#jython IRC Log


IRC Log for 2014-05-19

Timestamps are in GMT/BST.

[0:25] * paolo (~paolo@unaffiliated/paolo) Quit (Quit: Leaving)
[4:03] <fwierzbicki> agronholm: I'm building them now, it is still the weekend for me :)
[4:11] <fwierzbicki> ok soft release is out
[5:35] * robbyoconnor (~wakawaka@guifications/user/r0bby) has joined #jython
[7:30] * helix84 (~ctenar@ Quit (Remote host closed the connection)
[7:45] * paolo (~paolo@unaffiliated/paolo) has joined #jython
[8:10] * ebarrett (~edd@host-78-148-183-214.as13285.net) Quit (Ping timeout: 252 seconds)
[9:03] * lheuer1 (~Adium@g227051026.adsl.alicedsl.de) has joined #jython
[9:04] * lheuer1 (~Adium@g227051026.adsl.alicedsl.de) Quit (Client Quit)
[10:06] <agronholm> fwierzbicki: I'll consider it released when it appears on maven :)
[10:08] <agronholm> fwierzbicki: just noticed something...2.7b2 is twice as large on maven as b1 was...how come?
[10:16] <paolo> hi, I'm having a trouble with subprocess, can you have a look please?
[10:16] <paolo> https://gist.github.com/paolodina/b98c3f3a159024584e13
[10:16] <paolo> using cpython there are no issues
[10:17] <agronholm> paolo: have you checked what args go to subprocess?
[10:18] <paolo> not yet, for that I have to change Lib/subprocess.py and set a breakpoint with pdb
[10:18] <paolo> I'm going to do that
[10:18] <agronholm> or werkzeug/serving.py
[10:19] <paolo> ah ok
[10:19] <agronholm> why are subprocesses involved here anyway
[10:19] <agronholm> ohh
[10:19] <agronholm> the reloader
[10:19] <paolo> yes!
[10:28] <paolo> mmm args is None
[10:29] <agronholm> find out why
[10:38] <paolo> I was wrong saying args is None. args is defined as:
[10:38] <paolo> args = [sys.executable] + sys.argv
[10:38] <paolo> which evaluates to
[10:38] <paolo> [None, 'app/hello.py']
[10:39] <agronholm> paolo: how are you launching jython?
[10:39] <paolo> java -jar jython-standalone-2.7-b2.jar app/hello.py
[10:39] <agronholm> you're doing it wrong
[10:40] <agronholm> don't use the standalone
[10:40] <agronholm> install jython on a directory and use the command line launcher
[10:40] <paolo> ok
[10:41] <agronholm> the standalone jar is for those who embed it to existing JVMs
[11:02] <paolo> I downloaded Jython 2.7beta2 - jython.jar Installer, but how do I "install" it?
[11:03] <paolo> java -jar jython-2.7-b2.jar doesn't run an installation script/gui but starts the jython console.
[11:04] <agronholm> paolo: download the installer instead
[11:04] <agronholm> http://repo1.maven.org/maven2/org/python/jython-installer/2.7-b2/jython-installer-2.7-b2.jar
[11:04] <paolo> hope not to be blind but I can't find it looking at http://jython.org/downloads.html
[11:05] <agronholm> paolo: the links seem to be wrong on the home page
[11:06] <agronholm> ugh, someone already tweeted about the python 2.7b3 release :/
[11:12] <paolo> is that bad? I see it's been already tagged on bitbucket
[11:13] <agronholm> yeah but b2 ended up being "soft released" several times :/
[11:13] <agronholm> until it finally made its way to maven
[11:13] <paolo> hmmm.. understand
[12:29] * ivan` (~ivan@unaffiliated/ivan/x-000001) Quit (Quit: ERC Version 5.3 (IRC client for Emacs))
[12:31] * ivan` (~ivan@unaffiliated/ivan/x-000001) has joined #jython
[13:09] <paolo> $ virtualenv -p ./jython/2.7b2/bin/jython baz
[13:09] <paolo> Running virtualenv with interpreter ./jython/2.7b2/bin/jython
[13:09] <paolo> Cannot find file /home/paolo/jdevel/golap-workspace/jython/2.7b2/Include (bad symlink)
[13:10] <paolo> do you know what's wrong?
[13:36] <peke> agronholm: fwierzbicki: i was also somewhat confused with various beta2 releases.
