Skip to content
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

Unblock documentation build #10807

Merged

Conversation

milesgranger
Copy link
Contributor

@milesgranger milesgranger commented Jan 16, 2024

Just restricting the newish slew of releases from sphinxcontrib which made the sphinx>=5 start failing. Feel free to merge this now, while I still poke at who's causing the <=4.5.0 restriction.


Caused by dask-sphinx-theme pinned to sphinx-book-theme==0.2.0 which requires sphinx<5.
xref dask/dask-sphinx-theme#68

@milesgranger
Copy link
Contributor Author

@phofl I'm not sure if we want 7c14d41, but as it's not a library, pinning everything seems like an easy way to protect against future breakages from others upgrading their dependencies.

@milesgranger milesgranger force-pushed the milesgranger/10805-doc-build-broken branch from 5ffb71e to 7c14d41 Compare January 16, 2024 13:47
@milesgranger
Copy link
Contributor Author

Tried with mamba: https://readthedocs.org/projects/dask/builds/23132383/
but not happy. Can try again in a follow-up, but using pip with the pinned packages works for now at least.

environment.yml
name: dask-docs
channels:
  - conda-forge
dependencies:
  - python=3.11
  - cloudpickle>=1.5.0
  - pandas>=1.4.0
  - sphinx>=4.0.0
  - numpydoc
  - dask-sphinx-theme>=3.0.0
  - sphinx-click
  - sphinx-copybutton
  - sphinx-remove-toctrees
  - sphinx-design
  - jupyter_sphinx
  - toolz
  - fsspec
  - scipy
  - pytest
  - pytest-check-links
  - requests-cache
  - ipython
  - ipykernel<6.22.0
  - pip
  - pip:
      - sphinx-tabs
      - sphinx_autosummary_accessors
      - git+https://github.com/dask/distributed

Copy link
Contributor

github-actions bot commented Jan 16, 2024

Unit Test Results

See test report for an extended history of previous test failures. This is useful for diagnosing flaky tests.

     15 files  ±0       15 suites  ±0   3h 27m 54s ⏱️ - 14m 37s
 12 941 tests ±0   11 593 ✅ +3     929 💤 ±0  419 ❌  - 3 
160 329 runs  ±0  143 023 ✅ +3  16 887 💤 ±0  419 ❌  - 3 

For more details on these failures, see this check.

Results for commit 57569d8. ± Comparison against base commit ca42dcf.

♻️ This comment has been updated with latest results.

@crusaderky
Copy link
Collaborator

-1 from me about pip-compile. Past experience with exact pins has been unexpectedly painful for me.
Can we just add the 5 new restrictions on sphinxcontrib to requirements-doc.txt?

@milesgranger
Copy link
Contributor Author

milesgranger commented Jan 16, 2024

-1 from me about pip-compile. Past experience with exact pins has been unexpectedly painful for me.
Can we just add the 5 new restrictions on sphinxcontrib to requirements-doc.txt?

Sure thing, no problem. Out of curiosity, do you remember what your pains have been? W/o pinning application requirements I find myself doing this little number every now and again. My experience with fully pinned requirements outside of libs have been rather pleasant.

Copy link
Member

@jrbourbeau jrbourbeau left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @milesgranger! One minor comment about a comment

docs/requirements-docs.txt Show resolved Hide resolved
@jrbourbeau jrbourbeau changed the title Bandaid fix for docs Unblock documentation build Jan 16, 2024
@jrbourbeau jrbourbeau merged commit 69d209f into dask:main Jan 16, 2024
24 of 26 checks passed
@milesgranger milesgranger deleted the milesgranger/10805-doc-build-broken branch January 16, 2024 19:49
@milesgranger milesgranger mentioned this pull request Jan 17, 2024
@crusaderky
Copy link
Collaborator

crusaderky commented Jan 17, 2024

Sure thing, no problem. Out of curiosity, do you remember what your pains have been? W/o pinning application requirements I find myself doing this little number every now and again. My experience with fully pinned requirements outside of libs have been rather pleasant.

My experience, specifically with conda, is that builds can be deleted by the author or by maintenance bots and it's happened to me many times that an exact-pin conda environment (output of conda env export) will just stop working after a while because one maintainer of one of the 200, mostly obscure, packages in the environment has been rebuilt by the author for whatever reason and the previous build has been deleted.

I don't have first-hand experience with pip.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants