#pypy IRC log for Monday, 2010-11-22

Ademan (~dan@ left irc: Quit: leaving00:02
Trundle (~andy@python/site-packages/trundle) left irc: Remote host closed the connection00:07
Scorchin (~Scorchin@host86-163-124-29.range86-163.btcentralplus.com) left irc: Quit: Scorchin00:17
Ademan (~dan@63.sub-72-109-145.myvzw.com) joined #pypy.00:33
stakkars (~tismer@i577B525B.versanet.de) left irc: Ping timeout: 240 seconds00:33
kenaan03dan 10r7932614 pypy/branch/psycopg2compatibility/pypy 05A(call_logger.py) 07M(api.py intobject.py pyobject.py pystate.py stubs.py test_pystate.py): Importing my changes from before. ...00:35
Ademan (~dan@63.sub-72-109-145.myvzw.com) left irc: Read error: Connection reset by peer00:37
stakkars (~tismer@i577B6F8F.versanet.de) joined #pypy.00:40
Ademan (~dan@63.sub-72-109-145.myvzw.com) joined #pypy.00:42
Ademan (~dan@63.sub-72-109-145.myvzw.com) left irc: Read error: Connection reset by peer01:05
savio (~thc@189001130056.usr.predialnet.com.br) joined #pypy.01:23
savio (~thc@189001130056.usr.predialnet.com.br) left irc: Remote host closed the connection01:50
savio (~thc@189001130056.usr.predialnet.com.br) joined #pypy.01:53
gutworth (~benjamin@64-131-30-120.usfamily.net) left irc: Ping timeout: 276 seconds02:03
gutworth (~benjamin@64-131-30-120.usfamily.net) joined #pypy.02:04
savio (~thc@189001130056.usr.predialnet.com.br) left irc: Ping timeout: 250 seconds02:52
kursor (~kursor@p57BB0B41.dip0.t-ipconnect.de) joined #pypy.03:05
Alex_Gaynorgutworth: no, no I am not a Church fan.03:12
gutworthor mathematician at heart?03:15
Alex_GaynorIt's not even the semantics, its the fact that the syntax is so obtuse, completely obscures the meanning.03:17
gutworthmathematicians love obscure syntax03:29
mwhudsongoogling for "church language" sure finds a bizarre bunch of stuff03:36
Alex_Gaynormwhudson: lambda calculus03:42
mwhudson (~mwh@canonical/launchpad/mwhudson) left irc: Ping timeout: 276 seconds03:43
panni_ (hannes@ip-178-203-101-205.unitymediagroup.de) left irc: Quit: ( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )03:45
gutworthI wouldn't complain. I have to write a paper on webern...03:56
Ademan-remote (~dan@63.sub-72-109-145.myvzw.com) joined #pypy.04:06
Ademan-remoteis we_are_translated() true at import time during translation?04:06
Ademan-remoteor only in a translated PyPy?04:07
htoothrot (~mux@66-169-185-121.dhcp.ftwo.tx.charter.com) joined #pypy.04:09
gutworthimport time is never translated04:10
linq (~c@bas1-kitchener08-1176182993.dsl.bell.ca) joined #pypy.04:10
santagada (~anonymous@ left irc: Quit: santagada04:12
Ademan-remoteof course but it would be useful to know if we are about to/are being translated, so I wasn't sure if we_are_translated() was true then too...04:12
Ademan-remote(despite the name)04:13
Ademan-remoteI don't suppose you know of a function like that do you?04:21
gutworthyou can't know04:25
Ademan-remotewhy not? at worst  if __main__.__name__ == 'translate', and surely you could be more accurate than that04:36
gutworthyou shouldn't decide things like that at module level04:38
exarkunwhy do you even want that information?04:42
gutworth (~benjamin@64-131-30-120.usfamily.net) left irc: Ping timeout: 252 seconds04:58
Ademan-remoteexarkun: I have a decorator I want to pass through if we're being translated05:03
exarkunAdeman-remote: That doesn't really answer "why"05:05
exarkun"Why are you drilling a hole in your grandmother's head?"  "So that there is a way for the blood to come out."05:05
Ademan-remoteexarkun: because the decorator uses **args and that doesn't translate...05:07
exarkunLots of things don't translate.05:09
exarkunUsually the answer is "don't do that".05:09
exarkunWhat's special about this?05:09
Ademan-remotebecause it drastically improves the debugability of CPyExt which is otherwise a total bear...05:14
Ademan-remoteI just eliminated the kwargs though, when used with cpyext05:15
Ademan-remote (~dan@63.sub-72-109-145.myvzw.com) left irc: Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org05:19
Ademan (~dan@63.sub-72-109-145.myvzw.com) joined #pypy.05:21
nettok (~quassel@ joined #pypy.05:27
mwhudson (~mwh@canonical/launchpad/mwhudson) joined #pypy.05:33
kursor (~kursor@p57BB0B41.dip0.t-ipconnect.de) left irc: Quit: kursor05:49
kursor (~kursor@p57BB0B41.dip0.t-ipconnect.de) joined #pypy.05:59
kursor (~kursor@p57BB0B41.dip0.t-ipconnect.de) left irc: Client Quit06:00
kursor (~kursor@p57BB0B41.dip0.t-ipconnect.de) joined #pypy.06:01
kursor (~kursor@p57BB0B41.dip0.t-ipconnect.de) left irc: Quit: kursor06:14
Ademanthat was a lot of work to rpython-ize that...06:22
fijal (~fijal@41-134-41-226.dsl.mweb.co.za) joined #pypy.06:30
fijalmwhudson: ping06:34
mwhudsonfijal: hi06:35
fijalmwhudson: do you want a short explanation of memory consumption?06:35
fijaldo you know how pypy handles cycles with __del__s?06:36
mwhudsonfijal: please :)06:36
mwhudsonuh... i forget06:36
fijalessentially it arbitrarily breaks cycles and then peals one layer of dels at a collect time06:36
fijalnow we collect not very often (400 major collects per translation run)06:37
fijalwhich in itself eats a bit of memory (we should think about trading this somehow)06:37
fijalbut also that means that cycles with dels get collected slowly06:37
fijaland generators used to introduce cycles with dels06:37
fijal(we also consider interp-level dels the same way as app-level)06:38
Alex_Gaynorisn't app-level del just implemented as a base object class who's __del__ calls the app level one?06:38
mwhudsonfijal: collecting less often results in not using so much memory?06:38
fijalAlex_Gaynor: it's slightly more complex06:39
fijalmwhudson: eh, no06:39
fijalcollecting less often means using more memory06:39
fijalbut also collecting less often means dels gets collected slowly06:39
mwhudsonah right06:39
fijaland C generation is using generators06:39
mwhudsonso part of the fix was to collect more often?06:40
fijalnow they don't introduce cycles, which make them disappear faster06:40
fijalno, we didn't get to that part yet :)06:40
mwhudsonoh, so just not creating many cycles was enough06:40
fijalprecisely with generators06:40
mwhudsonmust have been a surprising result :)06:41
fijalyes, it's more by chance06:41
[mathias] (~mathias@ left irc: 06:41
mwhudsondel + cycles is just an evil combination06:42
