[00:18] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[00:24] _hannes (n=yuuhu@i577B69FB.versanet.de) left irc: Read error: 104 (Connection reset by peer)
[00:34] hpk (n=hpk@193.91.97.17) left #pypy ("Leaving").
----- silence for 34 minutes ----- [01:08] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 110 (Connection timed out)
----- silence for 22 minutes ----- [01:30] rhymes (n=rhymes@host131-75.pool8254.interbusiness.it) left irc: "supercalifragiliblahblah"
----- silence for 1 hr and 8 minutes ----- [02:38] Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) joined #pypy.
----- silence for 23 minutes ----- [03:01] sanxiyn (n=tinuviel@211.104.100.240) joined #pypy.
----- silence for 2 hr and 40 minutes ----- [05:41] brian (i=nobody@unaffiliated/brian) joined #pypy.
----- silence for 1 hr and 58 minutes ----- [07:39] Shoragan (n=shoragan@d072.apm.etc.tu-bs.de) joined #pypy.
----- silence for 28 minutes ----- [08:07] Gromit (n=klix@does-d9b91925.pool.mediaWays.net) joined #pypy.
[08:20] GabeW (n=gwachob@pdpc/supporter/professional/GabeW) left irc: "leaving"
----- silence for 50 minutes ----- [09:10] Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) left irc: Excess Flood
[09:11] Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) joined #pypy.
[09:22] Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) left irc: Excess Flood
[09:23] Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) joined #pypy.
[09:32] rhymes (n=rhymes@host131-75.pool8254.interbusiness.it) joined #pypy.
[09:32] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
----- silence for 16 minutes ----- [09:48] rhymes (n=rhymes@host131-75.pool8254.interbusiness.it) left irc: "supercalifragiliblahblah"
[10:02] arre (n=ac@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[10:08] sanxiyn (n=tinuviel@211.104.100.240) left #pypy ("Bye").
[10:09] auc (n=auc@logilab.net2.nerim.net) joined #pypy.
[10:10] Nick change: cfbolz_food -> cfbolz
[10:17] cfbolz (n=carlson@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 104 (Connection reset by peer)
[10:17] mwh (n=mwh@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 104 (Connection reset by peer)
[10:17] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 104 (Connection reset by peer)
[10:17] lac (n=lac@pdpc/supporter/gold/lac) left irc: Read error: 104 (Connection reset by peer)
[10:17] jacob22|home (n=jacob@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 104 (Connection reset by peer)
[10:17] arre (n=ac@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 104 (Connection reset by peer)
[10:17] cfbolz_ (n=carlson@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[10:17] stakkars (n=tismer@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 104 (Connection reset by peer)
[10:17] -ChanServ (ChanServ@services.)- You do not have channel operator access to [#pypy-sync]
[10:17] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[10:17] mwh (n=mwh@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[10:21] <mwh> morning
[10:23] adim (n=adim@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[10:24] johahn (n=johahn@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[10:25] tic (n=vision@c-8faae255.019-35-67626717.cust.bredbandsbolaget.se) joined #pypy.
[10:39] arre (n=ac@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[10:50] <braintone> cfbolz - r20876 - plans for today
----- silence for 18 minutes ----- [11:08] arigo (n=arigo@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[11:11] <braintone> nik - r20877 - (ale, nik) for now raise OSError from C code on socket errors. more sophisticated error tests.
[11:14] <braintone> mwh - r20878 - ll_os_lseek now returns a long long. this required changing far too many places in the code...
[11:28] lac2 (n=lac@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[11:28] Nick change: lac2 -> lac
----- silence for 21 minutes ----- [11:49] _hannes (n=yuuhu@i577B69FB.versanet.de) joined #pypy.
[11:49] <braintone> nik - r20879 - (ale, nik) error handling for socket creation complete (but convoluted).
[11:56] <braintone> mwh - r20880 - implement converting Signeds to SignedLongLongs. be more honest about .knowntype on SomeIntegers. (no-one else has noticed that we broke translate_pypy yet...)
[12:04] <braintone> mwh - r20881 - (mwh, johahn) Copy test over from the rpython tests (...) and made it pass by implementing a couple more operations in int.h
[12:14] Gromit (n=klix@does-d9b91925.pool.mediaWays.net) left irc: Read error: 110 (Connection timed out)
[12:16] <braintone> cfbolz - r20882 - (pedronis, cfbolz, arigo) Support for 'direct_call': introduced GraphState, changed the way the new graphs are patched around, pushed, pulled, etc.
