cfbolz changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://botbot.me/freenode/pypy/ ) | use cffi for calling C | the secret reason for us trying to get PyPy users: to test the JIT well enough that we're somewhat confident about it
tbodt has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
marr has quit [Ping timeout: 260 seconds]
lritter has joined #pypy
dddddd has quit [Read error: Connection reset by peer]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
tbodt has quit [Client Quit]
lritter has quit [Quit: Leaving]
adamholmberg has joined #pypy
adamholmberg has quit [Ping timeout: 248 seconds]
jcea has quit [Quit: jcea]
<bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/3870 [mattip: force build, release-pypy2.7-6.x]
Hasimir has quit [Ping timeout: 240 seconds]
Hasimir has joined #pypy
energizer has quit [Remote host closed the connection]
energizer has joined #pypy
inad922 has joined #pypy
energizer has quit [Ping timeout: 240 seconds]
asmeurer__ has quit [Quit: asmeurer__]
asmeurer has joined #pypy
energizer has joined #pypy
energizer has quit [Remote host closed the connection]
energizer has joined #pypy
mcyprian has joined #pypy
mcyprian has quit [Client Quit]
<bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-s390x/builds/844 [mattip: force build, release-pypy3.5-6.x]
<bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-s390x/builds/845 [mattip: force build, release-pypy2.7-6.x]
inad922 has quit [Ping timeout: 240 seconds]
jaffachief has joined #pypy
jaffachief has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kenaan has quit [Ping timeout: 256 seconds]
danchr has quit [Ping timeout: 264 seconds]
asmeurer has quit [Quit: asmeurer]
asmeurer_ has joined #pypy
illume has joined #pypy
asmeurer_ has quit [Quit: asmeurer_]
energizer has quit [Ping timeout: 255 seconds]
marr has joined #pypy
inad922 has joined #pypy
nimaje has quit [Ping timeout: 276 seconds]
nimaje has joined #pypy
antocuni has joined #pypy
lazka has joined #pypy
antocuni has quit [Ping timeout: 256 seconds]
danchr has joined #pypy
kcr has quit []
kcr has joined #pypy
DRMacIver has quit []
DRMacIver has joined #pypy
Alex_Gaynor has quit []
Alex_Gaynor has joined #pypy
inad922 has quit [Ping timeout: 248 seconds]
<lesshaste> it's pretty amazing that seaborn works now in pypy!
<lesshaste> although you get a huge number of warnings when you import pandas like https://bpaste.net/show/52820e43becf
inad922 has joined #pypy
nimaje has quit [Remote host closed the connection]
nimaje has joined #pypy
rguillebert has quit []
rguillebert has joined #pypy
fijal has quit []
fijal has joined #pypy
<mattip> lesshaste: did you use Cython 0.28.2 and the latest nightly PyPy ?
<lesshaste> mattip, yes to 0.28.2
<lesshaste> Python 2.7.13 (d044ce2d7d53, Mar 27 2018, 22:00:13)
<lesshaste> #
<lesshaste> has much changed since then?
<lesshaste> in pypy are sum(np.array([1,3,4,4])) and np.sum(np.array([1,3,4,4])) much different under the hood?
jcea has joined #pypy
<mattip> lesshaste: np.sum should be faster, it is written in C, better written as np.array().sum()
<lesshaste> mattip, ok but there is also the pypy + numpy overhead, right?
graingert has quit []
graingert has joined #pypy
<lesshaste> speed wise. is there any difference between list comprehensions and for loops in pypy?
<lesshaste> I mean is one just translated into the other?