[13:37] <agronholm> paolo: no
[13:37] <agronholm> peke: I'd rather wait for beta3 to test that
[13:39] <peke> agronholm: i assume the reason b2 id so much bigger than b1 is that it contains more/bigger dependencies.
[13:39] <agronholm> peke: but doubling in size from beta1 doesn't compute
[13:39] <peke> i think jimbaker added some jar for unicode handling that was about 10mb alone
[13:39] <agronholm> really??
[13:41] <peke> i think it was this one: http://icu-project.org/download/4.0.html
[13:42] <peke> well, most likely some newer release but that project anyway.
[14:20] <fwierzbicki> agronholm: I know there are new dependencies in b2 vs b1 - but I didn't think it would make it twice as large??? weird
[14:20] <agronholm> I hope that size increase at least buys us CJK support
[14:21] <fwierzbicki> it could be??? jimbaker would know
[14:21] <agronholm> fwierzbicki: anyway, what's the situation with b3
[14:21] <fwierzbicki> Jim wanted to do a soft release first - maybe it is too confusing to do that - though I only announce that to devs for that reason. This one should not have so many tries
[14:22] <fwierzbicki> maybe just the one
[14:26] <fwierzbicki> the soft releases are to avoid things like the way trunk wasn't running on windows ??? a beta with that level of a problem is no good.
[14:27] <agronholm> that only helps if someone is actually testing the beta on windows...
[14:27] <agronholm> in my experience the only reason to get people to test it is to release it
[14:27] <agronholm> s/reason/way/
[15:28] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) has joined #jython
[15:51] * enebo (~enebo@c-75-73-8-169.hsd1.mn.comcast.net) has joined #jython
[15:53] <fwierzbicki> agronholm: luckily these days we have an active core dev that works primarily on windows (Jeff Allen) so it works out
[15:54] <fwierzbicki> In any case I think the last group was an anomaly - it was a year between betas. This one should go quicker and smoother I bet.
[16:44] <jimbaker> fwierzbicki, agronholm - right, icu4j gets us full unicodedata support
[16:44] <jimbaker> plus related things like idna
[16:45] <jimbaker> we used to support some of unicodedata, but not CJK aspects for example
[16:48] <jimbaker> so the additions are icu4j (10M) bouncycastle (2M), netty (1M)
[16:51] <jimbaker> ideally once we look at maven packaging we can reduce ship dependencies. but *compatibility first*
[17:00] <agronholm> what's in icu4j that takes so much space
[17:14] <jimbaker> well, it's a pretty comprehensive library for managing all sorts of internationalization type issues. we only use a fraction of what it supports of course, but it's not available in just the stuff we need
[17:16] <jimbaker> it would be reasonably straightforward to write just we need in unicodedata, but i don't see this as an immediate priority - it's only 10M for icu4j
[17:16] <jimbaker> contributions are of course welcome ;)
[17:16] * robbyoconnor (~wakawaka@guifications/user/r0bby) Quit (Excess Flood)
[17:17] * robbyoconnor (~wakawaka@guifications/user/r0bby) has joined #jython
[17:28] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) Quit ()
[17:29] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) has joined #jython
[17:40] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) Quit ()
[17:51] * robbyoconnor (~wakawaka@guifications/user/r0bby) Quit (Ping timeout: 265 seconds)
[17:55] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) has joined #jython
[18:02] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) Quit ()
[18:03] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) has joined #jython
[19:11] * enebo (~enebo@c-75-73-8-169.hsd1.mn.comcast.net) Quit (Quit: enebo)
[19:55] * Arfrever (~Arfrever@apache/committer/Arfrever) Quit (Ping timeout: 258 seconds)
[19:57] <paolo> can't figure out why os.umask gives AttributeError: 'module' object has no attribute 'umask'
[19:57] <paolo> on linux
[19:57] <paolo> any idea?
[20:00] * agronholm_ (~agronholm@2001:1bc8:102:6f29:69bc:2b76:488:afb7) has joined #jython
[20:00] * ChanServ sets mode +o agronholm_
[20:02] * agronholm (~agronholm@2001:1bc8:102:6f29:6d7b:a7c2:d087:3f3e) Quit (Quit: ZNC - http://znc.in)
[20:02] * agronholm_ is now known as agronholm
[20:06] * enebo (~enebo@mobile-198-228-233-150.mycingular.net) has joined #jython
[20:28] * robbyoconnor (~wakawaka@guifications/user/r0bby) has joined #jython
[20:45] * robbyoconnor (~wakawaka@guifications/user/r0bby) Quit (Remote host closed the connection)
[20:46] * Arfrever (~Arfrever@apache/committer/Arfrever) has joined #jython
[20:57] * enebo (~enebo@mobile-198-228-233-150.mycingular.net) Quit (Quit: enebo)
[21:28] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) Quit (Ping timeout: 255 seconds)
[21:29] * enebo (~enebo@c-75-73-8-169.hsd1.mn.comcast.net) has joined #jython
[22:07] <jimbaker> paolo, maybe it's because JNR is not available?