fijalwell, we're suffering slightly more06:42
fijalsince our interp-level objects can have dels and cycles06:43
fijalabout collecting more often - right now we have a set of 1.82 memory eaten by the GC of live objects06:43
mwhudsonit evil in any situation06:43
fijalthat's not a lot if heap is small06:43
fijalbut a lot if it's large06:43
mwhudsonmaybe slightly more evil in pypy-c than cpython, but still a difference of degrees06:43
mwhudsonfijal: sorry, i don't understand "right now we have a set of 1.82 memory eaten by the GC of live objects" at all06:44
fijalif the GC sees that we have 1G of objects alive06:45
fijalit'll schedule next collection at 1.82G06:45
fijalwhich is a bit of a waste if your heap is large06:47
fijalmwhudson: do I make any sense before 1st coffee?06:49
mwhudsonah right06:49
mwhudsonyes, that makes sense06:49
mwhudsoni am given to understand that tuning this sort of parameter is still a bit of a black art in the java world06:50
exarkunwhy 1.82?  Why not 1.81 or 1.83?06:52
Ademanit seems I spoke too soon, still not rpython...06:53
Ademanis print >>sys.stderr, astring not rpython? (if so what should I do instead?)06:54
exarkunwell, + "\n" perhaps06:56
Alex_Gaynordoes print not flush?06:57
fijalAlex_Gaynor: if your stdout is a terminal it's line-buffered by default06:58
fijalbut has nothing to do with print06:58
Alex_Gaynorright, I always assumed pritn also called flush06:58
fijalexarkun: my point is that any constant parameter would be wrong06:59
Action: exarkun nods06:59
fijaland it *is* black art06:59
fijalmwhudson: I claim we have different approach then java where it is "eat all your RAM for performance, buy more servers"06:59
fijalwhat are escaping rules for href targets in html?07:03
Alex_Gaynorfijal: same as for everything else, &, <, >, ', "07:03
fijaldo I have to escape &07:03
exarkunthe rules _are_ the same as for everything else07:04
exarkunbut the rules are "if you're doing it yourself, you're doing it wrong"07:04
kenaan03fijal 10r7932714 07M(pypy/build/bot2/pypybuildbot/pypylist.py): Properly escape URLs ...07:05
fijalexarkun: my question was more like "are there any rules for hrefs?"07:05
fijalwhy it's a hand crafted html remains to be discussed07:05
mwhudsonfijal: you could make yourself sound more sophisticated by asking what the rules are for attribute values in general :-)07:06
fijalsurprisingly enough, pypy devs are quite backward when it comes to we technology07:07
Ademanhrm no stderr.write() blew up similarly http://paste.pocoo.org/show/294463/ you can see it's referring to the result from the stderr.write()07:13
fijalAdeman: you're mixing types somehow07:14
fijalare you sure you don't pass a combination of string and low-level string?07:15
GasbaKid (~GasbaKid@ joined #pypy.07:15
fijalbtw, there is always debug_print which does stuff :)07:15
fijalthere is an RPython facility for logging07:15
enkrav (~enkrav@77-58-146-192.dclient.hispeed.ch) joined #pypy.07:15
Ademanfijal: I'll have to take a look at that... haha07:16
Ademanfor my own sanity though, does the type mixing have to occur on that line? it seems so since it's that particular variable07:17
fijalit does not have to07:17
fijalwhat's logstr?07:17
Ademanstr else I messed up badly :-)07:18
Ademanyeah I don't see how it could ever be anything else07:19
fijaleh ok07:19
fijalyou don't use stderr.write07:19
fijalyou can use os.write07:20
fijalif you want to07:20
fijalhowever, I strongly suggest looking into rpython logging07:20
Ademansure, that's what I'll do07:21
Ademaner more like rlib.debug.debug_print ?07:23
Ademando I have to set an environment variable for debug_print to go anywhere?07:26
fijalI think when untranslated the default is to go to stderr07:29
fijalyou can set PYPYLOG07:29
fijalI want jquery in minidom....07:35
Alex_Gaynorfijal: use pyquery07:38
Ademanfijal: jitviewer looks excelent by the way07:38
Alex_Gaynorits a jQuery like UI on top of lxml07:38
fijalAlex_Gaynor: I don't want to use lxml07:38
Alex_Gaynorwell you don't use the lxml API with it07:38
Alex_Gaynorit just uses the internals07:38
fijalit doesn't work on pypy07:38
fijalscrew that07:38
Alex_Gaynorah, well there's that :)07:38
fijalI'm cereal07:45
AdemanI wish there was an ElementTree compatible interface that supported XPath well07:47
AdemanI don't think I've ever been able to use ElementTree to run an XPath query without writing some of the logic in Python myself...07:47
kenaan03fijal 10r7932814 07M(pypy/trunk/pypy/tool/release/make_release.py): Fix for new html ...07:48
Alex_Gaynorfijal: did you ever backport my sys.path fix?07:52
fijalto release?07:52
fijalI think so07:52
Alex_Gaynorr79161, /me goes to check07:53
Alex_Gaynorfijal: no, I guess this was a seonc commit, can you backport it?07:54
Alex_Gaynorthis is necessary for `pip install <anything>` to work :)07:54
enkrav (~enkrav@77-58-146-192.dclient.hispeed.ch) left irc: Ping timeout: 240 seconds07:56
fijalsvn: File not found: revision 79328, path '/svn/pypy/trunk'07:58
fijalAlex_Gaynor: can you merge yourself? I don't have time now and I'll forget07:58
Alex_Gaynorok, I'm not on the right OS ATM, but I can do it tomorrow (probably by asking someone here how to merge)07:59
Ademan (~dan@63.sub-72-109-145.myvzw.com) left irc: Remote host closed the connection08:05
Ademan (~dan@63.sub-72-109-145.myvzw.com) joined #pypy.08:05
hruske (~Gasper@89-212-104-197.dynamic.dsl.t-2.net) joined #pypy.08:07
Ademanshould I be worried about [annrpython:WARNING] <FunctionGraph of (pypy.module.cpyext.api:77)format_args at 0x1293cb94> block@0 op=3/ no precise annotation supplied for iter(SomeInstance(can_be_None=False, classdef=pypy.rlib.unroll.unrolling_iterable),) ?08:13
fijalI don't think08:13
Ademanas best I can tell my usage is fine   for i in unrolling_iterable(range(...)):08:14
mvt (~mvantelli@ joined #pypy.08:15
Ademanhrm, just exploded, related to the same line...08:15
__name__ (~quassel@sburn/devel/name) joined #pypy.08:20
kenaan03dan 10r7932914 07M(pypy/branch/psycopg2compatibility/pypy/module/cpyext/api.py): Trying to RPython-ize CPyExtCallLogger ...08:20
verte (~verte@ joined #pypy.08:23
hruske (~Gasper@89-212-104-197.dynamic.dsl.t-2.net) left irc: Quit: Leaving08:26
GasbaKid_ (~GasbaKid@ joined #pypy.08:29