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
<mattip> python sum() is worse, it calls __iter__ which calls __getitem__
<mattip> release candidates are up https://bitbucket.org/pypy/pypy/downloads
<mattip> any fixes or comments are welcome
tayfun26 has joined #pypy
<mattip> lesshaste: np.ndarray.sum is all in c, no python iteration involved
<lesshaste> cool! (re: release notice)
<lesshaste> and thanks
nimaje has quit [Remote host closed the connection]
nimaje has joined #pypy
<lesshaste> distruibute <- typo
<lesshaste> emissions <- typo
<lesshaste> "The Matplotlib TkAgg backend now works with PyPy, as do pygame and pygobject." for me matplotlib just works. Maybe make that clear? As I don't have to do anything at all to make it work with pypy
<lesshaste> not everyone will know what tkagg is and if it is the default
marky1991 has quit [Ping timeout: 240 seconds]
mitsuhiko has quit []
mitsuhiko has joined #pypy
exarkun has joined #pypy
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
<exarkun> is there a pypy-stm-based concurrent programming framework that's well suited to current high-end datacenter servers (perhaps ~32 cores) for I/O-heavy workloads?
<exarkun> Maybe "I/O-heavy" is somewhat misleading. Perhaps I simply care about CPU-bound processes. I said "I/O-heavy" because I am thinking of the case where you need multiple cores to be able to saturate whatever I/O bus is involved.
marky1991 has joined #pypy
<exarkun> Oh, darn, a huge chunk of http://doc.pypy.org/en/latest/stm.html is out of date. :(
<lazka> lesshaste, matplotlib defaults to "agg" here
<lazka> v3 might gain auto backend selection: https://github.com/matplotlib/matplotlib/pull/979
irclogs_io_bot has quit [Remote host closed the connection]
raynold has quit [Quit: Connection closed for inactivity]
nimaje has quit [Ping timeout: 255 seconds]
nimaje has joined #pypy
<lesshaste> lazka, do you mean you have to do something to get matplotlib to work with pypy?
irclogs_io_bot has joined #pypy
<lazka> lesshaste, afaik it only supports specifying one backend atm, so if you have a different one specified in any of the matplotlibrc files or through MPLBACKEND then you have to tell it to use another one
nimaje has quit [Ping timeout: 276 seconds]
nimaje has joined #pypy
illume has joined #pypy
lritter has joined #pypy
glyph has quit [*.net *.split]
demonimin has quit [*.net *.split]
lesshaste has quit [*.net *.split]
michaelgreene has quit [*.net *.split]
Lightsword has quit [*.net *.split]
sthalik has quit [*.net *.split]
__pv has quit [*.net *.split]
ssbr has quit [*.net *.split]
mdash has quit [*.net *.split]
njs has quit [*.net *.split]
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
__pv has joined #pypy
Lightsword has joined #pypy
sthalik has joined #pypy
ssbr has joined #pypy
mdash has joined #pypy
michaelgreene has joined #pypy
njs has joined #pypy
demonimin has joined #pypy
lesshaste has joined #pypy
glyph has joined #pypy
danieljabailey has quit [Ping timeout: 260 seconds]
danieljabailey has joined #pypy
jaffachief has joined #pypy
illume has joined #pypy
marky1991 has quit [Ping timeout: 256 seconds]
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
<lesshaste> lazka, ok.. all I was saying was that for a normal user it might be helpful just to say "matplotlib works" and put the tkagg detail underneath
jaffachief has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
jaffachief has joined #pypy
jaffachief has quit [Client Quit]
<__pv> typo in release notes: distruibute
jaffachief has joined #pypy
kenaan has joined #pypy
<kenaan> cfbolz default a614238235c7 /pypy/doc/release-v6.0.0.rst: fix typos (thanks lesshaste and __pv)
jaffachief has quit [Client Quit]
jaffachief has joined #pypy
<Rotonen> exarkun: you'll do more by software architecture allowing for multiprocessing.pool with shared memory workers, pypy-stm is mostly a PoC (and hits dead ends and seeks new approaches - plenty of genuine research effort left there before anything usable for the context you seek in)
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
jaffachief has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<exarkun> Rotonen: Yea but multiprocessing.Pool is boring.