[12:23] <braintone> cfbolz - r20883 - remove this very very very old __main__. it's broken. it sucks. nobody should use it.
[12:24] xyz359 (n=xyz359@pool-68-238-168-248.tampfl.fios.verizon.net) left irc: Read error: 104 (Connection reset by peer)
[12:24] Nick change: cfbolz_ -> cfbolz
[12:31] <braintone> mwh - r20884 - (mwh, johahn) implement truncation of signed long long to signed long (need to do all this stuff for unsigned variants at some point too ... fun).
[12:34] <braintone> nik - r20885 - (ale, nik) implemented socket.fileno() and socket.close()
[12:35] nik__ (n=chatzill@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[12:35] Srekel3 (n=Srekel3@dhcp-5.idc.cs.chalmers.se) joined #pypy.
[12:35] <Srekel3> hi!
[12:35] <Srekel3> is the sprint underway?
[12:38] <Srekel3> I'm from Gothenburg and I've been thinking about visiting to see how things are going, if there's room. I can program Python, but I have little to no knowledge of its internals so I don't know if I can be of any help
[12:39] <Srekel3> regardless, it would be interesting to see how you are working :) especially if I learn something
[12:41] aleale (n=andersle@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[12:41] <arigo> Srekel3: hi!
[12:41] <arigo> sure, you are welcome to show up
[12:42] <lac> Götebergsgatan 22
[12:42] <lac> between Valand and Vasaplatsen tram stops
[12:43] <Srekel3> cool
[12:43] <lac> Götabergsgatan
[12:43] <Srekel3> I might do that tomorrow after school/work
[12:43] <lac> (cannot spell the name of my own street, shame)
[12:43] <Srekel3> are you there like constantly :) or do you go home at 22.00 or something?
[12:43] <lac> I live here.
[12:43] <lac> some others are staying here,
[12:43] <lac> some go to hotels.
[12:44] <brian> I don't get pypy.
[12:44] <braintone> rxe - r20886 - Proof of concept tasklet switching in rpython.
[12:44] <lac> you cannot get the sources, or you do not understand why we want to do this in the first place?
[12:44] <brian> I don't get it.
[12:45] <brian> Like, the purpose.
[12:45] <lac> we want a more flexible self-hosted version of python.
[12:45] <Srekel3> ok, so basically people go home when they (or you ;))feel like it? :)
[12:46] <lac> then we can do many things which you cannot do in the current version.
[12:46] <cfbolz> Srekel3: yes, something like that
[12:46] <Srekel3> alright
[12:46] <cfbolz> Srekel3: we are away for food/sleep sometimes
[12:46] <lac> Srekel, Saturday night we are all going to Silvis for dinner 1800
[12:47] <Srekel3> ok. how much does food cost there? (aboutish)
[12:47] <braintone> cfbolz - r20887 - (cfbolz, arigo, pedronis) Support for reading from structs. Hints about immutable structs.
[12:48] <cfbolz> Srekel3: depends on whether you are a vegetarian or not
[12:48] <Srekel3> not
[12:48] <lac> they are cooking a whole lamb for us. So it is not order off the menu.
[12:48] <Srekel3> ah, cool
[12:49] <lac> 450 SEK for lamb eaters. If you want to come, let me know now.
[12:49] <lac> or as soon as you know.
[12:49] <lac> They are marinating the lamb now ... :-)
[12:49] <brian> 'mmmm lamb
[12:49] <Srekel3> ouch. can't afford that I'm afraid
[12:49] <lac> This was just -- we son't be home Saturday night.
[12:49] Action: brian steals the lamb.
[12:50] <cfbolz> Srekel3: well, you can eat vegetarian, of course :-)
[12:51] <cfbolz> which is _a lot_ cheaper
[12:51] <Srekel3> true. :) how much cheaper would that be? (I might have other plans on saturday though, I'm not 100% sure)
[12:52] <cfbolz> no idea, but cheaper
[12:52] <Srekel3> is this it? http://www.restaurangguiden.se/gbg/linnesta/silvis/main.htm
[12:53] <cfbolz> probably
[12:53] Topic changed on #pypy by cfbolz!n=carlson@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se: PyPy, the flexible Snake (log at: http://tismerysoft.de/pypy/irc-logs/pypy/) food and development
[12:57] Action: Srekel3 leaves for a while - game design workshop :)
[13:06] stakkars (n=tismer@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[13:16] <braintone> mwh - r20888 - (mwh, johahn) fix up builtin_isinstance wrt the changes to how integers are annotated.
