| fprimex (~brent@fl-67-233-144-239.dhcp.embarqhsd.net) joined #pypy. | 00:11 | |
| arigato (~arigo@82.113.121.130) left irc: Quit: See you | 00:25 | |
| fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) joined #pypy. | 00:29 | |
| rokujyouhitoma (~rokujyouh@12.51.221.130) left irc: Ping timeout: 240 seconds | 00:29 | |
| rokujyouhitoma (~rokujyouh@12.51.221.130) joined #pypy. | 00:30 | |
| jterrace | i cant close my own pull request? | 00:30 |
|---|---|---|
| abki_ | my program running on top of zmq-ctypes is reallly slow, according to the profiling tool most of the time is spent in _call_functptr@function.py of ctypes | 00:35 |
| Alex_Gaynor | abki_: are you seeing any messages logged about declaring types? | 00:36 |
| abki_ | no messages at all | 00:37 |
| etrepum (~bob@c-76-27-62-239.hsd1.ut.comcast.net) left irc: Quit: etrepum | 00:37 | |
| abki_ | the function is _call_funcptr | 00:37 |
| abki_ | without `t` to func | 00:37 |
| Alex_Gaynor | no suggestions off the top of my head, you could try playing with the --jit trace_limit=<value> param | 00:37 |
| Alex_Gaynor | I can't recall what the default is (--help will show you), try increasing it | 00:37 |
| abki_ | Alex_Gaynor: thx for the tip | 00:39 |
| abki_ | but I found something else | 00:39 |
| jterrace | Alex_Gaynor: can you delete pull request #16 for me? looks like i cant close it myself | 00:39 |
| jterrace | just submitted the uint stuff on its own: https://bitbucket.org/pypy/pypy/pull-request/17 | 00:39 |
| Alex_Gaynor | jterrace: great, thanks so much, will review and merge this stuff later tonight | 00:39 |
| jterrace | thanks | 00:39 |
| jterrace | ill try to submit the str stuff tonight too | 00:40 |
| jterrace | once you merge those two, i can submit the fromstring | 00:40 |
| abki_ | it looks like the profiling tool trace different things with pypy et python | 00:40 |
| abki_ | with python the biggest box in runsnake is zmq.socket.send | 00:41 |
| abki_ | in pypy it's _call_funcptr but it's inside zmq.socket.send... so the profiling tools says the same thing actually | 00:42 |
| jterrace | actually, Alex_Gaynor: if i add the string stuff as a separate pull request, it might not automerge | 00:42 |
| jterrace | should i just wait until the other ones get merged? | 00:42 |
| jterrace | 15 and 17 should be good, ill submit more tomorrow if they get merged | 00:45 |
| jterrace (~Adium@dynamic-oit-vapornet-b-25.Princeton.EDU) left irc: Quit: Leaving. | 00:45 | |
| timotimo (~timo@195.191.196.6) left irc: Remote host closed the connection | 00:45 | |
| abki_ | which value is significant in time command output please ? | 00:46 |
| papercrane (~papercran@c-71-202-211-7.hsd1.ca.comcast.net) left irc: Ping timeout: 268 seconds | 00:50 | |
| papercrane (~papercran@c-71-202-211-7.hsd1.ca.comcast.net) joined #pypy. | 00:54 | |
| voidspace_ (~voidspace@87-194-212-65.bethere.co.uk) joined #pypy. | 00:55 | |
| voidspace_ (~voidspace@87-194-212-65.bethere.co.uk) left irc: Changing host | 00:55 | |
| voidspace_ (~voidspace@python/psf/voidspace) joined #pypy. | 00:55 | |
| voidspace (~voidspace@python/psf/voidspace) left irc: Ping timeout: 240 seconds | 00:57 | |
| Nick change: voidspace_ -> voidspace | 00:57 | |
| squiddy (~squiddy@f053082034.adsl.alicedsl.de) left irc: Quit: Leaving | 00:57 | |
| CIA-13 | 03mwh 07roundup * 10#906/CTRL-leftarrow and CTRL-rightarrow don't work in PyPy console: | 00:58 |
| CIA-13 | I think implementing it in pyrepl probably makes sense. Sorry for taking such a | 00:58 |
| CIA-13 | long time to reply... | 00:58 |
| CIA-13 | In the mean time I found that the input ... * 14https://bugs.pypy.org/issue906 | 00:58 |
| Nick change: smuxi -> panni_ | 01:03 | |
| zain__ (~textual@c-67-160-201-63.hsd1.ca.comcast.net) joined #pypy. | 01:05 | |
| zain (~textual@c-67-160-201-63.hsd1.ca.comcast.net) left irc: Ping timeout: 240 seconds | 01:06 | |
| Nick change: zain__ -> zain | 01:06 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/166 | 01:07 |
| papercrane (~papercran@c-71-202-211-7.hsd1.ca.comcast.net) left irc: Ping timeout: 268 seconds | 01:08 | |
| witulski (~stupsi@ip-176-198-244-135.unitymediagroup.de) joined #pypy. | 01:10 | |
| witulski (stupsi@ip-176-198-244-135.unitymediagroup.de) left #pypy. | 01:10 | |
| papercrane (~papercran@c-76-103-172-115.hsd1.ca.comcast.net) joined #pypy. | 01:10 | |
| derdon (~derdon@p5DE89AF6.dip.t-dialin.net) left irc: Remote host closed the connection | 01:12 | |
| stakkars | arigato: nice script to click oh "ok", maybe it is a better solution than mine, which costs more resources. | 01:14 |
| stakkars | arigato: and my solution is more versatile, but a bit harder to configure, and somehow screen/display dependant. | 01:15 |
| stakkars | arigato: I use the jython visuaal scripting of | 01:16 |
| stakkars | arigato: http://www.sikuli.org | 01:16 |
| stakkars | arigato: which you probably also teach to play bubs'n bros | 01:17 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/740 | 01:17 |
| fprimex (~brent@fl-67-233-144-239.dhcp.embarqhsd.net) left irc: Quit: ufc | 01:19 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) joined #pypy. | 01:35 | |
| berkerpeksag (~berkerpek@46.196.255.228) joined #pypy. | 01:35 | |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/1884 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/625 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/301 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/639 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1153 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/823 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1465 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/745 | 01:36 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/289 | 01:36 |
| berkerpeksag (~berkerpek@46.196.255.228) left irc: Client Quit | 01:36 | |
| abki_ | I did some benchmark with timeit inside my code it's still has 20% difference with the cpython version | 01:49 |
| abki_ | How can I compare both runs to see where the perfomance gap comes from ? | 01:49 |
| stakkars | what does the "-lrt" makefile option mean, and why do we need this? | 01:50 |
| Alex_Gaynor | stakkars: it links the rt library, whatever it is | 01:50 |
| stakkars | I am trying to find out if there is a windows equivalent. | 01:51 |
| asmeurer__ (~asmeurer@c-174-56-21-245.hsd1.nm.comcast.net) left irc: Quit: asmeurer__ | 01:51 | |
| stakkars | finding many testsd which cannot run on windows, and instead of skipping, they raise an explicit error. Boring! | 01:52 |
| stakkars | It feels like things are dumped on windows, and I have to solve it. | 01:53 |
| stakkars | Alex_Gaynor: ok, some real time stuff. Whatever it is, and no idea why it must be used. | 01:54 |
| stakkars | surprisingly, test_posix.TestMakefile tests literally if "-lrt" is in the makefile! | 01:56 |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) joined #pypy. | 01:56 | |
| stakkars | this is a test for every platform, provided it is Linux. Absolutely ignorant style, without any hint. Who did it? | 01:58 |
| Action: mikefc hides behind the couch | 01:58 | |
| exarkun | Windows is an absolutely hideous platform | 01:59 |
| stakkars | sorry about my rant, but I always must fight three platforms and get a bit upset by this. | 02:00 |
| Vorpal (~AnMaster@unaffiliated/anmaster) left irc: Ping timeout: 252 seconds | 02:01 | |
| stakkars | exarkun: but it is the biggest platform that the masses use, and I want to get all those users for PyPy. | 02:02 |
| exarkun | stakkars: Yes, and I hope you keep working on that. :) | 02:02 |
| stakkars | I hate this platform absolutely and suffer the fact that I once was a specialist for it ;-) | 02:03 |
| exarkun | But I'm not surprised that PyPy has lots of problems on Windows, owing to aforementioned hideousness. :) | 02:03 |
| exarkun | (I spent 6 hours today trying to find a workaround for a windows kernel memory management bug) | 02:03 |
| Alex_Gaynor | exarkun: have you filed it? can one even file a bug against windows? | 02:04 |
| exarkun | Alex_Gaynor: If you have a support contract with them, they'll let you file bugs. | 02:04 |
| Alex_Gaynor | sigh | 02:04 |
| exarkun | So, good luck to everyone else who ever calls WSAAddressToStringA! You're in for an adventure. | 02:04 |
| stakkars | why me? (yell, scream, &) well, I'm not complaining, being the happy janitor for PyPy | 02:05 |
| stakkars | it will all be good (trying to believe myself) :-) | 02:05 |
| stakkars | exarkun: thank you, you said the right thing at the right time | 02:06 |
| Action: stakkars falling asleep. will work that out, tomorrow | 02:07 | |
| exarkun | stakkars: :) | 02:07 |
| ojii (~ojii@22-14.62-81.cust.bluewin.ch) left irc: Quit: Leaving | 02:17 | |
| stakkars | ok, I will try to get more help. There are so many windows issues that I cannot hanle this alone. | 02:20 |
| stakkars | I want to build win64 pypy, but win32 is full of bugs which need to be cleaned first, to | 02:21 |
| stakkars | get something stable to compare against. | 02:21 |
| Guest48067 (~timo@gallifrey.wybt.net) joined #pypy. | 02:22 | |
| stakkars | I think to write this to the list, and into the blog. Windows is not supported enough, although | 02:22 |
| Nick change: Guest48067 -> timotimo | 02:22 | |
| stakkars | I wish to thank all who took care. Good night. | 02:22 |
| Insectoid (~Q@c-98-230-8-37.hsd1.fl.comcast.net) left irc: Ping timeout: 252 seconds | 02:36 | |
| rokujyouhitoma (~rokujyouh@12.51.221.130) left irc: Remote host closed the connection | 02:37 | |
| overminder (~overminde@wf103-072.ust.hk) joined #pypy. | 02:39 | |
| Shanita (~John@osbk-4db15917.pool.mediaWays.net) joined #pypy. | 02:39 | |
| Kaskuka (~John@osbk-4db06a58.pool.mediaWays.net) left irc: Ping timeout: 240 seconds | 02:40 | |
| rokujyouhitoma (~rokujyouh@12.51.221.130) joined #pypy. | 02:41 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/974 | 02:46 |
| int3__ (~int3__@dorm-57.subnet-204.amherst.edu) joined #pypy. | 02:49 | |
| fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) left irc: Quit: fzzzy | 02:53 | |
| gutworth (~benjamin@dhcp-128-36-84-91.central.yale.edu) left irc: Ping timeout: 248 seconds | 03:09 | |
| fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) joined #pypy. | 03:10 | |
| Rhyolite | Why are so many people obsessed with programming directly in RPython? | 03:23 |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Ping timeout: 240 seconds | 03:28 | |
| fenrrir (~fenrrir@189.70.241.215) joined #pypy. | 03:34 | |
| mikefc | the promise of speeeeeeeed | 03:36 |
| mikefc | and that the separation of all the components of pypy is very confusing and rpython seems to be the magic ingredient. | 03:36 |
| lucian | also, because of cython; some people want a python with static types for some reason | 03:37 |
| alistra | static types = best thing ever | 03:38 |
| lucian | in ML/haskell, maybe | 03:39 |
| Action: lucian goes to sleep | 03:39 | |
| alistra | way to initiate a flame | 03:40 |
| alistra | and go to sleep | 03:40 |
| Garen_ | Rhyolite: Something better than py2exe would be nice | 03:42 |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) left irc: Ping timeout: 240 seconds | 03:43 | |
| Nick change: Garen_ -> Garen | 03:46 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/1884 | 03:47 |
| fenrrir (fenrrir@189.70.241.215) left #pypy ("Saindo"). | 03:48 | |
| harrison_ (~quassel@adsl-69-209-194-17.dsl.chcgil.ameritech.net) left irc: Ping timeout: 255 seconds | 03:52 | |
| timotimo (~timo@gallifrey.wybt.net) left irc: Quit: Lost terminal | 04:05 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) left irc: Quit: Computer has gone to sleep. | 04:08 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) joined #pypy. | 04:10 | |
| Nick change: Gulaway -> Gulopine | 04:23 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) left irc: Quit: Computer has gone to sleep. | 04:26 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) joined #pypy. | 04:37 | |
| mikefc | any cython experts? | 04:38 |
| Alex_Gaynor | In here? Probably not. | 04:44 |