[22:08] <jimbaker> (JNR = java native runtime, could not be available because of processor architecture or because of your security manager)
[22:09] <paolo> jimbaker: it could be, how can I check?
[22:10] <jimbaker> paolo, i would start by looking at POSIXPythonHandler, which you should be able to look at from jython itself
[22:26] <paolo> jimbaker: I see that jython.jar contains org/python/modules/posix/POSIXPythonHandler.class
[22:27] <jimbaker> sure, the question is whether it can actually be used or if it just falls back to standard java-supported ops
[22:27] * enebo (~enebo@c-75-73-8-169.hsd1.mn.comcast.net) Quit (Quit: enebo)
[22:27] <paolo> I don't know
[22:41] * fwierzbicki (~Adium@99-106-170-162.lightspeed.sntcca.sbcglobal.net) Quit (Quit: Leaving.)
[23:00] * enebo (~enebo@c-75-73-8-169.hsd1.mn.comcast.net) has joined #jython
[23:02] * enebo (~enebo@c-75-73-8-169.hsd1.mn.comcast.net) Quit (Client Quit)
[23:09] * fwierzbicki (~Adium@85.sub-70-197-12.myvzw.com) has joined #jython
[23:10] <pjenvey> you can check os._native_posix, if it's False then it's the problem jimbaker is describing
[23:11] <pjenvey> paolo - ^^
[23:12] <paolo> >>> os._native_posix
[23:12] <paolo> False
[23:13] <paolo> pjenvey: thanks for the quick way to check the native posix stuff. Any suggestion to fix that?
[23:14] <paolo> I tried to disable verification of mixed code using java control panel but nothing changed
[23:18] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) has joined #jython
[23:19] <pjenvey> are you using the latest version, i think the jnr version was just reverted back
[23:21] <pjenvey> def try jython trunk
[23:23] <paolo> I'm using 2.7b2
[23:26] <paolo> I'll try the trunk
[23:27] <paolo> even better, I'll wait 2.7b3 is officially released
[23:27] * chrisseaton (~chrisseat@79-76-253-47.dynamic.dsl.as9105.com) Quit (Ping timeout: 252 seconds)
[23:32] <pjenvey> if you try trunk you can ensure that 2.7b3 will work fine for you =]
[23:33] <pjenvey> are you on a plain linux x86-64?
[23:36] <paolo> it still doesn't work :(
[23:36] <paolo> paolo@tk:~$ uname -a
[23:36] <paolo> Linux tk 3.2.0-61-generic-pae #93-Ubuntu SMP Fri May 2 21:46:08 UTC 2014 i686 i686 i386 GNU/Linux
[23:37] <paolo> (ubuntu 12.04)
[23:38] <pjenvey> can you also try this? run jython -vv, and see if a stack trace is displayed near a "Failed to load native POSIX impl"
[23:38] <pjenvey> message
[23:40] <paolo> yes sir, there is a stack trace
[23:40] <paolo> import: trying org.python.modules.posix.PosixModule as builtin modules in SysPathJavaLoader
[23:40] <paolo> import: 'posix' as org.python.modules.posix.PosixModule in builtin modules
[23:40] <paolo> Failed to load native POSIX impl; falling back on Java impl. Stacktrace follows.
[23:40] <paolo> java.lang.NoClassDefFoundError: Could not initialize class jnr.posix.POSIXFactory$LinuxLibCProvider$SingletonHolder
[23:47] <paolo> full stacktrace: https://gist.github.com/paolodina/2374900b94d841602339
[23:50] <pjenvey> now that's a funny one
[23:57] <paolo> not sure it is ;)
[23:59] <pjenvey> is there anything else odd about your environment, no security manager in use, are you not using oracle's jvm? i'm a bit stumped


These logs were automatically created by JythonLogBot on irc.freenode.net using a slightly modified version of the Java IRC LogBot.