#pypy IRC log for Tuesday, 2012-01-10

craigkerstiens (~craigkers@204.14.152.118) left irc: Quit: craigkerstiens00:01
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1506 [12fijal, better-jit-hooks]00:01
bbot24Failure: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/100800:03
bbot24Failure: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/19900:05
fijalwtf00:06
nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 255 seconds00:08
Alex_Gaynorfijal: a bunch of flat tests rae failing on default00:11
fijalalso benchmarks failed00:12
Alex_Gaynorprobably becaus eof the failing tests00:12
fijalfrom yesterday00:12
whitelynx (~whitelynx@75.111.197.204) left irc: Ping timeout: 268 seconds00:12
whitelynx (~whitelynx@75.110.189.230) joined #pypy.00:19
_habnabit (habnabit@python/site-packages/habnabit) left #pypy ("Channel buffer killed").00:19
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds00:23
nedbat (~nedbat@python/psf/nedbat) joined #pypy.00:36
antocuni (~antocuni@host157-123-dynamic.2-87-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds00:39
Nisstyre (~yours@c-208-90-102-250.netflash.net) left irc: Ping timeout: 240 seconds00:51
mfoord (~anonymous@87-194-212-65.bethere.co.uk) joined #pypy.00:53
mfoord (~anonymous@87-194-212-65.bethere.co.uk) left irc: Changing host00:53
mfoord (~anonymous@python/psf/voidspace) joined #pypy.00:53
Nisstyre (~yours@c-208-90-102-250.netflash.net) joined #pypy.01:05
aboudreault (~alanb@osgeo/member/aboudreault) left irc: Ping timeout: 240 seconds01:22
kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) left irc: Quit: Computer has gone to sleep.01:23
aboudreault (~alanb@osgeo/member/aboudreault) joined #pypy.01:24
squiddy (~squiddy@f053084228.adsl.alicedsl.de) left irc: Ping timeout: 252 seconds01:28
nettok (~quassel@190.148.27.164) joined #pypy.01:37
[Arfrever] (~Arfrever@apache/committer/Arfrever) left irc: Read error: Operation timed out01:42
[Arfrever] (~Arfrever@apache/committer/Arfrever) joined #pypy.01:43
bbot24Failure: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/77101:46
kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) joined #pypy.01:47
bbot2Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/191902:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/33302:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/67302:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/32202:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/67102:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/118902:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/85402:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/150702:00
bbot2Started: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/78302:00
mfoord (~anonymous@python/psf/voidspace) left irc: Quit: mfoord02:08
fermianyon (~lane@c-71-229-21-197.hsd1.al.comcast.net) left irc: Ping timeout: 240 seconds02:28
etrepum_ (~bob@accessnat4.mochimedia.net) joined #pypy.02:29
etrepum (~bob@accessnat4.mochimedia.net) left irc: Ping timeout: 240 seconds02:32
Nick change: etrepum_ -> etrepum02:32
etrepum (~bob@accessnat4.mochimedia.net) left irc: Client Quit02:33
mat^2 (~mathias@212.130.113.35) left irc: Ping timeout: 240 seconds02:53
ericflo (~ericflo@75.103.8.110) left irc: Quit: ericflo02:56
bbot24Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/191903:00
etrepum (~bob@75-101-96-144.dsl.static.sonic.net) joined #pypy.03:02
Kaskuka (~John@osbk-4db1589d.pool.mediaWays.net) left irc: Ping timeout: 240 seconds03:03
Kaskuka (~John@osbk-4db150f7.pool.mediaWays.net) joined #pypy.03:04
ernop (~ernie@114.250.129.15) left irc: Ping timeout: 268 seconds03:19
dracman (~draco@212.255.37.149) left irc: Ping timeout: 252 seconds03:19
bgola (~bgola@189.100.41.65) left irc: Ping timeout: 248 seconds03:20
bgola (~bgola@189.100.41.65) joined #pypy.03:21
mikefc_ (9665bcc5@gateway/web/freenode/ip.150.101.188.197) left irc: Quit: Page closed03:22
ernop (~ernie@114.250.129.15) joined #pypy.03:22
dracman (~draco@212.255.35.167) joined #pypy.03:25
ernop (~ernie@114.250.129.15) left irc: Quit: Leaving03:33
mikefc_ (9665bcc5@gateway/web/freenode/ip.150.101.188.197) joined #pypy.03:50
fermianyon (~lane@c-68-35-198-245.hsd1.al.comcast.net) joined #pypy.03:50
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.03:54
setmeaway (setmeaway3@119.201.52.190) left irc: Ping timeout: 240 seconds03:55
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Client Quit03:56
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/150704:04
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/85404:04
fermianyon (~lane@c-68-35-198-245.hsd1.al.comcast.net) left irc: Ping timeout: 240 seconds04:09
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/67104:19
anish_ (anish@nat/redhat/x-vpcjsncxisauzseb) joined #pypy.04:24
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/118904:34
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.04:38
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Client Quit04:39
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.04:48
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Client Quit04:48
mikefc_I want to add numpy.testing to numpypy (at applevel).  But if I just copy the numpy/testing dir into lib_pypy/numpypy/testing I can't actually access the module "numpypy.testing" until I do a translate?  Is this correct?  What am I doing wrong here?05:01
Action: mikefc_ doesn't have a machine beefy enough to do a translate05:01
nettok (~quassel@190.148.27.164) left irc: Remote host closed the connection05:02
nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 255 seconds05:03
setmeaway (stemearay@119.201.52.190) joined #pypy.05:05
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.05:10
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Client Quit05:11
setmeaway (stemearay@119.201.52.190) left irc: Ping timeout: 248 seconds05:12
gutworth (~benjamin@dhcp-128-36-84-221.central.yale.edu) left irc: Ping timeout: 276 seconds05:15
fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) left irc: Quit: fzzzy05:22
durin42 (~durin@adium/durin42) left irc: Read error: Operation timed out05:30
durin42 (~durin@adium/durin42) joined #pypy.05:34
whyking (~quassel@p5B3DD930.dip.t-dialin.net) joined #pypy.05:48
setmeaway (~setmeaway@119.201.52.190) joined #pypy.05:59
mikefc_ (9665bcc5@gateway/web/freenode/ip.150.101.188.197) left irc: Quit: Page closed06:06
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/67306:06
mat^2 (~mathias@212.130.113.35) joined #pypy.06:09
setmeaway (~setmeaway@119.201.52.190) left irc: Ping timeout: 252 seconds06:18
fermianyon (~lane@c-71-229-21-197.hsd1.al.comcast.net) joined #pypy.06:19
aleksi (~aleksi@85.235.191.82) joined #pypy.06:21
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.06:29
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Client Quit06:31
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.06:44
mat^2 (~mathias@212.130.113.35) left irc: 06:48
whitelynx (~whitelynx@75.110.189.230) left irc: Ping timeout: 252 seconds06:54
bbot24Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/78307:05
whitelynx (~whitelynx@75.111.197.204) joined #pypy.07:08
