Timestamps are in GMT/BST.
[0:53] * Arfrever (~Arfrever@apache/committer/Arfrever) Quit (Quit: Ex??re)
[1:27] * stakkars (~tismer@p5DDB706E.dip.t-dialin.net) Quit (Quit: hasta la vista)
[1:43] * ctismer is now known as zz_ctismer
[1:45] * zz_ctismer is now known as ctismer
[3:02] * ctismer (~ctismer@django.stackless.com) Quit (Quit: ZNC - http://znc.in)
[3:07] * ctismer (~ctismer@django.stackless.com) has joined #jython
[3:07] * ctismer (~ctismer@django.stackless.com) Quit (Client Quit)
[3:10] * zz_ctismer (~ctismer@django.stackless.com) has joined #jython
[3:14] * zz_ctismer is now known as ctismer
[3:36] * ctismer (~ctismer@django.stackless.com) Quit (Quit: ZNC - http://znc.in)
[3:40] * zz_ctismer (~ctismer@django.stackless.com) has joined #jython
[3:43] * zz_ctismer is now known as ctismer
[5:59] * lopex (u4272@gateway/web/irccloud.com/x-qhgzgfcoseodriog) has joined #jython
[6:35] * seletz (~seletz@business-178-015-118-087.static.arcor-ip.net) has joined #jython
[6:36] <seletz> Good morining all.
[6:36] <seletz> bah, whad a bad start.
[6:36] * seletz gives up
[6:40] <agronholm> most of the team are probably sleeping right now
[7:46] * lheuer (~Adium@unaffiliated/lheuer) has joined #jython
[10:00] * juneau001 (~juneau@50.44.13.106) Quit (Quit: juneau001)
[11:06] * juneau001 (~juneau@131.225.24.76) has joined #jython
[12:32] * wainersm (~wainersm@189.61.204.85) has joined #jython
[13:12] * juneau001 (~juneau@131.225.24.76) Quit (Quit: Take care...)
[13:47] * lopex (u4272@gateway/web/irccloud.com/x-qhgzgfcoseodriog) Quit (Write error: Connection reset by peer)
[14:00] * enebo (~enebo@65-128-218-235.mpls.qwest.net) has joined #jython
[14:41] * lopex (u4272@gateway/web/irccloud.com/x-fnzduftdiffeyojc) has joined #jython
[16:09] <fwierzbicki> sabi: are you planning on looking at __format__ anymore? If not I will pick it up - Adconion has a special interest in using __format__ with floats.
[16:17] <sabi> fwierzbicki: go right ahead
[16:17] <fwierzbicki> sabi: ok cool
[16:20] <sabi> take a look at how i did things in http://hg.python.org/jython/rev/78c49afd03c9 - you should be able to do stuff with DecimalFormat like that
[16:30] <jimbaker> pjenvey, over the weekend i took a look at id - i can make a concurrent impl that comes within 5% of what's done currently in the single-threaded case. there might be a case where better concurrency would matter, but most likely not
[16:34] <jimbaker> so the best bet is impl do some sort of IdentityDict, as we discussed (which really is just a wrapper of MapMaker.weakKeys().weakValues().makeMap()...)
[16:44] <fwierzbicki> sabi: oh thanks for the pointer
[17:22] <fwierzbicki> sabi: it looks like float is working reasonably well except for some python/jython disagreement on the number of digits to display after the decimal point - is the missing stuff more along the lines of the decimal and datetime type?