[13:25] <braintone> mwh - r20889 - implement __abs__ for our integer types.
[13:26] <braintone> mwh - r20890 - a r_longlong case in wrap()
[13:33] jacob22|home (n=jacob@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[13:34] <jacob22|home> The Pypy project needs some of these: http://www.cenqua.com/pairon/
[13:40] <mwh> we need some kind of quantum-accelerated supercomputer for translate_pypy
[13:40] <xorAxAx> if you parallelise it, you can run it on every cheap cluster
[13:41] <mwh> if we had some ham, we could have some ham and egg sandwiches, if we had some egss
[13:45] <Rhamphoryncus> mwh: why does that sentence remind me of the "a if b else c" syntax? :)
[13:45] <mwh> well, it's kind of the opposite :)
[13:46] <Rhamphoryncus> hmm indeed. "conditional expr conditional", rather than "expr conditional expr" :)
[13:46] <mwh> it's a fairly common phrase
[13:46] <mwh> (well, 262 hits for "If we had some ham, we could have ham and eggs, if we had some eggs" on google)
[13:46] <mwh> don't know the source
[13:46] Action: Rhamphoryncus hopes you don't code like that ;)
[13:47] <braintone> arigo - r20891 - Fixed arithmetic on bools.
[13:48] <braintone> rxe - r20892 - Try to track some obscure bug with list length and is true - but failed.
[13:49] <mwh> ok, we might have unbroken translate_pypy
[13:49] <mwh> (inlining now)
[13:50] <mwh> compilation will probably fail, but hopefully only in shallow ways :)
[13:51] <braintone> mwh - r20893 - (johahn, mwh) more operations for llongs
----- silence for 21 minutes ----- [14:12] Gromit (n=klix@does-d9b90af4.pool.mediaWays.net) joined #pypy.
[14:13] <braintone> mwh - r20894 - (mwh, johahn) more operations on long longs, enough for translate_pypy to work again. simple tests suggest os.lseek (the only function so far that returns a long long) works. woohoo!
[14:17] <braintone> arigo - r20895 - (pedronis, arigo) Fixed misusages of s.is_constant() for non-immutable objects -- in this case, s.const is valid as far as its identity is concerned, but s.const can still get mutated.
[14:24] <braintone> arigo - r20896 - (pedronis, arigo) Moved bits around this file, to hide them from geninterp.
----- silence for 29 minutes ----- [14:53] <braintone> cfbolz - r20897 - (cfbolz; arigo, pedronis floating around) Make the first arraytest pass by adding some operations
[15:03] arigo (n=arigo@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Remote closed the connection
[15:05] <braintone> nik - r20898 - (ale, nik) first steps implementing socket.connect. not rtyper support, yet. some mystery test failures remaining.
[15:07] arre (n=ac@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: "using sirc version 2.211+KSIRC/1.3.11"
[15:08] johahn (n=johahn@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc:
----- silence for 16 minutes ----- [15:24] aleale (n=andersle@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 110 (Connection timed out)
----- silence for 29 minutes ----- [15:53] <Srekel3> see you guys tomorrow (maybe)!
[15:53] Srekel3 (n=Srekel3@dhcp-5.idc.cs.chalmers.se) left irc:
----- silence for 1 hr and 51 minutes ----- [17:44] <braintone> tismer - r20900 - added backend optimizations. scheduler_run juggled 15 variables around, now it is down to 12. But I still think this is too much!
[17:54] <stakkars> crazy, crazy, phantastic!!!
[17:55] <stakkars> we gonna have basic coroutine support in RPython, I guess!
[18:02] Gromit_ (n=klix@does-d9b90ae5.pool.mediaWays.net) joined #pypy.
[18:03] <Gromit_> re
[18:03] Gromit (n=klix@does-d9b90af4.pool.mediaWays.net) left irc: Nick collision from services.
[18:03] Nick change: Gromit_ -> Gromit
[18:05] <stakkars> I'm going to vote for enabling the stackless feature all the time
[18:06] <stakkars> the advantages are manyfold.
[18:06] <stakkars> - we can do a real garbage collector by this
[18:06] <stakkars> - we va forget about clumsy state engines for iterators and stuff
[18:07] <stakkars> instead, just use coroutine switching.
[18:07] <stakkars> I just debugged though a longer Stackless example in C - phantastic!
[18:07] rhymes (n=rhymes@host209-75.pool8254.interbusiness.it) joined #pypy.
