cfbolz changed the topic of #pypy to: #pypy PyPy, the flexible snake https://pypy.org | IRC logs: https://quodlibet.duckdns.org/irc/pypy/latest.log.html#irc-end and https://libera.irclog.whitequark.org/pypy | so many corner cases, so little time
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
glyph has quit [Quit: End of line.]
glyph has joined #pypy
Joannah has joined #pypy
<Joannah> 10:00
<Joannah> Hi mattip , Carl recommended I contact you. What is the story with numpy/scipy/scikit-learn and PyPy? Am trying to pip install them using the ./pypy-3.9-c but they fail. Is there something am missing, like pypy doesnt support them?
<mattip> hi. You need to compile them, since pypy-3.9 is very new. But the compilation should JustWork on linux. What platform are you on?
<mattip> conda is intending to start a migration for pypy3.8 and pypy3.9 soon, which means you theoretically could use conda to get binary packages,
<mattip> but it will take a while to get it all going
<Joannah> Am on debian
lehmrob has joined #pypy
<mattip> weird. We regularly build numpy with HEAD, so that should work
<mattip> I recently built scipy too. I haven't tried scikit-learn. pypy3.7 has all of them on conda
<mattip> conda create -n pypy_test_env pypy3.7; conda activate pypy_test; conda install scikit-learn
Joannah has quit [Ping timeout: 256 seconds]
<mattip> joannah (for the logs): can you put the traceback on a pastebin somewhere?
<ctismer> Hi mattip! I observed that objects are not immediately collected after `gc.collect()`. I have to call it twice. Why? Can I avoid this?
<mattip> sometimes you have to call three times, it is to untangle reference loops
<ctismer> Thank you, I hoped for that answer. Then I can calm down my colleagues 😉
<mattip> I am pretty sure this is true on CPython as well
<ctismer> Yes, but normally we see immediate deletion. And I did not know how the PyPy GC differs.
<ctismer> So we needed to change many tests.
<LarstiQ> also the "normally" is not guaranteed on CPython
<mattip> it is much healthier in general to separate obj.close() from obj.__del__(), and use some kind of context management to call obj.close()
<mattip> rather than depending on refcount semantics to "automatically" call obj.__del__() -> obj.close()
<mattip> It seems we have a new advocate for PyPy in Anthony Shaw
<ctismer> LarstiQ: Yes, I was talking about our test cases which check object deletion and things work without cycles, where PyPy needs two GC calls. I was wondering about the second.
<cfbolz> Snow snow!
<LarstiQ> cfbolz: ah, had it disappeared?
<cfbolz> We didn't really have much this winter
<cfbolz> But now I am hiking for two days, and finding lots here in Hessen in the forest
<LarstiQ> cool :)
<LarstiQ> we've had heaps more than in recent years, lots of skiing and skating in daycare.
slav0nic has joined #pypy
slav0nic has quit [Remote host closed the connection]
slav0nic has joined #pypy
slav0nic has quit [Remote host closed the connection]
* ctismer thinks "as cool as snow usually is" 😎
slav0nic has joined #pypy
epony has quit [Ping timeout: 240 seconds]
lehmrob has quit [Quit: Konversation terminated!]
epony has joined #pypy
greedom has joined #pypy
greedom has quit [Remote host closed the connection]
greedom has joined #pypy
greedom has quit [Remote host closed the connection]
greedom has joined #pypy
epony has quit [Read error: Connection reset by peer]
epony has joined #pypy
Guest96 has joined #pypy
greedom has quit [Remote host closed the connection]
Guest96 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tazle_ has joined #pypy
lastmikoi_ has joined #pypy
tazle has quit [*.net *.split]
dustinm has quit [*.net *.split]
antocuni has quit [*.net *.split]
lastmikoi has quit [*.net *.split]
lastmikoi_ is now known as lastmikoi
antocuni has joined #pypy
dustinm has joined #pypy
slav0nic has quit [Ping timeout: 268 seconds]