| ROCK_SAP (i=kscoxksx@gateway/tor/x-0ede4ab804241578) joined #pypy. | 00:25 | |
| panni_ (i=hannes@f049005153.adsl.alicedsl.de) left irc: Read error: 110 (Connection timed out) | 00:28 | |
| simonp1 (n=simon@66.151.59.138) left irc: "Download Gaim: http://gaim.sourceforge.net/" | 00:32 | |
| Rhamphoryncus_ (n=rhamph@S010600146c3c91a1.ed.shawcable.net) joined #pypy. | 00:34 | |
| tuxella (n=cordiv_t@5aca29fd.bb.sky.com) left irc: Read error: 113 (No route to host) | 00:45 | |
| Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) left irc: Nick collision from services. | 00:45 | |
| Nick change: Rhamphoryncus_ -> Rhamphoryncus | 00:45 | |
| lene (n=chatzill@trillke.net) left irc: "ChatZilla 0.9.61 [Mozilla rv:1.7.7/20050414]" | 01:03 | |
| stakkars | arigato: hi! still up? You are yet not back into your time zone | 01:10 |
|---|---|---|
| arigato | hi! | 01:14 |
| arigato | I suppose I'm more or less back to my regular hours | 01:14 |
| arigato | I'm trying to make sure I sleep long enough, and using an alarm clock to not sleep too long | 01:15 |
| __doc__ (n=doc_pyth@84-73-209-154.dclient.hispeed.ch) left irc: "Leaving" | 01:17 | |
| fijal (n=fijal@84-73-71-163.dclient.hispeed.ch) left irc: Read error: 110 (Connection timed out) | 01:20 | |
| -christel (i=christel@freenode/staff/gentoo.christel) to $*- [Global Notice] Hi all! To those of you relying on tor for connections to freenode I am terribly sorry to inform you that non tor-gpg (http://freenode.net/irc_servers.shtml#tor) connections have been blocked. This block may be in place for at least a number of days while we try to deal with systematic abuse. Again, I apologise for the inconvenience and hope that we will find a solution soon. Thank you for using freenode and have a good day! | 01:24 | |
| stakkars | I have the opposite problem right now :-) | 01:25 |
| guillemborrell (n=guillem@84.77.173.114) left irc: Remote closed the connection | 01:37 | |
| arigato (n=arigo@c-91d570d5.022-54-67626719.cust.bredbandsbolaget.se) left irc: "See you" | 01:43 | |
| jiwon (n=jiwon@jiwon-home.Stanford.EDU) joined #pypy. | 01:55 | |
| panni_ (i=hannes@f048000095.adsl.alicedsl.de) joined #pypy. | 02:10 | |
| santagada (n=santagad@201.21.248.4) left irc: | 02:54 | |
| headius (n=headius@216-43-127-213.dsl.mcleodusa.net) left irc: | 03:36 | |
| headius (n=headius@216.160.3.79) joined #pypy. | 04:04 | |
| headius (n=headius@216.160.3.79) left irc: | 04:35 | |
| stakkars (n=tismer@208.77.235.7) left irc: Read error: 110 (Connection timed out) | 04:49 | |
| black_13 (i=black_13@cpe-76-187-203-35.tx.res.rr.com) joined #pypy. | 05:06 | |
| black_13 | stackless does not have a coroutine it instead has a tasklet and a channel? | 05:08 |
| headius (n=headius@216.160.3.79) joined #pypy. | 06:01 | |
| panni_ (i=hannes@f048000095.adsl.alicedsl.de) left irc: "( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )" | 06:04 | |
| black_13 | anyone stil alive | 06:08 |
| black_13 (i=black_13@cpe-76-187-203-35.tx.res.rr.com) left irc: Read error: 104 (Connection reset by peer) | 06:52 | |
| mwhudson (n=mwh@62-31-157-102.cable.ubr01.azte.blueyonder.co.uk) left irc: Read error: 110 (Connection timed out) | 07:35 | |
| Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) left irc: "Valid HTML! http://validator.w3.org/ | Support ISO 8601! http://www.cl.cam.ac.uk/~mgk25/iso-time.html" | 08:02 | |
| tuxella (n=cordiv_t@5aca29fd.bb.sky.com) joined #pypy. | 08:44 | |
| elmom (i=elmom@gateway/tor/x-36617a4ec2ed0106) left irc: Remote closed the connection | 09:09 | |
| ROCK_SAP (i=kscoxksx@gateway/tor/x-0ede4ab804241578) left irc: Remote closed the connection | 10:01 | |
| antocuni (n=antocuni@host81-60-dynamic.52-82-r.retail.telecomitalia.it) joined #pypy. | 10:05 | |
| antocuni | hi | 10:05 |
| angry-citizen (n=doc_pyth@84-73-209-154.dclient.hispeed.ch) joined #pypy. | 10:15 | |
| Nick change: angry-citizen -> __doc__ | 10:19 | |
| jewel (n=jewel@61.247.251.10) joined #pypy. | 11:00 | |
| haskelltone | arigo - r49237 - Some remaining usages of id(). | 11:13 |
