| eglp (ikea@trimix.eklof.eu) left irc: Ping timeout: 246 seconds | 00:03 | |
| eglp (ikea@trimix.eklof.eu) joined #pypy. | 00:05 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Quit: Gotsta go | 00:07 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 00:07 | |
| pedronis (~pedronis@73-53.195-178.cust.bluewin.ch) left irc: Ping timeout: 260 seconds | 00:08 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Client Quit | 00:09 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 00:10 | |
| pedronis (~pedronis@73-53.195-178.cust.bluewin.ch) joined #pypy. | 00:13 | |
| lesshaste (~lesshaste@2607:4000:200:12:213:72ff:fe3b:133e) left irc: Quit: Leaving | 00:20 | |
| nedbat (~nedbat@c-24-60-250-145.hsd1.ma.comcast.net) joined #pypy. | 00:26 | |
| nedbat (~nedbat@c-24-60-250-145.hsd1.ma.comcast.net) left irc: Changing host | 00:26 | |
| nedbat (~nedbat@python/psf/nedbat) joined #pypy. | 00:26 | |
| Nick change: Unhelpful_ -> Unhelpful | 00:30 | |
| unbit (~unbit@93-58-162-185.ip159.fastwebnet.it) left irc: Quit: unbit | 00:32 | |
| fprimex (~fprimex@fl-67-235-253-218.dhcp.embarqhsd.net) joined #pypy. | 00:37 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/1135 | 00:42 |
|---|---|---|
| umgeher (~umgeher@unaffiliated/umgeher) left irc: Quit: ... | 00:43 | |
| sysfault (~exalted@p3m/member/sysfault) left irc: Ping timeout: 260 seconds | 00:44 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/879 | 00:45 |
| sysfault (~exalted@p3m/member/sysfault) joined #pypy. | 00:49 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64-2/builds/142 | 00:53 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/cpython-2-benchmark-x86-64/builds/88 | 00:53 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/2171 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/434 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/862 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/512 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/792 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1331 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/964 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1624 | 01:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/913 | 01:00 |
| Kxepal (~Miranda@ppp91-79-2-250.pppoe.mtu-net.ru) joined #pypy. | 01:06 | |
| dmishe (~dmishe@c-98-206-42-112.hsd1.il.comcast.net) left irc: Quit: Textual IRC Client: http://www.textualapp.com/ | 01:08 | |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-03.NATPOOL.NYU.EDU) joined #pypy. | 01:30 | |
| ericflo (~ericflo@75.103.8.106) left irc: Quit: ericflo | 01:31 | |
| nettok (~quassel@190.148.36.184) joined #pypy. | 01:44 | |
| nettok (~quassel@190.148.36.184) left irc: Remote host closed the connection | 01:49 | |
| bentrofatter (~quassel@c-98-206-160-137.hsd1.il.comcast.net) left irc: Remote host closed the connection | 01:54 | |
| Moku (~John@osbk-4db0638c.pool.mediaWays.net) joined #pypy. | 02:04 | |
| Nick change: Moku -> Guest50369 | 02:04 | |
| Shanita (~John@osbk-4d08763f.pool.mediaWays.net) left irc: Ping timeout: 260 seconds | 02:04 | |
| kvda (~kvda@202.58.240.18) joined #pypy. | 02:15 | |
| gtaylor (~gtaylor@108-196-160-156.lightspeed.tukrga.sbcglobal.net) left irc: Quit: Konversation terminated! | 02:18 | |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-03.NATPOOL.NYU.EDU) left irc: Ping timeout: 246 seconds | 02:20 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Quit: Gotsta go | 02:34 | |
| Ademan (~dan@adsl-71-141-244-5.dsl.snfc21.pacbell.net) left irc: Ping timeout: 240 seconds | 02:35 | |
| aat (~aat@pool-108-41-149-224.nycmny.fios.verizon.net) left irc: Quit: Computer has gone to sleep. | 02:36 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/321 | 02:36 |
| oiig_ (~eohnik@112.161.134.227) joined #pypy. | 02:37 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 02:38 | |
| GregBowyer (~GregBowye@205.209.7.101) joined #pypy. | 02:44 | |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-02.NATPOOL.NYU.EDU) joined #pypy. | 02:52 | |
| aat (~aat@pool-108-41-149-224.nycmny.fios.verizon.net) joined #pypy. | 02:55 | |
| dmalcolm (~david@c-24-61-45-39.hsd1.ma.comcast.net) left irc: Quit: Leaving | 02:56 | |
| Nick change: chesttone -> waisttone | 02:57 | |
| nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 276 seconds | 03:08 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Read error: Connection reset by peer | 03:15 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 03:15 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Client Quit | 03:17 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 03:17 | |
| craigkerstiens (~textual@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Quit: Computer has gone to sleep. | 03:18 | |
| craigkerstiens (~textual@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy. | 03:28 | |
| FunkyBob (curtis@li45-68.members.linode.com) left #pypy. | 03:35 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/2171 | 03:39 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/2172 [py3k] | 03:39 |
| craigkerstiens (~textual@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Quit: Computer has gone to sleep. | 03:40 | |
| GregBowyer (~GregBowye@205.209.7.101) left irc: Ping timeout: 248 seconds | 03:48 | |
| aat (~aat@pool-108-41-149-224.nycmny.fios.verizon.net) left irc: Quit: Computer has gone to sleep. | 03:50 | |
| fprimex (~fprimex@fl-67-235-253-218.dhcp.embarqhsd.net) left irc: Quit: Leaving | 03:55 | |
| jesyspa (~s@s559738e3.adsl.wanadoo.nl) joined #pypy. | 04:03 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/512 | 04:06 |
| craigkerstiens (~textual@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy. | 04:10 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Quit: Gotsta go | 04:10 | |
| kvda (~kvda@202.58.240.18) left irc: Quit: x__x | 04:15 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 04:17 | |
| GregBowyer (~GregBowye@76.89.130.43) joined #pypy. | 04:25 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Quit: Gotsta go | 04:26 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 04:26 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Quit: Gotsta go | 04:32 | |
| kvda (~kvda@202.58.240.18) joined #pypy. | 04:35 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy. | 04:40 | |
| ericflo (~ericflo@c-50-131-58-97.hsd1.ca.comcast.net) joined #pypy. | 04:47 | |
| ericflo (~ericflo@c-50-131-58-97.hsd1.ca.comcast.net) left irc: Client Quit | 04:47 | |
| unbit (~unbit@93-58-162-185.ip159.fastwebnet.it) joined #pypy. | 04:50 | |
| Nick change: ctismer -> ctismer_afk | 04:57 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/964 | 04:58 |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1624 | 05:00 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1331 | 05:14 |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) joined #pypy. | 05:14 | |
| yumike (~yumike@213.154.168.130) joined #pypy. | 05:14 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/792 | 05:17 |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/862 | 05:19 |
| [mat^2] (~Mathias@212.130.113.35) left irc: Quit: Leaving | 05:40 | |
| bethebunny (~bunny@216-45-228-37-minneapolis-mn.hfc.comcastbusiness.net) left irc: Quit: leaving | 05:57 | |
| xcombelle (~xcombelle@AToulouse-551-1-84-80.w92-149.abo.wanadoo.fr) left irc: Read error: Connection reset by peer | 06:07 | |
| mattip (4fb376f1@gateway/web/freenode/ip.79.179.118.241) joined #pypy. | 06:11 | |
| mattip | fwiw: I "discovered" why the buildbot/master.py was modified by hand: | 06:13 |
| mattip | of the four win32 buildbots, only aurora is reliably connected, | 06:13 |
| mattip | SalsaSalsa is one I run and my family is forever rebooting the computer | 06:14 |
| mattip | snakepit32 hasn't been seen in a while | 06:14 |
| mattip | and bigboard cannot successfully translate: it hits the 2GB memory wall and dies. | 06:15 |
| mattip | anyone with admin rights on the build master around? | 06:15 |
| mattip | I created a win32-cleanup branch on the pypy/buildbot repository | 06:19 |
| mattip | if someone would be so kind as to review and merge ... | 06:19 |
| mattip | then retart the master | 06:19 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/434 | 06:19 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/913 | 06:20 |
| AlekSi (~AlekSi@85.235.191.82) joined #pypy. | 06:20 | |
| plainflavored (~headwiz@c-76-115-50-141.hsd1.or.comcast.net) joined #pypy. | 06:21 | |
| LarstiQ | mattip: I can't help I'm afraid. What timezone are you in? | 06:21 |
| mattip | israel | 06:21 |
| mattip | it's OK, I'm used to talking to myself. :) | 06:22 |
| ronny | lol | 06:22 |
| ronny | moin | 06:22 |
| ronny | hmm | 06:22 |
| vad (~vad@77.72.196.122) joined #pypy. | 06:24 | |
| ronny | mep | 06:25 |
| ronny | mattip: just took a look at buildbot master dir, back when done vomiting | 06:25 |
| mattip | hehe. | 06:25 |
| ronny | mattip: but yes, i can restart it later | 06:32 |
| mattip | ok. Should I merge the branch or will you? | 06:32 |
| kvda (~kvda@202.58.240.18) left irc: Quit: x__x | 06:34 | |
| ronny | mattip: done | 06:36 |
| mattip | :) | 06:36 |
| Var (~Varriount@pool-96-255-203-197.washdc.fios.verizon.net) left irc: Ping timeout: 260 seconds | 06:37 | |
| kenaan | 12RonnyPfannschmidt buildbot[win32-cleanup] 1134a513c1e44a 15/: close branch befor merge | 06:38 |
| kenaan | 12RonnyPfannschmidt buildbot 1164eeefddd046 15/bot2/pypybuildbot/: merge from win32-cleanup | 06:38 |