| mikefc | thought there might be someone working on bringing the two together. I'm trying to optimise some line drawing code... | 04:45 |
| mikefc | the pypy code is 30% faster than cython code. but i don't know that my cython typedef'ing is up to standard | 04:47 |
| Alex_Gaynor | you'll have much more luck in a cython channel I suspect | 04:47 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1465 | 04:49 |
| mikefc | yep. | 04:49 |
| kenaan | 12alex_gaynor default 113eba2ed546ad 15/pypy/module/micronumpy/: (jterrace) added tolist to various objects in numpy. | 04:50 |
| nedbat (~nedbat@python/psf/nedbat) left irc: Read error: Operation timed out | 04:51 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/823 | 04:57 |
| Insectoid (~Q@c-98-230-8-37.hsd1.fl.comcast.net) joined #pypy. | 04:58 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) left irc: Quit: Computer has gone to sleep. | 05:06 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/639 | 05:09 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1153 | 05:14 |
| int3___ (~int3__@st401-246.subnet-244.amherst.edu) joined #pypy. | 05:17 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/625 | 05:18 |
| int3__ (~int3__@dorm-57.subnet-204.amherst.edu) left irc: Ping timeout: 248 seconds | 05:20 | |
| linq (~ident@24-246-25-39.cable.teksavvy.com) left irc: Ping timeout: 248 seconds | 05:22 | |
| overminder (~overminde@wf103-072.ust.hk) left irc: Ping timeout: 252 seconds | 05:23 | |
| linq (~ident@24-246-25-39.cable.teksavvy.com) joined #pypy. | 05:25 | |
| kenaan | 12alex_gaynor default 11716c084b46a1 15/pypy/module/micronumpy/: (jterrace, alex): Expose unsigned integer types in numpy and fix a typo. Fixed up the handling of unsigned64... | 05:26 |
| linq (~ident@24-246-25-39.cable.teksavvy.com) left irc: Ping timeout: 240 seconds | 05:39 | |
| bbot2 | 3Success: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/745 | 06:18 |
| rinu (75c48a48@gateway/web/freenode/ip.117.196.138.72) joined #pypy. | 06:38 | |
| rinu | is this the pypy irc ? | 06:38 |
| tos9 | Yes | 06:39 |
| rinu | i am kinda not used to irc , will whatever i post be seen to everybody here ? | 06:39 |
| tos9 | It will and has been, yes :) | 06:40 |
| rinu | ok , does pypy convert the python code to machine level code ? | 06:41 |
| rinu | or is it optimised c code ? | 06:41 |
| rinu | hey? why is this slow ? | 06:43 |
| rinu | hello ? | 06:49 |
| rinu (75c48a48@gateway/web/freenode/ip.117.196.138.72) left irc: Quit: Page closed | 06:52 | |
| pjenvey | y u no | 06:54 |
| etrepum (~bob@69-9-57-57.static.logl001.digis.net) joined #pypy. | 07:18 | |
| int3___ (~int3__@st401-246.subnet-244.amherst.edu) left irc: Remote host closed the connection | 07:22 | |
| int3__ (~int3__@dorm-57.subnet-204.amherst.edu) joined #pypy. | 07:23 | |
| int3__ (~int3__@dorm-57.subnet-204.amherst.edu) left irc: Remote host closed the connection | 07:23 | |
| rokujyouhitoma (~rokujyouh@12.51.221.130) left irc: Ping timeout: 255 seconds | 07:42 | |
| rokujyouhitoma (~rokujyouh@12.51.221.130) joined #pypy. | 07:43 | |
| overminder (~overminde@143.89.167.3) joined #pypy. | 07:47 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/301 | 07:47 |
| CIA-13 | 03mikefc 07roundup * 10#959/numpypy array operation improperly jitted away: | 07:51 |
| CIA-13 | [new] * with pypy jit, array values get left at [0,0,0] instead of set to [128, 128, 128] (at | 07:51 |
| CIA-13 | about the 2000th iteration) | 07:51 |
| CIA-13 | * succeeds with CPyt ... * 14https://bugs.pypy.org/issue959 | 07:51 |
| overminder (~overminde@143.89.167.3) left irc: Ping timeout: 255 seconds | 07:51 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 07:58 | |
| CIA-13 | 03mikefc 07roundup * 10#959/numpypy array operation improperly jitted away: [chatting] updated to the latest nightly (16f29d272911). Still happening. * 14https://bugs.pypy.org/issue959 | 08:00 |
| Nick change: Gulopine -> Gulaway | 08:09 | |
| Nick change: Gulaway -> Gulopine | 08:13 | |
| Nick change: Gulopine -> Gulaway | 08:21 | |
| antocuni (~antocuni@host117-71-dynamic.58-82-r.retail.telecomitalia.it) joined #pypy. | 08:26 | |
| squiddy (~squiddy@f053082220.adsl.alicedsl.de) joined #pypy. | 08:31 | |
| arigato (~arigo@89.204.130.97) joined #pypy. | 08:46 | |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1154 [12arigo, SpecialisedTuples] | 08:54 |
| `fox` (~fox@host48-59-dynamic.48-82-r.retail.telecomitalia.it) joined #pypy. | 09:05 | |
| kenaan | 12hakanardo jit-targets 11cb17a24f8acd 15/pypy/jit/metainterp/compile.py: dissable fallback for now | 09:09 |
| kenaan | 12hakanardo jit-targets 11ba23d85f0f16 15/: hg merge default | 09:09 |
| tumbleweed | anyone interested in a build + test log of 1.7 on ia64? (the only debian porterbox with enough ram for me to be brave enough to try it) a fair number of test sugfaults, and lots of failures because it can't parse /proc/cpuinfo http://people.debian.org/~stefanor/pypy/pypy_1.7+dfsg-1_ia64.build | 09:09 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/626 [12hakanardo, jit-targets] | 09:11 |
| hruske (~Gasper@2001:15c0:66e9:14:8096:69ff:fe24:fcf7) joined #pypy. | 09:15 | |
| hruske (~Gasper@2001:15c0:66e9:14:8096:69ff:fe24:fcf7) left irc: Client Quit | 09:15 | |
| kkris (~kris@77.116.49.123.wireless.dyn.drei.com) joined #pypy. | 09:15 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/626 [12hakanardo, jit-targets] | 09:17 |
| arigato | hakanardo: can you write a little motivation in checkins like cb17a24f8acd? it's hard to guess why | 09:23 |
| hakanardo | arigato: sure, I realy just want to run the tests without it | 09:28 |
| arigato | ah | 09:28 |
| hakanardo | to see if it was the cause of all the failuers or if it was the merge | 09:28 |
| hakanardo | that's the two things that have happend since the tests last passed I think... | 09:29 |
| hakanardo | how either of them can cause an assertion error in optimize_STRLEN is a bit odd though :) | 09:30 |
| fijal | hi armin | 09:31 |
| arigato | hi | 09:32 |
| fijal | tumbleweed: what segfaults? | 09:32 |
| fijal | tests should not segfault | 09:32 |
| tumbleweed | agreed :P | 09:32 |
| tumbleweed | unfortunatly that didn't seem to show up much in the log | 09:32 |
| tumbleweed | I'll re-run some of those tests by hand and see what I can see... | 09:33 |
| fijal | can you find names of tests that segfaulted? | 09:33 |
| fijal | arigato: for what is worth, we're getting a *real* build, like one where we run tests | 09:33 |
| fijal | (-A, lib-python & test_pypy_c at least) | 09:34 |
| fijal | for debian | 09:34 |
| arigato | what is unreal about the builds we have so far? | 09:34 |
| arigato | e.g. tannit32 is debian too | 09:35 |
| fijal | I mean packaging | 09:36 |
| tumbleweed | arigato: you masked http://paste.pocoo.org/show/519228/ on darwin, but I'm seeing it on Debian too | 09:36 |
| fijal | our packages for PPA/fedora/gentoo etc. don't really run tests | 09:36 |
| fijal | arigato: fedora is still changing C flags I believe :/ | 09:37 |
| Vorpal (~AnMaster@unaffiliated/anmaster) joined #pypy. | 09:37 | |
| `fox` (~fox@host48-59-dynamic.48-82-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds | 09:38 | |
| arigato | write in a corner of our web site: linux-distribution-specific packages: (links to x y z) fedora: broken, don't use | 09:38 |
| arigato | with a bit of explanation | 09:38 |
| fijal | we should ask dmalcolm | 09:40 |
| fijal | when he's around | 09:40 |
| `fox` (~fox@host156-15-dynamic.4-87-r.retail.telecomitalia.it) joined #pypy. | 09:40 | |
| tumbleweed | given the time taken for translation, running tests is a no-brainer | 09:40 |
| fijal | tumbleweed: you would think so :)P | 09:40 |
| fijal | yet noone does that | 09:41 |
| arigato | anyone has a recent enough version of openssl to support mdc2 and whirlpool? | 09:45 |
| arigato | openssl dgst -whirlpool filename | 09:45 |
| tumbleweed | sure | 09:46 |
| arigato | open('x', 'wb').write("Nobody inspects the spammish repetition") | 09:47 |
| tumbleweed | whirlpool yes, MDC-2 no | 09:47 |
| arigato | ok, so at least the whirlpool of this file x? | 09:47 |
| tumbleweed | whirlpool(x)= 1a22b79fe5afda02c63a25927193ed01dc718b74026e597608ce431f9c3d2c9e74a7350b7fbb7c5d4effe5d7a31879b8b7a10fd2f544c4ca268ecc6793923583 | 09:47 |
| arigato | thank you :-) | 09:47 |
| tumbleweed | ah, got mdc2 on a machine running debian/stable. It was disabled in more recent openssl builds on debian | 09:50 |
| tumbleweed | no, it lies | 09:50 |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1154 [12arigo, SpecialisedTuples] | 09:55 |
| kenaan | 12arigo default 119de8b8b018cb 15/pypy/module/_hashlib/test/test_hashlib.py: Add a test for some of the nonstandard hashes (issue957) | 09:58 |
| kkris1 (~kris@93-82-34-223.adsl.highway.telekom.at) joined #pypy. | 10:00 | |
| fijal | uh | 10:00 |
| fijal | so nedbat's changes to sandbox rely on identity of 1 :/ | 10:00 |
| fijal | arigato: do we guarantee 1 is 1? | 10:01 |
| kkris (~kris@77.116.49.123.wireless.dyn.drei.com) left irc: Ping timeout: 252 seconds | 10:02 | |
| Da_Blitz (~Da_Blitz@203.56.250.63) left irc: Read error: Connection reset by peer | 10:05 | |
| Da_Blitz (~Da_Blitz@203.56.250.63) joined #pypy. | 10:05 | |
| arigato | fijal: right now, yes | 10:06 |
| fijal | ok | 10:06 |
| fijal | anyway, not the best thing to rely on | 10:06 |
| kenaan | 12fijal default 111b1197dcc86d 15/pypy/translator/sandbox/: Merge nedbat-sandbox branch, with one tweak, don't realy on identity of one, just use object() for identitiy checks. | 10:07 |
| arigato | yes | 10:07 |
| kenaan | 12fijal nedbat-sandbox 112eb520555de5 15/: close merged branch | 10:07 |
| fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) left irc: Quit: fzzzy | 10:09 | |
| kenaan | 12arigo SpecialisedTuples 111c9dc407966c 15/pypy/module/cpyext/methodobject.py: Dead import | 10:13 |
| kenaan | 12arigo SpecialisedTuples 115e1080705d7c 15/pypy/module/cpyext/sequence.py: Fix: make sure that PySequence_Fast() returns a W_ListObject or a W_TupleObject, not just some object of ... | 10:13 |
| kenaan | 12arigo SpecialisedTuples 110b382030a9e9 15/pypy/module/cpyext/sequence.py: Typo. | 10:13 |
| tumbleweed | so, what's the right thing to do in test__mercurial when I'm not building from a repo? get_repo_version_info returns "?" as the hg revision, but test_mercurial expects either '' or a hash | 10:21 |
| fijal | nice | 10:21 |
| fijal | tumbleweed: get_repo_version() should return a version at which is was released | 10:22 |
| tumbleweed | so I just patch the hell out of it :) | 10:22 |
| fijal | ff4af8f31882 | 10:22 |
| fijal | I believe | 10:22 |
| fijal | the tag is release-1.7 | 10:22 |
| tumbleweed | I can store that when I build my source tarball | 10:23 |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy. | 10:23 | |
| tumbleweed | but if I'm applying a few patches to pypy, that's no longer accurate | 10:23 |
| Action: jerith tries translating pypy. | 10:25 | |
| arigato | tumbleweed: use the same policy as CPython | 10:28 |
| tumbleweed | >>> sys.subversion | 10:30 |
| tumbleweed | ('CPython', '', '') | 10:30 |
| arigato | I'm thinking about the normal prompt in sys.version | 10:31 |
| tumbleweed | >>> sys.version | 10:31 |
| tumbleweed | '2.7.2+ (default, Nov 30 2011, 19:22:03) \n[GCC 4.6.2]' | 10:31 |
| tumbleweed | nothing about revision hashes | 10:31 |
| arigato | Python 2.7.2+ (2.7:fd9d7a8e45bc, Dec 3 2011, 18:50:12) | 10:32 |
| arigato | on my side | 10:32 |
| tumbleweed | ok, I'll return '' | 10:32 |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 10:36 | |
| arigato (~arigo@89.204.130.97) left irc: Read error: Connection reset by peer | 10:36 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 10:36 | |
| aleksi (~aleksi@109-184-62-134.dynamic.mts-nn.ru) joined #pypy. | 10:39 | |
| arigato (~arigo@82.113.99.33) joined #pypy. | 10:39 | |
| arigato | compute_digest_size() in module/_hashlib is broken? but I don't get how the test I just checked in passes | 10:40 |
