Make the output reproducible #82
Conversation
Whilst working on the Reproducible Builds effort [0] I noticed that sphinx-panels does not create reproducible output. This is because it uses Python's uuid.uuid4 to generate unique identifiers, but those numbers are random/nondeterminstic by design. This patch will seed these random numbers from SOURCE_DATE_EPOCH if it exists, otherwise it will revert back to random numbers. I originally filed this in Debian as bug #1017475 [1]. [0] https://reproducible-builds.org/ [1] https://bugs.debian.org/1017475
|
Thanks for submitting your first pull request! You are awesome! |
|
Heya, I'm afraid sphinx-panels is now deprecated for sphinx-design: https://sphinx-design.readthedocs.io/en/furo-theme/get_started.html#migrating-from-sphinx-panels |
thanks! in debian we have a handful of packages using sphinx-panels, with the 2 most prominent being matplotlib and scipy; are you aware of any plans from those projects to migrate / were they told to migrate over to sphinx-design? thanks! |
Whilst working on the Reproducible Builds effort [0] I noticed that sphinx-panels and sphinx-design do not create reproducible output. This is because it uses Python's uuid.uuid4 to generate unique identifiers, but those numbers are random/nondeterminstic by design. This patch will seed these random numbers from SOURCE_DATE_EPOCH if it exists, otherwise it will revert back to random numbers. I originally filed this in Debian as bug #1017475 [1], as well as within sphinx-panels [2] [0] https://reproducible-builds.org/ [1] https://bugs.debian.org/1017475 [2] executablebooks/sphinx-panels#82
|
Sure thing. Have also filed this over in sphinx-design: executablebooks/sphinx-design#90 |
Whilst working on the Reproducible Builds effort I noticed that sphinx-panels does not create reproducible output.
This is because it uses Python's
uuid.uuid4to generate unique identifiers, but those numbers are random/nondeterminstic by design. This patch will seed these random numbers fromSOURCE_DATE_EPOCHif it exists, otherwise it will revert back to random numbers.(I originally filed this in Debian as bug #1017475.)