Aethenelle has quit [Quit: Aethenelle]
KeyJoo has joined #jruby
<CharlesOliverNut> hmm
<CharlesOliverNut> seems like matrix integration died again
<CharlesOliverNut> that's two strikes
<CharlesOliverNut> Relevant replies to you @guizmaii:
<headius[m]> Testing
Aethenelle has joined #jruby
<CharlesOliverNut> Lovely, integration is only sending messages one way
<headius[m]> kares: gitter integration is only receiving from gitter and not transmitting to gitter 😦
KeyJoo has quit [Quit: KeyJoo]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
<travis-ci> jruby/jruby (spooky_fibers:9e3e94d by Charles Oliver Nutter): The build has errored. https://travis-ci.org/jruby/jruby/builds/524775090 [240 min 10 sec]
Aethenelle has quit [Quit: Aethenelle]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
<travis-ci> jruby/jruby (master:1654de7 by Charles Oliver Nutter): The build has errored. https://travis-ci.org/jruby/jruby/builds/524623150 [217 min 44 sec]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
<travis-ci> jruby/jruby (master:54b875d by Charles Oliver Nutter): The build has errored. https://travis-ci.org/jruby/jruby/builds/524640907 [215 min 29 sec]
<travis-ci> jruby/jruby (master:cba96aa by Marcin Mielzynski): The build has errored. https://travis-ci.org/jruby/jruby/builds/524596947 [223 min 41 sec]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
RomainManni-Buca has joined #jruby
<RomainManni-Buca> hello guys
<RomainManni-Buca> is it true that last jruby version dont have the compiler mode option as an optim for startups - was forcing jruby 1.9 before?
<RomainManni-Buca> also is there any hack other than https://github.com/jruby/jruby/wiki/Improving-startup-time to really boost the startup? with asciidoctorj (just a jruby exec of asciidoctor (in ruby)) it is likely 3s to create the instance which is way too much (~500ms would be good)
shellac has joined #jruby
shellac has quit [Quit: Computer has gone to sleep.]
rusk has joined #jruby
shellac has joined #jruby
shellac has quit [Client Quit]
shellac has joined #jruby
shellac has quit [Ping timeout: 245 seconds]
<kares[m]> headius: meh, that is beta-ish :)
<kares[m]> anyway its annoying maybe if it persist we should just remove it, till it gets some 💌
<headius[m]> Yeah I know it is early days but it breaks one benefit of this, that I don't have to have three clients open. When it's unreliable I still have to.
<headius[m]> I guess it would be more fair to say that's two strikes for the integration feature, not for matrix as a whole
<JulesIvanicGitte> @headius thanks for the info :)
<CharlesOliverNut> @rmannibucau There are some JVM and JRuby flags you can pass to improve startup. Can you show me an example of a simple use of asciidoctorj so I can have a go at it?
<CharlesOliverNut> We have improvements coming up to try to cut startup time, but it's a difficult problem with any jvm tooling
lucasb has joined #jruby
<kares[m]> test
<kares[m]> yeah its anooying ... tried re-adding the bridge - seems to be not working much
<kares[m]> (seemingly it does not work at all)
<headius[m]> I think we should just drop it for now then...I'm getting double notifications one direction and the other way doesn't work
<lopex> wtf
<lopex> jruby -e "include Java; p org.joda.time.LocalTime.new.getHourOfDay()"
<lopex> 13
<lopex> jruby -e "p Time.now"
<kares[m]> yy its pbly best to try later
<lopex> 2019-04-26 13:28:38
<headius[m]> I'm definitely eager to have a fully open chat room system but the integration with gitter and IRC is a bit too flaky right now
<lopex> how does java read time zones ?
<kares[m]> gitter bridge definitely has problems atm: `Bridge returned HTTP 500 Internal Server Error: this.client.delete is not a function`
<lopex> date -> Fri Apr 26 15:29:34
<kares[m]> lopex: should read it from the system
<kares[m]> while JRuby does check ENV (TZ which Java does not)
<kares[m]> got 15 from `org.joda.time.LocalTime.new.getHourOfDay()` :)
<kares[m]> lopex: try the Java time API maybe it will read fine (compared to JODA)
<lopex> kares[m]: jruby matches java
<lopex> its' two hours off to system date
<kares[m]> so `jruby -e "p java.time.LocalTime.now.getHour()"` is 15 for you?