| arigato (n=arigo@c-91d570d5.022-54-67626719.cust.bredbandsbolaget.se) joined #pypy. | 11:13 | |
| cfbolz (n=cfbolz@p54ABA44D.dip0.t-ipconnect.de) joined #pypy. | 11:14 | |
| cfbolz | morning! | 11:14 |
| xorAxAx | moin carl | 11:14 |
| arigato | morning! | 11:15 |
| cfbolz | I fixed the stats to disregard unicodedata | 11:15 |
| xorAxAx | disregarding the stdlib would be nice as well :) | 11:16 |
| xorAxAx | or fixeol | 11:16 |
| cfbolz | ah, good idea | 11:16 |
| xorAxAx | then you can start parsing the messages for pairprogramming hints :-) | 11:16 |
| xorAxAx | and finally realise that loc is a pretty annoying metric | 11:17 |
| cfbolz | no | 11:17 |
| xorAxAx | in a project for university last year, i was the one with 10 times more loc than any other person of the ~100 people | 11:17 |
| cfbolz | I don't want to hack the java code | 11:17 |
| cfbolz | a project with 100 people? | 11:18 |
| xorAxAx | because i ran fixeol on some packages they gave us | 11:18 |
| xorAxAx | cfbolz: no, 10x10 | 11:18 |
| xorAxAx | but we got the stats of all projects (anonymised, though) | 11:18 |
| cfbolz | right | 11:18 |
| arigato | argh | 11:22 |
| xorAxAx | cfbolz: you could sort by no. of commits, though | 11:23 |
| arigato | ah, maybe not completely argh | 11:24 |
| stargaming | So, if someone always commits small style fixes etc., he wins over the one writing big chunks of great code locally and commiting them eventually? | 11:25 |
| cfbolz | stargaming: yes | 11:25 |
| cfbolz | stargaming: which we want to encourage anyway | 11:25 |
| cfbolz | xorAxAx: I am not sure I can control the thing that well | 11:26 |
| stargaming | Ah, okay. | 11:26 |
| xorAxAx | cfbolz: hehe | 11:26 |
| cfbolz | xorAxAx: it's pretty much a black box process | 11:26 |
| xorAxAx | yes, logilab would lose a few positions after such a change :) | 11:26 |
| cfbolz | the heatmap is a very cool feature | 11:27 |
| cfbolz | arigato: why argh? | 11:28 |
| haskelltone | arigo - r49238 - Use again the invert of the address of the object as its default hash, for Boehm. | 11:28 |
| arigato | cfbolz: we have a test that for a prebuilt RPython instance 'x', hash(x) has the same value on CPython and after translation | 11:29 |
| cfbolz | and that failed? | 11:29 |
| arigato | no, but that's almost impossible to guarantee after a translation to an oo backend | 11:29 |
| arigato | the test passes because it works with llinterp+ootype | 11:29 |
| arigato | but then I realized that the property in question is not so essential for oo translations anyway | 11:30 |
| cfbolz | true | 11:30 |
| arigato | its purpose was to not break prebuilt dicts in ll backends | 11:30 |
| cfbolz | yes | 11:30 |
| arigato | btw, feel like looking at http://wyvern.cs.uni-duesseldorf.de/pypytest/49233/failed/rlib.test.test_rope.py.test_find_iterator_unicode.html? | 11:31 |
| haskelltone | arigo - r49239 - Port of r49238. | 11:33 |
| cfbolz | arigato: ah, yes. that's a python 2.4 bug | 11:33 |
| cfbolz | arigato: isn't the issue for ootype backends moot because they don't have proper prebuilt objects anyway? | 11:33 |
| arigato | ? | 11:34 |
| cfbolz | they have a function that creates the objects at startup, I think | 11:34 |
| arigato | that's a detail | 11:34 |
| arigato | the point is more, should an RPython program be able to assume something about the hash() of objects | 11:34 |
| cfbolz | ah | 11:35 |
| cfbolz | and I guess the answer is no | 11:35 |
| cfbolz | if you store hash(x) somewhere in a prebuilt instance it will be different from hash(x) at runtime for prebuilts | 11:35 |
| arigato | in the prolog interpreter, is the result of get_unify_hash() stored anywhere? | 11:36 |
| arigato | it seems to be stored on the Rule | 11:36 |
| arigato | looks like a perfect example of something that will be broken because of a changing hash across translation | 11:37 |
| arigato | (if there are prebuilt Rules) | 11:38 |
| cfbolz | there aren't | 11:40 |
| cfbolz | there shouldn't :) | 11:40 |
| arigato | maybe time to add a _freeze_ method that explodes with a nice error message :-) | 11:41 |