| fijal (~fijal@41.54.155.67) left irc: Ping timeout: 240 seconds | 10:40 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 10:41 | |
| arigato | well it's not called (first hint) | 10:41 |
| tumbleweed | not entirely sure what to make of any of these: http://paste.pocoo.org/show/519242/ | 10:42 |
| arigato | baaaaaaaaaaaah | 10:42 |
| arigato | bah**42 | 10:43 |
| arigato | nowadays py.test is broken in the sense that app-level tests can yield other tests, | 10:43 |
| arigato | but they are then run at intepr-lvel | 10:43 |
| arigato | so of course the tests pass | 10:44 |
| fijal (~fijal@41.54.155.67) left irc: Ping timeout: 252 seconds | 10:45 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 10:46 | |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 10:49 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 10:51 | |
| aleksi (~aleksi@109-184-62-134.dynamic.mts-nn.ru) left irc: Ping timeout: 255 seconds | 10:52 | |
| tumbleweed | here's what I'm applying to pypy/module/_ssl/test/test_ssl.py: http://paste.pocoo.org/show/519246/ | 10:53 |
| `fox` (~fox@host156-15-dynamic.4-87-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds | 10:54 | |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 10:55 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 10:55 | |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 10:59 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:01 | |
| hpk (~hpk@hq2.merlinux.eu) joined #pypy. | 11:04 | |
| fijal (~fijal@41.54.155.67) left irc: Ping timeout: 255 seconds | 11:05 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:06 | |
| fijal (~fijal@41.54.155.67) left irc: Ping timeout: 255 seconds | 11:11 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:12 | |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 11:15 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:15 | |
| kenaan | 12hakanardo jit-targets 113b4f0126fae2 15/pypy/jit/metainterp/: clone the short_boxes to allow it to be reused in some fallabck if the first athempt to optimize fails | 11:19 |
| kenaan | 12hakanardo jit-targets 11ebf413fc4faa 15/pypy/: hg merge default | 11:19 |
| fijal (~fijal@41.54.155.67) left irc: Ping timeout: 255 seconds | 11:20 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:21 | |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 11:24 | |
| kenaan | 12arigo default 110d499a1ce192 15/pypy/: Fix. The testing framework was happily accepting app-level methods that are generators, and then running the yield... | 11:24 |
| kenaan | 12arigo default 112e5b6dce5753 15/pypy/: Fix: _hashlib was just broken for any non-standard hash function. Small extra clean-ups. | 11:24 |
| CIA-13 | 03arigo 07roundup * 10#957/module hashlib is broken for non-standard digest algorithms: [resolved] Bah. 2e5b6dce5753 * 14https://bugs.pypy.org/issue957 | 11:24 |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:24 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy. | 11:26 | |
| overminder (~overminde@143.89.167.3) joined #pypy. | 11:27 | |
| Wraithan (~Wraithan@c-67-189-110-252.hsd1.or.comcast.net) joined #pypy. | 11:27 | |
| Wraithan | https://gist.github.com/1460132 Interesting numbers, was wondering if maybe I am not considering something in my timeit usage since sometimes background stuff can affect timeit | 11:28 |
| Wraithan | I expected that list comprehension to be a little bit faster, but not as much as that is. Also unexpected is cpython being faster at deepcopy. | 11:29 |
| fijal (~fijal@41.54.155.67) left irc: Ping timeout: 252 seconds | 11:29 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:30 | |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 11:33 | |
| fijal (~fijal@41.54.155.67) joined #pypy. | 11:33 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 240 seconds | 11:34 | |
| arigato | Wraithan: you don't give the JIT enough time to warm up. | 11:34 |
| Wraithan | Well in my practical case I wont need more to run more than 1200 times. | 11:35 |
| Wraithan | 1000 was just a nice round number to us | 11:35 |
| fijal (~fijal@41.54.155.67) left irc: Read error: Connection reset by peer | 11:36 | |
| arigato | runs of pypy=5ms instead of cpython=1ms are in the category "too bad, but we really don't care, and wonder why you do" | 11:39 |
| arigato | pypy has a JIT with a very slow warm-up (too slow for some use cases, admittedly) | 11:39 |
| arigato | I'm just saying that these numbers show nothing at all, please don't conclude from them "pypy is slow" | 11:39 |
| arigato | instead, try pypy on some complete program that runs several seconds in cpython | 11:40 |
| Wraithan | arigato: Oh, wasn't saying pypy was slow, just thought it was interesting that it was slower than cpython for that use case. I am using pypy regardless for the project I just use cpython as a control when I am doing testing. | 11:40 |
| Wraithan | More interesting to me was that listcomp+slice was faster than deepcopy by an order of magnitude in both interpreters. | 11:41 |
| arigato | there are known use cases where pypy is slower: when the JIT doesn't have time to warm-up, or when running test suites, for example | 11:41 |
| arigato | Wraithan: I agree with this conclusion in CPython, but for pypy you need to run it at least 0.1 second, preferrably 1 second, before you can reduce anything about performance | 11:42 |
| arigato | (although it's likely that listcomp+slice wins anyway) | 11:42 |
| arigato | s/reduce/deduce | 11:42 |
| hakanardo | translation is currently broken? http://paste.pocoo.org/show/519259/ | 11:43 |
| Wraithan | Well the competition I am in, aichallenge.org pypy 1.7.0 was about a 15x speed up over cpython2.7.2 | 11:43 |
| arigato | Wraithan: ok :-) | 11:43 |
| arigato | I just wanted to warn you that some programs scale differently than others in pypy, | 11:44 |
| arigato | e.g. a line using deepcopy has typically a much longer warm-up time than listcomp+slice, | 11:44 |
| Wraithan | The only people in it that are doing well with cpython are using cython and numpy. My pure python bot is nearing the same speeds as them my algorithms are just less good but improving | 11:44 |
| arigato | which means that if you run it not for long enough, you might miss the optimizations that will be applied later in the deepcopy case | 11:44 |
| arigato | hakanardo: that looks like micronumpy | 11:45 |
| arigato | just ignore it with --withoutmod-micronumpy | 11:46 |
| Wraithan | Even at 1m reps deep copy is still an order of magnitude slower, also cpython is taking forever to run it 1m times | 11:46 |
| arigato | ok | 11:46 |
| hakanardo | arigato: ok, can I pass that to the buildbot? | 11:47 |
| arigato | :-/ | 11:47 |
| arigato | no :-( | 11:47 |
| hakanardo | ok, I'll run some of the tests localy then | 11:47 |
| Wraithan | arigato: https://gist.github.com/1460132#comments yeah, pypy just needed time to get warmed up | 11:47 |
| Wraithan | heh | 11:47 |
| arigato | hakanardo: I think we need to adjust our policy | 11:48 |
| hakanardo | policy about commits? | 11:48 |
| arigato | e.g. "please do *all* changes to micronumpy in branches, and merge to default only when it's known to work" | 11:48 |
| arigato | yes | 11:48 |
| hakanardo | maybe that policy should not be limited to micronumpy? | 11:49 |
| arigato | well, it's a general question, but I think it's asking too much in general | 11:49 |
| arigato | it's just that micronumpy is actively developed and has a number of people not too familiar with rpython in i | 11:50 |
| arigato | in it | 11:50 |
| arigato | I would not like to make branches and test everything for every single small random change I do, for example :-) | 11:51 |
| hakanardo | :) | 11:51 |
| hakanardo | so how about a default-translating branch | 11:52 |
| hakanardo | that gets default merge in automaticlya every time the nightly run successfully translates | 11:52 |
| hakanardo | or I could just start merging latets translateing version instead of merging default :) | 11:53 |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) joined #pypy. | 11:58 | |
| arigato | :-) | 11:59 |
| kenaan | 12arigo SpecialisedTuples 11f07d5e9e9d7a 15/pypy/jit/tl/pypyjit_demo.py: Update the demo. | 11:59 |
| arigato | ok, got the crash on default too... | 11:59 |
| kenaan | 12arigo SpecialisedTuples 11b70295020ce8 15/: hg merge default | 11:59 |
| kenaan | 12arigo default 110808b0899c34 15/: Merge the SpecialisedTuples branch, started by Mark W. P. Add a number of interp-level classes for 'tuple', specia... | 11:59 |
| Wraithan | ~.p | 12:07 |
| Wraithan | oops | 12:07 |
| kenaan | 12arigo default 111c29f19cb8d7 15/pypy/module/micronumpy/: hg backout 3eba2ed546ad Does not translate. The issue is that to_builtin_type() is not RPython: the result of sel... | 12:10 |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy. | 12:15 | |
| rokujyouhitoma (~rokujyouh@12.51.221.130) left irc: Remote host closed the connection | 12:19 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 240 seconds | 12:24 | |
| kenaan | 12arigo default 11f0e6f9c06870 15/pypy/objspace/std/listobject.py: Hopefully fix test_pypy_c:test_call:test_stararg_virtual, by forcing getitems_fixedsize() to unroll small constant-... | 12:26 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1155 [12arigo] | 12:26 |
| arigato (~arigo@82.113.99.33) left irc: Quit: See you | 12:27 | |
| cheater | hi again | 12:34 |
| cheater | you guys said that you make xml.etree import a python version and not a C version of some sort... looking through the etree source, i notice it uses expat. In CPython, it is an .so. However in PyPy when I import the module and print it out it says "<module 'pyexpat' (built-in)>". Does this mean the pyexpat is somehow integrated into pypy? I have also noticed lib_pypy/pyexpat.py - is that the source for the expat that pypy is using?.. | 12:36 |
| cheater | the thing is, i have put a "raise Exception" at the top of lib_pypy/pyexpat.py and i can still import pyexpat from pypy. | 12:38 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 12:40 | |
| cheater | i guess pyexpat.py is just some sort of python-level interface that pypy uses | 12:40 |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy. | 12:52 | |
| stakkars | hakanardo: sounds like a good idea (default-translating) | 12:52 |
| stakkars | on the other hand, what is the default branch then good for? | 12:53 |
| nedbat (~nedbat@python/psf/nedbat) joined #pypy. | 12:53 | |
| amaury_ | cheater: pyexpat is indeed a builtin module in pypy | 12:55 |
| cheater | amaury_: do you have any experience using expat? | 12:55 |
| cheater | amaury_: i'm trying to fix a bug in etree, and i'm not sure if it's in etree itself or in expat | 12:56 |
| amaury_ | pypy uses the same Elementtree as CPython | 12:58 |
| gutworth (~benjamin@128.36.84.91) joined #pypy. | 13:00 | |
| stakkars | amaury_: do you too experience weird "access denied" errors on many tests? | 13:01 |
| stakkars | I even get the effect on trwo builds after another: | 13:02 |
| amaury_ | uh? with which kind of instructions? | 13:02 |
| amaury_ | ah os.remove(= | 13:02 |
| amaury_ | )? | 13:02 |
| stakkars | the first build goes well, but the second crashes with some "access denied" | 13:02 |
| stakkars | when trying a third time, it works again. | 13:02 |
| stakkars | could it be that there is a side effect of tortoisehg, maybe? | 13:03 |
| amaury_ | a file not closed, or a DLL still loaded? | 13:03 |
| amaury_ | ah, possible as well | 13:04 |
| aleksi (~aleksi@109-184-60-236.dynamic.mts-nn.ru) joined #pypy. | 13:04 | |
| stakkars | yes, the complaints look like something is not done, or to late. | 13:04 |
| stakkars | s/to/too/ | 13:04 |
| amaury_ | I know there was a way to exclude some directories | 13:04 |
| amaury_ | at least with tortoise SVN | 13:05 |
| Action: amaury_ away | 13:06 | |
| _aleksi (~aleksi@109-184-115-163.dynamic.mts-nn.ru) joined #pypy. | 13:08 | |
| fijal (~fijal@197.109.254.70) joined #pypy. | 13:08 | |
| aleksi (~aleksi@109-184-60-236.dynamic.mts-nn.ru) left irc: Ping timeout: 240 seconds | 13:10 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 244 seconds | 13:11 | |