asmeurer (~asmeurer@c-174-56-21-245.hsd1.nm.comcast.net) left irc: Quit: asmeurer07:11
kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) left irc: Quit: Computer has gone to sleep.07:23
kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) joined #pypy.07:24
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Ping timeout: 252 seconds07:25
mvt (~mvantelli@87.213.45.85) joined #pypy.07:26
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.07:29
JaRoel|4d (~jaroel|4d@sink.jaroel.nl) left irc: Remote host closed the connection07:32
fermianyon (~lane@c-71-229-21-197.hsd1.al.comcast.net) left irc: Ping timeout: 240 seconds07:32
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Ping timeout: 255 seconds07:57
zain (~textual@66.175.91.194) left irc: Quit: Computer has gone to sleep.07:59
antocuni (~antocuni@host157-123-dynamic.2-87-r.retail.telecomitalia.it) joined #pypy.08:10
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/33308:15
JaRoel (~jaroel|4d@office.fourdigits.nl) joined #pypy.08:17
asmeurer (~asmeurer@c-174-56-21-245.hsd1.nm.comcast.net) joined #pypy.08:17
bivab (~david@fwstups.cs.uni-duesseldorf.de) joined #pypy.08:17
Trundle (~andy@p578bfdcf.dip0.t-ipconnect.de) joined #pypy.08:20
Trundle (~andy@p578bfdcf.dip0.t-ipconnect.de) left irc: Changing host08:20
Trundle (~andy@python/site-packages/trundle) joined #pypy.08:20
JaRoel|4d (~jaroel|4d@office.fourdigits.nl) joined #pypy.08:26
JaRoel (~jaroel|4d@office.fourdigits.nl) left irc: Ping timeout: 240 seconds08:29
aboudreault (~alanb@osgeo/member/aboudreault) left irc: Ping timeout: 240 seconds08:30
kennethreitz (~kennethre@c-24-127-96-129.hsd1.va.comcast.net) left irc: Quit: Computer has gone to sleep.08:31
Arfrever (~Arfrever@apache/committer/Arfrever) left irc: Quit: Ex+re08:32
setmeaway (~setmeaway@119.201.52.190) joined #pypy.08:40
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/32208:40
squiddy (~squiddy@g224194229.adsl.alicedsl.de) joined #pypy.08:49
CIA-1703Twirrim 07roundup * 10#994/significantly slow joins: 09:08
CIA-17[new] I'm uncertain whether it's appropriate to raise this as a new bug. It seems09:08
CIA-17tied to 866 (maybe?), but against 1.7 release and nightly (201 ... * 14https://bugs.pypy.org/issue99409:08
G2P (~G2P@fw-asn1.ornis.com) joined #pypy.09:28
lucian (~lucian@93-97-174-114.zone5.bethere.co.uk) joined #pypy.09:46
kenaan12hager ppc-jit-backend 11dd765153417e 15/pypy/jit/backend/ppc/test/test_runner.py: add test to ensure that arguments are passed correctly09:47
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) joined #pypy.09:49
tumbleweedfijal: I assume I should skip the optimisations that are marked INTERNAL USE ONLY?09:58
lucian_ (NB8QM3MWVF@irc-2.usealice.org) joined #pypy.09:59
fijaltumbleweed: you can mention that the option is there09:59
fijalbut generally we don't want people to play with that list09:59
fijalit almost never leads to anything good :)09:59
tumbleweedrighto10:00
fijallet me put it in other words - if you don't know what potential values can go there don't use it :)10:00
fijalit's only for "does this optimization cause a problem"10:00
fijaleverything else can be used (at least in theory) to tweak performance10:00
tumbleweedanother option is to generate the manpage from source (like the api docs), but one often wants grammar differences. I'm happy to update it for releases10:01
fijalI think the manpage is short enough so not make it too complex?10:01
fijaldo you have a commit bit btw?10:01
tumbleweedon your repo, no :)10:02
tumbleweedwhich is why I file bugs10:02
tumbleweedyeah, it's short enough10:02
fijalwant commit access?10:02
fijalwe give it to pretty much anyone, don't feel special ;-)10:02
tumbleweedheh, in that case yes please10:03
fijalfeel free to put it in doc/manpage.1.rst or so10:03
tumbleweedrighto10:03
fijalit's funny10:03
fijalwe never ever had a problem with a newcomer commiting something silly to trunk10:03
fijalwe had tons of problems with people like me or armin commiting something silly to trunk :)10:04
tumbleweedclearly only sensible people are interested in pypy :)10:04
fijalusername10:04
fijal?10:04
tumbleweedstefanor10:04
fijalno, newcomers are careful10:04
fijalhere you go10:05
tumbleweedyeah, and the longtimers tend to get quite cavalier "what could possibly go wrong..."10:05
tumbleweedin my experince, the shortest, simplest patches that couldn't possibly need testing are usually the ones that break the most.10:05
fijalyes :)10:06
fijal"why bother running tests it's a typo"10:06
fijalI checked in syntax errors to pypy trunk10:06
CIA-1703timo 07roundup * 10#994/significantly slow joins: (log message trimmed)10:09
CIA-17[chatting] this is not about join being slow. in fact, this code:10:09
CIA-17start = time()10:09
CIA-17ir="0"10:09
CIA-17i=1 * 14https://bugs.pypy.org/issue99410:09
fijaltimotimo: we should enable stringjoin10:16
fijalI'm sick of it10:16
fijaleh, I broke stuff10:18
fijaltumbleweed: speaking of which...10:18
tumbleweedhrm?10:19
tumbleweedah10:19
kenaan12fijal default 11e6f379da6e7c 15/pypy/jit/backend/x86/regalloc.py: argh, I'm stupid, use the correct API10:24
CIA-1703timo 07roundup * 10#994/significantly slow joins: 10:28
CIA-17FWIW, issue926 might be interesting to you. If translating your pypy with the10:28
CIA-17strbuf objspace option gives you an immense speed boost, you'll kno ... * 14https://bugs.pypy.org/issue99410:28
timotimofijal: was that what you meant?10:28
timotimoi'm not even sure why it's not enabled by default. no message on that issue says what the drawbacks are10:29
fijaltimotimo: "someone has to test it"10:29
fijalis the main drawback10:29
timotimowhat's the kind of thing that could go wrong? are there too few test cases? segfaults lying in wait to jump out at the worst possible moment?10:30
timotimoor could there be speed regressions somehow?10:30
fijaltimotimo: someone has to create a branch, run tests and benchmarks (on buildbot)10:30
fijaland merge it10:30
fijalnothing special10:30
fijaltimotimo: feel like doing it?10:31
timotimowell, i could try, i suppose10:32
timotimolet me bring my local copy up to speed10:32
fijaltimotimo: you need to mostly kick buildbots :)10:35
fijaland also read a bit code I presume10:35
timotimowell, i heard reading code improves ones programming skills10:35
timotimomy starting point should be the target, right? where i would turn the option on by default. is that the only change that needs to happen before i can run the buildbots on that branch?10:36
fijalyes, pretty much10:36
kenaan12hager ppc-jit-backend 11768f640c18b7 15/pypy/jit/backend/ppc/test/test_runner.py: rename test, start with 2 arguments10:37
kenaan12hager ppc-jit-backend 11d1b7f8e3b929 15/pypy/jit/backend/ppc/ppcgen/regalloc.py: (bivab, hager): fix offset to stack parameters10:37
timotimoyou said strjoin, but it seems strbuf is "better" in some way10:38
timotimoat least strbuf claims to be "strjoin v2"10:39
fijalthis is precisely the work that has to be done :)10:39
fijaldetermining what's up10:40
timotimomhm10:41
kenaan12arigo default 11799b4c3164db 15/pypy/rlib/jit.py: Fix the docstrings.10:41
timotimothere is a really big amount of objspace "optimisations" in there10:42
fijalwe should remove some of them10:43