| haskelltone | arigo - r49240 - These checks are too precise, e.g. when llvm reuses the test. | 11:46 |
| pedronis (n=user@c-91d570d5.022-54-67626719.cust.bredbandsbolaget.se) joined #pypy. | 11:46 | |
| xorAxAx | on cpy-dev about cpy: | 11:48 |
| xorAxAx | "On both of my systems, using -O2 reduces execution time in pystone by 9% | 11:48 |
| xorAxAx | and in pybench by 8%. It's function inlining: "-O3 | 11:48 |
| xorAxAx | -fno-inline-functions" works just as well as "-O2". " | 11:48 |
| haskelltone | arigo - r49241 - This doesn't make much sense any more. | 11:50 |
| headius (n=headius@216.160.3.79) left irc: | 12:23 | |
| jiwon (n=jiwon@jiwon-home.Stanford.EDU) left irc: Read error: 110 (Connection timed out) | 12:24 | |
| jewel (n=jewel@61.247.251.10) left irc: Read error: 110 (Connection timed out) | 12:27 | |
| cfbolz | arigato: what's the proper way to check for python versions? | 12:45 |
| cfbolz | pedronis: morning! | 12:48 |
| pedronis | morning | 12:48 |
| pedronis | cfbolz: sys.version_info ? | 12:49 |
| cfbolz | ah, and then tuple comparison? | 12:49 |
| pedronis | depending what you need, yes | 12:50 |
| haskelltone | cfbolz - r49242 - the test hit a bug under 2.4 and 2.3 | 12:56 |
| xorAxAx | + if sys.version_info > (2, 5): | 13:00 |
| xorAxAx | umm | 13:00 |
| xorAxAx | untested test obviously :) | 13:01 |
| cfbolz | hum | 13:02 |
| haskelltone | cfbolz - r49243 - wrong comparison order - thanks alexander | 13:05 |
| Action: xorAxAx once managed to do a list/tuple comparison in a version check path showing a pretty confusing error message to users ("Your version 0.3 is not new enough, we need at least 0.2" :)) | 13:07 | |
| cfbolz | xorAxAx: I tested it with 2.5 and there it worked :-) | 13:08 |
| pedronis | tdd is hard | 13:09 |
| AntonK | good afternoon all | 13:17 |
| arigato | hi! | 13:19 |
| xorAxAx | hi AntonK | 13:21 |
| xorAxAx | AntonK: seems like chris doesnt think that bug 1828 is important :) | 13:21 |
| AntonK | it's not Chris | 13:22 |
| AntonK | loop stuff is mostly Devang's | 13:22 |
| AntonK | dunno, why he didn't comment on it:( | 13:22 |
| xorAxAx | ah | 13:23 |
| AntonK | I think, I'll ping him :) | 13:23 |
| AntonK | ping'ed :) | 13:31 |
| AntonK | will see | 13:31 |
| jewel (n=jewel@61.247.251.10) joined #pypy. | 13:37 | |
| AntonK (n=asl@asl.dorms.spbu.ru) left irc: Read error: 110 (Connection timed out) | 14:04 | |
| tav (n=tav@91.84.124.179) left irc: Read error: 110 (Connection timed out) | 14:07 | |
| cfbolz (n=cfbolz@p54ABA44D.dip0.t-ipconnect.de) left irc: "Leaving" | 14:23 | |
| jewel (n=jewel@61.247.251.10) left irc: Read error: 110 (Connection timed out) | 14:36 | |
| tverwaes_ | in python you can always do module.__dict__... | 15:27 |
| tverwaes_ | is there a way of doing this on the __main__? | 15:28 |
| Nick change: tverwaes_ -> tverwaes | 15:28 | |
| xorAxAx | globals()? | 15:28 |
| tverwaes | mmh right ....... and something on which I can call __dict__? | 15:29 |
| exarkun | __main__.__dict__ | 15:29 |
| exarkun | #python might be a better channel | 15:30 |
| tverwaes | mina is not defined | 15:30 |
| tverwaes | yeah.. sorry for that :) | 15:30 |
| exarkun | import it. | 15:30 |
| tverwaes | o..k.. :) | 15:31 |
| tverwaes | thnx | 15:31 |
| pedronis (n=user@c-91d570d5.022-54-67626719.cust.bredbandsbolaget.se) left irc: Read error: 113 (No route to host) | 15:38 | |
| antocuni (n=antocuni@host81-60-dynamic.52-82-r.retail.telecomitalia.it) left irc: "Leaving" | 15:56 | |
| exarkun | Hrm. | 16:01 |
| exarkun | So there's already an implementation of os.execve. | 16:01 |
| exarkun | It's disabled though. "work in progress". | 16:01 |
| exarkun | It has a test too. Not a great one, but an okay one. | 16:02 |
| xorAxAx | for very slow values of progress :) | 16:02 |
| exarkun | If I enable the function, the test passes. | 16:02 |
| exarkun | So is it actually a really terrible test? Or does execve really work? | 16:02 |
| xorAxAx | maybe blame the skip line and ask the author? :) | 16:02 |
| exarkun | who's afa? | 16:03 |
| xorAxAx | finger him | 16:03 |
| exarkun | uh, right | 16:03 |
| xorAxAx | ah, amaury | 16:04 |
