[01:03:06] <McSinyx[m]> pip -v is already too verbose for normal use
[01:05:51] <McSinyx[m]> IIRC there used to be a warning when the new resolver was an experimental feature
[01:06:56] <McSinyx[m]> still, a small line like ``Enabled features: 2020-resolver'' doesn't seem like a bad idea
[14:21:24] <graingert> I'm getting a strange interaction between tox and pip --use-feature=2020-resolver https://github.com/Teemu/pytest-sugar/pull/215/files ci is still running
[14:22:07] <graingert> but I'm expecting py38-pytest54-supported-xdist to install pytest-xdist<2.0.0 pytest~=5.4.3
[14:22:57] <graingert> but it install pytest-xdist==2.10 pytest==5.4.3 - which is incompatable
[14:23:12] <graingert> * correction: but it installs pytest-xdist==2.1.0 pytest==5.4.3 - which is incompatable
[14:23:52] <graingert> this is unexpected because pytest-xdist 2.0.0 and above install_requires = [..., "pytest>=6.0.0", ...]
[14:24:13] <graingert> I'm thinking this is most likely an tox big
[18:23:21] <tos9> before that there should be a release where we flip a coin and use the random resolver it returns
[19:58:39] <graingert> zbr: oh that would be handy for https://github.com/Teemu/pytest-sugar/pull/203/files#diff-b91f3d5bd63fcd17221b267e851608e8R9-R10
[22:21:11] <graingert> will `--use-feature=2020-resolver` ever be deprecated/removed?
[22:21:52] <graingert> eg say it's 2030 and it turns out P = NP, will the flag just be ignored?