby
<Antiarc>
interestingly it doesn't seem to really improve boot time on a full Rails app
_djbkd has quit [Remote host closed the connection]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] bjfish pushed 1 new commit to master: http://git.io/vLBuh
<JRubyGithub>
jruby/master aca23bd Brandon Fish: [Truffle] Add resolv and logger test to mri test index
JRubyGithub has left #jruby [#jruby]
<Antiarc>
Maybe a little bit, though not as much as that 10-15%. But that's probably expected.
<nirvdrum>
Antiarc: It looks like the respond_to? check is there, but broken :-/
<Antiarc>
oh, where?
_djbkd has joined #jruby
<Antiarc>
with the patch, 5 loads of my app average 16.06 +- 0.70 sec. Baseline is 16.28s +- 1.09 sec. Not sure that's statistically significant. But I think it's probably worth fixing anyhow?
<nirvdrum>
RubyClass#finvokeChecked
<nirvdrum>
RubyClass.java:627
<nirvdrum>
So, that's supposed to check if the object responds to to_ary
<nirvdrum>
Then if you look at RubyClass.checkFuncallRespondTo (RubyClass.java:664)
<nirvdrum>
me.isBuiltin seems to always be true, unless overridden by the user.
<nirvdrum>
Maybe that part is okay, but if it is, the respond_to? call is never made.
<nirvdrum>
I'm really not entirely sure what it's supposed to be doing. But it completely ignores the respond_to? call if it's considered to be builtin.
<Antiarc>
Well, Gem::Specification isn't builtin, is it?
<Antiarc>
That's what is receiving to_ary here
<nirvdrum>
No. I think built_in here means where respond_to? is defined.
<Antiarc>
ahhh
<Antiarc>
NOEX_BASIC seems to be gone in ruby-head. Hm.
<Antiarc>
still there in 2.2 though
<nirvdrum>
Yeah. I'm still tracing. It's strange behavior.
<nirvdrum>
Bah. I guess this is compliant :-/
<Antiarc>
huh, really?
<nirvdrum>
class X; def method_missing(*args); [:a, :b, :c]; end; end; p [X.new].flatten
dinfuehr has quit [Remote host closed the connection]
<nirvdrum>
So Rubinius seems to be broken here.
<Antiarc>
well, going back to backtrace suppression, is there any facility for doing that?
<Antiarc>
Since IIRC that's the slow thing here
<nirvdrum>
Can you see where the trace is generated from?
<nirvdrum>
I don't see it happening when stepping through the debugger.
<Antiarc>
Yeah, give me a sec
<Antiarc>
I just set a breakpoint in...let me find it
<Antiarc>
RaiseException.preRaise
<nirvdrum>
I wonder if Gem::Specification is throwing it.
<Antiarc>
It was coming from specification.rb, yeah
<Antiarc>
updated it with the Java track trace too
<nirvdrum>
So, the respond_to? check might be fine. It's weird, but it's what MRI does it seems.
<Antiarc>
well, in cases where to_ary is handled with method_missing, but doesn't provide respond_to? it'll fail, no?
<nirvdrum>
Once it gets past that case, it checks if method_missing is built-in and essentially no-ops if it is.
<Antiarc>
Ahh, okay
<nirvdrum>
So if you just did ["a"].flatten, you won't hit the exception.
<nirvdrum>
The problem is Gem::Specification defines method_missing, so it's called.
<nirvdrum>
But their method_missing essentially calls the parent method, which would be built-in.
rcvalle has quit [Quit: rcvalle]
subbu has quit [Ping timeout: 252 seconds]
<nirvdrum>
So, unfortunately, patching RubyArray seems like the wrong way to go.
<lopex>
I think one might turn of trace in fillInStackTrace conditionaly for internal purposes ?
<nirvdrum>
Patching RubyGems is one way to go.
<nirvdrum>
I don't know if IR is smart enough to catch the super call and take different behavior based upon the call context.
<Antiarc>
Yeah, patching rubygems would be fine, but map { ... }.flatten is common enough that it seems like it might be a potential performance optimization if it's fixable in the general case