[18:07] <Gromit> did you manage to improved the performance?
[18:08] <stakkars> I didn't care ATM.
[18:08] <stakkars> the chance tohave coroutinesat the primitive level is so phascinating
[18:08] <Gromit> IC, ok
[18:08] <stakkars> (I again need more space :)
[18:09] <stakkars> can you send me some?
[18:09] <Gromit> of course coroutines are cool, iterators are just degenerated coroutines :)
[18:10] <braintone> ericvrp - r20902 - test for playing with long running processes
[18:10] <stakkars> yeah. I will write one basic coroutine class which has a little less than Greenlets have
[18:10] <stakkars> and for every value type you wanttopassbetween coroutines, you derive a class, where we specialize upon.
[18:11] <stakkars> or maybe with out a class, specializing right away on the switch() call. :-))
[18:11] <stakkars> if you want to see the power, just run test_all.py test_tasklets.py
[18:12] <Gromit> I just had a similar experience:
[18:12] <stakkars> then use your C++ environment on the generated source and single step. It is a journey through ALice'sWonderland
[18:12] <Gromit> together with Florian Bösch i just invented a dead simple pythonic object code format :)
[18:13] <stakkars> what's that? How is it simpler than marshal?
[18:13] <Gromit> an object code format for machine code, i should add
[18:14] <Gromit> perhaps usable as rpython backend, if there is an assembler implementation in python
[18:15] <Gromit> but we didn't think about pypy, when we developed this weird idea
[18:15] <stakkars> for what did you need it?
[18:16] <Gromit> Florian had the desire to implemet a simple 3d lib in python
[18:17] <Gromit> and so we started to discuss how this should be in a straight forward manner
[18:17] auc (n=auc@logilab.net2.nerim.net) left irc: "Parti"
[18:18] <stakkars> well, I would know howto write this nowadays :)
[18:18] <Gromit> esay
[18:18] <stakkars> not in Catall.
[18:18] <Gromit> the same wy as you did it on your forth interperter
[18:18] <Gromit> this time in python
[18:19] <stakkars> use RPython and create an extension.
[18:19] <Gromit> may be
[18:20] <Gromit> basicly the idea is to use python classes to represent the memnoics and memory
[18:20] <Gromit> with such an infrastructure one could assemble directly to ram
[18:20] <Gromit> with the help of some ctypes this is easy
[18:21] <Gromit> to write this to a file, we simply pickle the whole stuff
[18:21] <stakkars> why that? Do you want to use SSE2 and friends?
[18:21] <braintone> ludal - r20903 - (adim,ludal) allow the AST tree to be exported to application level also export the grammar rules all attributes of all AST nodes are made available can install a callback with parser.install_compiler
[18:21] <Gromit> and run a simple recursive relaction method on the just loade dcode
[18:21] <stakkars> relaction? lactation?
[18:22] <Gromit> LOL
[18:22] <Gromit> relocation
[18:22] <stakkars> ay
[18:23] <Gromit> basicly it's just for fun
[18:23] <Gromit> florian had they idea to be independet of any linker or libelf or ehat so ever
[18:23] <Gromit> or whatever
[18:24] <Gromit> and we discussed writing something like libelf in python
[18:24] <Gromit> but that isn't neccessary if you only wnat to load you extensions from python
[18:26] <stakkars> well, I simplywrite what need in Python, even generating methods dynamically as they are needed, and then I run PyPy's compile and get the extension module loaded into the same process - no saving of stuff at all.
[18:27] <Gromit> of course, that's obvious, easy and professional way
[18:27] <Gromit> i am just talking about having some fun
[18:31] <Gromit> aren't there already assembler backends for pypy?
[18:31] <Gromit> or rpython
[18:36] <braintone> mwh - r20904 - almost enough operations for pypy-llvm to work again (the way in which it fails now confuses me, someone who actually knows about llvm should try it, i guess).
[18:38] <stakkars> I have togo home and think of the consequences of basic coroutines - wohah
[18:39] <stakkars> hi pypy sprinters: is somebody working on __del__ support?
[18:41] aleale (n=andersle@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[18:41] <xorAxAx> Gromit: yes
[18:41] <xorAxAx> Gromit: ppc
[18:42] <xorAxAx> and isnt llvm a kind of assembler as well? :)
[18:43] <Gromit> jup it is
[18:43] <stakkars> absolutely!
[18:44] <Gromit> bit AFAIK it's not used to assemble into ram
[18:44] <xorAxAx> what do you mean by assemble into ram?