| exarkun | like there are computers on the internet that still running finger daemons | 16:04 |
| xorAxAx | no, but finger runs locally as well :) | 16:04 |
| exarkun | okay he's not here now what :) | 16:04 |
| xorAxAx | good question. does the log message give away any reasons | 16:04 |
| xorAxAx | ...? | 16:04 |
| exarkun | it talks about win32 | 16:05 |
| xorAxAx | hmm | 16:05 |
| exarkun | but it doesn't mention execve specifically | 16:05 |
| xorAxAx | then execve should only be registered on windows | 16:05 |
| xorAxAx | +not | 16:05 |
| exarkun | it's registered with a hasattr() check | 16:06 |
| exarkun | it was disabled in a different revision as the test was skipped though | 16:06 |
| exarkun | earlier | 16:06 |
| exarkun | arigato disabled the registration | 16:07 |
| exarkun | arigato: awake yet? :) | 16:07 |
| exarkun | AppTest* means what, again? | 16:08 |
| AntonK (n=asl@asl.dorms.spbu.ru) joined #pypy. | 16:08 | |
| xorAxAx | exarkun: that it is executed on top of py.py | 16:08 |
| xorAxAx | (i.e. the pypy interpreter) | 16:08 |
| exarkun | but.. untranslated? | 16:09 |
| xorAxAx | yes, it doesnt translate | 16:09 |
| exarkun | an AppTest which calls os.execve doesn't actually test that pypy can ffi to execve(2) correctly | 16:09 |
| exarkun | it just tests that the rpython wrapper around the ffi call works? And then calls CPython's execve function? | 16:10 |
| xorAxAx | exarkun: that might be the case, in some cases it falls back to cpython, yes. you can try with a raise in the code | 16:10 |
| exarkun | if I put a raise in pypy.module.posix.interp_posix.execve right before the os.execve call, it gets raised | 16:14 |
| exarkun | ah | 16:14 |
| exarkun | a translation fails with Exception: unexpected prebuilt constant: <built-in function execve>, which means there is no ffi glue for execve, at least not registered, right? | 16:14 |
| exarkun | and indeed I see execv but not execve in ll_os.py | 16:15 |
| exarkun | oops, and test_ll_os has no test for execv | 16:16 |
| xorAxAx | so its missing :) | 16:19 |
| exarkun | I guess I have some idea of what needs to be done now. Thanks for the help. | 16:21 |
| xorAxAx | hmm, there is some magic involved here | 16:22 |
| xorAxAx | depending on the OS, execve might be even faked in some .py module from the stdlib i think | 16:22 |
| exarkun | I'm going to ignore windows :) | 16:22 |
| xorAxAx | hmm, nop | 16:22 |
| exarkun | os.py is really disgusting :( | 16:23 |
| xorAxAx | yep | 16:23 |
| exarkun | how do I get stdout for a test with py.test? | 16:24 |
| xorAxAx | -s | 16:24 |
| xorAxAx | (if you mean "stdout while testing" | 16:24 |
| xorAxAx | and you also get it if it fails i think | 16:24 |
| exarkun | yea, thanks | 16:24 |
| exarkun | indeed, but that's so much extra work :) | 16:24 |
| exarkun | are there any path manipulation apis in pypy or pylib? | 16:29 |
| xorAxAx | yes, on which level? | 16:30 |
| exarkun | high? :) | 16:30 |
| exarkun | something like FilePath(__file__).sibling('foo.py') would be nice | 16:30 |
| xorAxAx | well, do you need it in normal python code or a test? | 16:30 |
| exarkun | ah, normal python is fine | 16:30 |
| xorAxAx | py.path.local etc. | 16:30 |
| exarkun | hmm | 16:46 |
| exarkun | oh cool | 16:46 |
| black_13 (i=black_13@76.187.203.35) joined #pypy. | 16:48 | |
| black_13 | does stackless have a coroutine or does it implement coroutines as tasklets? | 16:49 |
| arigato | black_13: Stackless Python implements tasklets only. PyPy's stackless is based on a choice of coroutines, greenlets or tasklets | 16:50 |
| arigato | at the moment tasklets are expressed on top of coroutines but that's not essential and will eventually be fixed for better performance | 16:51 |
| black_13 | do you mean that stackless python and python will unify at some point | 16:51 |
| arigato | by Stackless Python I mean www.stackless.com | 16:52 |
| arigato | PyPy has the same features | 16:52 |
| arigato | the standard CPython has no coroutine nor tasklets | 16:52 |
| arigato | and is unlikely to ever have | 16:52 |