| bbot2 (~bbot2@wyvern.cs.uni-duesseldorf.de) left irc: Quit: buildmaster reconfigured: bot disconnecting | 06:39 | |
| bbot2 (~bbot2@wyvern.cs.uni-duesseldorf.de) joined #pypy. | 06:39 | |
| ronny | mattip: done | 06:39 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/2173 [py3k] | 06:40 |
| mattip | ronny: cool, thanks. | 06:40 |
| mattip | ronny: something's not right: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32 did not change. | 06:43 |
| mattip | hg pull? | 06:43 |
| ronny | mattip: forgot hg up | 06:44 |
| bbot2 (~bbot2@wyvern.cs.uni-duesseldorf.de) left irc: Quit: buildmaster reconfigured: bot disconnecting | 06:44 | |
| bbot2 (~bbot2@wyvern.cs.uni-duesseldorf.de) joined #pypy. | 06:45 | |
| mattip | that's better. | 06:45 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/513 [12mattip] | 06:53 |
| mattip (4fb376f1@gateway/web/freenode/ip.79.179.118.241) left irc: Quit: Page closed | 06:54 | |
| kenaan | 12arigo extradoc 111fc3d5240cd4 15/blog/draft/stm-apr2012.rst: Precision w.r.t. the supported platforms. | 06:56 |
| arigato (~arigo@adsl-89-217-225-244.adslplus.ch) joined #pypy. | 07:04 | |
| antocuni (~antocuni@host204-125-dynamic.11-79-r.retail.telecomitalia.it) joined #pypy. | 07:04 | |
| fijal | arigato: looks good | 07:08 |
| fijal | arigato: although as usual maybe a sentence of introduction would help here and there | 07:08 |
| fijal | like expand what it means "not done by far" | 07:09 |
| arigato | ok :-) | 07:09 |
| fijal | since I have no idea how can the GC be not quite done but working | 07:09 |
| fijal | well | 07:10 |
| fijal | I do | 07:10 |
| arigato | well, it's leaking a bit, as I said | 07:10 |
| fijal | but others might not | 07:10 |
| fijal | ok | 07:10 |
| fijal | so it's leaking a bit, because ... | 07:10 |
| fijal | and some expansion would not hurt | 07:10 |
| fijal | not because you have bugs, but because you have unsupported features right? | 07:10 |
| arigato | well I don't know how much details I should go into | 07:11 |
| arigato | not quite right | 07:11 |
| fijal | ok | 07:11 |
| arigato | what's going on is that when entering transactional mode, and between any two transactions, all objects become immortal | 07:11 |
| kvda (~kvda@202.58.240.18) joined #pypy. | 07:11 | |
| arigato | if your program only calls transaction.run() once and transactions don't create tons of surviving objects, then it's a slow leak | 07:12 |
| kenaan | 12fijal extradoc 11e2615b4dd640 15/blog/draft/stm-apr2012.rst: explain | 07:12 |
| fijal | also what' a pure python transaction module | 07:12 |
| arigato | link? | 07:12 |
| arigato | I mean | 07:12 |
| arigato | just following the link, that's what it is? | 07:12 |
| arigato | I can explain more | 07:13 |
| fijal | you can explain that it's an emulation for example | 07:13 |
| arigato | indeed | 07:14 |
| arigato | you have a point :-) | 07:14 |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) joined #pypy. | 07:17 | |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) joined #pypy. | 07:24 | |
| fijal (~fijal@178-36-39-69.adsl.inetia.pl) left irc: Ping timeout: 240 seconds | 07:30 | |
| lewq (~luke@127cr.net) joined #pypy. | 07:34 | |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) left irc: Ping timeout: 252 seconds | 07:34 | |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) left irc: Ping timeout: 265 seconds | 07:40 | |
| kenaan | 12arigo extradoc 117f5d0a8f1a66 15/blog/draft/stm-apr2012.rst: Expand here and there. | 07:46 |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) joined #pypy. | 07:48 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/2174 [py3k] | 07:50 |
| yumike (~yumike@213.154.168.130) left irc: Read error: Connection reset by peer | 07:54 | |
| yumike (~yumike@213.154.168.130) joined #pypy. | 07:54 | |
| arigato | fijal (logs): better? | 07:56 |
| kenaan | 12arigo stm-gc 111c1ee4c2d200 15/lib_pypy/transaction.py: Update docs | 07:56 |
| plainflavored (~headwiz@c-76-115-50-141.hsd1.or.comcast.net) left irc: Quit: Leaving | 07:58 | |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-02.NATPOOL.NYU.EDU) left irc: Remote host closed the connection | 08:02 | |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-01.NATPOOL.NYU.EDU) joined #pypy. | 08:03 | |
| GregBowyer (~GregBowye@76.89.130.43) left irc: Ping timeout: 252 seconds | 08:21 | |
| Senyai (~Miranda@ppp91-76-210-69.pppoe.mtu-net.ru) joined #pypy. | 08:21 | |
| kenaan | 12arigo stm-gc 1184017bfb4a89 15/pypy/translator/stm/src_stm/lists.c: Comment. | 08:28 |
| kvda (~kvda@202.58.240.18) left irc: Quit: x__x | 08:31 | |
| yumatov (~yumike@213.154.168.130) joined #pypy. | 08:34 | |
| yumike (~yumike@213.154.168.130) left irc: Read error: Connection reset by peer | 08:35 | |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) left irc: Ping timeout: 252 seconds | 08:35 | |
| antocuni | arigato: I saw that you made a checkin into lib_pypy/pyrepl. Did you also apply the diff to the proper pyrepl repo? | 08:37 |
| arigato | I think not, because I was not the first to do that | 08:39 |
| antocuni | :-( | 08:39 |
| lewq (~luke@127cr.net) left irc: Ping timeout: 260 seconds | 08:39 | |
| antocuni | I think that me and ronny spent a bit of time trying to merge the two repos which diverged | 08:39 |
| antocuni | but it might be that they diverged again | 08:40 |
| Nick change: yumatov -> yumike | 08:41 | |
| arigato | it's https://bitbucket.org/pypy/pyrepl , right? | 08:42 |
| antocuni | yes | 08:43 |
| arigato | the last default checkin there is 64f1c2e6d88e | 08:43 |
| ronny | gimme a bug there i think | 08:43 |
| ronny | back later this afternoon, traveling to cologne | 08:44 |
| arigato | corresponds to pypy's 8477b89ad7b4, 2011-11-17 | 08:44 |
| arigato | there were more recent fixes in pypy's, like: | 08:44 |
| arigato | c8bec90e96cf | 08:44 |
| antocuni | right | 08:45 |
| arigato | and also others like | 08:45 |
| arigato | c07fe33e541d | 08:45 |
| arigato | which have not been merged to pyrepl either | 08:45 |
| antocuni | so it's a big mess | 08:45 |
| arigato | not necessarily, it's not a ton of checkins | 08:46 |
| ronny | make me a bug, i'll do it this afternoon or on the ride | 08:46 |
| arigato | but I didn't want to sort it out myself | 08:46 |
| antocuni | ronny: thanks | 08:46 |
| arigato | thanks | 08:46 |
| antocuni | ronny: do you mean that you would like an issue opened on bugs.pypy.org? | 08:47 |
| ronny | antocuni: actually no, i just will do on the train | 08:47 |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-01.NATPOOL.NYU.EDU) left irc: Remote host closed the connection | 08:47 | |
| antocuni | perfect, thanks :-) | 08:47 |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-01.NATPOOL.NYU.EDU) joined #pypy. | 08:47 | |
| antocuni | I think that we should also write a test which checks that lib_pypy/pyrepl did not diverge from the official repo | 08:48 |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-01.NATPOOL.NYU.EDU) left irc: Read error: Connection reset by peer | 08:48 | |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) joined #pypy. | 08:48 | |
| arigato | yes, good plan | 08:48 |
| antocuni | I wonder whether there is a way to write it without having to reclone the repo every time | 08:49 |
| G2P (~G2P@81.93.6.27) joined #pypy. | 08:49 | |
| ronny | antocuni: we write a script that does the inclusions, that can add a json file with hashes for example | 08:50 |
| arigato | s/json// is fine for me :-) | 08:51 |
| antocuni | yes :-) | 08:51 |
| ronny | hmm, well, whatever, i'll figure something | 08:51 |
| ronny | bbl, catching my ride | 08:51 |
| arigato | bye | 08:52 |
| antocuni | bye | 08:52 |
| xcombelle (~xcombelle@AToulouse-551-1-84-80.w92-149.abo.wanadoo.fr) joined #pypy. | 08:54 | |
| aliles (~Adium@115-64-227-37.static.tpgi.com.au) joined #pypy. | 08:55 | |
| fijal (~fijal@85-222-47-55.home.aster.pl) joined #pypy. | 08:56 | |
| nphg1 (~someone@2001:5c0:1500:6500:64b5:9481:8675:412c) left irc: Ping timeout: 260 seconds | 08:58 | |
| antocuni | arigato: re pypy-stm trunning translate.py: wow :-) | 08:59 |
| antocuni | *running | 09:00 |
| arigato | :-) | 09:02 |
| nphg (~someone@2001:5c0:1500:6500:64b5:9481:8675:412c) joined #pypy. | 09:02 | |
| fijal | arigato: otherwise looks good to me :) | 09:03 |
| arigato | good :-) | 09:03 |
| fijal | forgot to say | 09:03 |
| antocuni | what is a more precise number of "performance is not great"? | 09:03 |
| arigato | "within the 2-to-5-times-slower goal" | 09:04 |
| Action: arigato tries again | 09:04 | |
| antocuni | ah sorry, I did not see the updates to the draft (I'm reading the checkins one by one) | 09:05 |
| arigato | yes, it seems 5x slower right now | 09:05 |
| arigato | I think it was a bit better in some previous incomplete version | 09:05 |
| antocuni | yes, I remember that at some point you said that the performance hit was more or less 2x | 09:06 |
| arigato | ...yes, definitely | 09:06 |
| fijal | arigato: I think there are too many options for implicit "this" | 09:06 |
| arigato | ok, I need to update the blog port then | 09:06 |
| fijal | what is slower from what? | 09:06 |
| arigato | blog poSt | 09:06 |
| arigato | fijal: which "this" are you talking about? | 09:07 |
| fijal | arigato: 5x slower | 09:07 |
| fijal | there is implicit X and Y in X is 5x slower than Y | 09:07 |