<lopex> 13
<kares[m]> heh
<lopex> it's under docker though
<lopex> which has - /etc/localtime:/etc/localtime:ro
<lopex> volume
<kares[m]> think its likely at UTC
<kares[m]> for whatever reason ...
shellac has joined #jruby
<headius[m]> hahah this.client.delete
<headius[m]> yeeeeah might want to look into that folks
<headius[m]> lopex: hey that MS950 guy got back to us
<headius[m]> On MRI it looks like it does somehow figure out to use CP950. Maybe we can just alias?
<lopex> oh there is ENC_REPLICATE("CP950", "Big5")
<headius[m]> Aha, and CP950 is there
<headius[m]> yeah
<headius[m]> so I saw we just add another entry for MS950 and call it done
<headius[m]> I say
<headius[m]> not I saw
<lopex> sp why the scripts didnt pick it up
<lopex> er
<headius[m]> MRI warns that it's an unknown encoding
<headius[m]> but it does CP950 after all
<lopex> I have no idea why would it work on mri
<headius[m]> 🤷🏽
<lopex> so there's some logic for it ?
<lopex> I'll grep for 950
<headius[m]> it warns but then does pretty much the right thing
<headius[m]> so I don't get it
<lopex> hah
<lopex> {"MS932",WINDOWS_31J},
<lopex> in nkf.c
<lopex> er, wait
<headius[m]> so there are some MS aliases already
<headius[m]> but not 950
<lopex> enc_name = rb_sprintf("CP%d", cp);
<lopex> win32ole.c
<headius[m]> ooo
<lopex> also
<lopex> localeinit.c
<lopex> #define CP_FORMAT(buf, codepage) snprintf(buf, sizeof(buf), "CP%u", (codepage))
<lopex> if (strncmp("CP", n, 2) == 0) { in file.c
<lopex> which path would that be ?
<headius[m]> ole_encoding2cp
<headius[m]> I see it
<lopex> headius[m]: there's more
<lopex> ^^
<headius[m]> so they are doing some parsing of CP entries and setting up the aliases a few different ways
<headius[m]> heh:
<headius[m]> spelling even
rusk has quit [Remote host closed the connection]
xardion has quit [Remote host closed the connection]
xardion has joined #jruby
<headius[m]> lopex: that hanging test
<headius[m]> it seems like it's in the teardown for the whole suite
<headius[m]> I have removed the last few files from the stdlib test run and it always hangs after the last thing
shellac has quit [Ping timeout: 250 seconds]
<lopex> headius[m]: I wonder if it could be replicated locally on a cpu stressed machine
<lopex> or a low priority vm
<headius[m]> yeah I'm stumped
<lopex> or maybe, how to force more frequent schedules ?
<headius[m]> I can punt the stdlib tests to allow_failures for now
<headius[m]> I don't have time at the moment to spin up a travis docker image and try to repro that way
<lopex> I mean, ousted ?
<lopex> er, preemption rather
<lopex> would it be any good ?
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:ebccbe2 by Charles Oliver Nutter): The build has errored. https://travis-ci.org/jruby/jruby/builds/525028141 [213 min 10 sec]
travis-ci has left #jruby [#jruby]
<headius[m]> lopex: so basically it twiddles the code to make it context switch at different places
<headius[m]> seems good...unsure if that can help us but it might be useful for something
<headius[m]> lopex: what do you think about MS950 thing?
<headius[m]> you don't want to manually add the alias because we generate it, yeah?
<lopex> yeah
<headius[m]> so there's some fallback logic I have not been able to find
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:44d0e0c by Charles Oliver Nutter): The build failed. https://travis-ci.org/jruby/jruby/builds/525033551 [226 min 30 sec]
travis-ci has left #jruby [#jruby]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
<travis-ci> jruby/jruby (master:44d0e0c by Charles Oliver Nutter): The build failed. https://travis-ci.org/jruby/jruby/builds/525033551 [251 min 51 sec]
<headius[m]> come on
<headius[m]> whatever I tagged that
<headius[m]> hmm
<headius[m]> I wonder if the $ is messing it up somehow
<lopex> cant find anything obovious and non static in these win32ole
<headius[m]> hmm
<lopex> actually how to repro that ?