<cfbolz> :-)
redj_ has joined #pypy
redj has quit [Ping timeout: 264 seconds]
adamholmberg has joined #pypy
<Rotonen> exarkun: by what you're doing, you've chosen engineering as a field, at least to some degree, go figure :-)
<arigato> right now your only choice is to run 32 processes, which costs a lot of RAM with pypy
<arigato> (potentially)
illume has joined #pypy
<simpson> Rotonen: 'boring' is maybe a misleading word. Friends don't let friends `import multiprocessing`.
<arigato> there are some vague thoughts about a solution that looks like independent processes but are actually implemented as theads and share memory
<arigato> but there is no progress on that
<Rotonen> 'here is a bit of code, here is an input queue, go at it' is a good use case to cater to
<Rotonen> there is also an undocumented a thread pool in multiprocessing
<Rotonen> simpson: it's not that bad so long as you have very small intputs and outputs so the IPC-via-pickling does not bite you, or you can design everything to use shared memory
<simpson> Rotonen: I understand your position. I'm just sticking to the #python party line; I don't want to wish insanity and suffering on anybody.
dddddd has joined #pypy
<Rotonen> i've done such things, including but not limited to RPG IV, multiprocessing is fine and fluffy :D
<exarkun> My choice isn't really between pypy-stm and multiprocessing, anyway. If pypy-stm had made a ton of advances since I last checked in on it, I would be interested in exploring it some more. If not, there's always Haskell.
jaffachief has joined #pypy
<cfbolz> exarkun: pypy-stm is pretty much dormant at this point, I think
jaffachief has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<lesshaste> __pv, I already reported that (and others) above :)
inad922 has quit [Ping timeout: 255 seconds]
marky1991 has quit [Read error: Connection reset by peer]
jaffachief has joined #pypy
marky1991 has joined #pypy
<exarkun> cfbolz: good to know (though sad), thanks
jaffachief has quit [Client Quit]
Rhy0lite has joined #pypy
adamholmberg has quit [Remote host closed the connection]
adamholmberg has joined #pypy
adamholmberg has quit [Read error: Connection reset by peer]
jaffachief has joined #pypy
inad922 has joined #pypy
adamholmberg has joined #pypy
marky1991 has quit [Ping timeout: 260 seconds]
antocuni has joined #pypy
inad922 has quit [Ping timeout: 256 seconds]
jaffachief has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<lazka> lesshaste, agreed, a screenshot/gif might make it even more clear
inad922 has joined #pypy
<kenaan> mattip pypy.org[extradoc] a1da300b39d3 /source/download.txt: update sha256 for pypy2.7
antocuni has quit [Ping timeout: 264 seconds]
jeroud has quit []
jeroud has joined #pypy
glyph has quit [*.net *.split]
lesshaste has quit [*.net *.split]
demonimin has quit [*.net *.split]
michaelgreene has quit [*.net *.split]
Lightsword has quit [*.net *.split]
sthalik has quit [*.net *.split]
__pv has quit [*.net *.split]
ssbr has quit [*.net *.split]
mdash has quit [*.net *.split]
njs has quit [*.net *.split]
Hasimir has quit [*.net *.split]
pulkitg has quit [*.net *.split]
illume has quit [*.net *.split]
lazka has quit [*.net *.split]
dmalcolm has quit [*.net *.split]
norox has quit [*.net *.split]
rubdos has quit [*.net *.split]
blachance has joined #pypy
rubdos has joined #pypy
tos9 has joined #pypy
codehotter has joined #pypy
Graypup_ has joined #pypy
ulope has joined #pypy
rubdos has quit [Max SendQ exceeded]
michaelgreene has joined #pypy
lesshaste has joined #pypy
glyph has joined #pypy
demonimin has joined #pypy
ssbr has joined #pypy
njs has joined #pypy
sthalik has joined #pypy
mdash has joined #pypy
__pv has joined #pypy
Lightsword has joined #pypy
pulkitg has joined #pypy
dmalcolm has joined #pypy
norox has joined #pypy
illume has joined #pypy
lazka has joined #pypy
Hasimir has joined #pypy
rubdos has joined #pypy
agates has quit [Ping timeout: 245 seconds]
dash has quit [Ping timeout: 255 seconds]
abvi[m] has quit [Ping timeout: 268 seconds]
yuvipanda has quit [Ping timeout: 268 seconds]
pulkitg has quit [Ping timeout: 256 seconds]
Hasimir has quit [Changing host]
Hasimir has joined #pypy
bendlas has quit [Ping timeout: 256 seconds]
rubdos has quit [Client Quit]
rubdos has joined #pypy
inhahe__ has joined #pypy
tayfun26 has quit [Quit: tayfun26]
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
samth has quit []
marky1991 has joined #pypy
samth has joined #pypy
<nanonyme> Isn't stm-gc mostly dropped at this point in favour or just beating at PyPy until you can rip out the GIL?