| aleksi (~aleksi@95-37-53-115.dynamic.mts-nn.ru) joined #pypy. | 13:13 | |
| fijal (~fijal@197.109.254.70) left irc: Ping timeout: 255 seconds | 13:13 | |
| _aleksi (~aleksi@109-184-115-163.dynamic.mts-nn.ru) left irc: Ping timeout: 252 seconds | 13:13 | |
| derdon (~derdon@pD9E1DE68.dip.t-dialin.net) joined #pypy. | 13:16 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds | 13:17 | |
| jterrace (~Adium@dynamic-oit-vapornet-b-25.Princeton.EDU) joined #pypy. | 13:22 | |
| _aleksi (~aleksi@95-37-43-151.dynamic.mts-nn.ru) joined #pypy. | 13:24 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1155 [12arigo] | 13:26 |
| aleksi (~aleksi@95-37-53-115.dynamic.mts-nn.ru) left irc: Ping timeout: 248 seconds | 13:27 | |
| cheater | hi again | 13:29 |
| arigato (~arigo@fwstups.cs.uni-duesseldorf.de) joined #pypy. | 13:29 | |
| cheater | i think i have fixed etree's xpath handling, but i'm not sure how to execute the tests. can someone help me figure that part out? | 13:29 |
| khs (~khs@151.43.202.84.customer.cdi.no) joined #pypy. | 13:32 | |
| cheater | i have tried running "python /usr/lib/python2.6/python2.6-2.6.5/Lib/test/test_xml_etree.py" but it's complaining about failed imports.. | 13:33 |
| lotrpy (~lotrpy@202.120.58.116) left irc: Ping timeout: 252 seconds | 13:36 | |
| derdon (~derdon@pD9E1DE68.dip.t-dialin.net) left irc: Remote host closed the connection | 13:39 | |
| cheater | ah, i think i got that stuff executing. | 13:40 |
| arigato | cheater: it looks like you want #python | 13:40 |
| lotrpy (~lotrpy@202.120.58.116) joined #pypy. | 13:45 | |
| kenaan | 12arigo default 1120cf388f70b8 15/pypy/objspace/fake/objspace.py: Wrap(r_singlefloat) is not supported by the std obj space either. | 13:45 |
| kenaan | 12arigo default 11b56d8ea76b88 15/pypy/module/_continuation/test/test_translated.py: fix | 13:45 |
| kenaan | 12arigo SpecialisedTuples 11a48a35a5c618 15/: close branch | 13:47 |
| kenaan | 12hakanardo jit-targets 114c3b1c0a071a 15/pypy/jit/metainterp/: a bit more cloning to get a short_preamble for the fallback that does not contain ops from the failed optim... | 13:56 |
| kenaan | 12hakanardo jit-targets 11988827261d3c 15/pypy/: hg merge default | 13:56 |
| kenaan | 12hakanardo jit-targets 11853be60223ac 15/pypy/: hg merge default | 13:56 |
| harrison (~quassel@adsl-69-209-215-176.dsl.chcgil.sbcglobal.net) joined #pypy. | 13:56 | |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1156 [12hakanardo, jit-targets] | 13:58 |
| tito (~tito@pdpc/supporter/bronze/tito) left irc: Ping timeout: 244 seconds | 14:02 | |
| overmind_ (~overminde@143.89.167.3) joined #pypy. | 14:06 | |
| harrison (~quassel@adsl-69-209-215-176.dsl.chcgil.sbcglobal.net) left irc: Ping timeout: 240 seconds | 14:07 | |
| tito (~tito@pdpc/supporter/bronze/tito) joined #pypy. | 14:09 | |
| overminder (~overminde@143.89.167.3) left irc: Ping timeout: 244 seconds | 14:09 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: Operation timed out | 14:11 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy. | 14:13 | |
| harrison (~quassel@69.209.222.34) joined #pypy. | 14:15 | |
| overmind_ (~overminde@143.89.167.3) left irc: Ping timeout: 248 seconds | 14:16 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Ping timeout: 240 seconds | 14:18 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy. | 14:18 | |
| jimbaker (~jbaker@canonical/jimbaker) left irc: Ping timeout: 255 seconds | 14:18 | |
| tito (~tito@pdpc/supporter/bronze/tito) left irc: Remote host closed the connection | 14:31 | |
| fijal (~fijal@197.106.125.102) joined #pypy. | 14:39 | |
| fijal (~fijal@197.106.125.102) left irc: Read error: Connection reset by peer | 14:43 | |
| overmind_ (~overminde@wf099-182.ust.hk) joined #pypy. | 14:46 | |
| nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 258 seconds | 14:52 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1156 [12hakanardo, jit-targets] | 14:56 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 14:56 | |
| jterrace | Alex_Gaynor: the 2nd of 3 of the broken pieces of the original pull request: https://bitbucket.org/pypy/pypy/pull-request/18 | 14:58 |
| arigato | jterrace: sorry, I had to revert 3eba2ed546ad | 15:01 |
| arigato | see 1c29f19cb8d7 | 15:01 |
| arigato | I think it would be better if this was done on a branch (and then you can checkin directly into the same repository, too) | 15:02 |
| jterrace | oh hmm | 15:02 |
| arigato | well, I'll leave Alex_Gaynor or fijal to decide | 15:02 |
| jterrace | that wasnt in my pull request | 15:06 |
| jterrace | i think alex must have changed it | 15:06 |
| jterrace | i had them in each type | 15:06 |
| jterrace | so for bool, it was tolist = interp2app(W_BoolBox.descr_nonzero), | 15:06 |
| lmatos | just popping in to say I love pypy <3 Have a happy weekend. | 15:07 |
| jterrace | arigato: i think if it went back to the way i did it, it should work since i didnt use to_builtin_type | 15:07 |
| _aleksi (~aleksi@95-37-43-151.dynamic.mts-nn.ru) left irc: Remote host closed the connection | 15:13 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Ping timeout: 240 seconds | 15:18 | |
| Guest48507 (~timo@gallifrey.wybt.net) joined #pypy. | 15:20 | |
| Nick change: Guest48507 -> timotimo | 15:20 | |
| stakkars_ (~tismer@82.113.121.83) joined #pypy. | 15:33 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) joined #pypy. | 15:37 | |
| ojii (~ojii@172.114.63.81.cust.bluewin.ch) joined #pypy. | 15:44 | |
| stakkars_ | arigato: hi! | 15:53 |
| arigato | hi | 15:53 |
| stakkars_ | I was just reading a comment from you to Rotung concerning warm-up time. | 15:53 |
| stakkars_ | right now I understand the tracing strategy as something that waits for n executions of a function before it is considered for the jit. | 15:55 |
| stakkars_ | is that a kind of binary decision, like "we compile it" or "we don't compile it"? | 15:56 |
| stakkars_ | or is it possible to do something gradually? | 15:56 |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Ping timeout: 240 seconds | 15:56 | |
| stakkars_ | I'm thinking of my latest strategy for Psyco, not really related, and unfortunately not implemented: | 15:58 |
| nedbat (~nedbat@python/psf/nedbat) joined #pypy. | 15:58 | |
| stakkars_ | the approach was to compile everything without generating myriads of code by first | 15:58 |
| stakkars_ | compiling a function at level 1, that means with not inlining and little specializing. | 15:59 |
| stakkars_ | after code becomes eligible (executed often enough), it becomes eligible for more specialization | 16:00 |
| stakkars_ | and the next level of inlining. And so on. | 16:01 |
| stakkars_ | I would like to study its effects, still, to know how efficient that is. | 16:01 |
| stakkars_ | is that somehow applicable to PyPy, like a gear shift, that improves the code after it is eligible | 16:02 |
| stakkars_ | to do more optimisations, inlining at cetera, when it is used more often, and stepping back, otherwise? | 16:03 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 16:03 | |
| stakkars_ | moreover, while we don't make compilation persistent yet, can we maybe save the outcome of | 16:04 |
| stakkars_ | the heuristic decision about jitting make persistent, and become self-learning over sessions? | 16:05 |
| stakkars_ | -end of brain dum(p|b) | 16:05 |
| antocuni (~antocuni@host117-71-dynamic.58-82-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds | 16:07 | |
| arigato | the current approach is quite different, I fear | 16:08 |
| arigato | what you describe may work for Psyco | 16:08 |
| arigato | pypy has a tracing jit that focuses on compiling loops | 16:09 |
| kenaan | 12arigo jit-simplify-backendintf 1111ca1186b82f 15/: hg merge jit-targets | 16:09 |
| stakkars_ | yes I know. But maybe instead of long warm-up phase and then total compilation of something, maybe there | 16:10 |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy. | 16:10 | |
| stakkars_ | are cheaper possibilities that could be applied much earlier, befor it gets to the next stage? | 16:10 |
| arigato | I see | 16:11 |
| stakkars_ | something that is always applied as a quich start with little speedup, maybe about 1.5 to two? | 16:11 |
| arigato | yes, maybe it makes sense too | 16:12 |
| stakkars_ | without lots of computations and little specialization, maybe we can have a large set of | 16:13 |
| stakkars_ | pre-computed chean macro sets which produce fair code in common cases. | 16:14 |
| arigato | some Java experience is that you win by using a method jit first, and a tracing jit as the next stage, but that's a mess because it requires two backends, basically | 16:14 |
| stakkars_ | s/chean/cheap/ | 16:14 |
| tito (~tito@pdpc/supporter/bronze/tito) joined #pypy. | 16:14 | |
| stakkars_ | method jit is one like psyco? | 16:15 |
| arigato | "method jit" is the standard java jit, like hotspot | 16:15 |
| stakkars_ | maybe that fills a gap. Hmm, dunno how they work. | 16:16 |
| arigato | psyco is its own category, but indeed, it's more method-jit-like by many aspects | 16:16 |
| arigato | the point is that the pypy backends are not usable for method jits | 16:16 |
| arigato | as they are focused on compiling linear traces | 16:17 |
| stakkars_ | ha! a simple first step could do the simple thing!!! | 16:17 |
| stakkars_ | turn the interpreted code into compiled code. Like unladen-swallow, or all | 16:18 |
| linq (~ident@24-246-25-39.cable.teksavvy.com) joined #pypy. | 16:18 | |
| davidpk (~davidpk@5e0f1559.bb.sky.com) joined #pypy. | 16:18 | |
| stakkars_ | the myriads of other things that claim to be fast, just by removing the interpreter loop. | 16:18 |
| arigato | well yes, but precisely, it's a mess because we don't have a backend for producing this kind of machine code | 16:18 |
| stakkars_ | That almost always gives a 1.7-2.0 speedup. | 16:18 |
| arigato | maybe it's possible to also have a tracing-jit-like approach to this problem, but it needs to be investigated | 16:19 |
| davidpk | can anyone point me in the direction of a well-commented piece of RPython code making extensive use of the FFI? like, a C library binding or something? | 16:20 |
| stakkars_ | just saying that I'm very much interested in this, given that I ever get out of win64 woes ;-) | 16:20 |
| arigato | davidpk: which FFI? pypy.rpython.lltypesystem.rffi? | 16:21 |
| davidpk | arigato: yes, that one (i assumed there was only one, since that's the only one that shows up in the documentation) | 16:22 |
| arigato | there are various at various levels :-) | 16:22 |
| arigato | davidpk: any preference for the library? e.g. if you know well sockets, there is rlib/_rsocket_rffi.py and rsocket.py | 16:23 |
| arigato | rlib/rzlib.py is simpler, if you know zlib | 16:24 |
| fox__ (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 16:24 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Read error: Connection reset by peer | 16:24 | |
| arigato | rmmap.py is a bit messy but should work too | 16:25 |
| davidpk | arigato: i'll take a look. (for context: i want to bind to the Oniguruma API, http://dpk.org.uk/X -- yes i know about the re module but this code has to work with legacy regexps written in Ruby style with imx modifiers) | 16:25 |
| arigato | note that the usual comment applies: consider just using ctypes in normal Python, rather than writing something at RPython | 16:26 |
| stakkars_ | arigato: an old idea (sorry, some rare activity in my stuttering b-b-brain) | 16:27 |
| davidpk | ah, but i'm writing an interpreter in RPython, so that would be too slow, no? | 16:27 |
| arigato | davidpk: ah, then sure, you don't have normal Python at all | 16:27 |
| stakkars_ | remember the genc approach from Bill Tutt, around 1999? | 16:27 |
| arigato | no? | 16:28 |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) left irc: Quit: Computer has gone to sleep. | 16:28 | |
| stakkars_ | the python code for any function was turned into a compiled C code in one step and then executed. Gave a factor of almost two, just my removing the interpreter loop. | 16:29 |
| stakkars_ | ot maybe also only 1.4, but still. | 16:29 |