| arigato | dunno where you got the impression that CPython and Stackless Python would unify - that idea has been rejected long ago | 16:53 |
| xorAxAx | which reminds me at the callback-triggers-stack-unwind-issue :) | 16:53 |
| exarkun | is {str: str} even close to a valid element in the arg list extdef() takes? | 16:54 |
| arigato | exarkun: I think so, yes | 16:55 |
| exarkun | cool | 16:55 |
| exarkun | I wonder where my dict is going then | 16:55 |
| black_13 | arigato, i was just asking a question | 16:56 |
| arigato | black_13: sorry, maybe I was unclear then | 16:57 |
| black_13 | arigato, really what i am trying to understand how coroutines work and if they are usefull in implementing state machines | 16:57 |
| arigato | coroutines are basically "better generators" | 16:58 |
| black_13 | arigato, i have been reading wikipedia artictles (which may or may not be valid) that coroutines lead to better implementation of state machines | 16:58 |
| exarkun | coroutines are the opposite of explicit state machines | 16:58 |
| exarkun | they let you express state in stack frames and function locals | 16:58 |
| arigato | yes, it seems to me that coroutines are useful precisely because you can solve problems with them without having to write state machines | 16:59 |
| black_13 | exarkun, hum can you think of an example i am only recently doing python | 16:59 |
| exarkun | black_13: have you used generators in python? | 16:59 |
| black_13 | exarkun, i have only used python with wxpython to for small utility apps | 17:00 |
| black_13 | kind of fast mfc | 17:00 |
| black_13 | remember MFC | 17:00 |
| black_13 | http://en.wikipedia.org/wiki/Microsoft_Foundation_Classes | 17:01 |
| black_13 | win32 | 17:01 |
| black_13 | i found that i can write small throw away apps very quickly | 17:02 |
| black_13 | with python | 17:02 |
| exarkun | black_13: http://www.python.org/dev/peps/pep-0255/ discusses python generators, how and why you might use them | 17:03 |
| black_13 | hey i have seen a lot of these "pep" does what does pep stand for | 17:03 |
| exarkun | black_13: much of the information applies to coroutines as well, since as arigato mentioned, coroutines are basically better generators. | 17:03 |
| black_13 | ok | 17:04 |
| black_13 | does the yield idea apply to a tasklet | 17:04 |
| black_13 | from what i have seen you create a tasklet with a function | 17:04 |
| black_13 | similar to O/S thread has a thread function/proc | 17:05 |
| Action: arigato discovers /dev/full | 17:05 | |
| black_13 | full or null? | 17:05 |
| arigato | full... | 17:05 |
| black_13 | really never heard of that none | 17:05 |
| black_13 | what is full | 17:05 |
| arigato | a pseudo-device that pretends to be full... simply | 17:06 |
| black_13 | ok | 17:06 |
| arigato | black_13: you should try to write small examples of generators, first | 17:06 |
| black_13 | maybe your correctly | 17:06 |
| black_13 | right i mean | 17:06 |
| arigato | tasklets are a slightly different (older) concept from Stackless Python | 17:06 |
| black_13 | something | 17:06 |
| arigato | tasklets are a bit more like microthreads | 17:07 |
| arigato | generators are also a bit like microthreads, but with language syntax that make them look different | 17:07 |
| black_13 | going back to the state machine | 17:07 |
| black_13 | i have seen in C/C++ that state machine come in variety of table driven ideas | 17:08 |
| black_13 | or pointers to state objects (an i knowyou stated previous about a stack frame) but ... | 17:09 |
| black_13 | well i dont full understand the stack frame ... | 17:09 |
| arigato | state machines like this are not something you need every day | 17:09 |
| black_13 | games? | 17:10 |
| arigato | they only make sense in specific contexts, like writing a regular expression engine | 17:10 |
| black_13 | no i dont need that | 17:10 |
| arigato | for games, they are typically two approaches (not state machines): | 17:10 |
| black_13 | arigato, please tell | 17:10 |
| arigato | "regular" object-oriented programming, where you store each entity's state as attributes of the object | 17:10 |