<nanonyme> pypy-stm even
phlebas has quit []
<arigato> nobody is doing that
phlebas has joined #pypy
inhahe__ has quit [Ping timeout: 255 seconds]
<nanonyme> Oh, I must've been misled then
<nanonyme> That's the impression I got from latest STM-related blog posts
<Rotonen> nanonyme: i took that more as a noble aim for now as there is no hot lead (or industry desire via funding) on the STM front
inhahe__ has joined #pypy
<nanonyme> Rotonen, I read that that STM is dead, its money may be repurposed into GIL removal if there's interest from commercial partners
<nanonyme> But yeah, on better reading it's quite clear no one has committed to any work on that front either
antocuni has joined #pypy
altendky_ has joined #pypy
bjs_ has joined #pypy
agronholm_ has joined #pypy
_aegis__ has joined #pypy
bjs has quit [*.net *.split]
agronholm has quit [*.net *.split]
altendky has quit [*.net *.split]
_aegis_ has quit [*.net *.split]
hexa- has quit [*.net *.split]
bjs_ is now known as bjs
altendky_ is now known as altendky
<arigato> the outcome of that blog post was that almost nobody was interested
tbodt has joined #pypy
hexa has joined #pypy
<nanonyme> Yes, I guessed as much
lritter has quit [Remote host closed the connection]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
dash has joined #pypy
tbodt has joined #pypy
kirma has quit [Ping timeout: 265 seconds]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
Hasimir has quit [Ping timeout: 240 seconds]
tbodt has quit [Client Quit]
marky1991 has quit [Remote host closed the connection]
jaffachief has joined #pypy
tbodt has joined #pypy
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
tbodt has quit [Client Quit]
marky1991 has joined #pypy
<ronan> mattip: that annoying Cython warning is because PyType_Type.tp_basicsize is set to sizeof(PyHeapTypeObject)
tbodt has joined #pypy
inad922 has quit [Ping timeout: 268 seconds]
Hasimir has joined #pypy
tbodt has quit [Remote host closed the connection]
tbodt has joined #pypy
lazka has quit [Quit: Leaving]
lazka has joined #pypy
lazka has quit [Remote host closed the connection]
<antocuni> ronan: which is correct, AFAIK
<ronan> Cython seems to disagree
<antocuni> if you look at the code generated by cython, you see that it does a special check for pypy
<ronan> yes, I've hust found it
<antocuni> because pypy used to be buggy
<ronan> *just
<antocuni> now pypy is correct, but cython doesn't know 😅
<arigato> not the first time, I remember this pattern of cython working around a pypy bug in a way that breaks when the pypy bug is fixed
abvi[m] has joined #pypy
yuvipanda has joined #pypy
agates has joined #pypy
pulkitg has joined #pypy
bendlas has joined #pypy
<ronan> https://github.com/cython/cython/pull/2165 should have fixed it though
asmeurer_ has joined #pypy
<antocuni> are you sure that the C code you are compiling has been generated with cython 0.28.2?