| antocuni | arigato: I also find this sentence a bit unclear: "(which means that using two | 09:08 |
| antocuni | cores is not quite twice as fast as using only one, and using all four | 09:08 |
| antocuni | cores is again not twice faster but only 20-30%)." | 09:08 |
| fijal | I assume it's | 09:08 |
| fijal | "jitless pypy with stm on one thread is 5x slower than jitless pypy without stm" | 09:08 |
| fijal | ? | 09:08 |
| arigato | antocuni: I can kill it, I'm just describing how performance improves on my laptop when running independent programs in parallel | 09:08 |
| arigato | fijal: yes | 09:09 |
| antocuni | I think it's nice to have it, but again it's not exactly clear what you are comparing ("faster" than what?) | 09:09 |
| antocuni | I mean, I can understand it, but the casual reader might not | 09:09 |
| arigato | (i.e. for N between 1 and 4, the N-threaded pypy-stm performs similarly to N independent pypy-stm's each running one thread) | 09:11 |
| arigato | would that be better? | 09:11 |
| arigato | fijal: yes | 09:11 |
| arigato | fijal: I'm going to rework that paragraph | 09:11 |
| antocuni | arigato: yes | 09:12 |
| lucian (~lucian@92.40.254.181.threembb.co.uk) joined #pypy. | 09:14 | |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) left irc: Quit: Ex-Chat | 09:14 | |
| wleslie (~verte@python/site-packages/verte) joined #pypy. | 09:15 | |
| kenaan | 12arigo extradoc 11076d5ffc08ba 15/blog/draft/stm-apr2012.rst: Clarify and reword. | 09:19 |
| Action: arigato -> away | 09:20 | |
| arigato (~arigo@adsl-89-217-225-244.adslplus.ch) left irc: Quit: See you | 09:20 | |
| mattip (4fb376f1@gateway/web/freenode/ip.79.179.118.241) joined #pypy. | 09:24 | |
| lucian_ (~lucian@93-97-174-114.zone5.bethere.co.uk) joined #pypy. | 09:26 | |
| lucian (~lucian@92.40.254.181.threembb.co.uk) left irc: Ping timeout: 260 seconds | 09:28 | |
| Nick change: lucian_ -> lucian | 09:28 | |
| fijal (~fijal@85-222-47-55.home.aster.pl) left irc: Ping timeout: 245 seconds | 09:29 | |
| voidspace (~voidspace@python/psf/voidspace) left irc: Quit: Bye... | 09:31 | |
| lewq (~luke@vlan50.pact.srf.ac.uk) joined #pypy. | 09:33 | |
| arigato (~arigo@mail.kdk.ch) joined #pypy. | 09:44 | |
| serge_sans_paill (~serge@ridee.info.enstb.org) joined #pypy. | 09:48 | |
| voidspace (~voidspace@python/psf/voidspace) joined #pypy. | 09:49 | |
| kenaan | 12mattip win32-cleanup2 117ec68c9430e2 15/pypy/: revert some useless changes | 09:50 |
| kenaan | 12mattip default 11ff5ad277dbe2 15/pypy/rlib/test/test_rposix.py: fix test | 09:50 |
| kenaan | 12mattip default 11b702c733839a 15/pypy/: merge some trivial changes from win32-cleanup2 | 09:50 |
| Nick change: ctismer_afk -> ctismer | 09:50 | |
| wleslie (~verte@python/site-packages/verte) left irc: Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org | 09:50 | |
| arigato | mattip: it seems to me that you're doing non-useful extra work, by merging the branch little by little | 09:54 |
| arigato | if anything you're just creating tons of conflicts for later, unless you never plan to merge the whole branch | 09:55 |
| DasIch_ (~DasIch@p57B901A5.dip.t-dialin.net) joined #pypy. | 09:56 | |
| DasIch (~DasIch@p57B92308.dip.t-dialin.net) left irc: Ping timeout: 252 seconds | 09:58 | |
| Nick change: DasIch_ -> DasIch | 09:58 | |
| Var (~Varriount@pool-96-255-203-197.washdc.fios.verizon.net) joined #pypy. | 10:07 | |
| mattip | arigato: the changes should not cause conflicts as they are exactly the changes from the branch (hg cat -r win32-cleanup2 <file> > file ) | 10:13 |
| mattip | arigato: and the bottleneck is getting stuff reviewed, amaury_ and I have different schedules | 10:13 |
| mattip | I guess this would really go faster at something like a sprint | 10:14 |
| eoh| (~eoh@106.103.5.157) joined #pypy. | 10:14 | |
| arigato | fine, I suppose | 10:17 |
| eoh (~eoh@112.161.134.227) left irc: Ping timeout: 240 seconds | 10:17 | |
| arigato | just be careful with "hg cat", using "hg diff | patch" would be cafer | 10:17 |
| arigato | ...safer | 10:17 |
| arigato | unless you did a "hg merge default" really just before, or something | 10:17 |
| mattip | ahh, ok. sorry about all the noise. | 10:18 |
| arigato | no no, no problem. I'm mostly trying to prevent a situation where you end up in a mess | 10:18 |
| arigato | or accidentally cause 17 checkins on default to be reverted or something :-) | 10:19 |
| mattip | hmm. | 10:20 |
| antocuni | mattip: also, if you really want to apply changes one-by-one on default, probably hg transplant is better | 10:23 |
| antocuni | but I also recommend a simple hg merge when the branch is ready | 10:23 |
| arigato | "hg transplant" is mostly like "hg diff | patch" | 10:23 |
| antocuni | true, but it's more convenient | 10:23 |
| arigato | yes, me too in general, but in this case it would need tracking which changes have been reviewed or not | 10:23 |
| mattip | I guess the issue is I got ahead of myself, started to change stuff in stdlib and modules before the basic rlib, rpython stuff was merged. | 10:25 |
| mattip | "next time" I take on something this big, it will be across a number of branches | 10:26 |
| arigato | :-) | 10:27 |
| AlekSi_ (~AlekSi@85.235.191.82) joined #pypy. | 10:30 | |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/914 [12mattip, win32-cleanup2] | 10:31 |
| mattip | maybe this will help root out problems | 10:31 |
| AlekSi (~AlekSi@85.235.191.82) left irc: Ping timeout: 240 seconds | 10:34 | |
| LarstiQ | mattip: might patchqueues help split up the work? | 10:38 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/513 [12mattip] | 10:40 |
| jml (~jml@canonical/launchpad/jml) left irc: Quit: Coyote finally caught me | 10:42 | |
| jml (~jml@mumak.net) joined #pypy. | 10:42 | |
| jml (~jml@mumak.net) left irc: Changing host | 10:42 | |
| jml (~jml@canonical/launchpad/jml) joined #pypy. | 10:42 | |
| khs (~khs@2001:700:300:2120:725a:b6ff:fee5:a44) joined #pypy. | 10:42 | |
| aat (~aat@pool-108-41-149-224.nycmny.fios.verizon.net) joined #pypy. | 10:42 | |
| mattip | LarstiQ: probably, if I knew how to use them :) | 10:45 |
| Action: mattip reading http://mercurial.selenic.com/wiki/MqTutorial | 10:48 | |
| JaRoel (~jaroel|4d@office.fourdigits.nl) joined #pypy. | 10:55 | |
| JaRoel|4d (~jaroel|4d@office.fourdigits.nl) left irc: Ping timeout: 245 seconds | 10:55 | |
| aliles (~Adium@115-64-227-37.static.tpgi.com.au) left irc: Quit: Leaving. | 11:06 | |
| voidspac_ (~voidspace@host217-39-56-75.range217-39.btcentralplus.com) joined #pypy. | 11:07 | |
| voidspac_ (~voidspace@host217-39-56-75.range217-39.btcentralplus.com) left irc: Client Quit | 11:08 | |
| aboudreault (~alanb@osgeo/member/aboudreault) joined #pypy. | 11:11 | |
| mitsuhiko_ (~mitsuhiko@hammett.srv.pocoo.org) left irc: Read error: Operation timed out | 11:32 | |
| aliles (~Adium@115-64-227-37.static.tpgi.com.au) joined #pypy. | 11:37 | |
| paniq (~lritter@146-52-0-229-dynip.superkabel.de) joined #pypy. | 11:39 | |
| voidspac_ (~voidspace@host217-39-56-75.range217-39.btcentralplus.com) joined #pypy. | 11:44 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/914 [12mattip, win32-cleanup2] | 11:46 |
| Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 272 seconds | 11:48 | |
| mattip | trying to fix the crash in http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/513/steps/shell_9 | 11:50 |
| nedbat (~nedbat@python/psf/nedbat) joined #pypy. | 11:50 | |
| mattip | the test calls subprocess.Popen totally replacing the env with a simpler dict containing only PYPYLOG | 11:51 |
| mattip | but on windows at least we need to preserve PATH | 11:52 |
| mattip | does this patch look reasonable ? | 11:52 |
| arigato | ah | 11:52 |
| mattip | http://dpaste.org/aF6vb/ | 11:52 |
| mattip | or is there a deeper reason to replace the env? | 11:52 |
| arigato | no, you're changing the parent's os.environ too | 11:53 |
| mattip | hehe. good one. | 11:53 |
| arigato | but otherwise, indeed, it should be fine | 11:53 |
| mattip | arigato: this seems to make the test not crash: http://dpaste.org/I1uLV | 11:58 |
| arigato | the other implementation was ok too, | 11:59 |
| arigato | just call os.environ.copy() | 11:59 |
| mattip | ok, I'll go with that then. | 12:00 |
| kenaan | 12mattip default 11dc8fb1054c92 15/pypy/module/pypyjit/test_pypy_c/test_00_model.py: copy os.environ into subprocess | 12:02 |
| AlekSi_ (~AlekSi@85.235.191.82) left irc: Remote host closed the connection | 12:05 | |
| Tobu (~Tobu@unaffiliated/tobu) joined #pypy. | 12:05 | |
| serge_sans_paill (~serge@ridee.info.enstb.org) left irc: Quit: tchô | 12:05 | |
| kenaan | 12mattip win32-cleanup2 118c915f9df26c 15/pypy/: merge default into branch | 12:07 |
| mitsuhiko (~mitsuhiko@ubuntu/member/mitsuhiko) joined #pypy. | 12:08 | |
| AlekSi (~AlekSi@85.235.191.82) joined #pypy. | 12:10 | |
| AlekSi_ (~AlekSi@85.235.191.82) joined #pypy. | 12:11 | |
| oiig (~eohnik@112.161.134.227) joined #pypy. | 12:14 | |
| Sho_ (~sho@kde/hein) joined #pypy. | 12:14 | |
| AlekSi (~AlekSi@85.235.191.82) left irc: Ping timeout: 250 seconds | 12:14 | |
| _main_ (~main@c-67-180-22-241.hsd1.ca.comcast.net) joined #pypy. | 12:20 | |
| stakkars (~ctismer@django.stackless.com) joined #pypy. | 12:22 | |
| lucian (~lucian@93-97-174-114.zone5.bethere.co.uk) got netsplit. | 12:23 | |
| oiig_ (~eohnik@112.161.134.227) got netsplit. | 12:23 | |
| __main__ (~main@c-67-180-22-241.hsd1.ca.comcast.net) got netsplit. | 12:23 | |