| arigato | or a version that uses generators | 17:11 |
| black_13 | i totally get tat | 17:11 |
| arigato | where an entity's behavior progresses by running a generator (or some other kind of microthread) dedicated to that entity | 17:11 |
| black_13 | i have seen the stackless tutorial take that approache where the state is contained in an object | 17:11 |
| Action: arigato -> away | 17:12 | |
| black_13 | arigato, http://members.verizon.net/olsongt/stackless/why_stackless.html this talks about that | 17:12 |
| black_13 | arigato, still here? ( you said away) | 17:13 |
| arigato | about to leave :-) | 17:14 |
| arigato | sorry, talk to you later | 17:14 |
| black_13 | when are you on next? | 17:14 |
| black_13 | again? | 17:14 |
| black_13 | approximately | 17:14 |
| arigato | I'm often around, but not at fixed hours | 17:15 |
| arigato | will be around tomorrow, European time | 17:15 |
| black_13 | well take care thanks for the pointers | 17:15 |
| black_13 | i am central us | 17:16 |
| black_13 | texas | 17:16 |
| black_13 | what time where you are at? | 17:16 |
| arigato | it's evening, 6pm | 17:16 |
| xorAxAx | 5pm | 17:17 |
| arigato | see you! | 17:17 |
| black_13 | later then | 17:17 |
| black_13 | i wonder can yield be combined with a tasklet? | 17:17 |
| arigato (n=arigo@c-91d570d5.022-54-67626719.cust.bredbandsbolaget.se) left irc: Remote closed the connection | 17:17 | |
| tav (n=tav@91.84.164.61) joined #pypy. | 17:28 | |
| exarkun | Hmm | 17:30 |
| exarkun | should str2charp really accept strings with \0 in them? | 17:31 |
| xorAxAx | if you mean it should check that? | 17:34 |
| xorAxAx | makes some sense | 17:35 |
| xorAxAx | even though it couldnt use memcpy then | 17:35 |
| exarkun | does it use memcpy now? it looks like a for loop | 17:36 |
| xorAxAx | no, it doesnt | 17:37 |
| xorAxAx | pypy has a raw_memcopy function | 17:38 |
| xorAxAx | that could be used | 17:38 |
| black_13 (i=black_13@76.187.203.35) left irc: Read error: 110 (Connection timed out) | 17:42 | |
| santagada (n=santagad@201.21.248.4) joined #pypy. | 17:44 | |
| panni_ (i=hannes@f048000095.adsl.alicedsl.de) joined #pypy. | 18:10 | |
| haskelltone | exarkun - r49252 - llimpl for os.execve | 18:13 |
| xorAxAx | cool | 18:14 |
| exarkun | humm | 18:17 |
| exarkun | debug: OperationError: | 18:17 |
| exarkun | debug: operror-type: AttributeError | 18:17 |
| exarkun | debug: operror-value: 'module' object has no attribute 'name' | 18:17 |
| exarkun | did I do that? | 18:17 |
| fijal (n=fijal@84-73-93-20.dclient.hispeed.ch) joined #pypy. | 18:17 | |
| fijal | hi | 18:18 |
| exarkun | fijal: 'lo | 18:18 |
| fijal | exarkun: hi | 18:18 |
| xorAxAx | exarkun: you need to use a makefile target that defines enough debug flags and try again | 18:24 |
| haskelltone | niko - r49253 - update the debug info we print out: now the line number corresponds to the line in the file, and we emit local variable debug information with the name of the variable from the rpython source. this h | 18:24 |
| xorAxAx | then you will see where the exception comes from | 18:26 |
| xorAxAx | exarkun: was that a normal pypy-c? | 18:29 |
| exarkun | what does normal mean | 18:30 |
| xorAxAx | ignore that word :) | 18:30 |
| xorAxAx | i guess you tried to compile pypy with your changes and then it crashed at startup? | 18:30 |
| exarkun | hmm | 18:30 |
| exarkun | it doesn't crash at startup | 18:31 |
| xorAxAx | when? | 18:31 |
| exarkun | I didn't think it crashed at all, but now that you mention that, I suspect it could be the child process which is crashing | 18:31 |
| exarkun | and the stderr passing through | 18:31 |
| xorAxAx | yep, thats it | 18:32 |
| xorAxAx | the above lines are generated by the outer most rpython level exception handler | 18:32 |
| xorAxAx | hmm, its weird that its not catched | 18:32 |
| xorAxAx | because it is an operror which is the wrapper for interp-exceptions | 18:33 |
| xorAxAx | exarkun: what are you executing? | 18:35 |
| exarkun | the tests in twisted.test.test_process | 18:36 |
| exarkun | so it's a bit annoying to try to narrow down | 18:36 |