[18:48] <stakkars> create machine code into RAM and exec
[18:48] <Gromit> jup
[18:48] <stakkars> xorAxAx: should be clear for *you*
[18:49] <xorAxAx> llvm does that as well
[18:49] <xorAxAx> it just matters what you call RAM
[18:49] <xorAxAx> the layer doesnt matter ;-)
[18:49] <stakkars> same process space
[18:49] <Gromit> such functionality is neccessary for implementing JIT
[18:49] <xorAxAx> and LLVM has a jit?
[18:49] <stakkars> cook and eat in one go :-)
[18:49] <xorAxAx> s/.$//
[18:50] <xorAxAx> but its not very suitable for python, i guess
[18:50] <stakkars> it is
[18:50] <Gromit> it's certainly an option but currently i am after a self contained implementation
[18:50] <xorAxAx> Gromit: did you know that the .net jit speeds up brainfuck by factor 2 compared to an optimising C compiler?
[18:51] <Gromit> so we should generate -net assemblies?
[18:51] <Gromit> why not
[18:51] <Rhamphoryncus> nah, generate brainfuck, and use that to generate .net ;)
[18:52] <xorAxAx> Gromit: cf. ironpython
[18:52] <Gromit> jup
[18:53] <Gromit> xorAxAx: are an active pypy developer?
[18:53] <xorAxAx> Gromit: no idea
[18:54] <Gromit> LOL
[18:57] <stakkars> ironpython?
[18:57] Action: xorAxAx wonders if any deliverable states that "py.py shalt be able to run on pypy-c"
[18:58] <xorAxAx> stakkars: the python implementation written in c#, sponsored by ms. but i am very sure that you already know it
[19:02] <stakkars> sure I know. I meant if you think they are related to pypy. They are not, they discussed with us wether they might use the annotatr.
[19:03] <xorAxAx> hmm, the annotator?
[19:04] <xorAxAx> i thought they just want to implement an interpreter
[19:04] <xorAxAx> (and were less standards-compliant than pypy when i checked last time)
[19:05] <Gromit> xorAxAx: The only use for the annotator ironpython would have is to introducte slight incompatibilities
[19:06] <Gromit> just of enough of them to silence ms' marketing droids
[19:07] <stakkars> xorAxAx: they have implemented a quite compliant compiler. But AFAIKno type inferencing,yet.
[19:07] <braintone> adim - r20905 - added a small high-level test for compile hooks
[19:08] <xorAxAx> hmm, they should update www.ironpython.com %-)
[19:09] <xorAxAx> "New built-in modules: collections (deque), gc, thread" :-)
[19:14] <xorAxAx> wow, the threading test passes in ironpython
[19:14] <xorAxAx> s/test/tests/
[19:16] adim (n=adim@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc:
[19:17] <stakkars> bye
[19:22] <Gromit> cya
----- silence for 16 minutes ----- [19:38] stakkars (n=tismer@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 110 (Connection timed out)
[19:42] aleale (n=andersle@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 110 (Connection timed out)
----- silence for 19 minutes ----- [20:01] rhymes (n=rhymes@host209-75.pool8254.interbusiness.it) left irc: Read error: 104 (Connection reset by peer)
[20:02] rhymes (n=rhymes@host209-75.pool8254.interbusiness.it) joined #pypy.
[20:05] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc:
[20:11] thingie24 (n=asdf@87.237.32.254) left irc: Read error: 110 (Connection timed out)
[20:16] <xorAxAx> ironpython always reminds at java - when you look at the working set
[20:16] <xorAxAx> but a pypy jit wont conserve memory either i guess :)
[20:19] <Rhamphoryncus> That's something I've wondered about. Dynamic specialization (aka psyco) can improve runtime, but can it do anything about memory layout?
[20:19] <xorAxAx> Rhamphoryncus: "memory layout"?
[20:19] <Rhamphoryncus> unboxing
[20:19] <Gromit> yup
[20:19] <Rhamphoryncus> how?
[20:20] <xorAxAx> Rhamphoryncus: not for the sake of memory constraints of course
[20:20] <xorAxAx> Rhamphoryncus: psyco increases memory usage, quite heavily if you are using full() with big apps
[20:20] <xorAxAx> but thats no surprise
[20:20] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
[20:20] <Rhamphoryncus> that's just the compiled code though, right?
