Three days until release and a bug in the Linux kernel has turned a dozen buildbots red...
It's already been fixed in the kernel, but will take some time to bubble up. We'll skip that test for relevant kernel versions in the meantime.
https://buildbot.python.org/#/release_status
#Python #Python314
Release day!
First off, check blockers and buildbots.
A new release-blocker appeared yesterday (because of course) but it can wait until 3.14.1.
https://github.com/python/cpython/labels/release-blocker
Three deferred-blockers are also waiting until 3.14.1.
https://github.com/python/cpython/labels/deferred-blocker
A new tier-2 buildbot failure appeared yesterday (because of course) but it had previously been offline for a month and will need some reconfiguration. Can ignore.
https://buildbot.python.org/#/release_status
OK, let's make a Python!
Next up, merge and backport the final change to What's New in Python 3.14 to declare it latest stable.
https://github.com/python/cpython/pull/139631
Now start run_release.py, the main release automation script, which does a bunch of pre-checks, runs blurb to create a merged changelog, bumps some numbers, and pushes a branch and tag to my fork. It'll go upstream at the end of a successful build.
Then kick off the CI to build source zips, docs and Android binaries.
https://github.com/python/release-tools/actions/runs/18308460797
(That's actually the second CI attempt, we had to update some script arguments following an Android test runner update.)
This build takes about half an hour.
I've also informed the Windows and macOS release managers about the tag and they will start up installer builds.
This takes a few hours, so I've got time to finish up the release notes.
PEP 101 is the full process, but much is automated and we don't need to follow it all manually.
The Windows build has been started.
The jobs with profile-guided optimisation (PGO) build once, then collect a profile by running the tests, and then build again using that profile, to see how "real" code executes and optimises for that.
https://dev.azure.com/Python/cpython/_build/results?buildId=164907&view=results
Meanwhile, the docs+source+Android build has finished and the artifacts have been copied to where they need to go with SBOMs created.
macOS installer done, next on the final publishing and announcing steps.
🚀 It's out!
🥧 Please install and enjoy Python 3.14!
https://discuss.python.org/t/python-3-14-0-final-is-here/104210?u=hugovk
And the last few tasks: announce also on the blog & mailing lists, update the PEP & downloads landing page, fix Discourse post links, unlock the 3.14 branch for the core team to start landing PRs that didn't need to be in the RC, and eat the pie.
https://blog.python.org/2025/10/python-3140-final-is-here.html
https://www.python.org/downloads/
https://peps.python.org/pep-0745/
A HUGE thanks to @sovtechfund Fellowship for allowing me to dedicate my time on getting this out 🎉
https://hugovk.dev/blog/2025/im-excited-to-join-the-sovereign-tech-fellowship/
![ERROR: test_aead_aes_gcm (test.test_socket.LinuxKernelCryptoAPI.test_aead_aes_gcm)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/buildbot/buildarea/3.14.cstratak-fedora-stable-x86_64/build/Lib/test/test_socket.py", line 7071, in test_aead_aes_gcm
op.sendall(plain)
~~~~~~~~~~^^^^^^^
OSError: [Errno 22] Invalid argument](/proxy/post_attachment/136666/1aa0f4411b.png "ERROR: test_aead_aes_gcm (test.test_socket.LinuxKernelCryptoAPI.test_aead_aes_gcm)
----------------------------------------------------------------------
Traceback (most recent call last):
File \"/home/buildbot/buildarea/3.14.cstratak-fedora-stable-x86_64/build/Lib/test/test_socket.py\", line 7071, in test_aead_aes_gcm
op.sendall(plain)
~~~~~~~~~~^^^^^^^
OSError: [Errno 22] Invalid argument")