fijaland enable some others10:43
fijalcode trimming is a good idea10:43
timotimoyes10:44
fijal<glyph> make it work, make it right, make it fast10:44
fijal<glyph> pypy doesn't do things in that order usually, I guess10:44
fijal<glyph> make it crazy, throw away some branches, wait for armin to have an even crazier idea, then check it in without warning10:44
fijal<glyph> that is the pypy way yes10:44
timotimoit seems like i have to figure out how to work hgs commandline now10:46
lizardo (~lizardo@189.2.128.130) joined #pypy.10:49
kenaan12timo_jbo strbuf_by_default 11211606889b44 15/pypy/config/pypyoption.py: turn on the strbuf (strjoin v2) objspace optimisation by default10:49
fijalwe normally use - instead of _ by custom10:50
fijal(note for the next time)10:50
timotimooh, all right10:50
timotimoi'm on own-linux-x86-32 buildbot page, would i now fill out "force build" with all i know? do i have to fill out everything, or will it, for instance, take the newest revision of the branch i supply if i leave out the revision?10:51
fijalyou usually fill "who" and "branch"10:51
timotimoall right10:51
fijalthere is a script that posts it for you10:51
fijalbut read it before running10:51
fijalforce-builds.py or so10:51
bbot2Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/1920 [12Timo Paulssen, strbuf_by_default]10:51
timotimoah, should have waited a few seconds :)10:52
timotimowell, now i'll do it manually10:52
fijalso10:52
bbot2Started: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/784 [12Timo Paulssen, strbuf_by_default]10:52
fijalown app and jit10:52
fijalno need to run both 32 and 6410:52
fijal(you'll wait longer)10:52
kenaan12fijal look-into-thread 11a72a6f955660 15/pypy/module/thread/ll_thread.py: don't look into a function that does add_memory_pressure. We should fix it one day10:52
fijaland then when they're clean, benchmark10:52
timotimooh, ok, i'll cancel it.10:53
bbot213Exception: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/784 [12Timo Paulssen, strbuf_by_default]10:53
fijalah10:53
timotimopypy-c-jit-linux-x86-32 � this is the one you mean, yes?10:53
fijalwhen did you branch?10:53
timotimoa few minutes ago10:53
fijalhas to be after e6f379da6e7c10:53
fijalbecause I broke tests10:53
timotimoi'll have a look10:54
timotimothat's the commit my branch is based off of, yes10:54
fijalok10:55
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-win-x86-32/builds/542 [12timo, strbuf_by_default]10:57
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1190 [12timo, strbuf_by_default]10:57
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/323 [12timo, strbuf_by_default]10:57
timotimothat's easier indeed.10:57
fijal:]11:00
nirbheek (~nirbheek@gentoo/developer/flyingspaghettimonster/nirbheek) left irc: Read error: Connection timed out11:08
nirbheek (~nirbheek@gentoo/developer/flyingspaghettimonster/nirbheek) joined #pypy.11:11
lekma (~lekma@a83-163-86-26.adsl.xs4all.nl) joined #pypy.11:12
anish_ (anish@nat/redhat/x-vpcjsncxisauzseb) left irc: Ping timeout: 252 seconds11:12
lekmahi all11:13
fijallekma: hi11:14
lekmaI need a little bit of help in understanding lltype,11:15
lekmaI'm working on https://bugs.pypy.org/issue991 (even though mikefc called it moronic (care to elaborate))11:16
anish_ (anish@nat/redhat/x-rvkkvrhvfmtyrqkn) joined #pypy.11:16
lekmaand I'd like to understand how I can pass a NULL pointer to a function11:16
fijaluh11:17
fijalok :)11:17
fijalthis is head first to deep water11:17
fijalbut fine11:17
fijallltype.nullptr(TYPE_OF_PTR)11:17
lekmaok so if I have something like : @cpython_api([PyObject, PyObject], PyObject)11:18
kenaan12hager ppc-jit-backend 11308dd2d5e89f 15/pypy/jit/backend/ppc/ppcgen/: (bivab, hager): fix off-by-one bug in computation of offset to stack locations11:19
lekmaand the second argument is null: it should be: lltype.nullptr(PyObject.TO)11:19
fijalyes11:20
fijalprecisely11:20
lekmaok. Now how can I convert an lltype.nullptr(rffi.CCHARP.TO) to a lltype.nullptr(PyObject.TO)11:21
lekma?11:21
Action: lekma lost in application types/interpreter types/python types...11:22
lekmafijal: is there some documentation I should read on how to handle the types in rpython?11:26
fijalI doubt11:28
fijalrffi.cast11:28
fijalis the way to cast them11:29
fijalno why do you need to do it?11:29
whyking (~quassel@p5B3DD930.dip.t-dialin.net) left irc: Ping timeout: 240 seconds11:30
lekmafijal: I think I got carried away a little bit in trying to understand how it works...11:30
fijalok11:31
lekmafijal: anyway the idea is to implementd PyErr_SetFromErrnoWithFilenameObject11:31
fijalfeel free to ask questions, but we lack documentation on the subject11:31
stakkars_ (~tismer@i59F74D8E.versanet.de) joined #pypy.11:31
fijalyes11:31
lekmaand use this to implement PyErr_SetFromErrnoWithFilename and PyErr_SetFromErrno11:31
fijalit's also undocumented11:32
fijalin CPython11:32
fijalI think it's the wrong way around11:32
fijalit's much easier to implement the latter two IMO11:32
fijalbecause you have to deal with PyObject less11:32
fijaland you can just write it in RPython11:33
fijalno?11:33
lekmayep (and they are already inplemented)11:33
lekma:)11:33
fijalok11:34
lekmaproble is one of the user of one of my extension wants to build it in pypy11:34
fijalmaybe you should not use undocumented API in your extension?11:34
fijal;-)11:34
lekmawell it is still an exported (and useful) function :-)11:36
kenaan12hager ppc-jit-backend 11e1dea1c15227 15/pypy/jit/backend/ppc/ppcgen/ppc_assembler.py: fix wrong initialisation of StackLocation in regalloc_push/regalloc_pop11:37
lekmaand as you probably know not every useful thing is documented in the cpython api11:38
fijalI actually don't 11:38
fijalbut I would expect :)11:38
fijalok11:38
lekmawell, anyway, thanks for your help11:41
anish_ (anish@nat/redhat/x-rvkkvrhvfmtyrqkn) left irc: Ping timeout: 276 seconds11:43
anish_ (anish@nat/redhat/x-bfibenxaaxyuuwzf) joined #pypy.11:43
witulski (~stupsi@fwstups.cs.uni-duesseldorf.de) joined #pypy.11:43
witulski (stupsi@fwstups.cs.uni-duesseldorf.de) left #pypy.11:44
stakkars_ (~tismer@i59F74D8E.versanet.de) left irc: Quit: hasta la vista11:47
mfoord (~anonymous@python/psf/voidspace) joined #pypy.11:47
anish_ (anish@nat/redhat/x-bfibenxaaxyuuwzf) left irc: Quit: Leaving11:51
stakkars_ (~tismer@i59F74D8E.versanet.de) joined #pypy.11:51
witulski (~stupsi@fwstups.cs.uni-duesseldorf.de) joined #pypy.11:55
witulski (stupsi@fwstups.cs.uni-duesseldorf.de) left #pypy.11:55
kenaan12hager ppc-jit-backend 1146750704d638 15/pypy/jit/backend/ppc/ppcgen/ppc_assembler.py: adjust _build_propagate_exception_path to new interface12:04
bbot24Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/1920 [12Timo Paulssen, strbuf_by_default]12:06
timotimooh my12:08
fijal:)12:08
timotimoi suppose now i'm responsible for making it work? :)12:08
fijalif you want to merge it yes :)12:09
fijalhttp://buildbot.pypy.org/summary/longrepr?testname=TestW_StdObjSpace.%28%29.test_fastpath_isinstance&builder=own-linux-x86-32&build=1920&mod=objspace.std.test.test_stdobjspace12:09
fijalthis one is not a bug, you can just fix the test12:09
fijaltwo others seem to be bugs12:09