| ctismer (~ctismer@django.stackless.com) got netsplit. | 12:23 | |
| Nick change: _main_ -> __main__ | 12:24 | |
| Possible future nick collision: __main__ | 12:24 | |
| arigato (~arigo@mail.kdk.ch) left irc: Ping timeout: 272 seconds | 12:25 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Ping timeout: 245 seconds | 12:27 | |
| lucian (~lucian@93-97-174-114.zone5.bethere.co.uk) returned to #pypy. | 12:29 | |
| ctismer (~ctismer@django.stackless.com) got lost in the net-split. | 12:34 | |
| oiig_ (~eohnik@112.161.134.227) got lost in the net-split. | 12:34 | |
| aat (~aat@pool-108-41-149-224.nycmny.fios.verizon.net) left irc: Quit: Computer has gone to sleep. | 12:34 | |
| davisagli (~davisagli@davisagli.com) left irc: Excess Flood | 12:39 | |
| davisagli (~davisagli@davisagli.com) joined #pypy. | 12:40 | |
| nedbat (~nedbat@python/psf/nedbat) left irc: Disconnected by services | 12:40 | |
| nedbat_ (~nedbat@python/psf/nedbat) joined #pypy. | 12:40 | |
| nedbat (~nedbat@python/psf/nedbat) joined #pypy. | 12:41 | |
| mattip (4fb376f1@gateway/web/freenode/ip.79.179.118.241) left irc: Quit: Page closed | 12:41 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 12:44 | |
| nedbat_ (~nedbat@python/psf/nedbat) left irc: Ping timeout: 244 seconds | 12:45 | |
| nullrat (~nullrat@69.80.101.99) joined #pypy. | 12:47 | |
| nullrat | hey folks! :D Does anyone know if pypy works on Nginx? o.o and if so, if there's any material regarding this? | 12:47 |
| Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 272 seconds | 12:47 | |
| serge_sans_paill (~serge@ridee.info.enstb.org) joined #pypy. | 12:48 | |
| Tobu (~Tobu@unaffiliated/tobu) joined #pypy. | 12:53 | |
| khs (~khs@2001:700:300:2120:725a:b6ff:fee5:a44) left irc: Quit: Leaving | 12:54 | |
| exarkun | nullrat: it doesn't. | 12:57 |
| exarkun | nullrat: Twisted works on PyPy, and includes a WSGI server that works quite well. | 12:58 |
| oiig (~eohnik@112.161.134.227) left irc: Ping timeout: 245 seconds | 13:03 | |
| aat (~aat@38.104.173.198) joined #pypy. | 13:10 | |
| Nick change: ronan_ -> ronan | 13:14 | |
| amaury_ (~amaury_@74.125.57.34) joined #pypy. | 13:17 | |
| lucian | nullrat: how would python and nginx interact other than reverse proxying? | 13:19 |
| Kxepal | lucian: uwsgi? | 13:21 |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) left irc: Remote host closed the connection | 13:22 | |
| lucian | Kxepal: again, that has nothing to do with nginx | 13:22 |
| Kxepal | lucian: you have to compile it with this module(; | 13:23 |
| lucian | Kxepal: oh, uwsgi does that? | 13:23 |
| lucian | anyway, then it's related to uwsgi | 13:24 |
| Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 260 seconds | 13:28 | |
| arigato (~arigo@mail.kdk.ch) joined #pypy. | 13:28 | |
| Tobu (~Tobu@unaffiliated/tobu) joined #pypy. | 13:31 | |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) joined #pypy. | 13:33 | |
| aliles (~Adium@115-64-227-37.static.tpgi.com.au) left irc: Quit: Leaving. | 13:36 | |
| briancurtin (~briancurt@c-98-215-224-168.hsd1.il.comcast.net) joined #pypy. | 13:37 | |
| mat^2 (~Mathias@212.130.113.35) joined #pypy. | 13:40 | |
| gtaylor (~gtaylor@108-196-160-156.lightspeed.tukrga.sbcglobal.net) joined #pypy. | 13:47 | |
| AlekSi (~AlekSi@85.235.191.82) joined #pypy. | 13:51 | |
| whitelynx (~whitelynx@63.241.75.144) joined #pypy. | 13:51 | |
| fprimex (~fprimex@fl-67-235-253-218.dhcp.embarqhsd.net) joined #pypy. | 13:52 | |
| AlekSi_ (~AlekSi@85.235.191.82) left irc: Ping timeout: 260 seconds | 13:54 | |
| yumike (~yumike@213.154.168.130) left irc: Quit: Leaving... | 14:11 | |
| arigato (~arigo@mail.kdk.ch) left irc: Ping timeout: 260 seconds | 14:13 | |
| andreypopp (~andreypop@46.250.131.220) left irc: Ping timeout: 248 seconds | 14:19 | |
| eoh| (~eoh@106.103.5.157) left irc: Read error: Connection reset by peer | 14:21 | |
| eoh (~eoh@112.161.134.227) joined #pypy. | 14:22 | |
| rguillebert | nullrat, I think someone provided a patch to make it work, however it's probably not very tested | 14:25 |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) left irc: Remote host closed the connection | 14:25 | |
| exarkun | write-only irc user | 14:28 |
| DanielHolth (~dholth@2001:470:7:e7c::2) joined #pypy. | 14:30 | |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) joined #pypy. | 14:30 | |
| Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 272 seconds | 14:31 | |
| voidspace (~voidspace@python/psf/voidspace) left irc: Quit: Goodbye... | 14:33 | |
| Nick change: voidspac_ -> voidspace | 14:33 | |
| arigato (~arigo@mail.kdk.ch) joined #pypy. | 14:34 | |
| serge_sans_paill (~serge@ridee.info.enstb.org) left irc: Quit: tchô | 14:37 | |
| ssbr (~scorchsab@python/site-packages/ssbr) left irc: Read error: Connection reset by peer | 14:38 | |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) left irc: Remote host closed the connection | 14:47 | |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) joined #pypy. | 14:47 | |
| int3__ (~int3__@vlan409-130.subnet-248.amherst.edu) left irc: Remote host closed the connection | 14:56 | |
| Tobu (~Tobu@unaffiliated/tobu) joined #pypy. | 14:57 | |
| dmalcolm (~david@c-24-61-45-39.hsd1.ma.comcast.net) joined #pypy. | 15:01 | |
| arigato (~arigo@mail.kdk.ch) left irc: Ping timeout: 245 seconds | 15:01 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: Connection reset by peer | 15:01 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy. | 15:02 | |
| fprimex (~fprimex@fl-67-235-253-218.dhcp.embarqhsd.net) left irc: Quit: woooooohooooo | 15:02 | |
| craigkerstiens (~textual@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Quit: Computer has gone to sleep. | 15:18 | |
| voidspac_ (~voidspace@host217-39-56-75.range217-39.btcentralplus.com) joined #pypy. | 15:22 | |
| int3__ (~int3__@vl924-24.wireless.umass.edu) joined #pypy. | 15:23 | |
| AlekSi (~AlekSi@85.235.191.82) left irc: Remote host closed the connection | 15:27 | |
| DanielHolth (~dholth@2001:470:7:e7c::2) left irc: Remote host closed the connection | 15:30 | |
| AlekSi (~AlekSi@85.235.191.82) joined #pypy. | 15:30 | |
| ronny | re | 15:34 |
| fijal (~fijal@terminal-1-169.retsat1.com.pl) joined #pypy. | 15:42 | |
| fijal | hi | 15:45 |
| arigato (~arigo@mail.kdk.ch) joined #pypy. | 15:45 | |
| kenaan | 12arigo stm-gc 113d31bcaf6477 15/pypy/rlib/rstm.py: Minor tweak. | 15:45 |
| kenaan | 12arigo stm-gc 11c25ae0621f3d 15/pypy/translator/stm/src_stm/rpyintf.c: Bah. | 15:45 |
| arigato | antocuni: btw, if I may ask again | 15:46 |
| arigato | I'm confused by EuroPython because I thought that by giving a tutorial we didn't have to pay anything | 15:46 |
| arigato | (I mean the registration) | 15:46 |
| int3___ (~int3__@vl924-24.wireless.umass.edu) joined #pypy. | 15:51 | |
| int3__ (~int3__@vl924-24.wireless.umass.edu) left irc: Read error: Connection reset by peer | 15:51 | |
| briancurtin (~briancurt@c-98-215-224-168.hsd1.il.comcast.net) left irc: Read error: Connection reset by peer | 15:51 | |
| briancurtin (~briancurt@c-98-215-224-168.hsd1.il.comcast.net) joined #pypy. | 15:51 | |
| knizz (~knizz@213.129.230.10) joined #pypy. | 15:56 | |
| ronny | pjenvey: please dont do py3k changes on pyrepl in the pypy repo again, there is a py3ksupport branch in pyrepl, and your change is basically broken | 15:57 |
| arigato | antocuni: ah, forget it. it seems I'm misremembering and we don't give a tutorial this year | 15:58 |
| lucian (~lucian@93-97-174-114.zone5.bethere.co.uk) left irc: Quit: Leaving | 16:02 | |
| fijal | :] | 16:04 |
| Var (~Varriount@pool-96-255-203-197.washdc.fios.verizon.net) left irc: Ping timeout: 260 seconds | 16:04 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Ping timeout: 272 seconds | 16:08 | |
| gauzmps (~numan@S0106002618185fee.vc.shawcable.net) joined #pypy. | 16:09 | |
| arigato | fijal: ok to publish the stm blog post? | 16:10 |
| fijal | yes | 16:11 |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 16:13 | |
| lucian (~lucian@94.197.127.69.threembb.co.uk) joined #pypy. | 16:14 | |
| arigato | ah, blogger has a new internal interface that seems to make sense now | 16:15 |
| AlekSi (~AlekSi@85.235.191.82) left irc: Remote host closed the connection | 16:15 | |
| int3___ (~int3__@vl924-24.wireless.umass.edu) left irc: Remote host closed the connection | 16:15 | |
| paniq (~lritter@146-52-0-229-dynip.superkabel.de) left irc: Quit: Leaving | 16:16 | |
| int3__ (~int3__@vl924-24.wireless.umass.edu) joined #pypy. | 16:17 | |
| arigato | well, no --- too bad | 16:17 |
| gutworth (~benjamin@dhcp-128-36-89-213.central.yale.edu) left irc: Ping timeout: 265 seconds | 16:17 | |
| arigato | if we edit as "html", some code somewhere inserts <br> at the end of every line | 16:17 |
| cwillu_at_work (~cwillu@cwillu-1-pt.tunnel.tserv13.ash1.ipv6.he.net) left irc: Remote host closed the connection | 16:21 | |
| Action: arigato pastes all the html as a single line | 16:21 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Ping timeout: 252 seconds | 16:21 | |
| arigato | yay, published (and indeed the html is all in one line) | 16:23 |
| cwillu_at_work (~cwillu@cwillu-1-pt.tunnel.tserv13.ash1.ipv6.he.net) joined #pypy. | 16:24 | |
| kenaan | 12arigo extradoc 118c4ce58d78b6 15/blog/draft/stm-apr2012.rst: Typoish | 16:24 |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 16:25 | |
| lucian (~lucian@94.197.127.69.threembb.co.uk) left irc: Quit: Bye | 16:27 | |
