| 1 |
Hi, everyone. |
| 2 |
|
| 3 |
TL;DR: |
| 4 |
|
| 5 |
1. Gentoo eclasses now support Python 3.11 in ~arch. Please start |
| 6 |
testing your packages. (It's going to be stabilized once it leaves |
| 7 |
beta, i.e. sometime Nov-Dec.) |
| 8 |
|
| 9 |
2. We're switching to Python 3.10 in the beginning of June, so if you |
| 10 |
haven't ported your packages yet, please hurry. |
| 11 |
|
| 12 |
3. While updating your packages, please switch to DISTUTILS_USE_PEP517 |
| 13 |
build. |
| 14 |
|
| 15 |
For help, please consult our docs: |
| 16 |
https://projects.gentoo.org/python/guide/ |
| 17 |
|
| 18 |
|
| 19 |
=================================== |
| 20 |
Python 3.11 now supported in Gentoo |
| 21 |
=================================== |
| 22 |
Python 3.11.0b1 has been released recently. This means upstream feature |
| 23 |
freeze (i.e. technically no breaking changes, unless they do some last- |
| 24 |
minute reverts like with 3.10.0 final). This in turn we can start |
| 25 |
testing packages and marking them as 3.11 happy. |
| 26 |
|
| 27 |
We've already have ported some initial packages (which nowadays means |
| 28 |
tons of packages). My experience so far is that Python 3.11 is rather |
| 29 |
painless, compared to previous releases. What usually breaks are gross |
| 30 |
hacks or deprecations ignored for years (both common in Python ecosystem |
| 31 |
but not that common). |
| 32 |
|
| 33 |
Bad news is that for one useful package, you may end up having to deal |
| 34 |
with half a dozen unmaintained, broken NIH dependencies. Good news is |
| 35 |
that we're not alone with this, and neither we're the first ones to |
| 36 |
fight Python 3.11. So if something's broken, then there's a good chance |
| 37 |
that either it's already fixed in the upstream VCS repository or someone |
| 38 |
has made a pull request/patch for it. |
| 39 |
|
| 40 |
My usual recommendation is to prefer fixing problems over skipping |
| 41 |
tests. However, if something is important enough and the test failures |
| 42 |
are minor compared to the package's functionality, you can use |
| 43 |
EPYTEST_IGNORE and EPYTEST_DESELECT to skip broken tests conditionally |
| 44 |
to Python version. Sometimes it also makes sense to skip some |
| 45 |
dependency (python_gen_cond_dep) and possibly tests depending on it |
| 46 |
(using has_version). You can find some examples in the Guide, as well |
| 47 |
as by grepping the existing ebuilds. |
| 48 |
|
| 49 |
One package we haven't ported yet (and that we would really prefer not |
| 50 |
to port) is dev-python/nose. It's unmaintained for years, we're already |
| 51 |
carrying a ton of patches to keep it working and with Python 3.11, it's |
| 52 |
once again a mess. Plus, the code itself is a mess full of hacks. Good |
| 53 |
news is: 3.11 broke it completely, so upstreams can't just ignore |
| 54 |
the problem ("it works in my virtualenv"), so there's some hope they |
| 55 |
will be for switching to another test runner. That is, unless you're |
| 56 |
dealing with yet another unmaintained NIH dependency of an unmaintained |
| 57 |
NIH dependency. But there's some chance that you should be able to get |
| 58 |
it running with pytest easily or with a reasonably small patch. |
| 59 |
|
| 60 |
I'd also like to ask you to update your packages to EAPI 8 (if possible) |
| 61 |
and DISTUTILS_USE_PEP517 while touching them. Setuptools has deprecated |
| 62 |
`setup.py install` that the legacy builds are using long time ago. It |
| 63 |
doesn't look like it's going to stop working anytime soon but it's |
| 64 |
better to avoid running into last minute porting. |
| 65 |
|
| 66 |
The new flags are currently stable-masked on all profiles. Python |
| 67 |
3.11.0 final release is planned for October 2022. We're probably going |
| 68 |
to stabilize it sometime after, then work on unmasking the flags. |
| 69 |
|
| 70 |
|
| 71 |
============================================= |
| 72 |
Python 3.10 coming as the default interpreter |
| 73 |
============================================= |
| 74 |
As announced before, we're going to be switching to Python 3.10 |
| 75 |
in the beginning of June 2022. There shouldn't be any unplanned delay, |
| 76 |
and all the most important packages are ready. |
| 77 |
|
| 78 |
If you haven't ported your packages, please do so ASAP. |
| 79 |
|
| 80 |
If something isn't suitable for Python 3.10, it's probably a good idea |
| 81 |
to give the users an early warning that it's likely to be lastrited |
| 82 |
at some point. |
| 83 |
|
| 84 |
If you need help, please don't hesitate to ask on #gentoo-python. |
| 85 |
We can help, point you at the right docs or good examples of ebuilds. |
| 86 |
|
| 87 |
|
| 88 |
TIA for all you help and have fun with new Pythons! |
| 89 |
|
| 90 |
-- |
| 91 |
Best regards, |
| 92 |
Michał Górny |
Archives