Ademanhttps://codespeak.net/viewvc/pypy/branch/psycopg2compatibility/pypy/module/cpyext/api.py?revision=79329&view=markup line 79 I really don't understand why that wouldn't be ok... "Block cannot succeed" or something to that effect08:30
GasbaKid (~GasbaKid@ left irc: Ping timeout: 240 seconds08:31
fijalAdeman: which line is 79?08:32
Ademanfijal: for i in unrolling_iterable(range(len(argtypes))):08:33
fijalyou can't call unrolling_iterable in the middle of your functyion08:33
fijalit has to be prepared before08:34
mwhudson (~mwh@canonical/launchpad/mwhudson) left irc: Ping timeout: 260 seconds08:34
fijalwhat's CPyExtCallLogger?08:35
fijalan API from C?08:35
Ademanah, would moving it up into log_cpyext_call() be "before" enough?08:35
fijal"before" means at translation time08:35
fijalso I guess so08:35
fijalnote that it has to be in 'if we_are_not_translated'08:36
Ademanfijal: no it logs calls into PyPy from C code, amaury seemed to like it well enough (before things were uglier :-) )08:36
GasbaKid_ (~GasbaKid@ left irc: Quit: Leaving08:39
Ademanthanks fijal, it makes sense that this should work now... heh08:39
GasbaKid (~GasbaKid@ joined #pypy.08:39
kenaan03hakanardo 10r7933014 pypy/branch/jit-unroll-loops/pypy/jit/metainterp 07M(unroll.py test_greenfield.py): beeing more carefull about moving getitem ops ...08:39
hakanardo (~hakan@h59ec009e.seluldx.dyn.perspektivbredband.net) left irc: Quit: Leaving08:40
hpk (~hpk@HSI-KBW-109-193-198-175.hsi7.kabel-badenwuerttemberg.de) joined #pypy.08:41
fijalhpk: hi08:42
fijalI think we should upgrade py in pypy08:42
fijalor remove it08:42
fijalit's getting slowly into a mess08:42
fijalwell, it's hard to have pypy checkout on PYTHONPATH08:43
fijaland py somewhere else08:43
fijalwith new version08:43
fijalhpk: I'm trying to have both py 2.0 installed and pypy checkout08:44
fijaland they bite each other if I don't remove py from pypy checkout08:44
Ademan (~dan@63.sub-72-109-145.myvzw.com) left irc: Remote host closed the connection08:45
hpklet's see to move on there, i have a couple of local fixes 08:45
hpkwe can open a branch (svn still i guess) and see to get going, during this week 08:45
fijalI'm not sure I'll find time for that though08:45
fijalbut let's see08:46
Ademan (~dan@63.sub-72-109-145.myvzw.com) joined #pypy.08:46
antocuni (~antocuni@fwstups.cs.uni-duesseldorf.de) joined #pypy.08:52
verte (~verte@ left irc: Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org08:53
nettok (~quassel@ left irc: Ping timeout: 240 seconds08:59
fijalantocuni: how do I know what pdb++ can do?09:02
antocunifijal: I started writing a more comprehensive readme09:03
fijalcool :)09:03
Action: fijal insists on sphinx docs09:03
antocunifijal: as you can see by missing links, it's still in-progress :-)09:04
Ademanfijal: I think `help` will list pdb++ specific features under "undocumented" or whatever09:04
Ademanbut they'll show up09:04
antocuniAdeman: note that most of them *are* documented :-)09:05
antocuniin the sense of having a docstring that shows up if you type e.g. "help longlist"09:05
Ademanantocuni: ah, sorry :-)09:06
AdemanI really hope this specializes correctly, you can't specify on one arg and one argtype right?09:07
antocuniAdeman: "this" == ?09:08
Ademanantocuni: a call logger I'm trying to RPython-ize09:08
AdemanI'm taking a bit of a break from ootype09:08
Ademanmy head is fairly spun around :-)09:09
antocuniAdeman: ah, I saw it in the commits09:09
Ademanugh, somehow I have something annotated as SomePBC (I have to assume object space) and a pointer09:10
savio (~thc@189001130056.usr.predialnet.com.br) joined #pypy.09:13
hruske (~Gasper@89-212-104-197.dynamic.dsl.t-2.net) joined #pypy.09:20
ErKa (~keryell@APuteaux-154-1-37-100.w83-199.abo.wanadoo.fr) joined #pypy.09:33
ErKa (~keryell@APuteaux-154-1-37-100.w83-199.abo.wanadoo.fr) left irc: Read error: Connection reset by peer09:34
amaury_ (5bc3480e@gateway/web/freenode/ip. joined #pypy.09:37
kenaan03afa 10r7933114 pypy/branch/fast-forward/pypy/module/_io 07M(interp_textio.py test_textio.py): Test and fix ...09:41
amaury_fijal: Hi09:42
Ademan (~dan@63.sub-72-109-145.myvzw.com) left irc: Quit: leaving09:54
pedronis (~pedronis@253-109.104-92.cust.bluewin.ch) joined #pypy.09:55
fijalamaury_: hi10:06
savio (~thc@189001130056.usr.predialnet.com.br) left irc: Ping timeout: 265 seconds10:21
arigato (~arigo@fwstups.cs.uni-duesseldorf.de) joined #pypy.10:29
fijalarigato: hi10:30
amaury_fijal: did you see the 1.4 builds for win32?10:36
lizardo (~lizardo@ joined #pypy.10:36
fijalamaury_: I did not10:36
amaury_on tannit, /home/amaury/pypy-1.410:37
DasIch_ (~DasIch@p5DC5F712.dip.t-dialin.net) joined #pypy.10:38
ErKa (~keryell@APuteaux-154-1-37-100.w83-199.abo.wanadoo.fr) joined #pypy.10:39
DasIch (~DasIch@p5DC5F853.dip.t-dialin.net) left irc: Ping timeout: 276 seconds10:41
fijalamaury_: thanks10:44
fijalamaury_: are they stripped?10:44
amaury_well, this is on Windows10:44
amaury_what does "stripped" mean?10:45
fijalheh :)10:45
fijalwithout debug symbols I believe10:45
amaury_They are compiled with -O2, without debug symbols10:45
amaury_and win32 binaries don't store paths to .DLL, only the base name10:45
arigato(reminder to cfbolz: branch/jit-starargs broke at least pypy.module._stackless.test.test_coroutine10:46
amaury_I tried to unpack the .zip to some test directory, and pypy.exe seems to work correctly10:47
amaury_sys.path looked good10:47
fijalamaury_: good, thank you10:47
kenaan03arigo 10r7933214 07M(pypy/trunk/pypy/module/sys/test/test_initialpath.py): Fix test. ...10:49
kenaan03arigo 10r7933314 07M(pypy/trunk/pypy/tool/release/test/test_package.py): Fix test. ...10:52
kenaan03arigo 10r7933414 07M(pypy/trunk/pypy/translator/goal/test2/test_app_main.py): Fix tests. ...10:54
fijalamaury_: -O2 or -Ojit?10:57
fijalanyway, do we really want nojit versions any more?10:57
kenaan03arigo 10r7933514 07M(pypy/branch/jit-free/pypy/translator/driver.py): * Rename gc-collect-task into translation-task. ...10:57
santagada (~anonymous@ joined #pypy.10:59
fijalso we have all stuff in place11:02
fijalalex wanted his change in place11:02
fijalmaybe let's ignore it, it's beta after all?11:03
arigatowhich change?11:05
Action: arigato -> lunch11:05
fijalarigato: 7916111:06
Burgundy (Burgundy@ joined #pypy.11:11
ousado_ (~ousado@p5B3C0ADF.dip0.t-ipconnect.de) joined #pypy.11:14
