Log file Viewer

[08:48:18] <cdunklau> howdy folks, i'm getting 503s on pypi. https://pypi.org/search/?q=unrar-cffi and https://pypi.org/help/

[08:48:52] <cdunklau> i did manage to load the /help/ endpoint since getting a 503 on it

[08:49:09] <cdunklau> search still appears borked though

[09:12:45] <McSinyx[m]> I can also reproduce

[09:13:00] <McSinyx[m]> not anymore

[09:13:06] <McSinyx[m]> yay it's working again

[09:13:24] <cdunklau> seems to be transient

[09:13:47] <cdunklau> sometimes https://pypi.org/project/unrar-cffi/ works, sometimes 504

[09:13:50] <cdunklau> err 503

[09:14:27] <McSinyx[m]> I was only able to reproduce on search and help though

[09:14:42] <cdunklau> seems that the project/* things are more stable

[09:51:32] <dmerejkowsky> !logs

[09:51:32] <pmxbot> http://kafka.dcpython.org/channel/pypa

[10:24:30] <boegel> I'm getting "Gateway Time-out" issues when uploading a package to PyPI using twine:

[10:24:30] <boegel> HTTPError: 504 Server Error: Gateway Time-out for url: https://upload.pypi.org/legacy/

[10:25:58] <kroitor> Me too

[10:26:21] <kroitor> opened an issue here: https://github.com/pypa/warehouse/issues/7642

[10:26:26] <kroitor> not sure if that is the right place

[10:26:36] <cdunklau> i reported to the python infra channel

[10:29:47] <kroitor> python status page says all systems operational, though, so it's hard to know if they're aware

[10:36:10] <cdunklau> indeed

[11:01:48] <graingert> https://pypi.org/account/register/ is 503ing for me

[11:02:47] <graingert> test is broken too https://test.pypi.org/account/register/

[11:07:16] <graingert> di_codes: ^

[11:16:30] <dmerejkowsky> looks like they are aware now

[11:41:20] <graingert> dmerejkowsky: who are they?

[11:43:07] <graingert> status.python.org not updated yet

[11:50:29] <yoctozepto> hey folks, pypi index also down

[11:51:31] <yoctozepto> it looks like it made our pip calls to try download latest versions for packages instead of pinning to py2-compatible (yeah, I know, but we are supporting previous openstack releases)

[12:21:07] <cdunklau> yoctozepto: the powers that be are on it

[12:29:44] <dmerejkowsky> fixed : https://github.com/pypa/warehouse/issues/7643#issuecomment-604392508

[17:14:24] <omegak> Hi there, I'm having some weird behavior when installing a package with `pip install -e`. Within an activated virtualenv, `python --version` returns `2.7.17` and `which python` `.../.venv/bin/python`. Still, the script generated in `.venv/bin` contains `#!/usr/local/bin/python3` as shebang. Any idea what may be causing this? Worth mentioning, this only happens when run from within a Makefile.

[17:50:38] <sumanah> hi techalchemy - how's it going?

[17:55:52] <pradyunsg> omegak: hi! A few follow up questions: What version of virtualenv was used to create that environment? Or was it created with venv? Also, which package is this? Is the package's setup.py publicly available?

[17:59:56] <omegak> hi @pradyunsg! The venv is created with pipenv and the setup.py is this: https://github.com/indico/indico/blob/master/setup.py#L91

[18:02:15] <pradyunsg> omegak: that helps!

[18:06:56] <omegak> if I could know where in pip the shebang gets generated I can debug what is wrong in the environment

[18:09:20] <sumanah> [nudge techalchemy ]

[18:18:21] <pradyunsg> omegak: one sec, lemme grab a link.

[18:20:03] <pradyunsg> https://github.com/pypa/pip/blob/20.0.2/src/pip/_internal/operations/install/wheel.py#L77

[18:20:11] <pradyunsg> omegak: ^

[18:23:12] <omegak> but it seems to be installed with setuptools

[18:23:21] <omegak> I get this right below the shebang

[18:23:31] <omegak> # EASY-INSTALL-ENTRY-SCRIPT: 'indico','console_scripts','indico'

[18:24:57] <pradyunsg> omegak: Ahhh! Because setup.py develop!

[18:25:16] <pradyunsg> I hadn't noticed the `-e`.

[18:25:25] <techalchemy> hey sumanah apologies, just sorting out a prescription refill

[18:26:07] <sumanah> techalchemy: :-) so how goes the branch merge?

[18:26:43] <omegak> found it here https://github.com/pypa/setuptools/blob/master/setuptools/command/easy_install.py#L2156

[18:29:07] <omegak> any advice on debugging that? I tried printing and pdb, but no luck

[18:29:58] <dmerejkowsky> omegak: can you post the contents of the Makefile ?

[18:30:16] <dmerejkowsky> I think the bug is likely to be there

[18:30:32] <techalchemy> phew all sorted out; sumanah - had some tests fail but not too many- 17 / 230 or so

[18:31:04] <sumanah> techalchemy: cool! so you need to finish that next, right?

[18:31:12] <techalchemy> yup

[18:32:42] <omegak> dmerejkowsky, I don't think so, the Makefile is empty with a single `install` target

[18:32:46] <omegak> install:

[18:32:48] <omegak> pip install --editable indico

[18:33:19] <dmerejkowsky> try replacing `pip` with `python -m pip`

[18:34:57] <omegak> same result

[18:38:16] <pradyunsg> Hmm... Could be the shell - maybe the one that make is running is different from the one you have set up?

[18:39:29] <dmerejkowsky> yeah like if you're using zsh but make is using bash and they don't have the same $PATH

[18:39:37] <dmerejkowsky> but it seems far-fetched

[18:42:56] <omegak_> the $PATH seems correct when printed from the Makefile target...

[18:43:05] <dmerejkowsky> oh and do you run make from the virtualenv or is the virtualenv activated in the makefile ?

[18:43:28] <omegak_> the virtualenv is activated outside

[18:45:16] <dmerejkowsky> time to think about bad ideas ...

[18:45:32] <omegak_> well, I'm ready to debug and see where the value comes from

[18:46:58] <omegak_> > any advice on debugging that? I tried printing and pdb, but no luck

[18:50:28] <omegak_> alright, I managed to see values with --verbose

[18:50:52] <omegak_> I'll go rest for today and drink a cup of sanity

[18:51:04] <omegak_> thanks for the help!

[18:51:34] <dmerejkowsky> take care :)