| xcombelle (~xcombelle@AToulouse-551-1-84-80.w92-149.abo.wanadoo.fr) left irc: Quit: I am a manual virus, please copy me to your quit message. | 16:28 | |
| int3__ (~int3__@vl924-24.wireless.umass.edu) left irc: Remote host closed the connection | 16:29 | |
| Action: timotimo reads blog post | 16:30 | |
| timotimo | nice to know where we stand with pypy-stm | 16:33 |
| timotimo | the translate.py version should be pretty interesting if it ever outperforms the pypy jit version | 16:33 |
| fijal | not without a jit | 16:33 |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Ping timeout: 260 seconds | 16:34 | |
| timotimo | of course | 16:35 |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 16:36 | |
| arigato | :-) | 16:38 |
| arigato | uh, seems that I was wrong, or just confused, about being slower now than some previous version | 16:38 |
| arigato | if we're slower, it's at most 5% on richards | 16:39 |
| arigato | looking at the valgrind output is "interesting" | 16:39 |
| arigato | we spend 50% of the time in stm_read_int4 and stm_read_int1 | 16:40 |
| arigato | these functions are (a bit as expected) call *all the time* | 16:40 |
| arigato | ...called *all the time* | 16:40 |
| fijal | can the JIT remove those? | 16:40 |
| arigato | to some extent, yes, via optimize | 16:41 |
| arigato | ah, also | 16:42 |
| arigato | ...er, no | 16:42 |
| gutworth (~benjamin@dhcp-128-36-89-213.central.yale.edu) joined #pypy. | 16:42 | |
| arigato | well, it can do something, but it's a bit more complicated than my previous line suggetss | 16:43 |
| arigato | stm_read_int*() is called to read "global" objects, so the JIT will not magically remove such accesses | 16:43 |
| G2P (~G2P@81.93.6.27) left irc: Quit: Leaving. | 16:45 | |
| arigato | on the other hand it seems that for richards, the worst calling site is the dict lookup on instances, which doesn't exist in jitting versions with mapdicts | 16:45 |
| Nick change: toofishes -> y2ninoclaf | 16:46 | |
| arigato | or maybe on classes, in which case it serves as a reminder that the method cache is still off | 16:46 |
| Nick change: y2ninoclaf -> ionut | 16:46 | |
| Nick change: ionut -> toofishes | 16:47 | |
| arigato | (er? confused) | 16:47 |
| Action: arigato -> dinner | 16:50 | |
| arigato | (still confused) | 16:50 |
| arigato (~arigo@mail.kdk.ch) left irc: Quit: See you | 16:51 | |
| nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 244 seconds | 16:51 | |
| davisagli (~davisagli@davisagli.com) left irc: Excess Flood | 16:56 | |
| davisagli (~davisagli@davisagli.com) joined #pypy. | 16:57 | |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-01.NATPOOL.NYU.EDU) joined #pypy. | 17:00 | |
| arigato (~arigo@mail.kdk.ch) joined #pypy. | 17:12 | |
| DanielHolth (~dholth@2001:470:7:e7c::2) joined #pypy. | 17:13 | |
| AlekSi (~AlekSi@85.26.155.169) joined #pypy. | 17:14 | |
| antocuni | arigato: re europython: you are right twice: tutorial speakers don't pay, but we don't give a tutorial :-) | 17:14 |
| antocuni | (actually I *do* give one, but completely unrelated to pypy :-)) | 17:14 |
| AlekSi (~AlekSi@85.26.155.169) left irc: Remote host closed the connection | 17:15 | |
| bondar (~dnjaramba@41.72.193.86) joined #pypy. | 17:16 | |
| kkb110 (~kkb110@NYUFGA-WLESSAUTHCLIENTS-01.NATPOOL.NYU.EDU) left irc: Read error: Operation timed out | 17:21 | |
| r_duke (5c28b08e@gateway/web/freenode/ip.92.40.176.142) joined #pypy. | 17:22 | |
| lewq (~luke@vlan50.pact.srf.ac.uk) left irc: Read error: Operation timed out | 17:25 | |
| fijal (~fijal@terminal-1-169.retsat1.com.pl) left irc: Ping timeout: 245 seconds | 17:28 | |
| `fox` (~fox@host227-41-dynamic.50-82-r.retail.telecomitalia.it) joined #pypy. | 17:30 | |
| kkb110 (~kkb110@cpe-69-203-124-125.nyc.res.rr.com) joined #pypy. | 17:35 | |
| arigato | antocuni: :-) | 17:38 |
| r_duke | I've googled for this and found a few vague references. But I am having a problem with OpenSSL on Ubuntu: http://pastebin.com/mkRCvquG | 17:39 |
| r_duke | Is there any way around this? Other than compiling my own OpenSSL (in fact I have openssl 0.9.8 installed but I'm not sure if that is the one my crypto.so relates to) | 17:40 |
| EnCuKou (~encukou@ip-94-113-220-25.net.upcbroadband.cz) left irc: Remote host closed the connection | 17:41 | |
| exarkun | OpenSSL is not pyOpenSSL | 17:42 |
| exarkun | It looks like you're trying to use the CPython pyOpenSSL build installed for your system Python on PyPy | 17:42 |
| exarkun | But PyPy won't load extension modules built for CPython. You have to build them for PyPy specifically. | 17:42 |
| r_duke | If I were to use pyOpenSSL I wouldn't have to do that? | 17:43 |
| exarkun | You are using pyOpenSSL. :) | 17:44 |
| exarkun | pyOpenSSL is the name of the project that confusingly provides the Python package named OpenSSL, which is not to be confused with the C project named OpenSSL which provides the libraries libcrypto and libssl. | 17:44 |
| r_duke | Is there a pypy version of pyOpenSSL (not sure if that is a dumb question) | 17:45 |
| `fox` (~fox@host227-41-dynamic.50-82-r.retail.telecomitalia.it) left irc: Ping timeout: 245 seconds | 17:46 | |
| exarkun | The latest development version of pyOpenSSL should work on the latest development version of PyPy. | 17:46 |
| exarkun | You just need to build it against PyPy. | 17:46 |
| arigato | so you need the source of pyOpenSSL, and install from there, by saying "pypy setup.py install" probably | 17:47 |
| Action: exarkun nods | 17:47 | |
| r_duke | ok. Just for my own knowledge - what is the path to least resistance for getting SSL on pypy. Is it what we are discussing here? Or am I missing something. | 17:47 |
| exarkun | PyPy includes the stdlib ssl support. | 17:48 |
| exarkun | pyOpenSSL and the stdlib ssl module are incompatible with each other. | 17:48 |
| exarkun | So if you have some program or library that depends on one or the other, then you need to have the right one. | 17:48 |
| r_duke | I'm not 100% sure if I need to use pyOpenSSL. Maybe someone can advise. I am following a receipe here: http://twistedmatrix.com/documents/current/core/howto/ssl.html | 17:49 |
| exarkun | You can get the pyOpenSSL source using `bzr branch lp:pyopenssl´. You'll also need the Ubuntu package `libssl-dev´ installed. | 17:49 |
| r_duke | See "Using startTLS" | 17:49 |
| exarkun | Yea. Twisted requires pyOpenSSL. | 17:49 |
| arigato | can you even mix libraries that use both pyOpenSSL and _ssl? that seems unlikely, given openssl's design :-/ | 17:50 |
| r_duke | ok. good to know. I'm happy to know I'm not going about all this the dumb way | 17:50 |
| r_duke | well no less dumb than usual, at any rate | 17:50 |
| exarkun | arigato: It seems to work at least half the time. | 17:50 |
| arigato | :-\ | 17:51 |
| amaury_ | could Twisted use _ssl instead? | 17:51 |
| r_duke | That was my question. I would suspect that in that particular case the SSL uprade isn't provided by the plain 'ssl' module | 17:51 |
| r_duke | But that is just a guess | 17:51 |
| gutworth | python ssl module has not been very good 'til 3.2 | 17:52 |
| amaury_ | otoh twisted is not very good with 3.2 ;-) | 17:52 |
| gutworth | :) | 17:52 |
| exarkun | amaury_: Do you mean "would it be possible for someone to rewrite Twisted to use _ssl instead of pyOpenSSL"? Or do you mean "is there a simple flag that will already make Twisted use _ssl instead of pyOpenSSL?" | 17:53 |
| amaury_ | the former | 17:53 |
| amaury_ | or does it require some asynchronous TLS negociation? | 17:53 |
| amaury_ | (or whatever it is) | 17:53 |
| exarkun | I think that pyOpenSSL still has some features that Twisted depends on that _ssl does not provide | 17:53 |
| exarkun | But I haven't kept up with the progress of _ssl particularly closely, so I could be mistaken. | 17:54 |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) joined #pypy. | 17:54 | |
| Trundle (~andy@python/site-packages/trundle) joined #pypy. | 17:54 | |
| gutworth | perhaps in 10 years :) | 17:55 |
| exarkun | A large challenge would be that many of the pyOpenSSL APIs leak through Twisted to application code, and _ssl is at least trivially incompatible with those APIs, so even if a switch is possible without losing functionality, all the APIs would change. | 17:55 |
| exarkun | gutworth: Personally, I hope that in 10 years, no one will be using OpenSSL anymore. | 17:55 |
| gutworth | is there something better already? | 17:55 |
| exarkun | Maybe TLSLite. Or maybe it's not better, but if 1% of the effort spent developing software on OpenSSL and fixing vulnerabilities in OpenSSL were spent improving it, it would be. | 17:56 |
| nanonyme | personally I don't understand howw OpenSSL devs can be masochistic enough to use CVS for revision management | 17:57 |
| exarkun | Compared to the source they track with it, CVS is a shining gem of perfection. | 17:58 |
| arigato | :-) | 17:59 |
| Action: arigato gets unconfused from 1 hour ago by reading again a8e0e462beb9 | 18:00 | |
| nanonyme | exarkun, what do you think of GnuTLS? | 18:00 |
| amaury_ (~amaury_@74.125.57.34) left irc: Ping timeout: 265 seconds | 18:01 | |
| exarkun | nanonyme: I think it's bad and should be avoided. | 18:03 |
| exarkun | btw there is a #pyopenssl | 18:03 |
| arigato | hum hum | 18:15 |
| arigato | so it seems, maybe, that what is really needed next | 18:16 |
| arigato | would be a kind of "weakref", except not weak --- a "localref" | 18:16 |