| exarkun | particularly since stderr and stdout are not synchronized | 18:36 |
| exarkun | oh hey, there it is again | 18:37 |
| exarkun | in a nice convenient place too | 18:37 |
| exarkun | woah, hm | 18:38 |
| exarkun | ah, hm | 18:38 |
| exarkun | it might be pty related | 18:38 |
| xorAxAx | you can synchronize them by using the same stream :) | 18:38 |
| exarkun | xorAxAx: well yea, like I said, annoying :) | 18:39 |
| exarkun | it is happening after the test code in the child process finishes, I think | 18:39 |
| exarkun | at least, if it causes the interpreter to exit, it must be | 18:40 |
| exarkun | since the test code is running | 18:40 |
| exarkun | this test fails http://twistedmatrix.com/trac/browser/trunk/twisted/test/test_process.py#L1675 | 18:40 |
| exarkun | http://twistedmatrix.com/trac/browser/trunk/twisted/test/process_tty.py is the child process it runs | 18:41 |
| exarkun | instead of getting "hello world!\r\nhello world!\r\n" | 18:41 |
| exarkun | oh, crud | 18:41 |
| exarkun | no, I'm wrong, the code isn't running at all | 18:42 |
| exarkun | hmm, maybe something environment related | 18:44 |
| jewel (n=jewel@61.247.251.10) joined #pypy. | 18:44 | |
| xorAxAx | PYTHONSTARTUP maybe? :) | 18:45 |
| exarkun | oh | 18:45 |
| exarkun | or maybe it's just -u | 18:45 |
| xorAxAx | or -u :) | 18:45 |
| exarkun | yea okay I guess that's it | 18:45 |
| exarkun | not implemented at all? or..? | 18:45 |
| xorAxAx | it is, but it was broken | 18:46 |
| xorAxAx | and it doesnt to be tested | 18:46 |
| xorAxAx | :) | 18:46 |
| xorAxAx | its app_main.py in translator/goal | 18:47 |
| xorAxAx | and the os in there is just christians nanos | 18:47 |
| xorAxAx | exarkun: hmm, there is a test that doesnt fail, hmm | 19:01 |
| exarkun | it seems like testing anything in app_main would be pretty difficult | 19:01 |
| exarkun | at least, I have no idea what the execution context is actually like for that code | 19:01 |
| exarkun | so I don't know how you'd set up a test | 19:02 |
| xorAxAx | there is a test | 19:02 |
| xorAxAx | test2/test_app_main | 19:02 |
| exarkun | but if it passes, it must not be a valid test ;) | 19:02 |
| xorAxAx | yes | 19:03 |
| xorAxAx | ahhh! | 19:03 |
| xorAxAx | there is an os import at the bottom | 19:03 |
| xorAxAx | ok | 19:04 |
| xorAxAx | hmm | 19:04 |
| Action: xorAxAx has an idea | 19:06 | |
| haskelltone | niko - r49254 - fix math.modf in the jvm: 1. the code in PyPy.java was slightly wrong 2. the static signature cannot reference generated objects, so when they are returned we must use jObject instead and insert a dow | 19:06 |
| haskelltone | niko - r49255 - use the new is_generated() test in InterlinkFunction to determine when the rewrite the signature to use Object, since it applies there as well -- before we just checked if the type was a reference typ | 19:10 |
| Action: xorAxAx fixes it | 19:16 | |
| xorAxAx | exarkun: fixed | 19:22 |
| haskelltone | xoraxax - r49256 - app_main does not use the normal os module now anymore. Nanos can expose its module even in scenarios without a pypy interpreter - this makes app_main testable again. This should fix the broken -u opt | 19:22 |
| exarkun | neat | 19:22 |
| jewel_ (n=jewel@61.247.251.10) joined #pypy. | 19:22 | |
| jewel (n=jewel@61.247.251.10) left irc: Read error: 110 (Connection timed out) | 19:44 | |
| Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) joined #pypy. | 19:45 | |
| haskelltone | xoraxax - r49257 - I think these files are not needed anymore. | 19:49 |
| xorAxAx | if not irreg: | 19:52 |
| xorAxAx | self.log.info("All exceptblocks seem sane") | 19:52 |
| xorAxAx | and no else case ... | 19:52 |
| xorAxAx | i guess nobody would spot that a line would be missing in case of a problem | 19:52 |
| exarkun | indeed | 19:54 |
| haskelltone | xoraxax - r49258 - Remove two status messages in the regular code path. | 19:56 |
| fijal | xorAxAx: I would | 19:59 |
| gordonh (n=gordonh@c-24-218-151-219.hsd1.ma.comcast.net) left irc: | 20:15 | |
| gordonh (n=gordonh@c-24-218-151-219.hsd1.ma.comcast.net) joined #pypy. | 20:15 | |