[20:20] <xorAxAx> yeah, i guess so. like in java
[20:21] <Rhamphoryncus> The only way I can think of to reduce memory usage is to store 31bit integers in pointers. That's got nothing to do with dynamic specialization though
[20:22] <xorAxAx> Rhamphoryncus: you could pysizer to estimate how much you would gain
[20:22] <xorAxAx> Rhamphoryncus: i would say - not much :)
[20:22] <xorAxAx> thats rather suggested for speed purposes AFAIK
[20:22] <Rhamphoryncus> eh, depends on what you're looking at :)
[20:23] <xorAxAx> yeah, sure
[20:23] <Rhamphoryncus> type, refcount, value = 12 bytes, plus the 4 bytes for the pointer. You could drop that to just 4 bytes, so 75% less memory
[20:23] <xorAxAx> but anything with many ints would use numeric etc. anyway
[20:23] <Rhamphoryncus> no, only specialized code would
[20:24] <xorAxAx> "specialized"?
[20:24] <Rhamphoryncus> well, manually specialized :)
[20:25] <Rhamphoryncus> using numeric rather than a list of ints, just for performance benefits.. that's manual specialization
[20:26] <xorAxAx> yeah, a pretty sane one :)
[20:28] <Rhamphoryncus> bah :)
[20:29] Action: Rhamphoryncus hides his secret desire to implement strings as a list of ints, and have the compiler optimize them ;)
[20:31] Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) left irc: "My vapourware can beat up your vapourware! | Support ISO 8601! http://www.cl.cam.ac.uk/~mgk25/iso-time.html";
[20:41] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) left irc: Read error: 110 (Connection timed out)
[20:46] <sabi> man, yay long long support
[20:46] Action: sabi yanks out his workaround
[20:47] Action: lac smiles at sabi
[20:50] _hannes (n=yuuhu@i577B69FB.versanet.de) left irc: Read error: 104 (Connection reset by peer)
[20:54] <Gromit> n8
[20:56] Gromit (n=klix@does-d9b90ae5.pool.mediaWays.net) left irc: "taking a walk"
[21:03] <xorAxAx> hmm, integer arithmetics in ironpython is 3 times slower
[21:03] <xorAxAx> (ok, long arithmetics)
[21:07] <xorAxAx> hmm, it doesnt implement the int/long PEP and shows partly strange semantics:
[21:07] <xorAxAx> >>> fact(30000)
[21:07] <xorAxAx> -9223372036854775808L
[21:07] <xorAxAx> good that pypy exists :)
----- silence for 43 minutes ----- [21:50] thingie24 (n=asdf@87.237.32.254) joined #pypy.
[21:50] Srekel (i=Srekel@c-ce12e055.1222-1-64736c10.cust.bredbandsbolaget.se) joined #pypy.
[21:51] Srekel (i=Srekel@c-ce12e055.1222-1-64736c10.cust.bredbandsbolaget.se) left irc: Read error: 104 (Connection reset by peer)
[21:51] Srekel (i=Srekel@c-ce12e055.1222-1-64736c10.cust.bredbandsbolaget.se) joined #pypy.
[21:54] ericvrp (n=eric@c-51c6e055.1321-1-64736c11.cust.bredbandsbolaget.se) joined #pypy.
----- silence for 31 minutes ----- [22:25] hpk (n=hpk@svs-heathrow.demon.co.uk) joined #pypy.
[22:25] Action: hpk wonders if there are going to be any sprint reports
[22:29] <ericvrp> hpk: we plan to do that tomorrow
[22:36] <mwh> hpk: hello
[22:36] <hpk> mwh: hi!
[22:39] <mwh> today went well, then?
[22:40] <hpk> yes
[22:40] <hpk> i'd say so
[22:40] <hpk> see you some time (i am off to the sao-paolo plane now)
[22:40] <hpk> have fun
[22:40] <mwh> i hear you're about to fly :)
[22:40] hpk (n=hpk@svs-heathrow.demon.co.uk) left irc: "Leaving"
[22:40] <mwh> and you!
[22:50] _hannes (n=yuuhu@i577B5C49.versanet.de) joined #pypy.
----- silence for 40 minutes ----- [23:30] dialtone (n=dialtone@host213-8.pool80116.interbusiness.it) left irc:
[23:32] dialtone (n=dialtone@host213-8.pool80116.interbusiness.it) joined #pypy.
----- silence for 24 minutes ----- [23:56] <braintone> ericvrp - r20911 - Some refactoring to getting closer to running long running processes.
[00:00] --- Fri Dec 9 2005