[Python-Dev] [Release] Python 3.11.0b4 is still blocked

[Pablo Galindo Salgado]

Hi everyone,

A small update since the last communication from the release team regarding the status of Python 3.11.0b4.

Unfortunately, even if we have fixed most of the original release blockers and 4 more that appear during this week, we still have a bunch of release blockers to deal with. One of them has been reported today.

I would like to release the next beta next week if everything looks good, but there are also some items that need discussion:.

https://github.com/python/cpython/issues/93910

https://github.com/python/cpython/issues/93516

Ideally we should reach consensus as a team on how to proceed in these issues. In particular, we should decide collectively what is an acceptable slowdown, specially looking to the release candidate.

If releasing the next betas is further delayed, I will consider delaying the full release schedule to accommodate for the delay so users have the appropriate time to test and validate every release.

Please do not hesitate in reaching out if you have any questions or concerns or if there is any issue you think we should include in the next release.

Thanks everyone for your help and understanding. I'm sure 3.11 is going to be an outstanding release thank to all of you :)

Cheers from cloudy London,

Pablo Galindo Salgado 

[Miro Hrončok]

On 24. 06. 22 14:25, Pablo Galindo Salgado wrote:
> Hi everyone,
>
> A small update since the last communication from the release team regarding the
> status of Python 3.11.0b4.
>
> Unfortunately, even if we have fixed most of the original release blockers and
> 4 more that appear during this week, we still have a bunch of release blockers
> to deal with. One of them has been reported today.
>
> I would like to release the next beta next week if everything looks good, but

> there are also some items that need discussion...

I was thinking. Are all release blockers automatically blocking the next beta?
Or does it mean this should not be released in final (and hence neither rc)
versions? Would it make sense to release 3.11.0b4 with some not-yet-fixed
blockers? Assuming those are not regressions that happened after 3.11.0b3 was
released.

--
Miro Hrončok
--
Phone: +420777974800
IRC: mhroncok

_______________________________________________
Python-Dev mailing list -- pytho...@python.org
To unsubscribe send an email to python-d...@python.org
https://mail.python.org/mailman3/lists/python-dev.python.org/
Message archived at https://mail.python.org/archives/list/pytho...@python.org/message/MHTM5MREGJCCJPHLZZQF6W4FWDWGAIDU/
Code of Conduct: http://python.org/psf/codeofconduct/

[Pablo Galindo Salgado]

Hi Miro,

>> Are all release blockers automatically blocking the next beta?

Yes.

>> Or does it mean this should not be released in final (and hence neither rc)

versions?

Release blockers block also beta releases (if the RM decides so).

>> Would it make sense to release 3.11.0b4 with some not-yet-fixed

blockers? 

No, the reason is that fixes can introduce more regressions and those need to be fixed. If these fixes

are pretty big we would be risking big changes in the RC phase, which we want to avoid. The idea is that

the fixes to critical problems reported on beta x can be tested on beta x+1.

At the end of the day, this is all subjected to the judgement of the release manager, and given how many

release blockers we have been getting and how many of these have been reported past week *after* several

attempts to release the next beta, I have decided to wait.

Additionally, I am considering pushing the full release some months in the future to allow for more betas, given

how unstable 3.11 is currently.

Pablo Galindo Salgado

[Miro Hrončok]

On 04. 07. 22 18:53, Pablo Galindo Salgado wrote:
> Hi Miro,
>
> >> Are all release blockers automatically blocking the next beta?
>
> Yes.
>
> >> Or does it mean this should not be released in final (and hence neither rc)
> versions?
>
> Release blockers block also beta releases (if the RM decides so).
>
> >> Would it make sense to release 3.11.0b4 with some not-yet-fixed
> blockers?
>
> No, the reason is that fixes can introduce more regressions and those need to
> be fixed. If these fixes
> are pretty big we would be risking big changes in the RC phase, which we want
> to avoid. The idea is that
> the fixes to critical problems reported on beta x can be tested on beta x+1.
>
> At the end of the day, this is all subjected to the judgement of the release
> manager, and given how many
> release blockers we have been getting and how many of these have been reported
> past week *after* several
> attempts to release the next beta, I have decided to wait.

Thanks. Understood.

> Additionally, I am considering pushing the full release some months in the
> future to allow for more betas, given
> how unstable 3.11 is currently.

Some months sounds pretty big to me. Once the current beta is released, I'd be
great to see some updated release schedule. We have just updated the main
Python version Fedora 37 to 3.11 and we have some deadlines I'd like not to miss.

https://fedorapeople.org/groups/schedule/f-37/f-37-key-tasks.html
2022-08-23 - Fedora 37 Beta Freeze
2022-10-04 - Fedora 37 Final Freeze

It would be really great to get something ABI stable at Beta Freeze and at
least an RC at the Final Freeze. If that is not realistic, we would need to
consider a revert.

--
Miro Hrončok
--
Phone: +420777974800
IRC: mhroncok

_______________________________________________
Python-Dev mailing list -- pytho...@python.org
To unsubscribe send an email to python-d...@python.org
https://mail.python.org/mailman3/lists/python-dev.python.org/

Message archived at https://mail.python.org/archives/list/pytho...@python.org/message/EP37TDGLFXE6TKQH6YBIQWSUGKWHC32U/

[Petr Viktorin]

Worse than a one-time revert. With the current schedule, the projects'
testing phases overlap so Fedora can test rebuilding all its Python
software with Python's alphas/betas. If the schedule is adjusted or made
unreliable, Fedora might need to add a six-month delay and rebuild with
final releases -- and find bugs much later.

_______________________________________________
Python-Dev mailing list -- pytho...@python.org
To unsubscribe send an email to python-d...@python.org
https://mail.python.org/mailman3/lists/python-dev.python.org/

Message archived at https://mail.python.org/archives/list/pytho...@python.org/message/FRAKKZNIVUL46JLPMRR76H24RSYRQMP7/

[Pablo Galindo Salgado]

Hi Petr and Miro,

Thanks for communicating your concerns regarding Fedora and the release schedule.

>> It would be really great to get something ABI stable at Beta Freeze and at

least an RC at the Final Freeze. If that is not realistic, we would need to
consider a revert.

I am not sure if there is some typo, but the beta freeze is the first beta. The ABI is frozen

in the first release candidate, not in beta freeze. We won't get stable ABI until the first RC

is released.

>> Some months sounds pretty big to me. Once the current beta is released, I'd be

great to see some updated release schedule. We have just updated the main
Python version Fedora 37 to 3.11 and we have some deadlines I'd like not to miss.

I understand that in the (still not decided) case that the release is delayed will be quite

inconvenient for Fedora and other Linux distributions. I will surely take this into account

when making a decision and I will try to avoid having to fall into this, but please understand

that the Release Team's responsibility is ensuring a stable release, and given the events in the

In past weeks, I do not feel comfortable with the current level of testing so we may require more

betas.

Thanks for your understanding,

Pablo Galindo Salgado