| gordonh (n=gordonh@c-24-218-151-219.hsd1.ma.comcast.net) left irc: Remote closed the connection | 20:15 | |
| gordonh (n=gordonh@c-24-218-151-219.hsd1.ma.comcast.net) joined #pypy. | 20:16 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:17 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:18 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:18 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:19 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:19 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:21 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:21 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:23 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:23 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Connection reset by peer | 20:24 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:24 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:25 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:25 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:25 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:26 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:29 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:29 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:34 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:34 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:34 | |
| santagada (n=santagad@201.21.248.4) left irc: | 20:34 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:35 | |
| headius (n=headius@216.160.3.79) joined #pypy. | 20:40 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:40 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:41 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 20:53 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 20:53 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 21:05 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 21:05 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 21:11 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 21:11 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 21:11 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 21:12 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 21:13 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 21:13 | |
| stakkars_ (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Connection reset by peer | 21:14 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 21:15 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 104 (Connection reset by peer) | 21:20 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 21:20 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Success | 21:21 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 21:22 | |
| jewel_ (n=jewel@61.247.251.10) left irc: Connection timed out | 21:43 | |
| black_13 (i=black_13@76.187.203.35) joined #pypy. | 21:44 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) left irc: Read error: 110 (Connection timed out) | 21:44 | |
| stakkars (n=tismer@adsl-71-133-157-82.dsl.irvnca.pacbell.net) joined #pypy. | 22:15 | |
| black_13 (i=black_13@76.187.203.35) left irc: Read error: 110 (Connection timed out) | 22:18 | |
| stephenjudkins (n=stephen@c-76-115-137-222.hsd1.or.comcast.net) joined #pypy. | 22:28 | |
| fijal_ (n=fijal@84-73-93-20.dclient.hispeed.ch) joined #pypy. | 22:39 | |
| fijal (n=fijal@84-73-93-20.dclient.hispeed.ch) left irc: Read error: 104 (Connection reset by peer) | 22:39 | |
| Nick change: haskelltone -> skulltone | 22:40 | |
| black_13 (i=black_13@cpe-76-187-203-35.tx.res.rr.com) joined #pypy. | 22:40 | |
| stephenjudkins_ (n=stephen@c-76-115-137-222.hsd1.or.comcast.net) joined #pypy. | 22:53 | |
| fijal_ (n=fijal@84-73-93-20.dclient.hispeed.ch) left irc: Read error: 110 (Connection timed out) | 23:01 | |
| stephenjudkins (n=stephen@c-76-115-137-222.hsd1.or.comcast.net) left irc: Read error: 110 (Connection timed out) | 23:10 | |
| black_13 (i=black_13@cpe-76-187-203-35.tx.res.rr.com) left irc: Read error: 104 (Connection reset by peer) | 23:20 | |
| --- Sun Dec 2 2007 | 00:00 | |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!