<lopex> I have mri compiled on cygwin
<headius[m]> I'm not sure
<headius[m]> I tried to force LANG=MS950 locally and it doesn't break
<headius[m]> maybe file.encoding
<lopex> I mean, how to force the warning on mri
<headius[m]> probably doesn't actually need to be any content in the file
<headius[m]> the warning is happening before it reads
<lopex> ah, missed that
<headius[m]> $ jruby -J-Dfile.encoding=MS950 -e 'File.read("pom.rb", encoding: "MS950")'
<headius[m]> -e:1: warning: Unsupported encoding MS950 ignored
<headius[m]> WORKSFORME
<headius[m]> so I think the problem then is that we're not doing the same thing when booting up
<headius[m]> maybe this is easier than we thought
<lopex> starting from unsupported_encoding in io.c
<headius[m]> I thought that snippit also failed on JRuby
<headius[m]> ugh tagged but not added
<headius[m]> fixed
<headius[m]> lopex: I can't get it to fail on startup at all in any case
<headius[m]> OP can't even get JRuby to boot it seems
<headius[m]> lopex: did you get your riot account to link to your freenode nickserv account?
<lopex> no
<headius[m]> ok I guess I'm just seeing your irc messages then
<lopex> I end up #<Encoding:CP852>
<headius[m]> well that's interesting
<lopex> on jruby I get #<Encoding:Windows-1250>
<lopex> which is just default_external
<headius[m]> yeah I don't have a traditional chinese windows system to repro on
<headius[m]> presumably this person is from Taiwan
<lopex> but I'm kind of lost here
<lopex> requiring win32ole also doesnt change anything
<headius[m]> looks like parseModeEncoding is the method we use for File.read
<headius[m]> EncodingUtils
<lopex> maybe rubyinstaller bahaves differently ?
<lopex> we dont even know what MRI he uses
<headius[m]> no
<lopex> looking at #ifdef _WIN32 in io also doesnt give me anything obvious
<headius[m]> woah what's this
<headius[m]> his stack trace mentions JRubyParser
<headius[m]> we don't have a class named JRubyParser
<lopex> maybe it's theirs
<lopex> I dont know, I think we match mri and that's all
<headius[m]> lopex: yeah I think this might be something else, like OP is getting an older JRuby or something
<lopex> or well the diff is CP852 vs windows-1250
<headius[m]> oh!
<headius[m]> it's called JRubyParser in the external parser project
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:bc418e5 by Charles Oliver Nutter): The build was fixed. https://travis-ci.org/jruby/jruby/builds/525053378 [218 min 40 sec]
travis-ci has left #jruby [#jruby]
<headius[m]> lopex: let it go for now
<headius[m]> there's no way that JRubyParser should be in his stack unless he's doing something weird
<headius[m]> and it would be totally unsurprising if it's handling unusual encodings bad;y
<headius[m]> badly
<headius[m]> I just commented saying such
<headius[m]> ok that's one more task for today I can call done-ish
<headius[m]> and build is green again finally
<lopex> where to I link that freenode user ?
<lopex> on bridge options ?
<headius[m]> I don't know, I have just seen some vague posts about being able to do it
<lopex[m]> hmm
<headius[m]> And they mention it in their security breach postmortem... "if you gave your nickserv password to the bridge you should reset it" etc
<headius[m]> I have not figured out how to do that in the first place
<headius[m]> it would be nice to have it message as me instead of matrix me
subbu is now known as subbu|lunch
subbu|lunch is now known as subbu
jrafanie has joined #jruby
drbobbeaty has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]
jrafanie has quit [Quit: Textual IRC Client: www.textualapp.com]
lucasb has quit [Quit: Connection closed for inactivity]
drbobbeaty has joined #jruby
drbobbeaty has quit [Client Quit]