ousado (~ousado@p5B3C01A8.dip0.t-ipconnect.de) left irc: Read error: Operation timed out11:16
Nick change: ousado_ -> ousado11:26
__name__ (~quassel@sburn/devel/name) left irc: Ping timeout: 255 seconds11:34
amaury_fijal: -O2 is the compiler option11:54
amaury_this is the default on all builds (except when you run "make debug")11:54
fijalarigato, amaury_, antocuni: can you try maybe downloading linux builds or os x builds and tell me if we're fine releasing beta?11:59
fijal(and what would be the procedure for that)11:59
lmoura (~lauromour@ joined #pypy.12:01
arigatoeither it's helpful, or it needs to be killed :-)12:02
fijalarigato: we don't have a procedure for beta :)12:02
fijalno, it's helpful12:02
fijalbut it could be amended with 'if you wish to...'12:02
arigatoI'm sure the procedure works for beta too :-)12:03
fijalok :-)12:03
fijalso are we fine to just go ahead?12:03
arigatoI think so12:03
__name__ (~quassel@ joined #pypy.12:03
__name__ (~quassel@ left irc: Changing host12:03
__name__ (~quassel@sburn/devel/name) joined #pypy.12:03
Action: antocuni downloads pypy-1.4-linux12:04
fijalnote I just renamed stuff to -beta12:05
arigatocan you put the "beta" closer to the "1.4"?12:05
antocuniah, I started pypy-1.4 and I got "PyPy 1.1.0beta released: http://codespeak.net/pypy/dist/pypy/doc/release-1.1.0.html" as a irc topic :-)12:05
arigato(sorry about that :-)12:05
fijalarigato: sure12:07
antocunithe "beta" does not show up near [PyPy 1.4.0]12:07
antocuninot sure that I care12:07
arigatofijal: see "pypy-1.1.0beta.tar.bz2"12:08
arigatowell I don't really care enough either, I'm fine with release "1.4beta" and then "1.4.1"12:08
arigatocontaining e.g. the jit-free branch12:08
fijalantocuni: well, our nightly pypy-c also shows pypy 1.4.012:09
antocunibtw, pypy-1.4 seems to work fine for me12:09
fijalso I don't care12:09
arigato(pypy-linux64-jit works nicely)12:09
antocuni(32bit for me)12:10
fijalso we have downloads, I'll update the website and send announcements12:10
fijalI think we should strip nojit versions12:12
fijalare you fine with saying "this is download page for 1.4beta which is a recommended version, see 1.3 download page for older"12:13
arigato"strip" in the sense "remove"?  or "/usr/bin/strip" ?12:13
fijalstrip as in not display on download page12:13
arigatowe have them, let's keep them I suppose12:13
arigatofwiw the stackless should be called "stackless-nojit" nowadays, just to be as clear as possible12:14
aleale (~aleale@ joined #pypy.12:14
fijalyes, possibly12:14
fijalarigato: that's from builder name12:14
fijalok, I guess there was no confusion so far with number of download options we have12:15
fijalI think12:15
fijalNote on building a 32-bit version on 64-bit systems12:15
fijaldo we still care?12:15
fijal(if so, how about XXX on the website :)12:18
arigatowhat changed since 1.3 about this?12:18
fijalworking 64bit jit backend12:18
fijalthat's kind of a serious change12:18
hpkis it possible to disable jit-tracing by default? 12:18
hpk(at the process exit) 12:18
fijalhpk: you mean debug info?12:19
Action: hpk just trying out pypy-1.4-beta from pypy.org download12:19
fijal(we do this for releases)12:19
fijalusually, but not for beta :)12:19
hpk 1.4beta is a release :)12:19
hpkhi aleale, long time no see12:20
fijalI guess we can rebuild all the stuff12:20
fijalbut it's kind of boring12:20
alealeI've been lurking12:20
hpkfijal:  what do you mean? boring to make a build that doesn't default to showing debug output? 12:21
antocuni (~antocuni@fwstups.cs.uni-duesseldorf.de) left irc: Read error: Operation timed out12:21
alealeregarding 79319 - join only takes two arguments. 12:21
fijalhpk: it's boring (tedious) to rebuild all binaries12:21
arigato (~arigo@fwstups.cs.uni-duesseldorf.de) left irc: Read error: Connection reset by peer12:22
alealeat least that what the error says12:22
hpkit's just annoying to have virtualenv and other scripts presenting pypy's debug info, cluttering up output considerably - kind of makes it less nice to use as a default interpreter 12:22
arigato (~arigo@fwstups.cs.uni-duesseldorf.de) joined #pypy.12:22
fijalaleale: what kind of python do you have?12:23
hpkarigato:  maybe interesting for you as well: i find it annoying to have virtualenv and other scripts presenting pypy's debug info, cluttering up output considerably - kind of makes it less nice to use as a default interpreter12:23
hpkfor example, using "pip install ..." looks like  this: http://paste.pocoo.org/show/294548/12:25
alealefijal: python2.612:26
fijalaleale: works for me12:26
fijalos.path.join('a', 'b', 'c')12:26
alealei'll check again12:26
arigatohpk: I see, but I'm not caring enough to recompile all pypy's right now, I fear12:28
arigatohpk: I suppose it makes it clear that it's only a beta12:28
antocuni (~antocuni@fwstups.cs.uni-duesseldorf.de) joined #pypy.12:28
hpksure, don't want to press you. 12:29
hpkjust makes it less likely i run it as default interpreter which is a pity12:30
hpkthere is no runtime option to avoid the output, is there? 12:30
fijalhpk: well, you can recompile it yourself with --jit-debug=off12:30
fijalpy.test --jit debug=012:30
fijalpypy-c --jit debug=012:30
fijal(a runtime option)12:31
hpki am talking about virtualenv, pip, py.test, easy_install and tons of other scripts that invoke pypy-c themselves 12:31
hpki cannot influence all these locations 12:31
fijalhpk: there is both a runtime option and a compile-time default12:31
hpkso i guess an environment variable could help12:32
arigatoI think that it's ok to say "it's a debug version, if you want you can retranslate yourself and turn them off"12:32
hpkfijal:  we discussed this before, 2-3 weeks before already, i think 12:32
fijala 3rd way to get it...12:32
hpkarigato:  sure, jsut in the future i suggest to do it the other way round: i.e. say that people wanting a debug version should compile themselves12:32
fijalhpk: 1.3 already had this12:32
fijalit's just that we don't want to rebuild now12:33
hpksure, nm, i do a translate myself 12:33
__name__ (~quassel@sburn/devel/name) left irc: Ping timeout: 255 seconds12:33
hpkusing the pypy-1.4beta :) 12:34
arigatofijal: I think the release announcement should hint that another release is coming, too, with jit-free in particular12:34