sven_hager (~sven@fwstups.cs.uni-duesseldorf.de) joined #pypy.12:11
fijaloh nice, a JIT bug :/12:17
CIA-1703fijal 07roundup * 10#995/jit bug with threads: 12:19
CIA-17[new] On branch look-into-thread running x.py yields:12:19
CIA-17http://paste.pocoo.org/show/532768/ or sometimes a segfault * 14https://bugs.pypy.org/issue99512:19
Sinker (~thinker@123-194-52-90.dynamic.kbronet.com.tw) left irc: Remote host closed the connection12:20
tilgovi (~randall@couchdb/developer/tilgovi) left irc: Read error: Connection reset by peer12:21
nedbat (~nedbat@python/psf/nedbat) joined #pypy.12:25
thinker (~thinker@123-194-52-90.dynamic.kbronet.com.tw) joined #pypy.12:30
timotimofijal: without knowing more about pypys internals, i'm not sure how the assertion there should be fixed. should i add an or clause that checks if the class is an W_AbstractStringObject? or is the abstract one a superclass and issubclass should be used instead?12:44
fijaltimotimo: just change the check to W_AbstractStringObject and be happy12:44
timotimowill that break the test if one changes the objspace strbuf back to false in a local test or something?12:45
fijalno12:46
timotimoOK, thanks12:47
fijalit's a very specific whitebox test btw12:48
timotimoyeah, i kind of guessed that12:49
fijalAlex_Gaynor: ^^^ sudo review12:59
kenaan12fijal extradoc 1107cb0fa35b28 15/blog/draft/laplace.rst: add a draft12:59
kenaan12fijal extradoc 11642dcd49d458 15/: merge12:59
timotimofijal: in objspace/std/test/test_newformat.py on line 140 (and similar lines in the same class), could the error be caused by __format__ returning just an interp level string rather than something wrapped by self.s?13:01
timotimothat wouldn't explain, why the test worked before, though.13:01
aboudreault (~alanb@osgeo/member/aboudreault) joined #pypy.13:01
fijalno, that would break translation13:03
timotimoso the error is indeed not in the test, but somewhere buried in the depths of the strbuf implementation?13:03
fijalprobably13:03
fijalit looks like unicode/str casting does not work properly13:03
kenaan12timo_jbo strbuf_by_default 119014cd34145f 15/pypy/objspace/std/test/test_stdobjspace.py: fix whitebox test that checks for W_StringObject, rather than W_AbstractStringObject.13:03
fijaltimotimo: I can have a look when you get bored13:04
fijalanyway, break13:04
timotimohave fun :)13:05
timotimoindeed, in W_StringBufferObject i see no code to facilitate such a coercion13:06
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1190 [12timo, strbuf_by_default]13:06
fijalwell, magic :)13:06
fijalit coerces to string then to unicode13:06
fijalI bet13:06
gutworth (~benjamin@dhcp-128-36-84-221.central.yale.edu) joined #pypy.13:07
timotimopypy.module.pypyjit.test_pypy_c.test_string.TestString � this seems like it's very whiteboxy for the regular string object implementation and would have to be rewritten for strbuf, no?13:07
timotimohttp://buildbot.pypy.org/summary/longrepr?testname=TestString.%28%29.test_getattr_promote&builder=pypy-c-jit-linux-x86-32&build=1190&mod=pypy.module.pypyjit.test_pypy_c.test_string � here13:07
kenaan12bivab default 11e2f82a5d9f5e 15/pypy/rlib/clibffi.py: reintroduce changes done in b6390a34f261 to push_arg_as_ffiptr in clibffi.py, somehow lost in a731ffd298b413:11
kenaan12bivab ppc-jit-backend 11d094b25960ad 15/: merge default13:13
lekma (~lekma@a83-163-86-26.adsl.xs4all.nl) left irc: Quit: lekma13:15
nedbat (~nedbat@python/psf/nedbat) left irc: Disconnected by services13:18
nedbat_ (~nedbat@python/psf/nedbat) joined #pypy.13:19
Nick change: nedbat_ -> nedbat13:22
TheNoodle (~noodle@sarcasticsupport.com) joined #pypy.13:26
mikefclekma: i think 'moronic' was too strong a word.13:31
mikefclekma: just implementing an undocumented cpython api call in pypy seemed a bit .. um.. .weird13:32
nirbheek (~nirbheek@gentoo/developer/flyingspaghettimonster/nirbheek) left irc: Read error: Connection timed out13:32
nirbheek (~nirbheek@gentoo/developer/flyingspaghettimonster/nirbheek) joined #pypy.13:34
Rhy0lite (dje@nat/ibm/x-ajasoisxzgajnovr) joined #pypy.13:38
exarkunThe right thing to do is contribute documentation (and unit tests) for it to CPython13:38
exarkunAnd then implement it for PyPy13:38
bivab (david@fwstups.cs.uni-duesseldorf.de) left #pypy.13:39
bivab (~david@fwstups.cs.uni-duesseldorf.de) joined #pypy.13:39
DanielHolth (~dholth@2001:470:8:e7c:224:8cff:fe26:8c67) joined #pypy.13:54
PiotrSikora (~none@nginx/adept/piotrsikora) left irc: Excess Flood13:55
kenaan12stefanor default 11e8239b6167fa 15/pypy/doc/: Add pypy.1 manpage to sphinx docs13:56
PiotrSikora (~none@nginx/adept/piotrsikora) joined #pypy.13:57
whyking (~quassel@p4FFB644F.dip.t-dialin.net) joined #pypy.13:57
kenaan12stefanor default 112f90612495e2 15/pypy/doc/man/pypy.1.rst: pypy manpage: Format for multiple --jit arguments14:00
dmalcolm (~david@c-24-61-12-82.hsd1.ma.comcast.net) joined #pypy.14:01
CIA-1703stefanor 07roundup * 10#992/Manpage: [resolved] Applied in e8239b6167fa * 14https://bugs.pypy.org/issue99214:01
fprimex (~fprimex@brent-macbook.sc.fsu.edu) joined #pypy.14:15
stakkars_ (~tismer@i59F74D8E.versanet.de) left irc: Quit: schnarch14:24
lmoura (lmoura@gateway/indt/session) left irc: Ping timeout: 240 seconds14:28
espes (~espes@123-243-221-248.tpgi.com.au) left irc: Ping timeout: 240 seconds14:36
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/323 [12timo, strbuf_by_default]14:36
bivab_ (~david@fwstups.cs.uni-duesseldorf.de) joined #pypy.14:38
Trundle (~andy@python/site-packages/trundle) left irc: Ping timeout: 240 seconds14:40
lucian (~lucian@93-97-174-114.zone5.bethere.co.uk) left irc: Ping timeout: 240 seconds14:40
bivab (~david@fwstups.cs.uni-duesseldorf.de) left irc: Ping timeout: 240 seconds14:40
bgola (~bgola@189.100.41.65) left irc: Ping timeout: 240 seconds14:40
jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Ping timeout: 240 seconds14:40
G2P (~G2P@fw-asn1.ornis.com) left irc: Ping timeout: 240 seconds14:40
Nisstyre (~yours@c-208-90-102-250.netflash.net) left irc: Ping timeout: 240 seconds14:40
cwillu (~cwillu@cwillu.com) left irc: Ping timeout: 240 seconds14:40
EnCuKou (~encukou@ip-94-113-220-25.net.upcbroadband.cz) left irc: Ping timeout: 240 seconds14:40
Varriount (~Varriount@hnng.public-craft.com) left irc: Ping timeout: 240 seconds14:40
Nick change: bivab_ -> bivab14:40
jacob22 (~jacob@c-c4c4e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.14:40
xorAxAx (~alexander@moinmoin/coreteam/alexander) left irc: Ping timeout: 240 seconds14:40
Fluxid (~fluxid@schranz.io) left irc: Ping timeout: 240 seconds14:40
Fluxid (~fluxid@schranz.io) joined #pypy.14:40
G2P (~G2P@fw-asn1.ornis.com) joined #pypy.14:40
EnCuKou (~encukou@ip-94-113-220-25.net.upcbroadband.cz) joined #pypy.14:40
Nick change: Fluxid -> Guest9608414:40
lucian (~lucian@93-97-174-114.zone5.bethere.co.uk) joined #pypy.14:40
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds14:40
ronan (~ronan@host86-143-209-182.range86-143.btcentralplus.com) left irc: Ping timeout: 240 seconds14:40
_root_ (~Scalable@li252-14.members.linode.com) left irc: Ping timeout: 240 seconds14:40
cwillu (~cwillu@cwillu.com) joined #pypy.14:41
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) joined #pypy.14:41
