[PySide] Python 3.8 and 3.5

Cristián Maureira-Fredes Cristian.Maureira-Fredes at qt.io
Fri Oct 25 13:48:06 CEST 2019



On 10/25/19 11:10 AM, Florian Bruhin wrote:
> Hey Cristián,
> 
> On Fri, Oct 25, 2019 at 08:57:00AM +0000, Cristián Maureira-Fredes wrote:
>> Python 3.5 has become really old for most distributions,
>> so we would like to drop support for it, in favor of goodies
>> that appeared on 3.6.
>>
>> [...]
>>
>> Python 2.7 is another story,
>> but as mention in previous blog post
>> we are planning to deprecate it as soon as Qt6 is released
>> in favor of people still being unable to move forward.
> 
> I don't really understand this. Looking at the official support timeline for
> those Python versions, it seems backwards:
> 
> 3.5 end of life: 2020-09-13
> 2.7 end of life: 2020-01-01
> 
> In other words, you're dropping support for a Python version from 2015 which
> has upstream support for (almost) another year, but continue to support a
> version from 2010 which has upstream support for another two months.
> 
> Also note that Python 3.5 is shipped in Ubuntu Xenial (16.04 LTS) which is
> still supported until April 2021 and in Debian Stretch which is supported for
> even longer. While there is an upgrade path (18.04 LTS and Debian Jessie,
> respectively), people who are still on a supported and common distribution
> won't be able to use a newer PySide anymore.
> 
> Finally, what features does dropping 3.5 give you if you need to keep 2.7
> support anyways?
> 
> Florian
> 

Hey Florian,

the reason behind dropping support for 2.7 so late,
is just by similar situations as the email by Stephen (on this thread), 
many people is still unable to move to python 3 sadly,
that's why we decided to wait.

The 3.5 story is mainly to start dropping unused Python versions,
I personally was unaware that Ubuntu 16.04 had still such old version,
so I guess there is no reason then to speed up the drop of the support.

 From 3.6 there are many features we wanted to use, like:
* f-strings (and refactor our code base),
* asyncio stable API (we wanted to move forward and provide compatibility),
* Type annotation for function parameters,
* default utf-8 encoding on windows for file system paths (we had a 
couple of issues)
* The ability to check if the GIL is being held (to help us solve
a couple of problems with GIL and event loop),
* among others,

If it's so critical to you,
I think this could be postponed to the release of Qt6 too,
after all the final release 3.5.8 is scheduled for tomorrow,
and then it naturally will be unused.

So, hopefully next year we can continue our discussion.

Thanks for pointed that out.
Cheers


-- 
Dr. Cristian Maureira-Fredes
Software Engineer

The Qt Company GmbH
Erich-Thilo-Str. 10
D-12489 Berlin

Geschäftsführer: Mika Pälsi,
Juha Varelius, Mika Harjuaho
Sitz der Gesellschaft: Berlin,
Registergericht: Amtsgericht
Charlottenburg, HRB 144331 B
--


More information about the PySide mailing list