[17:25] <jimbaker> fwierzbicki, sabi - is this due to the shortest repr for a float? http://bugs.python.org/issue1580
[17:26] <fwierzbicki> wow that's a lot of commentary :) -- it probably is related - in the case of 1.0 cpython prints 1.000000 and jython prints 1.0
[17:39] <jimbaker> fwierzbicki, i'm probably not following this closely enough, for trunk jython and cpython 2.7, repr(1.000000000000001) == '1.000000000000001' and repr(1.00000000000000001) = '1.0'
[17:39] <jimbaker> so that looks like the proper rounding is being done
[17:40] <fwierzbicki> sounds good - I'm purely worrying about representation from __format__ for now - though it's nice to know the rounding is good
[17:42] <jimbaker> fwierzbicki, i believe rounding is not the right term here - more about the shortest repr such that it would result in the same internal floating point bits
[17:42] <jimbaker> but regardless, this part looks done :)
[17:58] <fwierzbicki> yep that's my reading as well
[18:03] <sabi> fwierzbicki: the float specific formatting is broken entirely
[18:03] <sabi> e.g.:
[18:03] <sabi> >>> '{:5.3f}'.format(5.)
[18:03] <sabi> '5.000'
[18:03] <sabi> vs on jython right now:
[18:03] <sabi> '5.0 '
[18:03] <fwierzbicki> ah - ok
[18:04] <sabi> the '5' is handled by common code, the '.3' isn't. :)
[18:04] <fwierzbicki> got it, thanks!
[18:51] * enebo (~enebo@65-128-218-235.mpls.qwest.net) Quit (Quit: enebo)
[21:22] * wainersm (~wainersm@189.61.204.85) Quit (Quit: Ex-Chat)
[21:27] * kherron (438b3072@gateway/web/freenode/ip.67.139.48.114) has joined #jython
[21:27] * kherron is now known as KevinH
[21:29] <KevinH> So... I've run into a JVM crash on Windows when calling doing "new PySystemState()"... https://gist.github.com/2209863
[21:29] <KevinH> Anybody seen this before?
[21:31] <sabi> aiee, isatty. i'm at least partly to blame for that
[21:31] <sabi> what windows version?
[21:31] <sabi> oh sorry, i see it at the bottom
[21:32] <KevinH> Now for some context: this would be the *2nd* call to PySystemState.init() in the lifecycle of this JVM...
[21:34] <KevinH> This happens when a user logs out and then back in in our client runtime, which creates a new ScriptManager, which creates a new PySystemState... then the crash.
[21:47] <sabi> ok, well i can't seem to reproduce that particular issue so it may be related to the environment you're in
[21:47] <sabi> if you can produce a standalone test case, that'd be ideal...
[21:47] <sabi> either in python or java
[21:49] <sabi> the most likely thing i think is that something weird is going on with stdin/stdout/stderr, maybe one of them is closed and we're not checking for it properly
[21:52] <KevinH> Ok, I'll see what I can do.
[21:54] <sabi> worst comes to worst, you could comment out the isatty check and ship a custom version of jython. hopefully it won't come to that. :)
[21:57] <KevinH> oh man, i just realized we're futzing around with PySystemState's stdout and stdin variables
[21:57] <KevinH> might have worked when we were on 2.1, but not on 2.5 apparently...
[21:58] <sabi> yeah that was all rewritten in 2.5
[21:58] <KevinH> crap, yeah in 2.1 we replace PySystemState's instances with our own bifurcating output streams
[21:58] <KevinH> I bet if I leave them alone it won't crash...
[22:00] <sabi> if you're overwriting them, then you can override isatty() to do whatever you want then. the only reason it looks there is to decide what the console's encoding should be
[22:01] <sabi> if you return false to isatty() then it will use the object's encoding instead.
[22:01] <sabi> http://hg.python.org/jython/file/9c21ad4e5818/src/org/python/core/PySystemState.java#l239
[22:06] <KevinH> Yikes, we're also replacing System.out and System.err with our own OutputStream's elsewhere in the system
[22:06] <KevinH> I'll do some more poking around and get back to you... thanks for pushing me in the right direction
[22:07] <sabi> sure, no problem. good luck :)
[22:44] <KevinH> sabi: What is the console encoding used for? If I remove PYTHON_CONSOLE_ENCODING from the registry properties am I setting myself up for serious failure somewhere down the road?
[23:00] <sabi> KevinH: printing non-ascii characters
[23:01] <sabi> (or reading them)
These logs were automatically created by JythonLogBot_ on irc.freenode.net using a slightly modified version of the Java IRC LogBot (github).