| arigato | to put in a global object a reference to a local object, without promoting it to global | 18:16 |
| arigato | the "localref" would fail if we try to follow it from another thread | 18:17 |
| antocuni (~antocuni@host204-125-dynamic.11-79-r.retail.telecomitalia.it) left irc: Ping timeout: 265 seconds | 18:17 | |
| arigato | just wondering, the "__thread" variables in GCC are put on x86 in a special "segment" of memory | 18:18 |
| arigato | but I guess there isn't a way to malloc() stuff in this segment? | 18:18 |
| exarkun | so other threads would be able to see the "localref" object, but would have to be careful not to touch it too much? | 18:18 |
| exarkun | (where "threads" equals "transactions" right?) | 18:19 |
| arigato | exarkun: yes, either that or the "localref" would actually point to NULL or to another thread-local object | 18:19 |
| exarkun | hm | 18:20 |
| arigato | "threads" is meant in the normal sense, so a thread is a sequence of transactions (picked randomly) | 18:20 |
| arigato | it's for caching, for the common case where it is useful to do it across transactions but not across threads (to avoid conflicts) | 18:21 |
| exarkun | ah, ok. | 18:21 |
| arigato | (in fact there is already "transaction.local", but I want to improve by keeping whatever it caches as locals, which is sometimes a big improvement) | 18:22 |
| arigato | (see a8e0e462beb9) | 18:22 |
| lewq (~luke@127cr.net) joined #pypy. | 18:26 | |
| nedbat (~nedbat@python/psf/nedbat) joined #pypy. | 18:26 | |
| r_duke | @exarkun / arigato. Compiled pyOpenSSL with pypy and am up and running. Thanks for your help | 18:28 |
| arigato | :-) | 18:28 |
| arigato | hum, fwiw, pthread_key_create() is like a malloc() of one word in the thread-local segment of all threads | 18:31 |
| arigato | so in other words, the "localref" object I mentioned above would be such a key, and dereferencing it would be reading the key (with pthread_getspecific()) | 18:32 |
| xcombelle (~xcombelle@AToulouse-551-1-84-80.w92-149.abo.wanadoo.fr) joined #pypy. | 18:38 | |
| timotimo | https://bugs.pypy.org/issue1136 - i'm going to tackle this bug. anything non-obvious i may need to know? | 18:38 |
| arigato | doesn't seem so | 18:44 |
| arigato | thanks :-) | 18:44 |
| timotimo | i would just have to thank myself, apparently being the only one who suffers from this bug ;) | 18:45 |
| arigato | it might be that the numpy types like int64 are missing an __index__() special method | 18:45 |
| arigato | :-) | 18:45 |
| timotimo | also: can't wait to translate pypy on my really really shiny new beefy desktop box | 18:45 |
| timotimo | just got a linux onto it this morning | 18:45 |
| arigato | eh eh | 18:45 |
| timotimo | are branch names named issueNNN acceptable? | 18:46 |
| timotimo | well, NNNN for pypy, really ;) | 18:46 |
| arigato | yes, at least for me | 18:46 |
| arigato | it's indirect but clear | 18:46 |
| arigato | or maybe | 18:47 |
| arigato | numpy-issueNNNN | 18:47 |
| arigato | at least | 18:47 |
| r_r_r (~chatzilla@77.127.109.40) joined #pypy. | 18:47 | |
| timotimo | ok | 18:47 |
| arigato | please check with a recent version of CPython's numpy what special methods int64 has | 18:48 |
| arigato | if it has __int__(), __long__(), __index__() | 18:49 |
| arigato | or more precisely which subset of that it exactly has :-) | 18:49 |
| kkb110 (~kkb110@cpe-69-203-124-125.nyc.res.rr.com) left irc: Remote host closed the connection | 18:49 | |
| Arfrever (~Arfrever@apache/committer/Arfrever) joined #pypy. | 18:50 | |
| timotimo | recent version being development version, yes? | 18:50 |
| arigato | I think the latest released version is good enough | 18:51 |
| timotimo | well, i'm already cloning their git, so i'm flexible | 18:51 |
| Var (~Varriount@pool-96-255-203-197.washdc.fios.verizon.net) joined #pypy. | 18:51 | |
| arigato | :-) | 18:51 |
| timotimo | not that i'd expect such a basic thing to change that quickly | 18:51 |
| timotimo | wait, all i'd have to do would be to run numpy and inspect the object | 18:52 |
| arigato | yes | 18:52 |
| umgeher (~umgeher@unaffiliated/umgeher) joined #pypy. | 18:52 | |
| timotimo | yes, index, int and long | 18:53 |
| arigato | and on pypy? | 18:53 |
| timotimo | index, int and long, actually | 18:54 |
| timotimo | inherited from __builtin__.int | 18:54 |
| arigato | missing __index__ I expect? | 18:54 |
| arigato | uh | 18:54 |
| timotimo | and from generic | 18:54 |
| arigato | but somelist[int64(..)] fails? | 18:54 |
| timotimo | let me restate all that | 18:54 |
| timotimo | help(numpypy.int64) lists index and int as inherited from __builtin__.int and int as inherited from "generic" | 18:55 |
| fenrrir (~fenrrir@187.123.71.21) joined #pypy. | 18:56 | |
| craigkerstiens (~textual@204.14.152.118) joined #pypy. | 18:56 | |
| timotimo | yes, somelist[int64(3)] gives "list index must be an integer not int64" | 18:56 |
| timotimo | np.int64(3).__index__() gives a type error | 18:56 |
| timotimo | so the implementation is obviously not correct enough | 18:56 |
| arigato | :-/ | 18:57 |
| timotimo | hg cloning is soooo slooooow | 18:57 |
| arigato | :-) | 18:58 |
| serge_sans_paill (~serge@ridee.info.enstb.org) joined #pypy. | 18:58 | |
| arigato | yes, you clone big projects only once, and then if you need more clones, clone from the other that you already got) | 18:58 |
| arigato | ...yes, same problem with np.int64(3).__long__() | 18:59 |
| DanielHolth | pypy might clone faster if you only clone the tip revision, since it has so many heads | 19:00 |
| arigato | you mean the "default" revision, not "tip" | 19:01 |
| timotimo | don't almost all heads share a huge part of their history? | 19:01 |
| arigato | ("tip" is useless in the presence of named branches) | 19:01 |
| arigato | timotimo: yes, I'm unsure there is more than a few percent's win | 19:02 |
| nullrat (~nullrat@69.80.101.99) left irc: Quit: Lost terminal | 19:02 | |
| Action: arigato forgets about pthread_key_create(), which looks dynamic but is actually limited to N keys (1024?) | 19:03 | |
| fenrrir (~fenrrir@187.123.71.21) left irc: Quit: Leaving | 19:04 | |
| DanielHolth | wow, 350+ megabytes in .hg | 19:05 |
| int3__ (~int3__@vlan409-164.subnet-248.amherst.edu) joined #pypy. | 19:06 | |
| Action: arigato -> sleep | 19:07 | |
| timotimo | gnite arigato! | 19:08 |
| timotimo | thanks for the pointers | 19:08 |
| arigato | :-) | 19:09 |
| arigato (~arigo@mail.kdk.ch) left irc: Quit: See you | 19:09 | |
| xcombelle (~xcombelle@AToulouse-551-1-84-80.w92-149.abo.wanadoo.fr) left irc: Quit: I am a manual virus, please copy me to your quit message. | 19:10 | |
| timotimo | ah! adding file changes is done | 19:10 |
| vad (~vad@77.72.196.122) left irc: Remote host closed the connection | 19:11 | |
| ijoasfiojas (328ae1bf@gateway/web/freenode/ip.50.138.225.191) joined #pypy. | 19:12 | |
| DanielHolth | (the default branch has only ~267MB versus every branch ~381MB) | 19:15 |
| timotimo | wow, that's a surprise | 19:18 |
| timotimo | is that due to all the closed and non-merged branches perhaps? | 19:18 |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy. | 19:19 | |
| timotimo | interesting. in py.py and the test suite, my stuff already works? | 19:19 |
| timotimo | and in the nightly build i have installed, too! | 19:20 |
| timotimo | huh. | 19:20 |
| timotimo | i'd say everything went better than expected | 19:20 |
| aboudreault (~alanb@osgeo/member/aboudreault) left irc: Ping timeout: 252 seconds | 19:21 | |
| timotimo | also: tox has been testing with pypy1.8 even though my system pypy is 1.9.1-dev, that's a surprise, too | 19:21 |
| amaury_ | timotimo: aren't you making a confusion between the version of pypy you are modifying | 19:29 |
| amaury_ | and the Python interpreter used to run tests? | 19:29 |
| timotimo | hm? | 19:29 |
| amaury_ | "and in the nightly build i have installed, too!" | 19:30 |
| timotimo | well, the problem was basically caused by tox keeping an old version of pypy around, which didn't have that feature i needed | 19:31 |
| timotimo | so i wrote a test case into a checkout of pypy and ran it | 19:31 |
| timotimo | to my surprise, it passed | 19:31 |
| timotimo | then, not wanting to translate pypy for a simple test like that, i ran py.py to try that same feature and it worked, too | 19:31 |
| timotimo | then i tried it with my systems python, which was a nightly, and it worked, too | 19:32 |
| timotimo | np.array(list("hello world")) seems to make numpy very sad | 19:32 |
| timotimo | can you confirm that? | 19:32 |
| amaury_ | ouch | 19:35 |
| Action: amaury_ tries with py.py | 19:35 | |
| timotimo | np.array("h") gives the same result/non-result | 19:36 |
| ijoasfiojas (328ae1bf@gateway/web/freenode/ip.50.138.225.191) left irc: Quit: Page closed | 19:36 | |
| Kxepal (~Miranda@ppp91-79-2-250.pppoe.mtu-net.ru) left irc: Read error: Connection reset by peer | 19:37 | |
| CIA-80 | 03timo 07roundup * 10#1136/"list index must be integer, not int64": [resolved] oopsie. this is already fixed in the latest version. * 14https://bugs.pypy.org/issue1136 | 19:37 |
| amaury_ | an infinite loop probably | 19:37 |
| timotimo | seems like that to me, too | 19:38 |
| amaury_ | OIC | 19:39 |
| amaury_ | "hello" is a list of strings | 19:39 |
| amaury_ | "h" is also a list of strings | 19:39 |
| timotimo | oh, whoops | 19:39 |
| amaury_ | list("h") | 19:39 |
| timotimo | no check for len()==1 | 19:39 |