| arigato | ah sure, sorry, I didn't know that Bill Tutt tried it (probably) first | 16:29 |
| arigato | it's the basic of Pyrex and Cython | 16:29 |
| stakkars_ | yes, and I discussed it a bit and worked with it. Somehow the start why I later initiated pypy things. | 16:30 |
| stakkars_ | yes. Would it be cheap to apply this always? | 16:31 |
| stakkars_ | generated that initial suboptimal code instead of interpreting at all? | 16:31 |
| arigato | yes, probably, but I don't have a clear idea about how to generate this without large amounts of hints in the pypy source code | 16:32 |
| stakkars_ | then PyPy would be faster in the first place, and _then_ it can work on factor five and more | 16:32 |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) joined #pypy. | 16:32 | |
| arigato | ...thinking out loud, I suppose it's possible, but requires some serious static analysis again, to determine automatically the boundary between the code that we compile and the residual code that we just write CALLs to | 16:34 |
| arigato | it's also a bit unclear if it can really work: | 16:35 |
| arigato | remember that pypy does not change anything about the semantic of the language | 16:35 |
| arigato | so the risk is that the machine code that we generate would be much more complicated than Pyrex/Cython | 16:35 |
| arigato | e.g. because we cannot possibly call "space.add()" without first making sure that the frame contains the correct current values | 16:36 |
| stakkars_ | yes, I do. | 16:36 |
| stakkars_ | hmmmm, es | 16:36 |
| arigato | so it's unclear that we would win 1.7x-2x | 16:36 |
| stakkars_ | yes | 16:36 |
| arigato | or else we use the same advanced techniques to reconstruct the frame if needed, but I fear that this has a big machine-code-generation-time cost | 16:37 |
| arigato | which in this case is kind of missing the point | 16:38 |
| jterrace (~Adium@dynamic-oit-vapornet-b-25.Princeton.EDU) left irc: Quit: Leaving. | 16:38 | |
| stakkars_ | essentially, I was probably thinking of geninterp for machine code generation. | 16:38 |
| stakkars_ | but yes, our interpreter in its whole generality would requiter a lot of steps to start with, and it is not cheap. | 16:39 |
| arigato | I was (of course) thinking that we would not write the JIT manually, but generate it from the interpreter source code --- otherwise there is little point | 16:40 |
| arigato | there is even no point at all, because I cannot fix a line in the interpreter without having to worry about the first-level JIT | 16:40 |
| stakkars_ | no no, I want to do it automatically, of course. | 16:41 |
| arigato | ok :-) | 16:42 |
| stakkars_ | ok, (I cannot stop), now thinking of geninterp again. | 16:43 |
| arigato | as a first very rough approximation: | 16:43 |
| arigato | edit module/pypyjit/policy.py | 16:43 |
| stakkars_ | Right now we are dealing with the app-level code, right? | 16:43 |
| arigato | to restrict the JIT to pypy/interpreter, and not e.g. pypy/objspace | 16:43 |
| arigato | and see how the result behaves | 16:43 |
| stakkars_ | what if we geninterp the code? Then we get interp-level code, which is equivalent but not interpreting. | 16:44 |
| arigato | it is not equivalent at all, only a rough approximation | 16:44 |
| stakkars_ | arigato: ah, oh-kay, yes that would be kind of first level. | 16:44 |
| fox__ (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Read error: Connection reset by peer | 16:44 | |
| stakkars_ | ok, needs more thought, but in fact, the idea to have something that can be done veeeeery quickly, with only a bit of speedup. | 16:46 |
| arigato | yes | 16:46 |
| arigato | fwiw, I've been thinking of the opposite end of the spectrum: | 16:47 |
| arigato | after running for a while machine code that we produced, | 16:47 |
| arigato | it may be worth collecting a whole bunch of them and send them to gcc or to llvm | 16:47 |
| stakkars_ | ok, in order to make that very fast, I'm thinking of a precompiled set of code pieces that can be easily parameterized and merged | 16:47 |
| stakkars_ | together to build some fair but simple code. | 16:48 |
| stakkars_ | gcc/llvm: absolutely, and no new idea, have discussed this long time ago. | 16:48 |
| stakkars_ | and the god thing: while we don't have STM, we can have a second thread which works on this with an unused core | 16:50 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 16:50 | |
| Action: stakkars_ has to stop, dinner with my son | 16:51 | |
| arigato | :-) | 16:51 |
| arigato | happy dinner | 16:51 |
| stakkars_ | son = autonomous replication unit, one of the few finished projects of mine :-) | 16:51 |
| stakkars_ | thanks | 16:51 |
| stakkars_ (~tismer@82.113.121.83) left irc: Quit: schnarch | 16:52 | |
| arigato | :-) | 16:56 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Ping timeout: 255 seconds | 17:08 | |
| etrepum (~bob@69-9-57-57.static.logl001.digis.net) left irc: Quit: etrepum | 17:08 | |
| overmind_ (~overminde@wf099-182.ust.hk) left irc: Ping timeout: 240 seconds | 17:09 | |
| kenaan | 12arigo default 11b2c92d4e8383 15/pypy/doc/faq.rst: Added a FAQ entry. | 17:09 |
| arigato (~arigo@fwstups.cs.uni-duesseldorf.de) left irc: Quit: See you | 17:10 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 252 seconds | 17:19 | |
| fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) joined #pypy. | 17:20 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 17:28 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy. | 17:34 | |
| mat^2 (~mathias@212.130.113.35) left irc: Read error: Connection reset by peer | 17:44 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Read error: Connection reset by peer | 17:44 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 17:44 | |
| mat^2 (~mathias@212.130.113.35) joined #pypy. | 17:44 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Ping timeout: 240 seconds | 17:49 | |
| RiaanFV (~riaanfv@41-132-169-26.dsl.mweb.co.za) joined #pypy. | 17:50 | |
| kkris (~kris@188-22-115-53.adsl.highway.telekom.at) joined #pypy. | 17:53 | |
| kkris1 (~kris@93-82-34-223.adsl.highway.telekom.at) left irc: Ping timeout: 252 seconds | 17:54 | |
| kenaan | 12alex_gaynor default 11cf00150bde7c 15/pypy/module/micronumpy/: Recommit the tolist() for numpy stuff, now that it translates. | 18:07 |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy. | 18:13 | |
| nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 258 seconds | 18:25 | |
| RiaanFV (~riaanfv@41-132-169-26.dsl.mweb.co.za) left irc: Ping timeout: 252 seconds | 18:37 | |
| DasIch_ (~dasich@p4FFDF269.dip.t-dialin.net) joined #pypy. | 18:37 | |
| DasIch (~dasich@p4FFDFC5D.dip.t-dialin.net) left irc: Ping timeout: 252 seconds | 18:41 | |
| khs (~khs@151.43.202.84.customer.cdi.no) left irc: Quit: Leaving | 19:01 | |
| mfoord (~anonymous@87-194-212-65.bethere.co.uk) joined #pypy. | 19:12 | |
| mfoord (~anonymous@87-194-212-65.bethere.co.uk) left irc: Changing host | 19:12 | |
| mfoord (~anonymous@python/psf/voidspace) joined #pypy. | 19:12 | |
| lhk (~lhk@p5B393913.dip.t-dialin.net) joined #pypy. | 19:26 | |
| lhk | hi | 19:26 |
| lhk | if I understood correctly the website says that pypy strives to implement python on top of other frameworks | 19:27 |
| lhk | sorry if I got something wrong but does this try to be an alternative to jython and ironpython | 19:27 |
| lhk | ?? | 19:27 |
| lucian | lhk: CPython first, jython/ironpython second | 19:28 |
| lhk | so its on the roadmap ? | 19:28 |
| lucian | i don't even know if the jvm/clr backends work anymore | 19:28 |
| lucian | i'm not a pypy dev, so don't take my word for it | 19:29 |
| Rhyolite | lhk: what do you think "on top of other frameworks" means? | 19:29 |
| lhk | I don't understood it therefore I asked | 19:29 |
| lhk | *didn't understand | 19:29 |
| Rhyolite | PyPy has its own JIT | 19:29 |
| Rhyolite | It strives to be compatible with CPython | 19:29 |
| lhk | so its faster cpython | 19:30 |
| Rhyolite | currently 2.7 and work underway for CPython 3.x | 19:30 |
| lucian | lhk: yeah, mostly | 19:30 |
| Rhyolite | sort of | 19:30 |
| Rhyolite | but Python is a language | 19:30 |
| lucian | lhk: but it has jv, and clr backends too | 19:30 |
| Rhyolite | CPython simply is one implementation | 19:30 |
| lucian | they aren't as complete | 19:30 |
| Rhyolite | PyPy currently has its own JIT for IA-32 and x86-64 | 19:30 |
| Rhyolite | with work on ARM and PPC | 19:31 |
| lucian | (and ARM, incomplete) | 19:31 |
| Rhyolite | there also have been experiments with LLVM, JVM, CLI | 19:31 |
| lucian | the llvm one has bitrotten long ago, afaik | 19:31 |
| lhk | so it takes python source code and compiles it to some kind of bytecode | 19:31 |
| Rhyolite | lucian: I am trying to be complete | 19:31 |
| lucian | lhk: not really, no | 19:31 |
| lhk | JVM CLI ? | 19:32 |
| Rhyolite | no | 19:32 |
| Rhyolite | it can translate it's own internal representation to connect with KVM and CLI | 19:32 |
| lhk | the JVM uses bytecode | 19:32 |
| Rhyolite | JVM and CLI | 19:32 |
| lucian | lhk: the PyPy translation toolchain takes an interpreter written in a subset of Python (called RPython) and generates a Virtual Machine out of it, including a GC and a JIT | 19:32 |
| lucian | lhk: pypy also has a python interpreter using the aforementioned translation toolchain | 19:33 |
| lhk | ok I think I got it | 19:33 |
| lucian | the translation toolchain has several backends, with the C/native one being worked on the most | 19:33 |
| lucian | so the pypy-c vm that results can run the full python language, but faster | 19:34 |
| lucian | since it has a JIT, automatically generated | 19:34 |
| lhk | you can use pypy as a "pipeline" with either its own RPython interpreter at the end or JVM and CLR | 19:34 |
| lhk | JIT makes the RPython interpreter fast | 19:34 |
| Rhyolite | lhk: sort of | 19:34 |
| Rhyolite | but you still are coming with too many preconceptions | 19:34 |
| Alex_Gaynor | basically :) | 19:34 |
| derdon (~derdon@p5DE89B26.dip.t-dialin.net) joined #pypy. | 19:34 | |
| lucian | lhk: even on jvm/clr, a VM runs on top | 19:35 |
| lhk | of course | 19:36 |
| Rhyolite | lhk: if "pipeline" means a static pipeline translating Python to different interpreters, then it does more than that | 19:36 |
| lhk | hm I thought so | 19:36 |
| lhk | what's it doing then | 19:36 |
| ronny | translation toolchain does whole programm transformations | 19:36 |
| lhk | let's assume you would advertise pypy | 19:37 |
| lhk | what would you say to your future customer | 19:37 |
| Rhyolite | ronny: that's just going to confuse lhk | 19:37 |
| ronny | k | 19:37 |
| Rhyolite | he's going to assume it means translating the user application | 19:37 |
| lhk | no I don't | 19:38 |
| Alex_Gaynor | I'd say that it's a Python VM that runs faster. | 19:38 |
| lhk | up to now I'd assume it would restructure the layout of the data fed to the vm | 19:38 |
| lhk | to make interpretation faster | 19:38 |
| lhk | though that is actually some kind of translation | 19:38 |
| Rhyolite | lhk: PyPy is a JIT | 19:38 |
| lhk | okay | 19:38 |
| Rhyolite | it interprets the user application | 19:38 |
| Rhyolite | and converts critical parts | 19:39 |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Ping timeout: 252 seconds | 19:39 | |
| Rhyolite | to machine code | 19:39 |
| Rhyolite | as one possible target | 19:39 |
| Rhyolite | but it also can run the interpreter on other backends, such as JVM and CLI | 19:39 |
| Rhyolite | utilizing JITs for those formats | 19:39 |
| ronny | lhk: what is your actual intent, so we can throw some tailored chunks fo information at you instead of the "random" data we give you now | 19:39 |
| kenaan | 12alex_gaynor extradoc 11dd2f021e2aa1 15/talk/iwtc11/licm.pdf: Put a compiled version of this paper in. | 19:40 |
| lhk | I'm currently working with python for my university course and was just interested in the python "world" | 19:40 |
| lhk | pypy seemed quite interesting | 19:40 |
| lhk | but I didn't know what to make of the website | 19:40 |