fijalarigato: yop12:34
arigatohpk: see?  it's still useful :-)12:34
fijalpypy-1.4beta does contain generator fixes?12:34
arigatothis means that hpk will hit the memory limit12:35
fijalhpk: then you probably can't translate it (unless you have tons of ram)12:35
fijalhpk: but you can get a nightly from jit-free branch12:35
hpki get AssertionError: Mixed configuration of the word size of the machine: anyway12:35
fijalhpk: obviously you're running 32bit pypy on 64bit12:36
fijalor so12:36
__name__ (~quassel@ joined #pypy.12:36
__name__ (~quassel@ left irc: Changing host12:36
__name__ (~quassel@sburn/devel/name) joined #pypy.12:36
fijalthat's not going to fly with cpython either :)12:36
arigatohpk: the issue is rather that "gcc" invokes a 32-bit version of gcc even if you are 64-bit or vice-versa12:36
hpkyou are probably both right, i downloaded the 32bit version12:37
antocuni (~antocuni@fwstups.cs.uni-duesseldorf.de) left irc: Read error: Operation timed out12:40
arigato (~arigo@fwstups.cs.uni-duesseldorf.de) left irc: Ping timeout: 276 seconds12:41
antocuni (~antocuni@fwstups.cs.uni-duesseldorf.de) joined #pypy.12:43
arigato (~arigo@fwstups.cs.uni-duesseldorf.de) joined #pypy.12:44
arigatogrumble, our network keeps turning on and off12:44
alealefijal: I get this error when translating :http://paste.pocoo.org/show/294554/12:45
fijalok, indeed12:47
kenaan03arigo 10r7933714 07M(pypy/trunk/pypy/module/sys/state.py): Translation fix on OS/X. ...12:47
arigatoaleale: "svn up", then it should work again12:47
arigatothe issue is not in the pypy-c, but only on the source code12:48
fijalarigato: obviously, otherwise how did pypy-c translate?12:48
arigatofijal: well the pypy-c used by aleale is older than this code12:48
arigatofijal: in his pypy-c, instead of this code, there is a hack in site.py12:49
arigatoin other word it's a bug introduced on "trunk" but not in "release-1.4.x"12:50
santagadaarigato: you know that problem with django and pip? Alex_Gaynor said he fixed it in trunk12:54
santagadaarigato: some time ago and maybe fijal or someone didn't merge it into the release branch12:55
arigatois that r79161?12:55
hpkhow much RAM usage can i expect translating pypy-jit with pypy-jit on 64bits? 12:59
Action: hpk stops translation, only has 6GBs currently13:00
hpkanyway got this: http://paste.pocoo.org/show/294561/13:00
kenaan03cfbolz 10r7933814 pypy/trunk/pypy/interpreter 07M(argument.py test_argument.py): fix test in module/_stackless. ...13:01
arigatohpk: ah, it's the bug we are tracking down13:01
arigatoshows up once every 4 or 5 translations13:01
arigatoso at least now we know that it's not specific to the jit-free branch13:02
arigatoif you got it on trunk with a pypy-1.4 too13:02
arigatowe never noticed before because only in the jit-free branch can we really run memory-efficient translations with pypy :-)13:02
hpkthis was from downloaded  pypy-1.4beta-linux64 and translated trunk13:03
alealearigato: thanks - see you13:03
aleale (~aleale@ left irc: Quit: aleale13:03
stakkars (~tismer@i577B6F8F.versanet.de) left irc: Quit: stakkars13:08
kenaan03fijal 10r7934014 05A(pypy/trunk/pypy/doc/release-1.4.0beta.txt): Write down a draft of 1.4beta release ...13:08
kenaan03fijal 10r7934114 07M(pypy/trunk/pypy/doc/release-1.4.0beta.txt): Disclaimer ...13:09
fijalarigato, antocuni, hpk: please have a look13:09
Action: antocuni does13:10
__name__ (~quassel@sburn/devel/name) left irc: Quit: http://quassel-irc.org - Chat comfortably. Anywhere.13:10
__name__ (~quassel@ joined #pypy.13:11
__name__ (~quassel@ left irc: Changing host13:11
__name__ (~quassel@sburn/devel/name) joined #pypy.13:11
stakkars (~tismer@i577B6F8F.versanet.de) joined #pypy.13:12
Action: arigato too13:15
arigatofijal: maybe mention cpyext ("pypy setup.py build"), and write a note about the summary printed to stderr at the end13:17
fijalwhat should I mention about cpyext?13:18
arigato"Regular CPython extension modules can be used too (to some extend).  You need to recompile them from their C source with "pypy setup.py build". "13:19
fijalwasn't it true for 1.3 as well?13:19
kenaan03fijal 10r7934214 07M(pypy/trunk/pypy/doc/release-1.4.0beta.txt): mention debugging msgs ...13:19
fijal(what changed)13:19
fijalfor most part it doesn't work13:19
arigatomaybe it was true, but the number of people around here that didn't know it shows something13:19
arigatomaybe mention explicitly that you usually need to fix a thing or two in the C sources13:20
arigatonotably places that abuse reference counting13:20
kenaan03fijal 10r7934314 07M(pypy/trunk/pypy/doc/release-1.4.0beta.txt): mention cpyext ...13:20
fijalfor example on 64bit we don't pass -fPIC to the compiler13:21
fijalanyway, mentioned13:21
arigatopoint to http://codespeak.net/svn/pypy/trunk/pypy/module/cpyext/patches/ for examples13:21
kenaan03fijal 10r7934414 07M(pypy/trunk/pypy/doc/release-1.4.0beta.txt): point to patches ...13:23
fijalarigato: ok, I'm gone soon, will do missing parts in the evening13:25
fijalfeel free to commit on this stuff13:26
fijal(incorporated all your comments)13:26
fijalbesides will add a note and a separate download page for beta13:26
fijalpost it on the blog python-dev, pypy-dev, python-announce13:27
fijalnot sure what more13:27
karyo (77cb72f9@gateway/web/freenode/ip. joined #pypy.13:34
cwillu_at_work (~cwillu@cwillu.com) joined #pypy.13:35
karyowould somebody who has CS101 level knowledge and 6 months worth python experience be able to contribute to pypy?13:35
arigatohi :-)13:36
arigatois there a more specific area where you would like to help?13:36
karyoi'm reading the "getting-started" pages that mention translators, interpreters, parsers and such. I understand what they do in concept, but I am very unfamiliar with the actual code. :-<13:42
arigatoyou need to start getting some familiarity by setting yourself a small subproject: for example, write a module for pypy, or help in the JIT backend13:43
arigatoor help with the general effort to port to Python 2.713:44
karyois there a central plan or page that I can consult regarding the 2.7 porting effort?13:45
arigatoit may not be entirely up-to-date, but that's the plan13:46
Burgundy (Burgundy@ left irc: Ping timeout: 250 seconds13:47
gutworth (~benjamin@64-131-30-120.usfamily.net) joined #pypy.13:47