Varriount (~Varriount@hnng.public-craft.com) joined #pypy.14:41
bgola (~bgola@189.100.41.65) joined #pypy.14:41
ronan (~ronan@host86-143-209-182.range86-143.btcentralplus.com) joined #pypy.14:41
Nisstyre (~yours@c-208-90-102-250.netflash.net) joined #pypy.14:42
_root_ (~Scalable@li252-14.members.linode.com) joined #pypy.14:42
xorAxAx (~alexander@moinmoin/coreteam/alexander) joined #pypy.14:42
espes (~espes@123-243-221-248.tpgi.com.au) joined #pypy.14:43
whitelynx|work (~whitelynx@63.241.75.144) joined #pypy.14:48
Alex_Gaynorfijal: you want me to review the blog post?14:54
Nick change: Guest96084 -> Fluxid215:04
rinu (~quassel@117.196.150.136) joined #pypy.15:05
nirbheek (~nirbheek@gentoo/developer/flyingspaghettimonster/nirbheek) left irc: Ping timeout: 240 seconds15:10
kenaan12hager ppc-jit-backend 11a51d6a2b3e1d 15/pypy/jit/backend/ppc/ppcgen/: adjust emit_guard_call_assembler and prepare_guard_call_assembler15:14
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) joined #pypy.15:16
arigato (~arigo@adsl-89-217-237-16.adslplus.ch) joined #pypy.15:22
dgl (~dgl@109.86.165.231) joined #pypy.15:25
aleksi (~aleksi@85.235.191.82) left irc: Ping timeout: 240 seconds15:28
mat^2 (~mathias@212.130.113.35) joined #pypy.15:28
aleksi (~aleksi@85.26.155.115) joined #pypy.15:29
aleksi (~aleksi@85.26.155.115) left irc: Ping timeout: 240 seconds15:37
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-win-x86-32/builds/542 [12timo, strbuf_by_default]15:41
rinu (~quassel@117.196.150.136) left irc: Remote host closed the connection15:48
kenaan12arigo default 1171d3d24c92d1 15/pypy/doc/extradoc.rst: Add two papers.15:54
xorAxAxhmm, arent HTM CPUs scheduled to be available this year?15:56
fijalAlex_Gaynor: yes15:56
fijalarigato: you feel like reviewing as well?15:56
fprimex (~fprimex@brent-macbook.sc.fsu.edu) left irc: Quit: http://www.fprimex.com15:57
fijalarigato: the numpy blog post to be precise15:58
Alex_Gaynorfijal: reviewing, I have tons of small changes ATM, so please don't push anything :)15:58
fijalok15:58
fijalAlex_Gaynor: I broke cast_int_to_float :/ fixed now15:59
Alex_Gaynorfijal: please don't do that :)15:59
Alex_Gaynorbreak it that is, we like when people fix things15:59
fijalyes15:59
fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) joined #pypy.16:03
kenaan12alex_gaynor extradoc 11a65754d300a3 15/blog/draft/laplace.rst: grammar changes all over16:05
lmoura (lmoura@gateway/indt/session) joined #pypy.16:05
Alex_Gaynorfijal: let me ask someone who isn't you or me to review this, to be sure it makes sense to the outside world16:06
fijalhow about armin?16:07
fijalor Rhy0lite 16:07
fijal?16:07
fijalI think Rhy0lite volunteered16:07
arigatoxorAxAx: HTM CPUs: this year?  I knew about next year16:07
fijalarigato: ?16:07
kenaan12alex_gaynor extradoc 115dc64fda0ea7 15/blog/draft/laplace.rst: another rewording16:08
arigatofijal: ok, I can review the numpy blog post16:08
fijalarigato: extradoc/blog/draft/laplace.rst16:08
fijalit's a bit of reading though16:08
arigato:-)16:08
arigatono problem16:08
xorAxAxarigato: haswell, in 2013, oops, indeed16:11
kenaan12stefanor default 110e67e4538c80 15/pypy/doc/Makefile: Rather use standard Sphinx 1.x target16:12
tumbleweedright, enough messing with manpages16:12
stakkars_ (~tismer@89.204.137.116) joined #pypy.16:13
fijaltumbleweed: pypy/doc/Makefile?16:16
fijalwhat's that supposed to do?16:17
amaury_ (amaury_@nat/google/x-ualhjpyadfbghfxb) joined #pypy.16:20
kenaan12hager ppc-jit-backend 11b5f5e48c3799 15/pypy/jit/backend/ppc/ppcgen/: (bivab, hager): StackLocations have now a value field which stores the offset to the SPP. It is used in reg...16:24
kenaan12hager ppc-jit-backend 11f04c600f8177 15/pypy/jit/backend/ppc/ppcgen/opassembler.py: (bivab, hager): we don't want to free the args here16:24
kenaan12arigo extradoc 11749fa78eeb73 15/blog/draft/laplace.rst: Rewrite to remove the emphasis on **per iteration** --- all the other numbers are also per iteration.16:24
arigatofijal: done reviewing.  it's missing the usual warning after "outperform C" at the end :-)16:24
fijal:)16:25
fijaldoes that mean we're uncareful or confident?16:25
kenaan12l.diekmann set-strategies 11498b6ee337e9 15/pypy/objspace/std/listobject.py: optimization fix16:25
fijalarigato: looking at assembler generated by C I'm confident we can outperform this particular example16:26
fijalsome time in near future16:26
arigatofijal: it means that you'll get as usual 14 replies in the blog comments attacking this sentence instead of focusing on what you're really saying16:26
kenaan12alex_gaynor extradoc 11fc2925740080 15/blog/draft/laplace.rst: less formal writing16:27
kenaan12alex_gaynor extradoc 1175aa1ba6d29f 15/blog/draft/laplace.rst: resolved merge16:27
fijalAlex_Gaynor: do you have opinions?16:27
Alex_Gaynorfijal: about beating C?16:27
arigatoit's fine if it's even true for once :-)16:27
fijalyes16:27
arigatoyou need to say it more precisely16:27
Alex_Gaynorlet me ask twitter :)16:27
mvt (~mvantelli@87.213.45.85) left irc: Quit: This computer has gone to sleep16:28
fijalarigato: ok, so precisely this loop can't be vectorized in C because C lacks a possibility to have aliasing information16:28
arigatoyes16:28
mattip (~chatzilla@bzq-79-180-115-39.red.bezeqint.net) joined #pypy.16:28
fijalwhile in our case, we already do track aliasing of arrays16:28
fijalmattip: hi16:28
fijalmagcius: what di you want me to do?16:28
arigatoand say that you're confident it can be faster than C after looking at the C result16:29
arigatoin assembler16:29
fijalok16:29
arigatoI think that it's an important point :-)16:29
fijalin this particular example16:29
arigatosure16:29
arigatoalso, C doesn't really lack a possibility to have aliasing information16:30
arigatothere is the __restrict keyword16:30
sven_hager (~sven@fwstups.cs.uni-duesseldorf.de) left irc: Quit: Leaving16:30
zain (~textual@66.175.91.194) joined #pypy.16:30
arigatoI never really understood it, and I guess I'm not alone, but I think it lets you somehow say that two vectors don't alias each other16:30
kenaan12fijal extradoc 11ad6f9cb35d27 15/blog/draft/laplace.rst: quantify "faster than C"16:31
fijalarigato: well16:31
fijalarigato: you would need to write by hand an if16:31
Alex_Gaynorno, there's the restrict keyword16:31
fijalbecause we track whether arrays are the same or different16:31
fijaland compile different assembler based on that16:31
arigatoindeed16:31
arigatojust saying that it should be possible to get this result in C too, with enough knowledge about C gore16:32
craigkerstiens (~craigkers@c-76-126-162-38.hsd1.ca.comcast.net) left irc: Quit: craigkerstiens16:32
arigatobut maybe that's also "in this particular example"16:32
fijalwell, gcc does not do it "right now"16:32
fijalI'm sure it's possible16:32
fijaland maybe even there is a magic incantation of gcc options16:33
arigatoyes, I think "gcc does not do it right now" is also part of the point16:33
fijalbut as far as I remember this is one of the reasons why people use fortran16:33
arigatoright16:33
fijalit's probably "hard"16:33
fijalanyway, how do you like the wording now?16:33
fijalI would like to avoid too many quantifiers16:33