| lhk | I understood that it offers a new interpreter and some optimization | 19:41 |
| lhk | but the statement that it would somehow implement interpreters on other backends confused me | 19:41 |
| Rhyolite | lhk: Just think of PyPy as Python, compatible with CPython, but faster | 19:41 |
| ronny | lhk: then its probably best if you consider it as python interpreter that offers a just in time compiler | 19:42 |
| lhk | I worked with .Net as well as with the JVM so backends for these are very appealing to me | 19:42 |
| ronny | lhk: the whole details of the translation toolchain are basically not relevant unless you want to toy/learn/experiment with exactly that | 19:42 |
| lhk | ok | 19:43 |
| lhk | one more thing: | 19:43 |
| lhk | I'm still unsure about support for JVM and CLR | 19:43 |
| lhk | does it work or is it obsolete or is it on the roadmap | 19:44 |
| amaury_ | nobody is working on it | 19:44 |
| ronny | afair its currently incomplete but basically working | 19:45 |
| nedbat (~nedbat@python/psf/nedbat) joined #pypy. | 19:45 | |
| lhk | do you have interest in further development for these platforms ? | 19:45 |
| lhk | please don't misunderstand me: | 19:45 |
| lhk | a faster CPython is very impressive | 19:45 |
| lhk | I really appreciate your work | 19:46 |
| lhk | still CLR and JVM would be important to me | 19:46 |
| rguillebert | the JIT isn't working anymore on the JVM but I think the interpreter still works | 19:46 |
| Nick change: Gulaway -> Gulopine | 19:46 | |
| MostAwesomeDude | Well, don't JVM and CLR have their own JITs? | 19:46 |
| rguillebert | yes but python code isn't jitted if you only build an interpreter | 19:47 |
| Rhyolite | MostAwesomeDude: PyPy's JIT does more optimization | 19:47 |
| Rhyolite | there really were two projects | 19:47 |
| Rhyolite | the initial one to translate the interpreter to JVM and CLR | 19:47 |
| rguillebert | PyPy's JVM JIT turns python code to java bytecode | 19:47 |
| Rhyolite | and a later one to dynamically generate CLI | 19:47 |
| MostAwesomeDude | Ah. | 19:47 |
| Rhyolite | which does incur a double JIT overhead | 19:47 |
| Rhyolite | there was a proposal to dynamically generate JVM | 19:47 |
| rguillebert | which can then be jitted by the jvm | 19:47 |
| Rhyolite | but that has not progressed | 19:48 |
| amaury_ | lhk: I think that none of the core developers has plans for CLR or JVM | 19:48 |
| amaury_ | but contributions are welcome | 19:48 |
| lhk | I don't have the knowledge and time | 19:48 |
| amaury_ | and we can also help | 19:48 |
| Rhyolite | lhk: it probably would be more interesting to write a JVM interpreter on CPython, if one wanted to mix Java and Python | 19:48 |
| rguillebert | also, I don't think you can interact with Java code | 19:49 |
| Rhyolite | sorry, I mean JVM interpreter on RPython | 19:49 |
| rguillebert | Rhyolite, but the two object models probably don't fit | 19:49 |
| rguillebert | also you have to handle JNI... | 19:50 |
| Rhyolite | rguillebert: see the discussion on the mailinglist about mixing Python 2.x and 3.x | 19:50 |
| rguillebert | right, but there's still JNI :) | 19:50 |
| lhk | thanks for the explanation I got to go | 19:51 |
| lhk (~lhk@p5B393913.dip.t-dialin.net) left irc: Quit: Verlassend | 19:51 | |
| rguillebert | however, embedding hotspot might work | 19:55 |
| lucian | rguillebert: that sounds suitably insane :)( | 19:59 |
| Rhyolite | lucian: fijal is embedding cpython within PyPy | 20:01 |
| Rhyolite | so why not? | 20:01 |
| lucian | Rhyolite: it just sounds insane, it does totally work | 20:02 |
| lucian | iirc, that's how the python lucene bindings works | 20:02 |
| ronny | it "just" needs a rpc protocol thats fit | 20:17 |
| rguillebert | calling Java code from python doesn't look hard but implementing Java interfaces in python... | 20:22 |
| mwhudson (~mwh@linaro/mwhudson) joined #pypy. | 20:22 | |
| lucian | rguillebert: jython is the only option, really | 20:23 |
| gsnedders | Jython works fairly well for that sort of stuff. | 20:24 |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy. | 20:26 | |
| rguillebert | except that it's not super fast like PyPy :) | 20:27 |
| gsnedders | Depends what you're doing. :) | 20:27 |
| arigato (~arigo@82.113.106.187) joined #pypy. | 20:28 | |
| kenaan | 12arigo default 11a83bd631dde3 15/pypy/jit/backend/llgraph/runner.py: fix comment | 20:33 |
| kenaan | 12arigo jit-simplify-backendintf 11ab500e92817c 15/pypy/jit/metainterp/: Fixed virtualizables. | 20:33 |
| kenaan | 12arigo jit-simplify-backendintf 11f9e61788fa85 15/pypy/jit/metainterp/: Fix for compile_tmp_callback(). | 20:33 |
| kenaan | 12arigo jit-simplify-backendintf 11c87a6c079a01 15/pypy/jit/backend/test/runner_test.py: Fix test. | 20:34 |
| kenaan | 12arigo jit-simplify-backendintf 116ea88ccf9778 15/pypy/jit/metainterp/test/test_tl.py: Fix test. | 20:34 |
| kenaan | 12arigo jit-simplify-backendintf 11874a608a3377 15/pypy/jit/: Translation fixes (front-end only so far). | 20:34 |
| kenaan | 12arigo jit-simplify-backendintf 113481aa4226dc 15/pypy/jit/metainterp/test/test_warmspot.py: Fix test. | 20:34 |
| kenaan | 12arigo default 116fb87770b5d2 15/pypy/module/micronumpy/: merge heads | 20:34 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 20:37 | |
| asmeurer__ (~asmeurer@dhcp-baca-230.resnet.nmt.edu) joined #pypy. | 20:47 | |
| Ademan (~dan@adsl-71-141-224-79.dsl.snfc21.pacbell.net) joined #pypy. | 20:53 | |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/746 [12hakanardo, jit-targets] | 21:01 |
| Nick change: Gulopine -> Gulaway | 21:17 | |
| kenaan | 12arigo jit-simplify-backendintf 1182684fd1231e 15/pypy/jit/: Switch to a less magical way of specializing execute_token(), one which has a chance to work in a ... | 21:18 |
| kenaan | 12arigo jit-simplify-backendintf 11c1bad6dee4e2 15/pypy/jit/backend/model.py: fix | 21:18 |
| kenaan | 12arigo jit-simplify-backendintf 112bc80fdea1c1 15/pypy/jit/: In-progress | 21:18 |
| kenaan | 12arigo jit-simplify-backendintf 119677e2ec7aef 15/pypy/jit/backend/x86/: Kill. | 21:18 |
| int3__ (~int3__@st401-246.subnet-244.amherst.edu) joined #pypy. | 21:20 | |
| aboudreault (~alanb@osgeo/member/aboudreault) joined #pypy. | 21:20 | |
| Action: arigato goes on a killing mission | 21:21 | |
| Alex_Gaynor | arigato: do we need to redo carl's terminator photoshop for you? | 21:22 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Read error: Connection reset by peer | 21:22 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) joined #pypy. | 21:22 | |
| arigato | :-) | 21:23 |
| jterrace (~jterrace@pu116187.student.Princeton.EDU) joined #pypy. | 21:28 | |
| jterrace | Alex_Gaynor: ping | 21:29 |
| Alex_Gaynor | jterrace: pong | 21:29 |
| jterrace | i think you broked my tolist patch | 21:29 |
| jterrace | it got pulled back out | 21:29 |
| Alex_Gaynor | I reccommited it | 21:29 |
| jterrace | oh nice | 21:30 |
| jterrace | you just had to add the to_builtin_type functions? | 21:31 |
| Alex_Gaynor | what do you mean? | 21:31 |
| jterrace | hmm just trying to understand | 21:31 |
| jterrace | so before i had it as tolist = interp2app for int/bool/float | 21:31 |
| arigato | I backed it out because it didn't translate | 21:32 |
| arigato | I also fixed ztranslation to show why :-) | 21:32 |
| jterrace | but did the original way i had it translate? | 21:32 |
| arigato | I cannot tell | 21:32 |
| Alex_Gaynor | jterrace: yes probably, I wanted them all to go through the same function the way it does on CPython | 21:33 |
| jterrace | trying to understand why you changed it from having tolist in each type to genericbox - just to make it easier for future types? | 21:33 |
| jterrace | oh ok | 21:33 |
| Alex_Gaynor | e.g. on CPython numpy.generic.tolist(numpy.int32(2)) works | 21:33 |
| jterrace | oh interesting | 21:34 |
| jterrace | i get it now | 21:34 |
| jterrace | thanks | 21:35 |
| jterrace | did you get a chance to look at #18? | 21:35 |
| etrepum (~bob@50.8.114.2) joined #pypy. | 21:40 | |
| -Martinp23 (martinp23@freenode/staff/martinp23) to $$*- [Global Notice] Hi everyone. Tonight should be our final night of upgrades of ircd-seven for the time being. The servers affected today are card, asimov, and verne. There'll be large netsplits. If you're on a server which will restart, I'll send you a message in a moment. Enjoy the ride - duration about 30 mins! :) | 21:47 | |
| mikefc | Alex_Gaynor: posted a bug report for the numpy array stuff. seems to be an array assignment operation being jitted out of existence. | 21:49 |
| Action: mikefc preps for work | 21:49 | |
| Alex_Gaynor | I doubt it's that simple, but I'll take a look in a bit | 21:49 |
| norox (~norox@2a00:801::2d0:b7ff:fea0:c753) got netsplit. | 21:54 | |
| apollo13 (~apollo13@unaffiliated/apollo13) got netsplit. | 21:54 | |
| PiotrSikora (~none@nginx/adept/piotrsikora) got netsplit. | 21:54 | |
| magcius (magcius@unaffiliated/magcius) got netsplit. | 21:54 | |
| asmeurer__ (~asmeurer@dhcp-baca-230.resnet.nmt.edu) got netsplit. | 21:54 | |
| kkris (~kris@188-22-115-53.adsl.highway.telekom.at) got netsplit. | 21:54 | |
| mat^2 (~mathias@212.130.113.35) got netsplit. | 21:54 | |
| lotrpy (~lotrpy@202.120.58.116) got netsplit. | 21:54 | |
| apanda (~apanda@c-24-7-101-62.hsd1.ca.comcast.net) got netsplit. | 21:54 | |
| __name__ (~name@sburn/devel/name) got netsplit. | 21:54 | |
| jell (jellonek@212.182.18.33) got netsplit. | 21:54 | |
| arkx (arkx@hilla.kapsi.fi) got netsplit. | 21:54 | |
| henning_ (~henning@unicorn.gnubo.de) got netsplit. | 21:54 | |
| thoughtpolice (~thoughtpo@li111-206.members.linode.com) got netsplit. | 21:54 | |
| aboudreault (~alanb@osgeo/member/aboudreault) got netsplit. | 21:54 | |
| derdon (~derdon@p5DE89B26.dip.t-dialin.net) got netsplit. | 21:54 | |
| harrison (~quassel@69.209.222.34) got netsplit. | 21:54 | |
| davisagli (~davisagli@davisagli.com) got netsplit. | 21:54 | |
| alexsuraci (u1962@gateway/web/irccloud.com/x-fmgkunpvlhywhrbp) got netsplit. | 21:54 | |
| Circlefusion (~cf@74-138-199-109.dhcp.insightbb.com) got netsplit. | 21:54 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) got netsplit. | 21:54 | |
| Rhyolite (~dje@pool-108-6-25-71.nycmny.fios.verizon.net) got netsplit. | 21:54 | |
| jterrace (~jterrace@pu116187.student.Princeton.EDU) got netsplit. | 21:54 | |
| Hodgestar (~quassel@gluon.za.net) got netsplit. | 21:54 | |
| kenaan (~kenaan@66.35.39.65) got netsplit. | 21:54 | |
| dimazest (~dimazest@balticom-202-59.balticom.lv) got netsplit. | 21:54 | |
| ec (~me@ell.io) got netsplit. | 21:54 | |
| exarkun (~exarkun@wordeology.com) got netsplit. | 21:54 | |
| asksol (~asksol@sanedefaults.net) got netsplit. | 21:54 | |
| price (~price@LINERVA.MIT.EDU) got netsplit. | 21:54 | |
| panni_ (smuxi@descent.fragstore.net) got netsplit. | 21:54 | |
| Dulak (~michael@unaffiliated/dulak) got netsplit. | 21:54 | |
| hakanardo__ (~ardo@putten.maths.lth.se) got netsplit. | 21:54 | |
| Wraithan (~Wraithan@c-67-189-110-252.hsd1.or.comcast.net) got netsplit. | 21:54 | |
| dmarkey (~dmarkey@dmarkey.xen.prgmr.com) got netsplit. | 21:54 | |
| brutal_chaos (~chaos@francium.electricgrey.com) got netsplit. | 21:54 | |
| GDorn (~gdorn@li223-147.members.linode.com) got netsplit. | 21:54 | |
| ivan` (~ivan@unaffiliated/ivan/x-000001) got netsplit. | 21:54 | |
| carljm (~carljm@spinach.threepines.org) got netsplit. | 21:54 | |
| hazmat (~hazmat@plone/hazmat) got netsplit. | 21:54 | |
| Taggnostr (~quassel@dyn57-215.yok.fi) got netsplit. | 21:54 | |
| Unhelpful (~quassel@rockbox/developer/Unhelpful) got netsplit. | 21:54 | |
| gsnedders (~gsnedders@mail.gsnedders.com) got netsplit. | 21:54 | |
| guibou (~guibou@kraken.insecable.net) got netsplit. | 21:54 | |
| espes__ (~niel@205.185.120.132) got netsplit. | 21:54 | |
| Eventh (evenwiik@caracal.stud.ntnu.no) got netsplit. | 21:54 | |
| idnar (~quassel@unaffiliated/idnar) got netsplit. | 21:54 | |
| lahwran (lahwran@unaffiliated/lahwran) got netsplit. | 21:54 | |
| Ormod (~valtha@ohmu.fi) got netsplit. | 21:54 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) got netsplit. | 21:54 | |
| hpk (~hpk@hq2.merlinux.eu) got netsplit. | 21:54 | |