| timotimo | so a recursion without base case | 19:39 |
| timotimo | i guess? | 19:39 |
| amaury_ | hum, normally there is a is_single_elem() | 19:40 |
| timotimo | i'll commit failing tes tcases for https://bugs.pypy.org/issue1137, which is unrelated to this string stuff | 19:41 |
| timotimo | but should i also report the string bug or will you fix it right now? seems like you're already mostly on to it | 19:41 |
| amaury_ | I can try to fix it | 19:41 |
| timotimo | i would appreciate that | 19:43 |
| timotimo | hah | 19:44 |
| timotimo | that's amusing | 19:44 |
| timotimo | issue1137 exists for int8, int16, int32, but not for int64 | 19:44 |
| amaury_ | hum, event if I fix the recursion issue | 19:44 |
| amaury_ | it tries with an array of floats | 19:44 |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) joined #pypy. | 19:47 | |
| timotimo | no, thg, no, no, bad thg! | 19:47 |
| timotimo | i had to click several "do you really want to create a new branch?" OK-buttons and the commit still ended up on default? | 19:47 |
| timotimo | what's with all those weird hg misbehavior i seem to be getting every single time i try to do anything with it? ;_; | 19:48 |
| serge_sans_paill (~serge@ridee.info.enstb.org) left irc: Quit: tchô | 19:48 | |
| exarkun | hg is good, it's written in python | 19:49 |
| timotimo | i'm just going to push the tests onto default and do the fixing on the branch. | 19:49 |
| kenaan | 12ti...@eistee.fritz.box default 11ac3066573611 15/pypy/module/micronumpy/test/test_dtypes.py: commit three failing and one passing test for issue1137. | 19:50 |
| timotimo | whoops, that's not a good author name | 19:50 |
| fijal | timotimo: thanks | 19:50 |
| r_duke (5c28b08e@gateway/web/freenode/ip.92.40.176.142) left irc: Quit: Page closed | 19:55 | |
| timotimo | so _single_item_result tries to figure out if whatever index object that has been plugged in will lead to the result being just a single item, rather than a slice of some sort | 19:56 |
| timotimo | and it checks for isinstance space.w_int, which int8, int16 and int32 fail | 19:56 |
| timotimo | that's why it tries to apply len to the int8/16/32: to figure out if it's a valid tuploid index | 19:57 |
| timotimo | i'm not sure if i should check for int8, int16 and int32 separately or maybe for some common trait | 19:58 |
| amaury_ | hum | 19:59 |
| amaury_ | string is also broken there it seems | 19:59 |
| timotimo | i could try to cast it to int :p | 19:59 |
| amaury_ | timotimo: space.int_w | 19:59 |
| amaury_ | i.e. the __int__ slot | 20:00 |
| timotimo | yes | 20:00 |
| timotimo | do you thik that's a good idea? | 20:00 |
| amaury_ | Do you know what numpy does here? | 20:00 |
| amaury_ | does it accept any class with __int__ or __index__? | 20:00 |
| timotimo | don't know, should i do some digging? | 20:00 |
| amaury_ | or test :) | 20:01 |
| timotimo | .o(and obviously, calling the index slot would be cleverer) | 20:01 |
| timotimo | digging would have meant starting with a test in a ipython session ;) | 20:01 |
| timotimo | it does use an __index__ slot if it's there | 20:02 |
| timotimo | if there's both index and int, index is prefered, it seems | 20:02 |
| timotimo | if there's only int, it will use the int slot | 20:03 |
| amaury_ | OK | 20:03 |
| amaury_ | I've also seen that the __int__ is not used to detect the type of the array | 20:03 |
| tilgovi (~randall@couchdb/developer/tilgovi) joined #pypy. | 20:03 | |
| amaury_ | np.array([2, A()]).dtype is object | 20:04 |
| timotimo | where did you see that? | 20:04 |
| amaury_ | np.array([2, A()], np.int).dtype is int64 | 20:04 |
| timotimo | is that numpy or numpypy? is that right or wrong? :) | 20:04 |
| amaury_ | an old numpy on top of Python2.6 | 20:05 |
| amaury_ | same with numpy-1.6.1 | 20:07 |
| timotimo | can i look-before-you-leap-figure-out if i can do index_w on an object? | 20:07 |
| amaury_ | well, with Python you can leap before you look | 20:10 |
| Var (~Varriount@pool-96-255-203-197.washdc.fios.verizon.net) left irc: Read error: Connection reset by peer | 20:10 | |
| timotimo | this is a tiny bit messy :\ | 20:11 |
| amaury_ | why? all operators work this way | 20:11 |
| timotimo | i mean this particular bit of code | 20:11 |
| timotimo | but never mind, it'll be all right | 20:11 |
| timotimo | oh, there is no index_w. so i'll have to space.invokemethod("__index__", w_idx" or something similar | 20:12 |
| timotimo | apparently it'll be space.call_method(w_idx, "__index__") | 20:12 |
| amaury_ | space.index :) | 20:12 |
| timotimo | ... or that | 20:12 |
| dmishe (~dmishe@c-98-206-42-112.hsd1.il.comcast.net) joined #pypy. | 20:14 | |
| timotimo | hey cool, all tests from test_dytpes pass with my alterations | 20:15 |
| timotimo | now for more tests | 20:15 |
| paniq (~lritter@g231090091.adsl.alicedsl.de) joined #pypy. | 20:16 | |
| timotimo | amaury_: i'm getting "operation cannot succeed" for getattr(space, ('index')), would that mean i have to add an index method to the micronumpy.compile.FakeSpace or did i do something else wrong? | 20:18 |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) left irc: Ping timeout: 276 seconds | 20:19 | |
| amaury_ | is is some translation error? | 20:19 |
| timotimo | it's in test_zjit, an annotation error | 20:19 |
| timotimo | AnnotatorErorr in test_zjit.TestNumpyJIt, the "add" test apparently | 20:20 |
| timotimo | i've added lines like idx = space.int_w(space.index(w_idx)) in two places, both make the annotator choke | 20:20 |
| amaury_ | do you have a test for this space.index? | 20:21 |
| amaury_ | oh right, sorry | 20:21 |
| timotimo | i didn't write one :| | 20:21 |
| amaury_ | in test_zjit a FakeSpace is used | 20:21 |
| timotimo | that's right, i said that ;) | 20:21 |
| amaury_ | which may miss some methods :) | 20:21 |
| timotimo | at least i thought so | 20:21 |
| timotimo | how do i implement that? it'd seem i'd have to implement call_method for that space :| | 20:22 |
| amaury_ | no, just implement index() | 20:24 |
| timotimo | yeah, but index is imlemented by calling the objects __index__ method, is that not correct? | 20:24 |
| amaury_ | IIUC, in FakeSpace, it is not implemented | 20:24 |
| timotimo | that's true | 20:25 |
| amaury_ | there is already a "def int" | 20:25 |
| amaury_ | you could add "index = int" | 20:25 |
| amaury_ | should be good enough for a simple space | 20:25 |
| timotimo | ... well, it's FakeSpace after all, right? | 20:25 |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) joined #pypy. | 20:25 | |
| timotimo | it wouldn't make tests work that make sure numpypy prefers __index__ over __int__ where __int__ fails or something obscure like that? | 20:26 |
| timotimo | oh, that's also incorrect. index returns a wrapped object, so i can't use int_w. i'll just implement it as self.wrap(self.int_w(obj)) | 20:27 |
| timotimo | makes tests pass, makes timo happy | 20:28 |
| ericflo (~ericflo@75.103.8.106) joined #pypy. | 20:29 | |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) left irc: Ping timeout: 255 seconds | 20:32 | |
| amaury_ | ? | 20:32 |
| amaury_ | timotimo: space.int_w(space.index(w_obj)) should work | 20:32 |
| timotimo | that's what i use in my code, yes | 20:33 |
| amaury_ | "index returns a wrapped object, so i can't use int_w" | 20:33 |
| timotimo | that was probably brainfart'd then | 20:33 |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) joined #pypy. | 20:33 | |
| timotimo | ah, no i know what i meant | 20:33 |
| amaury_ | index returns a wrapped object, so you must use int_w | 20:33 |
| timotimo | i can't put index = int_w into FakeSpace, because int_w returns an unwrapped int and index returns a wrapped int | 20:33 |
| timotimo | all micronumpy tests pass, is that good enough for me to commit something? | 20:34 |
| timotimo | additionally: should i write tests to see if micnorumpy prefers __index__ of custom classes to __int__? | 20:34 |
| kenaan | 12ti...@eistee.fritz.box numpypy-issue1137 116459d00ebd48 15/pypy/module/micronumpy/: use __index__ if available, or __int__. should fix this bug. | 20:36 |
| timotimo | bleh, didn't i change my user name? | 20:37 |
| Action: timotimo tries a full translation just for fun | 20:39 | |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) left irc: Ping timeout: 245 seconds | 20:43 | |
| umgeher_ (~umgeher@187.75.230.244) joined #pypy. | 20:51 | |
| umgeher (~umgeher@unaffiliated/umgeher) left irc: Ping timeout: 245 seconds | 20:54 | |
| `fox` (~fox@host227-41-dynamic.50-82-r.retail.telecomitalia.it) joined #pypy. | 20:58 | |
| knizz (~knizz@213.129.230.10) left irc: Ping timeout: 248 seconds | 21:01 | |
| DanielHolth (~dholth@2001:470:7:e7c::2) left irc: Quit: Leaving | 21:03 | |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) joined #pypy. | 21:06 | |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) left irc: Ping timeout: 260 seconds | 21:14 | |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) left irc: Quit: Leaving | 21:17 | |
| timotimo | amaury_: i'm now writing C code for my code in numpypy-issue1137, i'm rather certain it works. do you want to review and merge it before the next nightly build cycle starts? | 21:17 |
| CIA-80 | 03timo 07roundup * 10#1137/"TypeError: 'int32' has no length": | 21:19 |
| CIA-80 | [chatting] this was related to numpypy not trying the __index__ method on objects used as | 21:19 |