<ronan> I guess it hasn't
asmeurer_ has quit [Ping timeout: 240 seconds]
<Cheery> One of the features I'd desire would require that iterator could be cloned.
<Cheery> for list-derived iterators that probably is doable.. but
jaffachief has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Cheery> can it be done for sets?
<Cheery> or for dictionaries?
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
antocuni has quit [Ping timeout: 276 seconds]
tbodt has joined #pypy
lritter has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
energizer has joined #pypy
tbodt has joined #pypy
asmeurer has joined #pypy
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
Arfrever has joined #pypy
exarkun has quit [Read error: Connection reset by peer]
exarkun has joined #pypy
tbodt has joined #pypy
Garen has quit [Read error: Connection reset by peer]
Garen has joined #pypy
<kenaan> arigo py3.5 19216449b343 /pypy/module/signal/interp_signal.py: Merged in hroncok/pypy-3/hroncok/fix-multiprocessing-regression-on-newer--1524656522151 (pull request #607) Fix mult...
<kenaan> hroncok hroncok/fix-multiprocessing-regression-on-newer--1524656522151 610fde79e505 /pypy/module/signal/interp_signal.py: Fix multiprocessing regression on newer glibcs Starting w...
<cfbolz> Cheery: in rpython? or python?
<Cheery> cfbolz: in rpython
lritter has quit [Remote host closed the connection]
Rhy0lite has quit [Quit: Leaving]
hotpot33 has joined #pypy
inad922 has joined #pypy
marky1991 has quit [Read error: Connection reset by peer]
jaffachief has joined #pypy
jaffachief has quit [Client Quit]
jaffachief has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
adamholmberg has quit [Remote host closed the connection]
tbodt has joined #pypy
inad922 has quit [Ping timeout: 260 seconds]
<bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-s390x/builds/845 [mattip: force build, release-pypy2.7-6.x]
marky1991 has joined #pypy
jaffachief has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
tbodt has quit [Read error: Connection reset by peer]
cjwelborn has quit [Remote host closed the connection]
<mattip> any reason not to release 6.0?
cjwelborn has joined #pypy
<ronan> mattip: looks good to me
<mattip> cool
<ronan> I checked pandas on pypy3, there's one segfault but it's a pandas bug that I already reported
<mattip> thanks. And no nasty warnings about mismatched sizes?
<ronan> well, there's one, but I don't know where it comes from
<ronan> pretty sure it's something that was cythonised with an old Cython
<ronan> ah, I found it thanks to python -v
<ronan> it's numpy.random.mtrand
<mattip> it seems to go away when I build from a source distribution, but appears if I do `pip install numpy`
<mattip> maybe there is some caching from pip?
<ronan> yes, I just checked: the zip includes files cythonised with 0.26.1
<mattip> where is the zip stored?
<ronan> on PyPI somewhere
<mattip> ronan: not somewhere in ~/.cache/pip ?
<ronan> mattip: no, it's the source dist you can see in https://pypi.org/project/numpy/#files
<mattip> ahh, numpy is providing a sdist with cythonized files? I should tell them, they are releasing 1.14.3
<ronan> yes, they should make sure to use Cython 0.28.2
<mattip> ok, it looks like I don't need to do anything, the CI says cython 0.28.2
<ronan> good, we'll just need to tell people to upgrade numpy
demonimin has quit [Ping timeout: 264 seconds]
demonimin has joined #pypy
demonimin has joined #pypy
<mattip> ronan: any chance you could look at the blog post draft ^^^
tbodt has joined #pypy
<ronan> mattip: LGTM
<kenaan> mattip pypy.org[extradoc] 4ce1878652ff /download.html: regenerate web site
<mattip> cool
<kenaan> mattip default 128340bf51b2 /pypy/doc/release-v6.0.0.rst: formatting