fijalespecially because it's clear we can't predict all future C compilers for example16:33
nedbat (~nedbat@python/psf/nedbat) left irc: Disconnected by services16:34
nedbat_ (~nedbat@python/psf/nedbat) joined #pypy.16:34
nedbat (~nedbat@python/psf/nedbat) joined #pypy.16:35
kenaan12fijal extradoc 111f530d01ba87 15/blog/draft/laplace.rst: add link16:35
nedbat_ (~nedbat@python/psf/nedbat) left irc: Read error: Connection reset by peer16:35
zain (~textual@66.175.91.194) left irc: Quit: Computer has gone to sleep.16:35
kenaan12arigo extradoc 119566e67df82c 15/blog/draft/laplace.rst: wording.16:36
arigatook16:36
fijaltumbleweed: not a huge fan of Makefile :/16:36
fijalarigato, Alex_Gaynor: go?16:37
Alex_Gaynorfijal: err, one sec16:37
fijalok16:37
fijalok, so I'll post it in some time, going off for a bit16:37
fijalRhy0lite: feel free to review as well16:37
fijalRhyolite: you too :)16:37
JaRoel|4d (~jaroel|4d@office.fourdigits.nl) left irc: Remote host closed the connection16:39
Alex_Gaynorfijal: ok, good by me whenever16:39
amaury_ (amaury_@nat/google/x-ualhjpyadfbghfxb) left irc: Ping timeout: 252 seconds16:41
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds16:43
bivab (~david@fwstups.cs.uni-duesseldorf.de) left irc: Quit: bivab16:49
Nick change: Fluxid2 -> Fluxid16:49
kenaan12arigo extradoc 1124ad6171712f 15/blog/draft/laplace.rst: typo16:53
G2P (~G2P@fw-asn1.ornis.com) left irc: Quit: Leaving.17:01
fzzzy (~donovan@76-198-130-19.lightspeed.mtvwca.sbcglobal.net) left irc: Quit: fzzzy17:07
JaRoel|4d (~jaroel|4d@82.161.43.49) joined #pypy.17:10
JaRoel|4d (~jaroel|4d@82.161.43.49) left irc: Ping timeout: 240 seconds17:14
idnarhey, should I reopen https://bugs.pypy.org/issue809 (and how do I do that?) or should I file a new bug report?17:16
idnarI guess it's a different bug17:16
idnarbut the script attached to that issue serves to demonstrate the issue17:17
arigatoyou can reopen the bug just by writing more17:17
arigatoeither choice is fine17:18
idnarI think I'll open a new bug, this seems to be unrelated17:18
idnarCREATE TABLE fails with "OperationalError: table foo already exists" now17:18
arigato:-/17:19
JaRoel|4d (~jaroel|4d@2001:980:7133:1:c838:5fe1:5f9c:b4de) joined #pypy.17:20
CIA-1703mithrandi 07roundup * 10#996/sqlite3 CREATE TABLE always fails: 17:20
CIA-17[new] Attaching a script to demonstrate the issue; the CREATE TABLE fails with "table17:20
CIA-17already exists" in an empty database. * 14https://bugs.pypy.org/issue99617:20
idnaroops17:21
idnarattached the wrong script, but it doesn't really matter17:21
fijal (~fijal@196-215-121-137.dynamic.isadsl.co.za) joined #pypy.17:24
arigatothanks17:26
arigatofor reporting it17:26
lucian (~lucian@93-97-174-114.zone5.bethere.co.uk) left irc: Ping timeout: 240 seconds17:32
fzzzy (~donovan@nat/mozilla/x-brrizadwoewrtlvu) joined #pypy.17:34
gmcabrita (gmc@slackwa.re) joined #pypy.17:39
mattip (~chatzilla@bzq-79-180-115-39.red.bezeqint.net) left irc: Ping timeout: 240 seconds17:42
mattip (~chatzilla@bzq-109-67-122-241.red.bezeqint.net) joined #pypy.17:43
craigkerstiens (~craigkers@204.14.152.118) joined #pypy.17:44
Rhy0litefijal: pong17:45
amaury_ (amaury_@nat/google/x-hhhqndsdxagumaay) joined #pypy.17:49
Arfrever (~Arfrever@apache/committer/Arfrever) joined #pypy.17:56
mattip (chatzilla@bzq-109-67-122-241.red.bezeqint.net) left #pypy.18:01
fijal (~fijal@196-215-121-137.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds18:03
Kaskuka (~John@osbk-4db150f7.pool.mediaWays.net) left irc: Read error: Connection reset by peer18:10
Kaskuka (~John@osbk-4db150f7.pool.mediaWays.net) joined #pypy.18:11
kennethreitz (~kennethre@204.14.152.118) joined #pypy.18:19
kenaan12edelsohn extradoc 110d508d74845b 15/blog/draft/laplace.rst: English language cleanups.18:22
gehho2 (~gehho2@ZJ040192.ppp.dion.ne.jp) left irc: Quit: Leaving...18:28
mfoord (~anonymous@python/psf/voidspace) left irc: Quit: mfoord18:38
kenaan12edelsohn extradoc 1115a3491e715a 15/blog/draft/laplace.rst: More English improvements and a few commas.18:45
`fox` (~fox@host130-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.18:49
amaury_ (amaury_@nat/google/x-hhhqndsdxagumaay) left irc: Ping timeout: 268 seconds18:50
asmeurer (~asmeurer@c-174-56-21-245.hsd1.nm.comcast.net) left irc: Ping timeout: 252 seconds18:51
stakkars_ (~tismer@89.204.137.116) left irc: Quit: schnarch18:52
asmeurer (~asmeurer@c-174-56-21-245.hsd1.nm.comcast.net) joined #pypy.19:02
amaury_ (amaury_@nat/google/x-foviueauuirfwitl) joined #pypy.19:05
mattip (~chatzilla@bzq-109-67-122-241.red.bezeqint.net) joined #pypy.19:06
nickik (~nick@217.150.234.222) joined #pypy.19:12
tlynn (~tlynn@212.69.41.1) joined #pypy.19:14
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) joined #pypy.19:14
fijalRhy0lite: pong19:15
Rhy0litefijal: done19:15
fijalRhy0lite: cool thanks19:16
fijalidnar: indeed19:16
tlynnwould I be right to assume that proper stackless + jit isn't planned for a long time, if at all?19:19
kenaan12fijal default 11cc4956b9891d 15/pypy/doc/coding-guide.rst: add a note about special methods19:19
kenaan12fijal default 116b798036894a 15/pypy/: merge19:19
fijaltlynn: depends what you mean by "planned" :)19:19
fijalAlex_Gaynor: ping?19:20
CIA-1703mithrandi 07roundup * 10#997/PyCrypto fails to build with pypy/cpyext: 19:23
CIA-17[new] PyCrypto is an (optional) Twisted dependency. I'll attach the build log19:23
CIA-17demonstrating the failure. * 14https://bugs.pypy.org/issue99719:23
`fox` (~fox@host130-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Ping timeout: 260 seconds19:26
tumbleweedfijal: nobody requires you to use the makefile. I added (in the last commit) the same rules that a more recent sphinx-quickstart would have added19:29
fijalI guess it's fine19:29
tlynn (~tlynn@212.69.41.1) left irc: Quit: Leaving19:30
fijalRhy0lite, arigato, Alex_Gaynor: published19:31
Rhy0litegreat19:31
Action: tumbleweed tripled the RAM in my machine today. Maybe I can now translate pypy and run a web browser at the same time :)19:32
exarkunas long as you don't open too many web pages19:32
Rhy0litetumbleweed: so demanding! ;-)19:32
tumbleweedyeah, give it a week. and chromium will realise there's another 16G if ram it could use...19:33
Rhy0litefijal: I hope that you are happy with the way it reads now19:35
fijalI think some details are buried19:36
fijalbut on the other hand I'm not really looking for people who won't read the whole thing19:36
fijal(not this time)19:36
Rhy0liteyes, you are waiting until you can say that NumPyPy crushes all other implementations19:37
amaury_ (amaury_@nat/google/x-foviueauuirfwitl) left irc: Ping timeout: 268 seconds19:37
fijalpretty much19:38
Rhy0lite:-)19:38
Alex_GaynorRhy0lite: we'd even settlefor "beats all other implmenations by a tiny bit"19:38
Rhy0liteAlex_Gaynor: I think you and Maciej have different goals :-)19:39
Action: Alex_Gaynor likes winning, but is flexible on the margin19:39
Rhy0litewithout automatic vectorization, I think PyPy needs some sort of "inlined assembly" feature like GCC to really kick ass on vectorizable numerical workloads19:40
Alex_Gaynorwell, automatic vectorization is on the horizon19:41
