-
Notifications
You must be signed in to change notification settings - Fork 286
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[CI Bot] environment lockfiles auto-update #5918
Conversation
⏱️ Performance Benchmark Report: 424d6adPerformance shifts
Full benchmark results
Generated by GHA run |
⏱️ Performance Benchmark Report: 424d6adPerformance shifts
Full benchmark results
Generated by GHA run |
⏱️ Performance Benchmark Report: 424d6adPerformance shifts
Full benchmark results
Generated by GHA run |
FYI I checked the imagetest locally and the failure is not concerning - some text has moved horizontally but only slightly. Note I'm not currently intending to fully address the failures here as I'm working elsewhere. |
Regarding the warning in the docs build:
This will be fixed when sphinx-gallery/sphinx-gallery#1289 is released, which sounds like it could be soon. Proposed action(Remember that any changes should be made in a new pull request based on this branch, replacing this pull request).
|
49f716d
to
c7692f0
Compare
sphinx-gallery 0.16 was released yesterday. Still waiting for action to get it onto conda-forge (sphinx-gallery-feedstock). |
c7692f0
to
c8ea737
Compare
⏱️ Performance Benchmark Report: 51da245Performance shifts
Full benchmark results
Generated by GHA run |
c8ea737
to
7ffdb11
Compare
⏱️ Performance Benchmark Report: 69432ccPerformance shifts
Full benchmark results
Generated by GHA run |
7ffdb11
to
b4fd1ed
Compare
⏱️ Performance Benchmark Report: c7e0782Performance shifts
Full benchmark results
Generated by GHA run |
UpdateSphinx#5953 demonstrates an effective fix for the Sphinx problem. Image hashThe imagehash failure is much worse unfortunately. The test result has indeed changed, but iris/lib/iris/tests/graphics/__init__.py Lines 202 to 207 in 25b685c
iris/lib/iris/tests/graphics/idiff.py Lines 147 to 148 in 25b685c
A possible fix would be to actually save the tests' iris/lib/iris/tests/graphics/__init__.py Lines 215 to 219 in 25b685c
|
I'm struggling to believe that it can possibly be that, since the whole point of a BytesIO is to act as a proxy for a file, and according to the C/Linux streams "everything is a stream" principle it really should be entirely equivalent. Is it not more likely that the local results on which idiff works are different from the ones in the CI ? |
No, I can replicate the test result locally. I even inserted a |
Note that the two are identical when running in the current 'official' environment spec that is on |
OK I guess we have to believe it! |
b4fd1ed
to
16d3838
Compare
⏱️ Performance Benchmark Report: 13721e9Performance shifts
Full benchmark results
Generated by GHA run |
UpdateThe graphical error is nothing to do with Subsequent saves do not hit the same bug. |
Lockfiles updated to the latest resolvable environment.
If the CI tasks fail, create a new branch based on this PR and add the required fixes to that branch.