https://pantsbuild.org/ logo
Join Slack
Powered by
I've got a bunch of repeated failures that don't s...
# development
g
I've got a bunch of repeated failures that don't seem like flakes anymore while backporting the formatting changes to 2.14 and 2.15 that I think needs an extra pair of eyes... they should be unrelated to my code, and I'm wondering if those branches are just rotten? 
deploy_jar_test.py
:
Copy code
Error: An unexpected error occurred while trying to open file dave.jar
package_pex_binary_integration_test.py
, 
test_plugin_resolver.py
:
Copy code
Forbidden skipped test - Skipped: Could not find python 3.6 on system. Skipping.
flake8/rules_integration_test.py
:
Copy code
E       assert 1 == 0
E         +1
E         -0
nodejs_test.py
: (might be an actual flake, no more output)
Copy code
Exceeded timeout of 240.0 seconds when executing local process: Run Pytest for src/python/pants/backend/javascript/subsystems/nodejs_test.py:tests
b
Thanks for the rustfmt efforts. It seems unlikely that those branches will be touched again (given I think we’re already thinking 2.16.1 will be the last of 2.16.x), so potentially we could say that it was worth trying but not worth investing significant effort into 2.14 and 2.15? If a further release is warranted, can always do the work at that point?
g
That's fine with me 😄 It's easy enough to reformat if we absolutely need to backport some Rust code, or just un-format the incoming code before cherry-picking.
c
yea, as Huon says, I think 2.14 and 2.15 is most likely closed by now. and 2.16 is not far away of being more or less closed as well. (i.e. my +1)
h
Agreed. No need to go past 2.16.x
I am doing one last rc on 2.16 for a bugfix
It is already done, actually
g
Then all PRs should be ready to go I believe, and we'll just close 2.14 and 2.15.
(Actually just broke one of them by clicking update instead of ready-for-review... Will fix before we merge them then.)
Ok; fixed that one -- @happy-kitchen-89482 when you have time, feel free to rebase-merge them
b
Is there a GitHub way to "lock" branches?
g
No, but explain why you'd want it here because it's not something I've considered.
b
Half a way of declaring "this branch exists as an archive" and half so people don't try and open PRs against it expecting the changes to go iny
g
Ah! So not related to the reformatting itself. Was worried I'd missed something here.
b
Well presumably we wouldn't need to reformat those branches, and you might've spent less time
g
Fair enough, yeah. Debugging flakes isn't super fun :p
w
Is there a GitHub way to "lock" branches?
Maybe branch protection rules?
✅ 1
b
Hmm maybe we start leveraging that. @happy-kitchen-89482 can you add this to the next Pants meetup? I'll speak
👍 1
h
Done
3 Views
Open in Slack
PreviousNext
Powered by