eamanu has quit [Ping timeout: 268 seconds]
eamanu has joined #pypy
epony has quit [Ping timeout: 240 seconds]
<
tumbleweed>
mattip: all the release branches should probably have 794eaac8974d merged
wleslie has quit [Quit: ~~~ Crash in JIT!]
Atque has joined #pypy
epony has joined #pypy
Atque has quit [Quit: ...]
larstiq_ has joined #pypy
Cheery_ has joined #pypy
lazka has quit [*.net *.split]
LarstiQ has quit [*.net *.split]
mattip has quit [*.net *.split]
Cheery has quit [*.net *.split]
dustinm has quit [*.net *.split]
Atque has joined #pypy
lazka has joined #pypy
dustinm has joined #pypy
mattip has joined #pypy
<
mattip>
tumbleweed: thanks
<
mattip>
mgorny: Py_GenericAlias is annoying, it was only documnted for 3.10
<
mattip>
also it seems forcing the locale on the buildbot was a mistake,
<
mattip>
it fixed the 4 failing tests but broke everything else
otisolsen70 has joined #pypy
otisolsen70 has quit [Remote host closed the connection]
otisolsen70 has joined #pypy
<
mgorny>
mattip: weird, the linked doc says 3.9
<
mattip>
right, they added it after the release
Cheery_ is now known as Cheery
<
mgorny>
classical python development ;-)
<
mgorny>
in the meantime, it seems that zoneinfo stuff is somewhat broken
<
mgorny>
i'll investigate in a minute
<
mattip>
cool, thanks
<
mgorny>
seems to be missing sysconfigdata
<
mattip>
I am adding py3.9 to the buildbot nightly runs and a benchmark run as well
slav0nic has joined #pypy
<
mgorny>
hmm, i think gevent has a test that relies on fork being the default multiprocessing context
<
cfbolz>
of course :-(
<
cfbolz>
I am still not closer with the bug so far
<
mgorny>
but it should be easy to switch context locally
<
mgorny>
so i'll go straight for a PR ;-)
<
cfbolz>
mgorny: what do they do on mac?
lritter has joined #pypy
<
mgorny>
no clue, maybe they don't run CI on mac
<
mgorny>
i'm more curious how the test works on windows ;-)
<
cfbolz>
mgorny: thanks
<
mgorny>
132 packages left to test
<
mgorny>
so far around 10 waiting for closer investigation
<
mgorny>
gevent tests are really stupid
<
mgorny>
they give you output like 'one of 7 following tests failed'
<
cfbolz>
mgorny: the asgiref failure is
*with* your multiprocessing spawn change, right?
<
cfbolz>
(because without it, it passes)
<
mgorny>
i'll test without
<
mgorny>
cfbolz: indeed it seems to be the cause
<
mgorny>
i'll check if cpython's default_context affects it too
<
cfbolz>
so what we should really do is fix the fork problem
<
cfbolz>
right, good question
<
mgorny>
after all, darwin
<
mgorny>
yeah, happens on cpython too after switching default method
<
mgorny>
i'll test if it works on osx
<
mgorny>
that is, if i manage to boot my vm :-f
<
cfbolz>
mgorny: the test is really for a fork bug, so with spawn it doesn't actually make sense
<
mgorny>
ok, i'll submit a PR to them in a minute
Julian has joined #pypy
<
mgorny>
cfbolz: nose's tests are also broken by spawn
<
mgorny>
although that's one package i wouldn't care to fix ;-)
<
cfbolz>
mgorny: I wonder whether these failures show that we are too far into this idea
<
mgorny>
i'm still wondering whether it doesn't affect darwin or people just don't run tests there
<
mgorny>
(still trying to fix my vm)
<
cfbolz>
don't have a mac around to test either, sorry
<
mgorny>
pexpect tests also fail with some pickling errors because of that
otisolsen70_ has joined #pypy
otisolsen70_ has quit [Remote host closed the connection]
otisolsen70 has quit [Ping timeout: 256 seconds]
greedom has joined #pypy
Corbin has quit [Ping timeout: 268 seconds]
Corbin has joined #pypy
greedom has quit [Remote host closed the connection]
greedom has joined #pypy
<
Cheery>
during the weekend I figured I take linear logic sequents and transform that into a bytecode.
<
Cheery>
to get something that doesn't completely stand on the way.
<
Cheery>
it looks like magic for person who doesn't know what's going on.
<
Cheery>
name5 (a;b) (c=d) foobar(c) ⊗ (x,y → d) a=x ⊗ b=y
mattip has quit [Ping timeout: 256 seconds]
greedom has quit [Remote host closed the connection]
<
mgorny>
cfbolz: so finally managed it -- and asgiref tests fail the same on osx
<
mgorny>
(and even more -- apparently osx is not tested at all)
dmalcolm_ has joined #pypy
sam_ has quit [Ping timeout: 256 seconds]
shodan45 has quit [Ping timeout: 256 seconds]
exarkun has quit [Ping timeout: 256 seconds]
habnabit_ has quit [Ping timeout: 256 seconds]
shodan45 has joined #pypy
habnabit_ has joined #pypy
dmalcolm__ has quit [Ping timeout: 256 seconds]
exarkun has joined #pypy
sam_ has joined #pypy
Guest96 has joined #pypy
Julian has quit [Quit: leaving]
<
cfbolz>
mgorny: right
<
mgorny>
tbh, at this poitn i'm wondering if projects are doing something unsupported, or maybe if it's a bug in python stdlib
otisolsen70 has joined #pypy
mattip has joined #pypy
slav0nic has quit [Ping timeout: 250 seconds]
otisolsen70 has quit [Quit: Leaving]
Techcable has quit [Ping timeout: 250 seconds]
lritter has quit [Quit: Leaving]