fprimex (~fprimex@brent-macbook.sc.fsu.edu) joined #pypy.19:42
fijalRhy0lite: it's not very hard for us to do SSE19:42
fijalI mean it's annoying, but not very hard19:42
exarkunmmmmmm19:42
fijalexarkun: ???19:43
exarkunautomatic vectorization19:43
exarkuntastes delicious19:43
Alex_Gaynorfijal: we need computation in the JIT on small sizes to get the most out of it though, I think, int8 + int8, you can have more iterations in vectorization for int8 vs WORD19:44
fijalAlex_Gaynor: probably19:44
fijalAlex_Gaynor: but you can cheat and have only very partial support for that in the JIT19:44
Alex_Gaynorfijal: actually, you can't vectorize non-WORD things at all, ATM19:45
fijalright19:46
etrepum (~bob@75-101-96-144.dsl.static.sonic.net) left irc: Quit: etrepum19:46
fijalwell19:46
fijalfloats19:46
Rhy0liteexarkun: and less filling!19:46
Alex_Gaynordoubles only, I guess19:46
antocuni (~antocuni@host157-123-dynamic.2-87-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds19:48
ericflo (~ericflo@75.103.8.110) joined #pypy.19:51
Trundle (~andy@python/site-packages/trundle) joined #pypy.19:51
kenaan12fijal default 115f2580ed4505 15/pypy/tool/release_dates.py: remove a horribly outdated script19:54
arigatofijal: re blog post:19:58
arigato"Word of warning: this work isn't done"19:58
arigatoI guess you don't mean "what we present below does not come from actual measures"19:58
arigatobut rather "this is work in progress"19:59
fijalok19:59
Rhy0litearigato: that's how I interpreted it19:59
Rhy0liteisn't complete19:59
arigatoah, ok.  sometimes things are confusing for native speakers, sometimes for non-native ones19:59
fijalI changed to in-progress20:00
Rhy0liteyeah.  As I have told Maciej in the past, I don't want to change his style too much.20:00
Rhy0liteThe post should sound like him.20:00
Rhy0liteonly better :-)20:00
Alex_GaynorRhy0lite: I already changed the style around a bunch to make it more casual :)20:00
fijalweb2py and web.py are two different things right?20:01
Alex_Gaynoryes20:01
Alex_Gaynorweb2py is the awful one20:02
Rhy0liteI mainly try to fix American English idioms that aren't used correctly20:02
fijalAlex_Gaynor: the one that uses execfile?20:02
Alex_Gaynorfijal: yes20:02
fijalmeh :/20:02
fijalit got some sort of award20:02
Alex_Gaynorjust ignore me20:03
fijal?20:03
Alex_Gaynorit*20:03
Alex_Gaynorignore it20:03
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) joined #pypy.20:03
Rhy0liteAlex_Gaynor: yes, casual but intelligent tone.20:03
Rhy0liteI'm probably more pedantic about English grammar than the rest of the people in the project20:04
Rhy0litebut I try to control myself when editing other people's writing20:04
fijalthat's only because mwhudson does not work with us any more :)20:04
fijalarigato: have you seen that btw: http://www.chipworks.com/en/newsroom/silicon-art-library/20:05
fijalhttp://www.chipworks.com/en/newsroom/silicon-art-library/silicon-art-gallery-120:06
mwhudson:)20:09
etrepum (~bob@accessnat4.mochimedia.net) joined #pypy.20:16
arigatofijal: :-)20:21
fermianyon (~lane@c-71-229-21-197.hsd1.al.comcast.net) joined #pypy.20:22
kenaan12fijal default 11425ed28a95d9 15/: update LICENSE file. For future readers, run:  hg churn -c -t "{author}" | sed -e 's/ <.*//' | sed -e 's/[0-9][0-9]...20:23
fijalok20:23
kenaan12fijal default 11ca3f367e84af 15/pypy/doc/tool/makecontributor.py: remove now completely useless file20:23
Alex_Gaynorfijal: enjoy: http://devblog.joindiaspora.com/2012/01/10/client-side-re-write-oh-my/20:27
amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy.20:28
fijalnice20:28
fijalpeople are already complaining20:29
Rhy0litefijal: there's always a critic20:29
mikefcfijal: nice benchmark report20:32
mikefc"... but why doesn't numpypy do THAT_THING_THAT_I_WANT yet?"20:32
mikefcpeople are complainers20:32
fijal"why did you make functions that *I* need not my priority"20:35
fijalin fact, it would be cool to document what we implemented20:36
fijalAlex_Gaynor: feel like writing about it?20:36
Alex_Gaynormaybe later, not today20:36
mikefc"I'm not donating any money or effort, but why isn't it done for me right NOW"20:36
fijalAlex_Gaynor: we need to start tutorial one day20:38
fijalamaury_: ping20:40
amaury_hi20:40
fijalamaury_: is ob_size public API?20:40
amaury_hum, no20:41
amaury_Py_SIZE()20:41
fijalcan you comment on the tracker?20:41
fijalhttps://bugs.pypy.org/issue99720:41
amaury_I've read id20:42
amaury_ob_digits is certainly not supported20:42
amaury_and won't20:42
Trundle (~andy@python/site-packages/trundle) left irc: Ping timeout: 240 seconds20:48
mattipfijal: ping20:49
Rhy0lite (dje@nat/ibm/x-ajasoisxzgajnovr) left irc: Quit: Leaving20:50
mattipI feel like I'm kind of drifting off20:51
kenaan12mattip numpypy-axisops 1120bbff5d323d 15/pypy/module/micronumpy/: fixed mean, added funky tests in ReduceSignature20:51
exarkunDo you have a peaceful, easy feeling mattip?20:51
aboudreault (~alanb@osgeo/member/aboudreault) left irc: Quit: Leaving20:51
mattipit would be nice if someone could take a look at the numpypy-axisops branch and bring me back to reality20:52
mattipexarkun: actually I have an uneasy feeling I will need to start all over when someone sees what I have done ;)20:52
mattipIt all passes tests (well, until I write more complicated tests)20:53
mattipbut is this really the Correct Way?20:53
fijalmattip: pong20:54
Alex_Gaynormattip: I'l try to make some time to review it20:54
fijalI can make some time20:56
fijalmattip: so, what bothers you?20:56
mattipfirst off, the zjit tets I wrote still has many ops in the inner loop.20:59
mattipis there a better way than staring at the --viewloops to clean it up?21:00
mattipsecond, I'm not sure I'm using the infrastructure (signatures, virtual arrays) properly.21:00
mattipI have a feeling ReduceSignature.eval() should be doing more21:01
fijalwhat do you mean "a better way"?21:01
fijalbetter like what?21:01
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) left irc: Quit: This computer has gone to sleep21:01
mattipand third, I just committed some code I'm not proud of but needed it to solve a problem.21:01
lizardo (~lizardo@189.2.128.130) left irc: Quit: Leaving21:02
kennethreitz (~kennethre@204.14.152.118) left irc: Quit: Computer has gone to sleep.21:02
mattipbetter like you do it for me :) (just kidding)21:03
fijalthe mean code is bogus21:03
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) joined #pypy.21:03
fijalyou can check21:03
fijalspace.is_true(w_dim, space.w_None)21:03
fijalfor checks against none21:03
Alex_Gaynorfijal: is_w21:04
fijalyeah, is_w21:04
fijalsorry21:04
tilgovi (~randall@75-149-58-169-SFBA.hfc.comcastbusiness.net) joined #pypy.21:04
tilgovi (~randall@75-149-58-169-SFBA.hfc.comcastbusiness.net) left irc: Changing host21:04
tilgovi (~randall@couchdb/developer/tilgovi) joined #pypy.21:04
fijalmattip: also, signature represents compiled assembler21:04
fijalyou can't do crazy stuff in signature.eval21:04
fijalit should not do any runtime checks21:05
fijalit should be different signature if it has to do something different21:05
mattipok. That's the kind of help I need. Keep going21:05
fijalso no ifs on arrays in signature.eval21:07
