hey folks: i'm on release duty this week, but am a...
# development
w
hey folks: i'm on release duty this week, but am around all weekend. let me know if there is anything that i should know!
c
There has been a lot of build instability / churn with the Python 3.11 upgrade and the numerous GitHub action removals we were racing (ex: https://github.com/pantsbuild/pants/issues/21616) I think we are over that now, but sorry if there is a hot potato still.
👍 1
h
One thing to be aware of is: https://pantsbuild.slack.com/archives/C0D7TNJHL/p1732127874202679 https://pantsbuild.slack.com/archives/C0D7TNJHL/p1733353732198409 Which is leading me to tackle the underlying issue, which is https://github.com/pantsbuild/pants/issues/18250 It’s unclear why this mypy error only happens sometimes, or why now (it has manifested when run locally on laptops for a long time, but I thought that might have been an artifact of running on the
--changed
subset of the repo; But now we’re seeing it in CI, and it’s blocking some PRs from being merged). Anyway, fixing that underlying issue has led to the discovery of a related possibly bug, which I’m currently trying to reason through, expose, and fix.
Actually, perhaps you know: Why does FrozenDict require that its values be hashable? Is it a proxy for requiring that they be immutable so they can be fingerprinted for caching?
That’s fine if so, but I don’t think we’d ever actually need to call
hash()
itself on those values?
Either way, I am working through the fix etc, so no need to dive into this, just wondering if you have context
c
Pre 2.24 release stamp request on :https://github.com/pantsbuild/pants/pull/21731 (had to do a manual cherry pick so I can't self approve)
w
🚢