| CIA-80 | indexes, nor trying the __int__ slot for non-integer ob ... * 14https://bugs.pypy.org/issue1137 | 21:19 |
| `fox` (~fox@host227-41-dynamic.50-82-r.retail.telecomitalia.it) left irc: Ping timeout: 260 seconds | 21:19 | |
| nedbat (~nedbat@python/psf/nedbat) left irc: Disconnected by services | 21:20 | |
| nedbat_ (~nedbat@python/psf/nedbat) joined #pypy. | 21:20 | |
| Nick change: nedbat_ -> nedbat | 21:26 | |
| Da_Blitz (~Da_Blitz@203.56.250.63) left irc: Read error: Connection reset by peer | 21:26 | |
| Da_Blitz (~Da_Blitz@203.56.250.63) joined #pypy. | 21:27 | |
| bondar (~dnjaramba@41.72.193.86) left irc: | 21:28 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Remote host closed the connection | 21:29 | |
| timotimo | i wonder what the thought process was behind limiting detect_number_of_processors to be 3 when there are 4 processors? | 21:29 |
| Senyai (~Miranda@ppp91-76-210-69.pppoe.mtu-net.ru) left irc: Ping timeout: 265 seconds | 21:29 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 21:29 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Max SendQ exceeded | 21:29 | |
| amaury_ | to leave one processor to system tasks | 21:29 |
| timotimo | translated pypy-c with jit in 2960 seconds | 21:29 |
| timotimo | hm, that may not be a bad idea | 21:29 |
| amaury_ | ssh, kill, and so on | 21:30 |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 21:30 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Max SendQ exceeded | 21:30 | |
| timotimo | right | 21:30 |
| timotimo | those are especially important | 21:30 |
| timotimo | so yeah, it translates fine, the code that was originally bugged is now functioning. what do you think? :) | 21:31 |
| Kxepal (~Miranda@ppp91-79-17-31.pppoe.mtu-net.ru) joined #pypy. | 21:32 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 21:34 | |
| nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 276 seconds | 21:34 | |
| jokey (~jokey@gentoo/developer/jokey) left irc: Quit: No Ping reply in 180 seconds. | 21:36 | |
| jokey (~jokey@gentoo/developer/jokey) joined #pypy. | 21:36 | |
| asmeurer__ (~asmeurer@dhcp-altamirano-65.resnet.nmt.edu) joined #pypy. | 21:37 | |
| lucian (~lucian@host-23-244-171-31.cloudsigma.com) joined #pypy. | 21:39 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) left irc: Max SendQ exceeded | 21:40 | |
| FACEFOX (~facefox@pool-74-111-197-200.lsanca.fios.verizon.net) joined #pypy. | 21:41 | |
| whitelynx (~whitelynx@63.241.75.144) left irc: Quit: Ex-Chat | 21:41 | |
| briancurtin (~briancurt@c-98-215-224-168.hsd1.il.comcast.net) left irc: Quit: sys.exit() | 21:41 | |
| gtaylor (~gtaylor@108-196-160-156.lightspeed.tukrga.sbcglobal.net) left irc: Quit: Konversation terminated! | 21:42 | |
| jesyspa (~s@s559738e3.adsl.wanadoo.nl) left irc: Quit: leaving | 21:42 | |
| Da_Blitz (~Da_Blitz@203.56.250.63) left irc: Ping timeout: 245 seconds | 21:48 | |
| Da_Blitz (~Da_Blitz@203.56.250.63) joined #pypy. | 21:49 | |
| timotimo | hm, what was the necessary magic again to use classes i defined at the top of the test file in a test case that's inside a test classe derived from BaseNumpyAppTest? :\ | 21:50 |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) joined #pypy. | 21:53 | |
| aat (~aat@38.104.173.198) left irc: Quit: Computer has gone to sleep. | 21:54 | |
| timotimo | i thought i could add it to the class and prefix the name with w_ and have it work, but apparently not. | 21:57 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/880 | 22:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/322 | 22:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/cpython-2-benchmark-x86-64/builds/89 | 22:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/1136 | 22:00 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64-2/builds/143 | 22:00 |
| timotimo | noooooooooo, i missed the cutoff (: | 22:00 |
| timotimo | :( | 22:00 |
| dmishe (~dmishe@c-98-206-42-112.hsd1.il.comcast.net) left irc: Quit: Textual IRC Client: http://www.textualapp.com/ | 22:01 | |
| umgeher (~umgeher@187.75.230.244) joined #pypy. | 22:06 | |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) left irc: Remote host closed the connection | 22:09 | |
| umgeher_ (~umgeher@187.75.230.244) left irc: Ping timeout: 260 seconds | 22:10 | |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) joined #pypy. | 22:11 | |
| knizz (~knizz@d86-32-133-50.cust.tele2.at) left irc: Quit: Ex-Chat | 22:27 | |
| oiig_ (~eohnik@112.161.134.227) joined #pypy. | 22:28 | |
| kenaan | 12timo_jbo numpypy-issue1137 11c1a1e296db16 15/pypy/module/micronumpy/test/test_numarray.py: add tests for behavior of objects with either or both int and index slots. | 22:34 |
| dmalcolm (~david@c-24-61-45-39.hsd1.ma.comcast.net) left irc: Quit: Leaving | 22:36 | |
| paniq (~lritter@g231090091.adsl.alicedsl.de) left irc: Quit: Leaving | 22:43 | |
| amaury_ | timotimo: you can kick the buildbot yourself | 22:45 |
| amaury_ | and get a "nightly" build anytime | 22:45 |
| timotimo | that's true, but it feels a bit like cheating :p | 22:46 |
| amaury_ | no, because you can do it on a branch | 22:46 |
| timotimo | additionally, i already got the build, so i don't really need the confirmation or the build | 22:46 |
| amaury_ | btw, array of strings will be very difficult | 22:46 |
| timotimo | it will? aaw :( | 22:47 |
| amaury_ | I fear I have to fix a bug in std objspace | 22:47 |
| timotimo | can you make it raise an exception instead of infinitelooping in the mean time, though? | 22:47 |
| amaury_ | well sure | 22:47 |
| timotimo | that would help my test suite :) | 22:47 |
| amaury_ | but try this: | 22:48 |
| amaury_ | x = np.str_('abc') | 22:48 |
| amaury_ | isinstance(x, str) True | 22:48 |
| amaury_ | that's ok | 22:48 |
| amaury_ | but try to print x | 22:48 |
| amaury_ | or call any str method :( | 22:49 |
| timotimo | oh, that's with regular numpy now, yes? | 22:49 |
| amaury_ | no, pypy only | 22:49 |
| ixokai (~ixokai@pdpc/supporter/student/ixokai) left irc: Quit: Bye! | 22:49 | |
| amaury_ | regular numpy works correctly | 22:49 |
| timotimo | ah, ok. i get "module object has no attribute str_" | 22:49 |
| timotimo | oh, it's an old pypy again, sorry about that | 22:50 |
| amaury_ | np.string_, then :) | 22:50 |
| timotimo | i see the problem :( | 22:50 |
| amaury_ | well, it's worse than that | 22:50 |
| amaury_ | in np.str_ is supposed to inherit from two types | 22:51 |
| amaury_ | str and numpy.character | 22:51 |
| amaury_ | and it's impossible in pypy at the moment | 22:51 |
| timotimo | oh my :\ | 22:52 |
| amaury_ | a built-in type cannot be mixed with another | 22:52 |
| amaury_ | (except w_object maybe) | 22:52 |
| timotimo | want me to create a bug so there's a central point for progress reports and discussion? | 22:53 |
| amaury_ | I think I narrowed the problem to typeobject.get_parent_layout() | 22:53 |
| amaury_ | yes please | 22:53 |
| Action: amaury_ -> bed anyway | 22:53 | |
| timotimo | "string arrays in numpypy don't work yet"? | 22:54 |
| amaury_ | not at all! | 22:54 |
| aat (~aat@pool-108-41-149-224.nycmny.fios.verizon.net) joined #pypy. | 22:54 | |
| amaury_ | x = np.str_('abc') is an example | 22:54 |
| timotimo | "string arrays in numpypy fundamentally don't work"? | 22:54 |
| amaury_ | it's even worse with unicode | 22:54 |
| amaury_ | fundamentally is the word | 22:54 |
| timotimo | i'm confused | 22:54 |
| timotimo | x = np.str_("abc") does "work" - it does something | 22:55 |
| timotimo | (as opposed to the infinite loop i was getting earlier) | 22:55 |
| amaury_ | str_ is not array | 22:55 |
| timotimo | oh, of course | 22:55 |
| timotimo | yes | 22:55 |
| amaury_ | str_ is probably the array item | 22:55 |
| CIA-80 | 03timo 07roundup * 10#1138/in numpypy, strings fundamentally don't work: | 22:57 |
| CIA-80 | [new] npp.array("foo") or even npp.array("f") will enter an infinite loop. | 22:57 |
| CIA-80 | npp.str_("test") will succeed, but the result is not printable and s ... * 14https://bugs.pypy.org/issue1138 | 22:57 |
| timotimo | there it is | 22:57 |
| EnCuKou (~encukou@ip-94-113-220-25.net.upcbroadband.cz) joined #pypy. | 23:01 | |
| int3___ (~int3__@vlan409-164.subnet-248.amherst.edu) joined #pypy. | 23:03 | |
| int3__ (~int3__@vlan409-164.subnet-248.amherst.edu) left irc: Read error: Connection reset by peer | 23:03 | |
| int3___ (~int3__@vlan409-164.subnet-248.amherst.edu) left irc: Remote host closed the connection | 23:03 | |
| oiig_ (~eohnik@112.161.134.227) left irc: Ping timeout: 265 seconds | 23:03 | |
| eoh| (~eoh@106.103.5.157) joined #pypy. | 23:05 | |
| eoh (~eoh@112.161.134.227) left irc: Ping timeout: 240 seconds | 23:08 | |
| cdash004 (~sdf@host-173-230-2-206.vanodod.clients.pavlovmedia.com) joined #pypy. | 23:08 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 246 seconds | 23:13 | |
| Tobu (~Tobu@unaffiliated/tobu) left irc: Remote host closed the connection | 23:14 | |
| Tobu (~Tobu@unaffiliated/tobu) joined #pypy. | 23:15 | |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) joined #pypy. | 23:30 | |
| fijal (~fijal@aqi122.internetdsl.tpnet.pl) left irc: Ping timeout: 246 seconds | 23:39 | |
| voidspac_ (~voidspace@host217-39-56-75.range217-39.btcentralplus.com) left irc: Quit: Goodbye... | 23:49 | |
| r_r_r (~chatzilla@77.127.109.40) left irc: Quit: ChatZilla 0.9.88.2 [Firefox 12.0/20120420145725] | 23:54 | |
| --- Sat Apr 28 2012 | 00:00 | |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!