fijalobviously then you don't need an if in invent_array_numbering21:08
fijalcan the class be name AxisReduce?21:09
fijalto avoid confusion21:09
fijalmattip: look, the entire ReduceSignature is made of giant ifs21:10
fijalwhy not make 2 classes?21:10
fijalthere is definitely functionality for 2 classes21:11
fijalthey don't even intersect21:11
`fox` (~fox@host130-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.21:13
lambacck (~chris@d24-150-124-118.home.cgocable.net) left irc: Ping timeout: 240 seconds21:22
kennethreitz (~kennethre@204.14.152.118) joined #pypy.21:24
mikefcI want to add numpy.testing to numpypy (at applevel).  But if I just copy the numpy/testing dir into lib_pypy/numpypy/testing I can't actually access the module "numpypy.testing" until I do a translate?  Is this correct?  What am I doing wrong here?]21:27
mikefcor do i have to do something like:  python ~/pypy/bin/py.py -S --withmod-micronumpy pytest.py ./lib_pypy/numpypy/testing/tests21:28
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds21:29
mattipfijal: thanks. That was the kick I needed to straighten it out. I will seperate into two.21:30
Guest___ (~textual@63-228-127-70.dia.static.qwest.net) joined #pypy.21:31
Nick change: Guest___ -> rekamso21:31
mattipI need to think about how I got so far away from the original ReduceSignature in the first place.21:34
arigato (~arigo@adsl-89-217-237-16.adslplus.ch) left irc: Quit: See you21:36
mattip"the next epiphany is just around the corner, after a good night's sleep."21:36
mikefc"during" ? :)21:39
`fox` (~fox@host130-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds21:39
etrepum (~bob@accessnat4.mochimedia.net) left irc: Quit: etrepum21:40
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) joined #pypy.21:41
etrepum (~bob@accessnat4.mochimedia.net) joined #pypy.21:42
fijalmikefc: you can access21:42
fijalmikefc: even more, you can just use compiled pypy-c and modify it21:42
mikefcahh. that makes sense. so i can develop in a compiled pypy-nightly, and then copy things across into the mercurial tree?21:42
Action: mikefc slaps forehead21:42
fijalyou can just pick up pypy-c from nightly21:43
fijaland put it in say <your-pypy-checkout>/compiled21:43
fijalit'll pick up the lib correctly21:43
mikefcok. trying..21:44
mikefcsweet sangria of santa maria!21:45
mikefcthanks fijal21:45
`fox` (~fox@host130-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.21:49
mattip (~chatzilla@bzq-109-67-122-241.red.bezeqint.net) left irc: Quit: ChatZilla 0.9.88 [Firefox 8.0/20111115183158]21:49
kennethreitz (~kennethre@204.14.152.118) left irc: Quit: Computer has gone to sleep.22:00
AbeEstrada (~AbeEstrad@187.158.153.110) joined #pypy.22:01
AbeEstrada (AbeEstrad@187.158.153.110) left #pypy.22:02
kennethreitz (~kennethre@204.14.152.118) joined #pypy.22:04
fprimex (~fprimex@brent-macbook.sc.fsu.edu) left irc: Quit: home!22:14
jae (~jae@dedicated.jaerhard.com) left irc: Read error: Operation timed out22:19
CIA-1703afa 07roundup * 10#997/PyCrypto fails to build with pypy/cpyext: 22:20
CIA-17[chatting] Unfortunately PyCrypto accesses the internal representation of PyLong objects, which pypy cannot expose. This part of22:20
CIA-17PyCrypto (in _ ... * 14https://bugs.pypy.org/issue99722:20
jae (~jae@dedicated.jaerhard.com) joined #pypy.22:23
`fox` (~fox@host130-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Ping timeout: 260 seconds22:32
mikefc_ (9665bcc5@gateway/web/freenode/ip.150.101.188.197) joined #pypy.22:33
Dulak (~michael@unaffiliated/dulak) left irc: Quit: Leaving22:42
whyking (~quassel@p4FFB644F.dip.t-dialin.net) left irc: Ping timeout: 240 seconds22:48
nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 276 seconds22:51
nickik (~nick@217.150.234.222) left irc: Quit: Verlassend22:58
bbot2Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/20023:00
bbot2Started: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/77223:00
bbot2Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/100923:00
bbot2Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64-2/builds/2823:00
cjd (~cjd@cpc10-belf9-2-0-cust714.2-1.cable.virginmedia.com) joined #pypy.23:13
Nick change: cjd -> Guest4426323:14
Guest44263Hi. I have some code I'd like to run with pypy if installed but fallback to cPython otherwise.  Is there a standard pattern for that?23:16
fijalGuest44263: not that I know about?23:16
asmeurer (~asmeurer@c-174-56-21-245.hsd1.nm.comcast.net) left irc: Quit: asmeurer23:17
Guest44263 (~cjd@cpc10-belf9-2-0-cust714.2-1.cable.virginmedia.com) left irc: Quit: IRC webchat at http://irc2go.com/23:23
mikefc_            f = sys._getframe(1)             package_path = f.f_locals.get('__file__', None)             assert package_path is not None23:28
mikefc_how different is pypy's _getframe() from cpython?23:28
fijalnot different23:29
whitelynx|work (~whitelynx@63.241.75.144) left irc: Quit: Ex-Chat23:29
mikefc_this numpy.testing assert is failing.  so the frame has no __file__ 23:29
mikefc_i'm just going to comment it out and see what happens :)23:30
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) left irc: Quit: This computer has gone to sleep23:31
mikefc_hmm. seems to be some fancy way of snooping if scipy is installed.23:31
mikefc_which seems a really bad thing to do if we're in numpy/testing/tests23:31
fenrrir (~rodrigo@189.70.232.240) joined #pypy.23:32
fijalpfff23:32
fijalmikefc_: I don't want to know ;-)23:32
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) joined #pypy.23:32
mikefc_ugh. the testing infrastructure for numpy (numpy.testing) has tests.  these tests test using some basic numpy things we don't have yet. so i can't validate the numpypy testing stuff works.23:43
mikefc_i.e. the testing stuff relies on the things that it is testing already working.23:44
fijalright23:48
fijalmikefc_: like what?23:48
fijalmikefc_: I can make this my priority if you want23:49
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) left irc: Quit: This computer has gone to sleep23:51
fijalmikefc_: ?23:52
mikefc_lots of things. some little (NZERO, PZERO), some bigger (astype, finfo, np.random)23:52
mikefc_sorry, was trawling through the code23:52
fijalmikefc_: if you provide me some summary what exactly and how much of the functionality I can spend a day doing that23:53
mikefc_np.nan np.float23:53
fijaljust send me a mail and I'm off to bed23:53
fijalmikefc_: deal?23:53
mikefc_fijal: sure.  thanks.  I'll try and do some of the easier stuff if I can.23:54
jterrace (~Adium@dynamic-oit-vapornet-b-618.Princeton.EDU) joined #pypy.23:54
mikefc_numpy testing also requires 'nose'23:54
fijalmikefc_: py.test should work23:54
fijalmikefc_: or we can tweak it23:54
amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 248 seconds23:55
mikefc_ok.  I'll hopefully get more of a chance to look tonight.  have to do actual work now :)23:55
fzzzy (~donovan@nat/mozilla/x-brrizadwoewrtlvu) left irc: Quit: fzzzy23:56
kennethreitz (~kennethre@204.14.152.118) left irc: Quit: Textual IRC Client: http://www.textualapp.com/23:59
fijal (~fijal@196-215-5-11.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds23:59
--- Wed Jan 11 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!