| getxsick (xsx@unaffiliated/getxsick) got netsplit. | 21:54 | |
| MostAwesomeDude (~simpson@osuosl/staff/mostawesomedude) got netsplit. | 21:54 | |
| masquerade (~rdeaton@galactica.lug.udel.edu) got netsplit. | 21:54 | |
| CIA-13 (~CIA@cia.atheme.org) got netsplit. | 21:54 | |
| __main__ (~main@adsl-99-173-15-158.dsl.pltn13.sbcglobal.net) got netsplit. | 21:54 | |
| Gulaway (~Gulopine@173-203-203-206.static.cloud-ips.com) got netsplit. | 21:54 | |
| JStoker (jstoker@unaffiliated/jstoker) got netsplit. | 21:54 | |
| Arnar (~arnarb@li187-148.members.linode.com) got netsplit. | 21:54 | |
| int3__ (~int3__@st401-246.subnet-244.amherst.edu) got netsplit. | 21:54 | |
| nedbat (~nedbat@python/psf/nedbat) got netsplit. | 21:54 | |
| durin42 (~durin@adium/durin42) got netsplit. | 21:54 | |
| pedronis (~pedronis@73-53.195-178.cust.bluewin.ch) got netsplit. | 21:54 | |
| kalenz (~kalenz@ping.kalenz.fr) got netsplit. | 21:54 | |
| thinker (~thinker@123-194-52-90.dynamic.kbronet.com.tw) got netsplit. | 21:54 | |
| xorAxAx (~alexander@moinmoin/coreteam/alexander) got netsplit. | 21:54 | |
| ThomasWaldmann (~twaldmann@moinmoin/coreteam/thomas) got netsplit. | 21:54 | |
| zk (~is@whatit.is) got netsplit. | 21:54 | |
| raydeo (~raydeo@merickel.org) got netsplit. | 21:54 | |
| Garen (garen@69.76.18.3) got netsplit. | 21:54 | |
| maxyz (~maxy@186.18.94.35) got netsplit. | 21:54 | |
| lvh (~lvh@python/psf/lvh) got netsplit. | 21:54 | |
| lmatos (lmatos@staff.linode.com) got netsplit. | 21:54 | |
| myf (~myf@li245-175.members.linode.com) got netsplit. | 21:54 | |
| linq (~ident@24-246-25-39.cable.teksavvy.com) got netsplit. | 21:54 | |
| timotimo (~timo@gallifrey.wybt.net) got netsplit. | 21:54 | |
| cheater (~cheater@ip-80-226-24-12.vodafone-net.de) got netsplit. | 21:54 | |
| tos9 (~tos9@unaffiliated/tos9) got netsplit. | 21:54 | |
| jerith (~jerith@chthon.slipgate.za.net) got netsplit. | 21:54 | |
| Husio (husiatyn@oceanic.wsisiz.edu.pl) got netsplit. | 21:54 | |
| tumbleweed (~stefanor@ubuntu/member/tumbleweed) got netsplit. | 21:54 | |
| etrepum (~bob@50.8.114.2) got netsplit. | 21:54 | |
| nopper (~nopper@ns.patched.biz) got netsplit. | 21:54 | |
| pjenvey (~pjenvey@underboss.org) got netsplit. | 21:54 | |
| ousado (~ousado@unaffiliated/ousado) got netsplit. | 21:54 | |
| DanielHolth (~dholth@ip98-180-34-112.ga.at.cox.net) got netsplit. | 21:54 | |
| elipsion_ (~elipsion@81-232-141-35-no217.tbcn.telia.com) got netsplit. | 21:54 | |
| tito (~tito@pdpc/supporter/bronze/tito) got netsplit. | 21:54 | |
| papercrane (~papercran@c-76-103-172-115.hsd1.ca.comcast.net) got netsplit. | 21:54 | |
| oal (u4126@gateway/web/irccloud.com/x-nmdnucfmlpdbesxp) got netsplit. | 21:54 | |
| bbot2 (~bbot2@wyvern.cs.uni-duesseldorf.de) got netsplit. | 21:54 | |
| WildChild (~wildchild@modemcable165.184-57-74.mc.videotron.ca) got netsplit. | 21:54 | |
| __pv (pav@xen.ihme.org) got netsplit. | 21:54 | |
| nomadium (miguel@68.68.97.189) got netsplit. | 21:54 | |
| htoothrot (~mux@unaffiliated/htoothrot) got netsplit. | 21:54 | |
| arigato (~arigo@82.113.106.187) got netsplit. | 21:54 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) got netsplit. | 21:54 | |
| Insectoid (~Q@c-98-230-8-37.hsd1.fl.comcast.net) got netsplit. | 21:54 | |
| danchr (~danchr@cl-848.chi-02.us.sixxs.net) got netsplit. | 21:54 | |
| Arach (~arach@2.95.83.234) got netsplit. | 21:54 | |
| legogris (~legogris@c80-217-207-157.bredband.comhem.se) got netsplit. | 21:54 | |
| espes (~espes@123-243-221-248.tpgi.com.au) got netsplit. | 21:54 | |
| cwillu_at_work (~cwillu@cwillu-1-pt.tunnel.tserv13.ash1.ipv6.he.net) got netsplit. | 21:54 | |
| kirma (~jkirma@kekkonen.cs.hut.fi) got netsplit. | 21:54 | |
| _habnabit (~habnabit@python/site-packages/habnabit) got netsplit. | 21:54 | |
| tote (me@h-61-213.a213.priv.bahnhof.se) got netsplit. | 21:54 | |
| brodie (~brodie@unaffiliated/brodie) got netsplit. | 21:54 | |
| iko (iko@eskarina.iko.pp.se) got netsplit. | 21:54 | |
| tunixman (~tunixman@irc.theunixman.com) got netsplit. | 21:54 | |
| apexo (~apexo@2a01:238:436b:8301:5054:ff:fe87:82fb) got netsplit. | 21:54 | |
| dju (~dju@fsf/member/dju) got netsplit. | 21:54 | |
| ojii (~ojii@172.114.63.81.cust.bluewin.ch) got netsplit. | 21:54 | |
| zain (~textual@c-67-160-201-63.hsd1.ca.comcast.net) got netsplit. | 21:54 | |
| stakkars (~tismer@p5DDB72FD.dip.t-dialin.net) got netsplit. | 21:54 | |
| hakanardo (~hakan@h59ec009e.seluldx.dyn.perspektivbredband.net) got netsplit. | 21:54 | |
| __root__ (~Scalable@li252-14.members.linode.com) got netsplit. | 21:54 | |
| abki_ (~Amirouche@hyper.pilotsystems.net) got netsplit. | 21:54 | |
| jae (~jae@dedicated.jaerhard.com) got netsplit. | 21:54 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) got netsplit. | 21:54 | |
| Ademan (~dan@adsl-71-141-224-79.dsl.snfc21.pacbell.net) got netsplit. | 21:54 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) got netsplit. | 21:54 | |
| gutworth (~benjamin@128.36.84.91) got netsplit. | 21:54 | |
| squiddy (~squiddy@f053082220.adsl.alicedsl.de) got netsplit. | 21:54 | |
| dracman (~draco@212.255.19.163) got netsplit. | 21:54 | |
| stargaming (stargaming@unaffiliated/stargaming) got netsplit. | 21:54 | |
| Alex_Gaynor (u1246@gateway/web/irccloud.com/x-rgtzudjklhivhojf) got netsplit. | 21:54 | |
| dD0T (~dD0T@unaffiliated/dd0t) got netsplit. | 21:54 | |
| rndbit (~rndbit@vexar.de) got netsplit. | 21:54 | |
| Arfrever (~Arfrever@apache/committer/Arfrever) got netsplit. | 21:54 | |
| rguillebert (~rguillebe@fec76-1-78-234-124-105.fbx.proxad.net) got netsplit. | 21:54 | |
| mikefc (~mikefc@60-242-240-196.static.tpgi.com.au) got netsplit. | 21:54 | |
| melinath (~melinath@li341-41.members.linode.com) got netsplit. | 21:54 | |
| tnelson (~trent@trent.me) got netsplit. | 21:54 | |
| mwhudson (~mwh@linaro/mwhudson) got netsplit. | 21:54 | |
| fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) got netsplit. | 21:54 | |
| Vorpal (~AnMaster@unaffiliated/anmaster) got netsplit. | 21:54 | |
| DanKluev (~root@78-24-225-196-dsl.vntc.ru) got netsplit. | 21:54 | |
| unbit (~unbit@93-58-162-185.ip159.fastwebnet.it) got netsplit. | 21:54 | |
| setmeaway (~setmeaway@119.201.52.190) got netsplit. | 21:54 | |
| sunoano (~sa@unaffiliated/sunoano) got netsplit. | 21:54 | |
| lambacck (~chris@d24-150-124-118.home.cgocable.net) got netsplit. | 21:54 | |
| [Arfrever] (~Arfrever@apache/committer/Arfrever) got netsplit. | 21:54 | |
| smuxi (smuxi@descent.fragstore.net) joined #pypy. | 21:54 | |
| Nick change: smuxi -> panni_ | 21:54 | |
| Possible future nick collision: panni_ | 21:54 | |
| Taggnostr (~quassel@dyn57-215.yok.fi) returned to #pypy. | 21:54 | |
| Taggnostr (~quassel@dyn57-215.yok.fi) left irc: Read error: Connection reset by peer | 21:54 | |
| dimazest (~dimazest@balticom-202-59.balticom.lv) returned to #pypy. | 21:54 | |
| Unhelpful (~quassel@rockbox/developer/Unhelpful) returned to #pypy. | 21:54 | |
| carljm (~carljm@spinach.threepines.org) returned to #pypy. | 21:54 | |
| gsnedder1 (~gsnedders@mail.gsnedders.com) joined #pypy. | 21:54 | |
| Eventh (evenwiik@caracal.stud.ntnu.no) returned to #pypy. | 21:54 | |
| espes___ (~niel@205.185.120.132) joined #pypy. | 21:54 | |
| brutal_c1aos (~chaos@francium.electricgrey.com) joined #pypy. | 21:54 | |
| etrepum (~bob@50.8.114.2) returned to #pypy. | 21:54 | |
| aboudreault (~alanb@osgeo/member/aboudreault) returned to #pypy. | 21:54 | |
| int3__ (~int3__@st401-246.subnet-244.amherst.edu) returned to #pypy. | 21:54 | |
| nedbat (~nedbat@python/psf/nedbat) returned to #pypy. | 21:54 | |
| derdon (~derdon@p5DE89B26.dip.t-dialin.net) returned to #pypy. | 21:54 | |
| kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) returned to #pypy. | 21:54 | |
| linq (~ident@24-246-25-39.cable.teksavvy.com) returned to #pypy. | 21:54 | |
| tito (~tito@pdpc/supporter/bronze/tito) returned to #pypy. | 21:54 | |
| timotimo (~timo@gallifrey.wybt.net) returned to #pypy. | 21:54 | |
| harrison (~quassel@69.209.222.34) returned to #pypy. | 21:54 | |
| hpk (~hpk@hq2.merlinux.eu) returned to #pypy. | 21:54 | |
| papercrane (~papercran@c-76-103-172-115.hsd1.ca.comcast.net) returned to #pypy. | 21:54 | |
| cheater (~cheater@ip-80-226-24-12.vodafone-net.de) returned to #pypy. | 21:54 | |
| tumbleweed (~stefanor@ubuntu/member/tumbleweed) returned to #pypy. | 21:54 | |
| lvh (~lvh@python/psf/lvh) returned to #pypy. | 21:54 | |
| durin42 (~durin@adium/durin42) returned to #pypy. | 21:54 | |
| pedronis (~pedronis@73-53.195-178.cust.bluewin.ch) returned to #pypy. | 21:54 | |
| kalenz (~kalenz@ping.kalenz.fr) returned to #pypy. | 21:54 | |
| davisagli (~davisagli@davisagli.com) returned to #pypy. | 21:54 | |
| tos9 (~tos9@unaffiliated/tos9) returned to #pypy. | 21:54 | |
| jerith (~jerith@chthon.slipgate.za.net) returned to #pypy. | 21:54 | |
| thinker (~thinker@123-194-52-90.dynamic.kbronet.com.tw) returned to #pypy. | 21:54 | |
| Husio (husiatyn@oceanic.wsisiz.edu.pl) returned to #pypy. | 21:54 | |
| nopper (~nopper@ns.patched.biz) returned to #pypy. | 21:54 | |
| myf (~myf@li245-175.members.linode.com) returned to #pypy. | 21:54 | |
| pjenvey (~pjenvey@underboss.org) returned to #pypy. | 21:54 | |
| ousado (~ousado@unaffiliated/ousado) returned to #pypy. | 21:54 | |
| oal (u4126@gateway/web/irccloud.com/x-nmdnucfmlpdbesxp) returned to #pypy. | 21:54 | |
| alexsuraci (u1962@gateway/web/irccloud.com/x-fmgkunpvlhywhrbp) returned to #pypy. | 21:54 | |
| xorAxAx (~alexander@moinmoin/coreteam/alexander) returned to #pypy. | 21:54 | |
| ThomasWaldmann (~twaldmann@moinmoin/coreteam/thomas) returned to #pypy. | 21:54 | |
| Circlefusion (~cf@74-138-199-109.dhcp.insightbb.com) returned to #pypy. | 21:54 | |
| bbot2 (~bbot2@wyvern.cs.uni-duesseldorf.de) returned to #pypy. | 21:54 | |
| WildChild (~wildchild@modemcable165.184-57-74.mc.videotron.ca) returned to #pypy. | 21:54 | |
| mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) returned to #pypy. | 21:54 | |
| zk (~is@whatit.is) returned to #pypy. | 21:54 | |
| Rhyolite (~dje@pool-108-6-25-71.nycmny.fios.verizon.net) returned to #pypy. | 21:54 | |
| raydeo (~raydeo@merickel.org) returned to #pypy. | 21:54 | |
| DanielHolth (~dholth@ip98-180-34-112.ga.at.cox.net) returned to #pypy. | 21:54 | |
| Garen (garen@69.76.18.3) returned to #pypy. | 21:54 | |
| __pv (pav@xen.ihme.org) returned to #pypy. | 21:54 | |
| maxyz (~maxy@186.18.94.35) returned to #pypy. | 21:54 | |
| elipsion_ (~elipsion@81-232-141-35-no217.tbcn.telia.com) returned to #pypy. | 21:54 | |
| getxsick (xsx@unaffiliated/getxsick) returned to #pypy. | 21:54 | |
| Arnar (~arnarb@li187-148.members.linode.com) returned to #pypy. | 21:54 | |
| JStoker (jstoker@unaffiliated/jstoker) returned to #pypy. | 21:54 | |
| Gulaway (~Gulopine@173-203-203-206.static.cloud-ips.com) returned to #pypy. | 21:54 | |
| __main__ (~main@adsl-99-173-15-158.dsl.pltn13.sbcglobal.net) returned to #pypy. | 21:54 | |
| CIA-13 (~CIA@cia.atheme.org) returned to #pypy. | 21:54 | |
| masquerade (~rdeaton@galactica.lug.udel.edu) returned to #pypy. | 21:54 | |
| MostAwesomeDude (~simpson@osuosl/staff/mostawesomedude) returned to #pypy. | 21:54 | |
| nomadium (miguel@68.68.97.189) returned to #pypy. | 21:54 | |
| lmatos (lmatos@staff.linode.com) returned to #pypy. | 21:54 | |
| htoothrot (~mux@unaffiliated/htoothrot) returned to #pypy. | 21:54 | |
| dmarkey (~dmarkey@dmarkey.xen.prgmr.com) returned to #pypy. | 21:55 | |
| jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) returned to #pypy. | 21:55 | |
| Insectoid (~Q@c-98-230-8-37.hsd1.fl.comcast.net) returned to #pypy. | 21:55 | |
| danchr (~danchr@cl-848.chi-02.us.sixxs.net) returned to #pypy. | 21:55 | |
| Arach (~arach@2.95.83.234) returned to #pypy. | 21:55 | |
| legogris (~legogris@c80-217-207-157.bredband.comhem.se) returned to #pypy. | 21:55 | |
| espes (~espes@123-243-221-248.tpgi.com.au) returned to #pypy. | 21:55 | |
| cwillu_at_work (~cwillu@cwillu-1-pt.tunnel.tserv13.ash1.ipv6.he.net) returned to #pypy. | 21:55 | |
| kirma (~jkirma@kekkonen.cs.hut.fi) returned to #pypy. | 21:55 | |