arigatoI should let amaury_ or gutworth comment on it (hi gutworth :-)  because they are doing more than me in the fast-forward branch13:48
karyothx, anyway I think I should read more(if not all) of the docs before proceeding any further. The wealth of documentation is very welcoming btw.13:48
arigatoit may be occasionally out-of-date, but thanks13:49
karyo (77cb72f9@gateway/web/freenode/ip. left irc: Quit: Page closed13:50
__DasIch__ (~DasIch@p5DC5FCE1.dip.t-dialin.net) joined #pypy.13:55
hruske (~Gasper@89-212-104-197.dynamic.dsl.t-2.net) left irc: Quit: Leaving13:56
DasIch_ (~DasIch@p5DC5F712.dip.t-dialin.net) left irc: Ping timeout: 276 seconds13:58
__DasIch__ (~DasIch@p5DC5FCE1.dip.t-dialin.net) left irc: Ping timeout: 240 seconds14:01
Ademan-remote (~dan@63.sub-72-109-145.myvzw.com) joined #pypy.14:04
Ademan-remoteamaury_: am I crazy or is there a CPyExt failure on trunk? (test_tp_dict)14:05
amaury_Ademan-remote: while checking "memory leaks"? yes14:06
Ademan-remoteamaury_: cool, it looks like I'm passing again then :-)14:07
amaury_really? wonderful!14:08
kenaan03antocuni 10r7934514 07M(pypy/trunk/pypy/doc/release-1.4.0beta.txt): fix virtualenv version ...14:08
amaury_what did you change?14:08
Ademan-remoteI mean PyPy tests, I was passing until last night when I tried to RPython-ize my logger... haha14:09
Ademan-remoteif you're not busy with the release, in the next couple of days I'd like help examining psycopg2 (threw my modifications up on github)14:10
Ademan-remoteI still don't have a dedicated internet connection in my new home though...14:11
Alex_Gaynorwhat incantation do I give SVN to backport a commit?14:13
Alex_Gaynorpreferably without needing a checkout of the release branch14:14
Ademan-remoteAlex_Gaynor: I'd crosspost to #svn :-)  but don't you just want to svn merge?14:14
Ademan-remotesvn merge [...]trunk/path/to/changes [...]/tags/release/path/to/changes ?14:16
Alex_Gaynorexcept I only want a single commit14:16
Ademan-remoteDISCLAIMER: I'm the farthest thing possible from an svn expert14:16
Ademan-remoteAlex_Gaynor: good because merge requires a range of revisions :-)14:17
gutworth (~benjamin@64-131-30-120.usfamily.net) left irc: Ping timeout: 265 seconds14:18
amaury_Alex_Gaynor: "svn merge -c12345 /from /to"14:18
Alex_Gaynoramaury_: That gives me "Can't specify revision range with two URLs"14:21
amaury_I suppose you need to "svn co" the target branch14:22
Ademan-remoteamaury_: I get a segfault in the tests if I import PyObject_Repr(space, an_arg)  when the corresponding api_func.argtypes entry is argtype and is_PyObject(argtype) is true 14:22
Ademan-remotethis is inside the wrapper() do I need to somehow re-wrap an_arg?14:23
amaury_IMO you should use api.PyObject_Repr() instead14:24
kenaan03agaynor 10r7934614 pypy/release/1.4.x 07M(/* /pypy/* optimizer.py* Makefile* intimg.c* intimgtst.c* intimgtst.py* loop.c* sum.c* sumtst.c* sumtst.py* test_array_old.py* rerased.py* test_rerased.py* app_main.py ...): Merged 79161  ...14:25
Alex_Gaynorwoohoo, backported14:26
Ademan-remoteamaury: sounds good, how do I get an api object?14:28
amaury_Ademan-remote: please look at other tests14:34
amaury_- first inherit from BaseApiTest14:34
amaury_- then just add "api" to the parameters list14:35
Ademan-remoteamaury_: nono this is from api.py in wrapper (defined in make_wrapper)14:36
amaury_then the best is to use from_ref(), and call PyObject_Repr with that14:37
Ademan-remoteah thanks, it no longer segfaults at least14:41
santagadaAlex_Gaynor: thanks, but I think the binaries are already out14:41
Alex_Gaynorsantagada: for the beta, they'll be recompiled for the final release14:41
santagadaAlex_Gaynor: I thought they would also rebuild the branch14:43
kursor (~kursor@p57BB146F.dip0.t-ipconnect.de) joined #pypy.14:45
kursor_ (~kursor@p57BB146F.dip0.t-ipconnect.de) joined #pypy.14:49
kursor (~kursor@p57BB146F.dip0.t-ipconnect.de) left irc: Ping timeout: 240 seconds14:53
Nick change: kursor_ -> kursor14:53
DasIch (~DasIch@p5DC5F239.dip.t-dialin.net) joined #pypy.15:10
Rhy0lite (~dje@nat/ibm/x-xfwfeowpydvovwar) joined #pypy.15:24
DasIch (~DasIch@p5DC5F239.dip.t-dialin.net) left irc: Ping timeout: 245 seconds15:25
kenaan03arigo 10r7934714 pypy/extradoc 07M(jit.txt planning.txt rpython-overview.txt jit_lightningtalk.py overview2.png* pretty.css): Fix an issue, open another one. ...15:28
arigatowell, now I've checked in all my old, modified files15:28
amaury_better late than never15:29
arigatothese were really worth an "svn revert"15:29
arigatobut too late15:30
DasIch (~DasIch@p5DC5E605.dip.t-dialin.net) joined #pypy.15:31
Burgundy (blabla214@ joined #pypy.15:34
DasIch (~DasIch@p5DC5E605.dip.t-dialin.net) left irc: Ping timeout: 255 seconds15:36
DasIch (~DasIch@p5DC5F55D.dip.t-dialin.net) joined #pypy.15:37
DasIch (~DasIch@p5DC5F55D.dip.t-dialin.net) left irc: Ping timeout: 276 seconds15:42
kursor_ (~kursor@p57BB146F.dip0.t-ipconnect.de) joined #pypy.15:46
kursor (~kursor@p57BB146F.dip0.t-ipconnect.de) left irc: Read error: Operation timed out15:48
Nick change: kursor_ -> kursor15:48
santagada (~anonymous@ left irc: Quit: santagada15:50
__name__ (~quassel@sburn/devel/name) left irc: Read error: Connection reset by peer15:51
dmalcolm (~david@nat/redhat/x-fmqfknrpxvuaitzs) joined #pypy.15:58
Trundle (~andy@python/site-packages/trundle) joined #pypy.16:10
jimbaker (~jimbaker@c-24-8-39-78.hsd1.co.comcast.net) joined #pypy.16:11
DasIch (~DasIch@p5DC5FCEA.dip.t-dialin.net) joined #pypy.16:28
kenaan03afa 10r7934814 pypy/branch/fast-forward/pypy/module/_io 07M(interp_textio.py test_textio.py): Looong code to implement TextIOWrapper.readline() ...16:29
santagada (~lsantagad@ joined #pypy.16:30
santagada (~lsantagad@ left irc: Remote host closed the connection16:30
santagada (~lsantagad@ joined #pypy.16:32
arigatoah, ah, ah16:43
mvt (~mvantelli@ left irc: Quit: This computer has gone to sleep16:46
mvt (~mvantelli@ joined #pypy.16:47
kenaan03arigo 10r7934914 07M(pypy/trunk/pypy/jit/metainterp/test/test_exception.py): (antocuni, arigo) ...16:47
Moku (~John@osbk-4d087bc1.pool.mediaWays.net) left irc: Read error: Connection reset by peer16:52
kenaan03arigo 10r7935014 pypy/trunk/pypy/jit/metainterp 07M(optimizer.py test_optimizeopt.py): (antocuni, arigo) ...16:53
arigatonow what?  we retranslate all pypy 1.4's?16:54
