00:33
thinkwell has joined #pypy
00:52
thinkwell31 has joined #pypy
00:53
thinkwell has left #pypy [#pypy]
00:55
thinkwell31 has quit [Quit: Connection closed]
02:15
dansan has quit [Read error: Connection reset by peer]
02:16
dansan has joined #pypy
02:44
dansan has quit [Read error: Connection reset by peer]
02:44
dansan has joined #pypy
02:45
dansan has quit [Read error: Connection reset by peer]
02:45
dansan has joined #pypy
02:48
dansan_ has joined #pypy
02:49
dansan has quit [Client Quit]
03:32
jcea has quit [Remote host closed the connection]
03:33
jcea has joined #pypy
06:01
_whitelogger has joined #pypy
06:44
travis-ci has joined #pypy
06:44
travis-ci has left #pypy [#pypy]
08:35
jcea has quit [Remote host closed the connection]
08:35
jcea has joined #pypy
08:41
lobo_ has quit [Ping timeout: 256 seconds]
08:41
lobo_ has joined #pypy
09:50
mattip has quit [Ping timeout: 256 seconds]
10:05
oberstet has joined #pypy
10:34
mattip has joined #pypy
10:48
Gustavo6046_ has joined #pypy
10:49
Gustavo6046 has quit [Ping timeout: 246 seconds]
10:50
Gustavo6046_ is now known as Gustavo6046
13:00
antocuni has joined #pypy
13:12
Gustavo6046 has quit [Ping timeout: 240 seconds]
13:13
Gustavo6046_ has joined #pypy
13:13
arigo has joined #pypy
13:15
Gustavo6046_ is now known as Gustavo6046
13:16
arigo is now known as arigato
13:20
tsaka__ has quit [Ping timeout: 256 seconds]
14:19
<
phlebas >
what's your plan wrt the new parser? will you write a rpython_generator.py and copy the grammar to put in rpython actions instead of C actions?
14:32
<
cfbolz >
phlebas: we aren't that far in our plans yet
14:32
<
cfbolz >
But yes, that's one option
14:35
<
cfbolz >
(seems we aren't mentioned in the change log, maybe we should open a pull request for that again 😒)
14:46
<
fijal >
seems I have no idea what I'm talking about - what wheels work on pypy?
14:46
<
fijal >
not the ones built using cpython right?
15:05
<
Dejan >
noarch wheels should defo work
15:52
<
mattip >
cfbolz: one solution to the accreditation issue would be to pro-actively push stuff to cpython
15:53
<
cfbolz >
mattip: Pablo promised to fix the news entry
15:54
<
cfbolz >
I don't necessarily want to push stuff to CPython, my C is really bad
15:58
marky1991 has joined #pypy
16:01
<
fijal >
so are cpython wheels good for pypy or not? how can I even tell?
16:02
<
antocuni >
fijal: try to run "pip wheel psutil"
16:02
<
antocuni >
on my machine, it creates a file called psutil-5.8.0-cp38-cp38-manylinux2010_x86_64.whl
16:02
<
antocuni >
"cp38-cp38" means cpython 3.8. No way it can work on pypy
16:03
<
fijal >
what's the first part of cp38 and the second?
16:03
<
antocuni >
but if you run pip wheel pytest (which is pure python), you get pytest-6.2.1-py3-none-any.whl
16:03
<
antocuni >
which means it works on any python3 impl, including pypy
16:04
<
fijal >
and where is the description of what means what?
16:04
<
fijal >
also, does that not mean that we are running an inefficient pyc file?
16:04
<
fijal >
or is that no longer a thing?
16:04
<
antocuni >
I don't think that pyc are included inside wheels
16:28
<
fijal >
antocuni: ok, is there any reason why numpy=1.19 is not in your wheels for pypy3.7?
16:29
<
antocuni >
yes, it's because my wheel repo is mostly unmantained by now :)
16:29
<
antocuni >
PRs welcome
16:29
<
fijal >
ok, I'm struggling to use it anyway
16:30
<
antocuni >
fijal: pypy -m pip install numpy should work out of the box nowadays, if you have all the required build time dependencies
16:30
<
fijal >
it does, but it takes forever
16:32
<
fijal >
ok, would be cool to upload the right wheel then
17:33
<
mattip >
for numpy, you can do
17:34
<
mattip >
to get the wheel that will be released by the end of Jan
17:34
<
mattip >
but I would really recommend you use conda for any serious pypy work with binary packages
17:34
marky1991 has quit [Ping timeout: 265 seconds]
17:35
<
mattip >
conda create -n pypy37 pypy python=3.7
17:36
<
mattip >
then "conda install" works with most of the scientific python stack
17:36
<
mattip >
the migration is ongoing, 588 packages are done
17:39
<
fijal >
mattip: I think numpy is a side dependency of something very unimportant
20:11
jcea has quit [Ping timeout: 260 seconds]
20:42
Gustavo6046 has quit [Ping timeout: 256 seconds]
20:46
Gustavo6046 has joined #pypy
21:35
jcea has joined #pypy
22:06
marky1991 has joined #pypy
22:49
oberstet has quit [Quit: Leaving]
23:14
lazka1 has joined #pypy
23:14
lazka has quit [Read error: Connection reset by peer]
23:14
lazka1 is now known as lazka