| _habnabit (~habnabit@python/site-packages/habnabit) returned to #pypy. | 21:55 | |
| tote (me@h-61-213.a213.priv.bahnhof.se) returned to #pypy. | 21:55 | |
| brodie (~brodie@unaffiliated/brodie) returned to #pypy. | 21:55 | |
| iko (iko@eskarina.iko.pp.se) returned to #pypy. | 21:55 | |
| apexo (~apexo@2a01:238:436b:8301:5054:ff:fe87:82fb) returned to #pypy. | 21:55 | |
| tunixman (~tunixman@irc.theunixman.com) returned to #pypy. | 21:55 | |
| dju (~dju@fsf/member/dju) returned to #pypy. | 21:55 | |
| Dulak (~michael@unaffiliated/dulak) returned to #pypy. | 21:55 | |
| exarkun (~exarkun@wordeology.com) returned to #pypy. | 21:55 | |
| #pypy: mode change '+o exarkun' by ChanServ!ChanServ@services. | 21:55 | |
| Nick change: gsnedder1 -> gsnedders | 21:55 | |
| Possible future nick collision: gsnedders | 21:55 | |
| GDorn (~gdorn@li223-147.members.linode.com) returned to #pypy. | 21:55 | |
| Wraithan (~Wraithan@c-67-189-110-252.hsd1.or.comcast.net) returned to #pypy. | 21:55 | |
| asmeurer__ (~asmeurer@dhcp-baca-230.resnet.nmt.edu) returned to #pypy. | 21:55 | |
| kkris (~kris@188-22-115-53.adsl.highway.telekom.at) returned to #pypy. | 21:55 | |
| mat^2 (~mathias@212.130.113.35) returned to #pypy. | 21:55 | |
| lotrpy (~lotrpy@202.120.58.116) returned to #pypy. | 21:55 | |
| apanda (~apanda@c-24-7-101-62.hsd1.ca.comcast.net) returned to #pypy. | 21:55 | |
| __name__ (~name@sburn/devel/name) returned to #pypy. | 21:55 | |
| jell (jellonek@212.182.18.33) returned to #pypy. | 21:55 | |
| arkx (arkx@hilla.kapsi.fi) returned to #pypy. | 21:55 | |
| thoughtpolice (~thoughtpo@li111-206.members.linode.com) returned to #pypy. | 21:55 | |
| henning_ (~henning@unicorn.gnubo.de) returned to #pypy. | 21:55 | |
| asksol_ (~asksol@sanedefaults.net) joined #pypy. | 21:55 | |
| norox (~norox@2a00:801::2d0:b7ff:fea0:c753) returned to #pypy. | 21:55 | |
| apollo13 (~apollo13@unaffiliated/apollo13) returned to #pypy. | 21:55 | |
| PiotrSikora (~none@nginx/adept/piotrsikora) returned to #pypy. | 21:55 | |
| magcius (magcius@unaffiliated/magcius) returned to #pypy. | 21:55 | |
| abki_ (~Amirouche@hyper.pilotsystems.net) returned to #pypy. | 21:56 | |
| jae (~jae@dedicated.jaerhard.com) returned to #pypy. | 21:56 | |
| jterrace (~jterrace@pu116187.student.Princeton.EDU) returned to #pypy. | 21:56 | |
| idnar (~quassel@unaffiliated/idnar) returned to #pypy. | 21:56 | |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) returned to #pypy. | 21:56 | |
| Ademan (~dan@adsl-71-141-224-79.dsl.snfc21.pacbell.net) returned to #pypy. | 21:56 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) returned to #pypy. | 21:56 | |
| gutworth (~benjamin@128.36.84.91) returned to #pypy. | 21:56 | |
| squiddy (~squiddy@f053082220.adsl.alicedsl.de) returned to #pypy. | 21:56 | |
| dracman (~draco@212.255.19.163) returned to #pypy. | 21:56 | |
| stargaming (stargaming@unaffiliated/stargaming) returned to #pypy. | 21:56 | |
| Alex_Gaynor (u1246@gateway/web/irccloud.com/x-rgtzudjklhivhojf) returned to #pypy. | 21:56 | |
| dD0T (~dD0T@unaffiliated/dd0t) returned to #pypy. | 21:56 | |
| rndbit (~rndbit@vexar.de) returned to #pypy. | 21:56 | |
| Arfrever (~Arfrever@apache/committer/Arfrever) returned to #pypy. | 21:56 | |
| rguillebert (~rguillebe@fec76-1-78-234-124-105.fbx.proxad.net) returned to #pypy. | 21:56 | |
| mikefc (~mikefc@60-242-240-196.static.tpgi.com.au) returned to #pypy. | 21:56 | |
| melinath (~melinath@li341-41.members.linode.com) returned to #pypy. | 21:56 | |
| tnelson (~trent@trent.me) returned to #pypy. | 21:56 | |
| #pypy: mode change '+o Alex_Gaynor ' by adams.freenode.net | 21:56 | |
| guibou (~guibou@kraken.insecable.net) returned to #pypy. | 21:57 | |
| Ormod_ (~valtha@ohmu.fi) joined #pypy. | 21:57 | |
| ojii (~ojii@172.114.63.81.cust.bluewin.ch) returned to #pypy. | 21:57 | |
| zain (~textual@c-67-160-201-63.hsd1.ca.comcast.net) returned to #pypy. | 21:57 | |
| stakkars (~tismer@p5DDB72FD.dip.t-dialin.net) returned to #pypy. | 21:57 | |
| hakanardo (~hakan@h59ec009e.seluldx.dyn.perspektivbredband.net) returned to #pypy. | 21:57 | |
| __root__ (~Scalable@li252-14.members.linode.com) returned to #pypy. | 21:57 | |
| jterrace | Alex_Gaynor: i think #18 shouldnt be any trouble, if you merge that one, i can submit my final patch for fromstring | 21:57 |
| lotrpy (~lotrpy@202.120.58.116) left irc: Ping timeout: 252 seconds | 21:58 | |
| CIA-13 | 03r.lobb 07roundup * 10#949/1.7 Sandbox problems: Many thanks for all that, Ned. I'll give your branch a try tomorrow. * 14https://bugs.pypy.org/issue949 | 21:58 |
| htoothrot (~mux@unaffiliated/htoothrot) left irc: Excess Flood | 21:58 | |
| Alex_Gaynor | jterrace: yeah, I'm just trying to see if there's a simpler way to do it, so more code is shared with __builtin__.int.__new__ | 21:58 |
| jterrace | yeah i was hoping coerce could just handle strings but i couldnt get space.int to play nicely with strings | 21:59 |
| `fox` (~fox@host74-251-dynamic.50-79-r.retail.telecomitalia.it) left irc: Ping timeout: 267 seconds | 21:59 | |
| lotrpy (~lotrpy@202.120.58.116) joined #pypy. | 21:59 | |
| Alex_Gaynor | I think space.call_function(space.w_int, w_item) would work, but then FakeSpace needs to be updated, grumble | 22:00 |
| jterrace | oh | 22:00 |
| jterrace | but then the exceptions have to be propagated too right? | 22:00 |
| jterrace | or do you get that for free | 22:00 |
| ec (~me@ell.io) returned to #pypy. | 22:01 | |
| Hodgestar (~quassel@gluon.za.net) returned to #pypy. | 22:01 | |
| bbot2 | 4Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/746 [12hakanardo, jit-targets] | 22:01 |
| Alex_Gaynor | should propagate fine | 22:01 |
| hakanardo__ (~ardo@putten.maths.lth.se) returned to #pypy. | 22:01 | |
| lambacck (~chris@d24-150-124-118.home.cgocable.net) returned to #pypy. | 22:01 | |
| setmeaway (oosool3@119.201.52.190) joined #pypy. | 22:01 | |
| [Arfrever] (~Arfrever@apache/committer/Arfrever) returned to #pypy. | 22:01 | |
| mwhudson (~mwh@linaro/mwhudson) returned to #pypy. | 22:01 | |
| DanKluev (~root@78-24-225-196-dsl.vntc.ru) returned to #pypy. | 22:01 | |
| jterrace | oh also i probably should have mentioned | 22:02 |
| ivan` (~ivan@unaffiliated/ivan/x-000001) returned to #pypy. | 22:02 | |
| jterrace | numpy.int32('1.2') throws a ValueError on cpython with a cryptic message | 22:02 |
| htoothrot (~mux@unaffiliated/htoothrot) joined #pypy. | 22:02 | |
| jterrace | the message is different in pypy | 22:02 |
| Alex_Gaynor | close enough I think | 22:03 |
| jterrace | numpy.int64('9223372036854775808') gives a ValueError on cpython instead of an OverflowError | 22:04 |
| jterrace | but it's OverflowError on pypy, which i think is a better exception, but if you want perfect compatibility, it could be changed | 22:04 |
| jterrace | it gives you a pretty useless error: ValueError: setting an array element with a sequence. | 22:04 |
| brutal_chaos (~chaos@francium.electricgrey.com) got lost in the net-split. | 22:05 | |
| kenaan (~kenaan@66.35.39.65) got lost in the net-split. | 22:05 | |
| Ormod (~valtha@ohmu.fi) got lost in the net-split. | 22:05 | |
| price (~price@LINERVA.MIT.EDU) got lost in the net-split. | 22:05 | |
| asksol (~asksol@sanedefaults.net) got lost in the net-split. | 22:05 | |
| espes__ (~niel@205.185.120.132) got lost in the net-split. | 22:05 | |
| lahwran (lahwran@unaffiliated/lahwran) got lost in the net-split. | 22:05 | |
| sunoano (~sa@unaffiliated/sunoano) got lost in the net-split. | 22:05 | |
| hazmat (~hazmat@plone/hazmat) got lost in the net-split. | 22:05 | |
| unbit (~unbit@93-58-162-185.ip159.fastwebnet.it) got lost in the net-split. | 22:05 | |
| Vorpal (~AnMaster@unaffiliated/anmaster) got lost in the net-split. | 22:05 | |
| fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) got lost in the net-split. | 22:05 | |
| arigato (~arigo@82.113.106.187) got lost in the net-split. | 22:05 | |
| Alex_Gaynor | god I love numpy | 22:05 |
| Alex_Gaynor | :) | 22:05 |
| jterrace | heh | 22:05 |
| Vorpal (~AnMaster@unaffiliated/anmaster) joined #pypy. | 22:05 | |
| jterrace | so i left it as overflowerror.. because that seems like a bug in numpy to me | 22:05 |
| Action: jterrace shrugs | 22:05 | |
| jterrace | you get that terrible exception with something like numpy.float32('23.4sdf') too | 22:06 |
| Alex_Gaynor | ValueError: invalid literal for float(): 23.4sdf | 22:06 |
| Alex_Gaynor | is what I get | 22:06 |
| jterrace | on cpython? | 22:06 |
| Alex_Gaynor | yes | 22:07 |
| jterrace | i get ValueError with numpy 1.5 | 22:07 |
| jterrace | are you on 2.0? | 22:07 |
| Alex_Gaynor | http://bpaste.net/show/20882/ | 22:07 |
| Alex_Gaynor | that pasted more than I intended | 22:07 |
| Alex_Gaynor | lastbi though | 22:07 |
| Alex_Gaynor | I'm on 1.6.0 | 22:07 |
| jterrace | oh | 22:07 |
| jterrace | they must have fixed it in 1.6 because 1.5.1 i get ValueError | 22:08 |
| PiotrSikora (~none@nginx/adept/piotrsikora) left irc: Excess Flood | 22:08 | |
| PiotrSikora (~none@nginx/adept/piotrsikora) joined #pypy. | 22:08 | |
| lahwran (lahwran@unaffiliated/lahwran) joined #pypy. | 22:09 | |
| jterrace | i wonder why ubuntu hasnt upgraded to 1.6 | 22:09 |
| Ademan_ (~dan@adsl-71-141-224-79.dsl.snfc21.pacbell.net) joined #pypy. | 22:09 | |
| mwhudson (~mwh@linaro/mwhudson) left irc: Ping timeout: 252 seconds | 22:10 | |
| jterrace | it's still 1.5.1 in debian unstable... weird | 22:11 |
| -Martinp23 (martinp23@freenode/staff/martinp23) to $$*- [Global Notice] Hi everyone. The upgrades are now complete. There was a bit of a hiccup when barjavel needed a kicking to connect to the network, but apart from that everything went well. Thanks for your patience through the upgrades. Keen for the next series? Keep an eye on the blog to get involved with testing future versions and to be aware of maintenance. Thanks for using freenode | 22:16 | |
| tumbleweed | 1.6 is in experimental | 22:18 |
| kenaan (~kenaan@cube.twistedmatrix.com) joined #pypy. | 22:18 | |
| kenaan | 12arigo jit-simplify-backendintf 11a4f5ba4d79b4 15/pypy/jit/backend/: In-progress. Three complicated and long functions are gone :-) | 22:18 |
| tumbleweed | this is insane. Kicked off a pair of translations together a couple of hours ago, and tehy are look like they are going through the unit-tests in lock-step. Pypy is too predicable | 22:19 |
| Ademan_ (~dan@adsl-71-141-224-79.dsl.snfc21.pacbell.net) left irc: Quit: leaving | 22:23 | |
| int3__ (~int3__@st401-246.subnet-244.amherst.edu) left irc: Remote host closed the connection | 22:31 | |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/167 | 22:35 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/741 | 22:35 |
| bbot2 | Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/975 | 22:35 |
| jaredw (~jared@122-59-91-107.jetstream.xtra.co.nz) joined #pypy. | 22:41 | |
| int3__ (~int3__@st401-246.subnet-244.amherst.edu) joined #pypy. | 22:52 | |
| kkris (~kris@188-22-115-53.adsl.highway.telekom.at) left irc: Quit: Leaving. | 23:02 | |
| etrepum (~bob@50.8.114.2) left irc: Quit: etrepum | 23:04 | |
| e1ven (~E1ven@SQ7/ProjectLead/E1ven) joined #pypy. | 23:09 | |
| int3__ (~int3__@st401-246.subnet-244.amherst.edu) left irc: Remote host closed the connection | 23:10 | |
| mwhudson (~mwh@linaro/mwhudson) joined #pypy. | 23:16 | |
| dracman (~draco@212.255.19.163) left irc: Ping timeout: 252 seconds | 23:20 | |
| amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 252 seconds | 23:21 | |
| ojii (~ojii@172.114.63.81.cust.bluewin.ch) left irc: Remote host closed the connection | 23:22 | |
| squiddy (~squiddy@f053082220.adsl.alicedsl.de) left irc: Quit: Leaving | 23:32 | |
| lucian (~lucian@cpc1-newc15-2-0-cust84.gate.cable.virginmedia.com) left irc: Remote host closed the connection | 23:44 | |
| JaRoel|4d (~jaroel|4d@sink.jaroel.nl) left irc: Remote host closed the connection | 23:50 | |
| stakkars_ (~tismer@p5DDB72FD.dip.t-dialin.net) joined #pypy. | 23:55 | |
| davidpk (~davidpk@5e0f1559.bb.sky.com) left irc: Quit: davidpk | 23:59 | |
| --- Mon Dec 12 2011 | 00:00 | |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!