Alex_Gaynorarigato: well, for final release we should, if we want pip to work :)16:55
Moku (~John@osbk-4d087bc1.pool.mediaWays.net) joined #pypy.16:56
[mathias] (~mathias@ joined #pypy.16:57
kenaan03arigo 10r7935114 pypy/branch/jit-free/pypy/jit/metainterp 07M(optimizer.py test_exception.py test_optimizeopt.py): Merge r79349-r79350 from trunk. ...16:57
kenaan03arigo 10r7935214 07M(pypy/branch/jit-free/pypy/jit/backend/llgraph/llimpl.py): Fix bug (op.fail_args can be None to mean []). ...16:58
__name__ (~name@sburn/devel/name) joined #pypy.17:00
arigatofijal (when you are back): we fixed the issue of getting occasional nonsense about exceptions on a pypy-jit (r79350).17:03
arigatofijal: if you agree, I'd like to really rebuild the pypy-1.4beta-*17:03
arigatoit's not like it's hard, it just takes time17:04
kenaan03arigo 10r7935314 pypy/release/1.4.x/pypy/jit/metainterp 07M(optimizer.py test_exception.py test_optimizeopt.py): Merge r79349-r79350 from trunk. ...17:06
kenaan03arigo 10r7935414 pypy/release/1.4.x 07M(site.py state.py test_initialpath.py): Revert the temporary fix in r79320, and add the proper fixes ...17:12
kenaan03arigo 10r7935514 07M(pypy/trunk/pypy/doc/release-1.4.0beta.txt): Typo. ...17:14
Shanita (~John@osbk-4d0877a3.pool.mediaWays.net) joined #pypy.17:15
Moku (~John@osbk-4d087bc1.pool.mediaWays.net) left irc: Ping timeout: 240 seconds17:15
kursor (~kursor@p57BB146F.dip0.t-ipconnect.de) left irc: Quit: kursor17:40
ErKa (~keryell@APuteaux-154-1-37-100.w83-199.abo.wanadoo.fr) left irc: Ping timeout: 245 seconds17:41
DasIch_ (~DasIch@p5DC5EC67.dip.t-dialin.net) joined #pypy.17:44
__DasIch__ (~DasIch@p5DC5FC3E.dip.t-dialin.net) joined #pypy.17:46
DasIch (~DasIch@p5DC5FCEA.dip.t-dialin.net) left irc: Ping timeout: 252 seconds17:47
DasIch_ (~DasIch@p5DC5EC67.dip.t-dialin.net) left irc: Ping timeout: 240 seconds17:48
Nick change: __DasIch__ -> DasIch17:57
kenaan03antocuni 10r7935614 pypy/trunk/pypy/jit/tool 05A(log-template.gnumeric log2gnumeric.py): a tool to automatically convert a logfile into a gnumeric spreadsheet with a graph ...18:00
kenaan03arigo 10r7935714 07M(pypy/branch/jit-free/pypy/jit/metainterp/jitprof.py): Fix for test_jitprof. ...18:10
kenaan03arigo 10r7935814 07M(pypy/branch/jit-free/pypy/jit/backend/test/test_random.py): Fix test_random. ...18:10
kenaan03arigo 10r7935914 07M(pypy/branch/jit-free/pypy/jit/backend/test/test_random.py): Use the official interface to keep the looptoken alive. ...18:12
arigato (~arigo@fwstups.cs.uni-duesseldorf.de) left irc: Quit: See you18:18
antocuni (~antocuni@fwstups.cs.uni-duesseldorf.de) left irc: Read error: Operation timed out18:20
mwhudson (~mwh@canonical/launchpad/mwhudson) joined #pypy.18:24
amaury_ (5bc3480e@gateway/web/freenode/ip. left irc: Quit: Page closed18:32
panni_ (hannes@ip-178-203-101-205.unitymediagroup.de) joined #pypy.18:32
Ademan-remote (~dan@63.sub-72-109-145.myvzw.com) left irc: Ping timeout: 255 seconds18:50
kenaan03david 10r7936114 pypy/branch/arm-backend/pypy/jit/backend/arm 05A(jump.py test_jump.py) 07M(assembler.py locations.py opassembler.py): Finish implementing the jump operation, based on the x86 jump code, adapted for the ARM backe ...18:52
DasIch_ (~DasIch@p5DC5FC08.dip.t-dialin.net) joined #pypy.18:59
Rhy0lite (~dje@nat/ibm/x-xfwfeowpydvovwar) left irc: Quit: Leaving19:01
DasIch (~DasIch@p5DC5FC3E.dip.t-dialin.net) left irc: Ping timeout: 260 seconds19:02
Ademan-remote (~dan@75.sub-97-156-41.myvzw.com) joined #pypy.19:10
hpk (~hpk@HSI-KBW-109-193-198-175.hsi7.kabel-badenwuerttemberg.de) left irc: Ping timeout: 276 seconds19:13
Ademan-remote (~dan@75.sub-97-156-41.myvzw.com) left irc: Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org19:33
__name__ (~name@sburn/devel/name) left irc: Remote host closed the connection19:38
cwillu_at_work (~cwillu@cwillu.com) left irc: Remote host closed the connection19:48
Scorchin (~Scorchin@host86-162-0-121.range86-162.btcentralplus.com) joined #pypy.19:48
cwillu_at_work (~cwillu@cwillu.com) joined #pypy.19:48
GasbaKid (~GasbaKid@ left irc: Read error: No route to host19:59
GasbaKid (~GasbaKid@ joined #pypy.20:00
jae (~jae@jaerhard.com) left irc: Ping timeout: 276 seconds20:02
mwhudson (~mwh@canonical/launchpad/mwhudson) left irc: Ping timeout: 240 seconds20:06
santagada (~lsantagad@ left irc: Quit: Ex-Chat20:11
Trundle (~andy@python/site-packages/trundle) left irc: Ping timeout: 250 seconds20:29
kenaan03fijal 10r7936214 07M(pypy/release/1.4.x/pypy/config/translationoption.py): As we're going to rebuild those anyway, let's turn off jit debug ...20:31
sasha (~sasha@static-72-89-161-34.nycmny.fios.verizon.net) joined #pypy.20:35
mwhudson (~mwh@ joined #pypy.20:41
mwhudson (~mwh@ left irc: Changing host20:41
mwhudson (~mwh@canonical/launchpad/mwhudson) joined #pypy.20:41
bsod1 (~osa1@ joined #pypy.20:43
tav (~tav@ left irc: Ping timeout: 245 seconds20:45
tav (~tav@ joined #pypy.20:50
lizardo (~lizardo@ left irc: Quit: Leaving20:52
sasha (~sasha@static-72-89-161-34.nycmny.fios.verizon.net) left irc: Quit: Colloquy for iPhone - http://colloquy.mobi21:01
jae (~jae@jaerhard.com) joined #pypy.21:08
bsod1 (~osa1@ left irc: Ping timeout: 265 seconds21:14
sasha (~sasha@static-72-89-161-34.nycmny.fios.verizon.net) joined #pypy.21:19
antocuni (~antocuni@dslb-094-220-136-007.pools.arcor-ip.net) joined #pypy.21:22
santagada (~anonymous@ joined #pypy.21:23
antocunifijal, mwhudson: fyi, I have just release pyrepl 0.8.2 on pypi21:23
mwhudsonantocuni: woo21:24
antocuniit contains the same version which is distributed with pypy21:24
jae (~jae@jaerhard.com) left irc: Ping timeout: 245 seconds21:24
antocunimwhudson: I needed the updated version to be on pypy so I can make this easy_installable :-):  https://bitbucket.org/antocuni/fancycompleter/src21:24
mwhudsonhm i could see about putting it in a ppa too i guess21:27
jae (~jae@jaerhard.com) joined #pypy.21:31
sasha (~sasha@static-72-89-161-34.nycmny.fios.verizon.net) left irc: Quit: Colloquy for iPhone - http://colloquy.mobi21:33
amaury_ (4dc555be@gateway/web/freenode/ip. joined #pypy.21:45
nettok (~quassel@ joined #pypy.21:47
antocuniis it normal that east_install does not install what I specified in install_requirements in my setup.py? (pip does)21:48
hruske (~Gasper@89-212-104-197.dynamic.dsl.t-2.net) joined #pypy.21:50
exarkuninstall_requirements aren't runtime dependencies21:52
exarkunthey're requirements for installation to work21:52
savio (~thc@189001130056.usr.predialnet.com.br) joined #pypy.21:52
antocuniexarkun: I know. Anyway, it was my fault: I updated my local version of setup.py, but I was still trying to install from pypi, that's why it didn't know about the requirement s:-)21:55
enkrav (~enkrav@77-58-146-192.dclient.hispeed.ch) joined #pypy.21:58
Ademan (~dan@75.sub-97-156-41.myvzw.com) joined #pypy.22:05
DasIch_ (~DasIch@p5DC5FC08.dip.t-dialin.net) left irc: Ping timeout: 255 seconds22:07
kenaan03afa 10r7936514 07M(pypy/branch/fast-forward/pypy/module/sys/version.py): Show GCC version in sys.version; fixes the test. ...22:07
[mathias] (~mathias@ left irc: 22:09
gutworth (~benjamin@64-131-30-120.usfamily.net) joined #pypy.22:13
kenaan03wlav 10r7936614 07M(pypy/branch/reflex-support/pypy/module/cppyy/executor.py): for rtyper: make code paths more consistent ...22:15
kenaan03wlav 10r7936714 07M(pypy/branch/reflex-support/pypy/module/cppyy/interp_cppyy.py): cosmetics and for rtyper: explicit str ...22:15
kenaan03wlav 10r7936814 07M(pypy/branch/reflex-support/pypy/module/cppyy/converter.py): for rtyper: use rffi cast rather than chr ...22:16
\toothrot (~mux@66-169-185-121.dhcp.ftwo.tx.charter.com) joined #pypy.22:16
htoothrot (~mux@66-169-185-121.dhcp.ftwo.tx.charter.com) left irc: Ping timeout: 240 seconds22:17
mwhudsonantocuni: i'm just updating the copyright info in the files in pyrepl, do you want to be added to any of the files?22:18
DasIch_ (~DasIch@p5DC5E7DD.dip.t-dialin.net) joined #pypy.22:18
antocunimwhudson: actually I didn't do much in pyrepl :-)22:19
antocunimost of the "new" work since 0.8.1 was done by armin when he wrote pyrepl.readline22:20
antocunibut yes, I'm fine if you want to put me in the CREDITS file22:20
mwhudsonoh of course22:20
enkrav (~enkrav@77-58-146-192.dclient.hispeed.ch) left irc: Quit: Sto andando via22:48
Ademan (~dan@75.sub-97-156-41.myvzw.com) left irc: Read error: Connection reset by peer22:51
Ademan (~dan@75.sub-97-156-41.myvzw.com) joined #pypy.22:51
mwhudsonantocuni: i've just pushed a bunch of changes to svn, i hope they don't conflict with anything you were doing22:56
Action: antocuni looks22:58
mwhudsonmostly metadataish stuff22:58
antocunimwhudson: wow, encopyright is advanced :-)23:00
mwhudsonyeah, i got a bit distracted there...23:00
antocunimwhudson: if you are done, I'll updload an updated version on pypi23:00
mwhudsonantocuni: i am done for now, i guess there should be a version bump though?23:01
antocunimwhudson: well, I think we can just overwrite 0.8.223:02
antocuninobody has downloaded it yet :-)23:02
mwhudsoni guess23:02
mwhudsonthat sort of thing is a massive no no in the debian world, but i guess there's no functional changes23:03
kenaan03afa 10r7937714 pypy/branch/fast-forward/pypy/objspace/std 07M(dictmultiobject.py dicttype.py model.py test_dictmultiobject.py): Start implementing dictionary views ...23:03
gutworthah, dict views I forgot about those23:04
amaury_not essential to 2.7 IMO23:04
amaury_but still23:04
antocunimwhudson: I agree in principle, but I don't care enough :-)23:04
antocunianyway, I updated it23:04
mwhudsonantocuni: are you uploading an sdist or an export?23:05
mwhudsonseems like it doesn't contain the LICENSE file23:05
mwhudsonor some of the tests23:05
mwhudsonat least, the sdist i made23:05
antocuniit's possible, I didn't really inspect the file23:05
antocuniI just tried to pip install it, and it worked23:05
Action: antocuni looks23:06
Alex_Gaynoramaury_: is there a way to use the socket module with apptests?23:09
gutworthimport socket23:09
antocunimwhudson: are you building the sdist from the svn dir or from the bzr one?23:10
mwhudsonantocuni: bzr23:10
Alex_Gaynorgutworth: gives an importerror23:10
gutworthfor _socket?23:10
antocuniah, then it's probably why you don't have license23:10
mwhudsonantocuni: that might make a difference i guess, if it tries to use a vcs to determine which files to include23:10
antocunisetuptools automatically find svn tracked files, but not bzr I think23:11
antocunithen good, the pypi version has everything :-)23:11
mwhudsonah ok, i should base my packaging on your tarball then23:12
Ademan (~dan@75.sub-97-156-41.myvzw.com) left irc: Quit: leaving23:15
amaury_Alex_Gaynor: yes should work if you configure the space with --withmod=_socket23:24
Action: Alex_Gaynor reboots into his dev environ23:24
amaury_of course, pypy/module/_socket/test/test_sock_app.py has everything23:24
amaury_mod.space = gettestobjspace(usemodules=['_socket', 'array'])23:25
amaury_the see the class AppTestSocket23:25
Alex_Gaynoramaury_: http://paste.pocoo.org/show/294828/23:29
Alex_Gaynorhttp://paste.pocoo.org/show/294830/ is the file23:30
mwhudsonantocuni: can you check in the PKG-INFO from the tarball?23:36
mwhudsoni guess i can too easily enough23:36
antocunimwhudson: ok23:37
antocunimwhudson: done23:38
antocuni(I never really understood what they are used for, though :-))23:38
mwhudsonme either23:39
mwhudsoni guess as it seems to be derived data, perhaps it shouldn't be in the vcs23:39
mwhudsonoh well23:39
dexarkun (~exarkun@135.sub-174-252-41.myvzw.com) joined #pypy.23:41
nettok (~quassel@ left irc: Ping timeout: 255 seconds23:48
__name__ (~name@sburn/devel/name) joined #pypy.23:49
amaury_Alex_Gaynor: I suggest to "import posix" before "import socket"23:52
dexarkun (~exarkun@135.sub-174-252-41.myvzw.com) left irc: Quit: dexarkun23:52
amaury_or use the _socket module directly23:52
Alex_Gaynorimport posix doesn't affect anything23:52
gutworthugg, melange23:53
GasbaKid (~GasbaKid@ left irc: Ping timeout: 245 seconds23:54
Alex_Gaynorgutworth: is the PSF in the code-in?23:54
Alex_Gaynordid you guys apply?23:54
Alex_Gaynorok, so it wasn't just the DSF23:55
gutworthtoo bad, I spent a lot of time coming up with issues I myself could do :)23:55
amaury_gutworth: thanks for the "No test?" comment on cpython-svn23:56
gutworthand now it seems you need whatever alpha super fast version of chrome just to get it to load23:56
antocuni (~antocuni@dslb-094-220-136-007.pools.arcor-ip.net) left irc: Quit: Leaving23:56
--- Tue Nov 23 201000:00

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