[Qgis-psc] Vote about release plan

[Marco Hugentobler]

Hi Hugo

Personally I don't have a strong opinion about 2.10 yes or no.

>About PyQT5 and Python3, naive question: is PyQT4 considered obsolete ?
>What are the good reasons to use PyQT5 ?
>I am pretty sure Python 2.X is not yet obsolete :)


The idea is to have API breaks as rare as possible. And if, then do the 
API breaking things in one release. So having composer API break in 
autumn release and PyQt5 break in a later release is not good (unless 
you want to delay the move to PyQt5 for years).

>About PyQT5 and Python3, naive question: is PyQT4 considered obsolete ?
>What are the good reasons to use PyQT5 ?
>I am pretty sure Python 2.X is not yet obsolete :)


Well, Qt3 is not obsolete and I'm still happy we moved to a later 
version at some point.

Regards,
Marco


On 16.04.2015 17:12, Hugo Mercier wrote:
> Hi,
>
> Sorry, I am not a PSC member, but the question here seems important ...
>
> I also think there is no urge to break all the plugins in the next release.
>
> What about Régis' comments ? There is an important issue here: a plan
> with 2.x releases have been advertised some time ago and people count on
> it to discuss new features, fund and make them developed. And they have
> to be part of the next stable non API-breaking release.
> We are currently developing new features that we expect initially to
> land in a 2.10. If the next release force our customers to rewrite all
> their plugins, then we might want to deliver new features in a
> non-official release, which is not a very good idea.
>
> So, why would we want a API break ? In my opinion, only for changes in
> situations where it is nearly impossible to do otherwise.
>
> Break the API for the removal of obsolete features ? I would be in favor
> of that. But we could state that removing obsolete features are OK and
> are not considered a "hard break" that would require to rewrite all the
> plugins ? After all, obsolete code is marked as such since a long time.
>
> About PyQT5 and Python3, naive question: is PyQT4 considered obsolete ?
> What are the good reasons to use PyQT5 ?
> I am pretty sure Python 2.X is not yet obsolete :)
>
> Same as other people: no problem for a new breaking version with last
> versions (PyQT5, Python3, C++11, etc.) but please announce it in
> advance, don't replace the expected next non-breaking release ...
>
> Le 16/04/2015 16:50, Marco Hugentobler a écrit :
>> Hi Nathan
>>
>> The idea was that QGIS 3 means Qt 5, PyQt5 and python 3 (bring
>> everything to recent versions without the need to provide backwards
>> compatibility). At some point we need to make that step, delaying it
>> does not make it easier.
>>
>> Opinions?
>>
>> Regards,
>> Marco
>>
>> On 16.04.2015 16:35, Nathan Woodrow wrote:
>>> Going for 3.0 also about when and why we break API with more of a lean
>>> towards the why.  For me the only thing really forcing us is PyQt5 and
>>> that is the only real driver, pretty much everything else can be retro
>>> fitted in.  Not nice but breaking API just just because and for a
>>> version number isn't cool.
>>>
>>> So with that said when are we going to be forced to go to PyQt5? If we
>>> aren't going to be forced for a while then I think we should just keep
>>> keeping on and do 3.0 later.
>>>
>>> If I had a vote it would be to stay with 2.x for as long as we can
>>> until forced with PyQt5, we can always write wrappers over hard to us APIs
>>>
>>> Nathan
>>>
>>>
>>> On Fri, 17 Apr 2015 12:24 am Marco Hugentobler
>>> <marco.hugentobler at sourcepole.ch
>>> <mailto:marco.hugentobler at sourcepole.ch>> wrote:
>>>
>>>      Hi
>>>
>>>       From the last psc meeting and the following discussion on the mailing
>>>      list (http://lists.osgeo.org/pipermail/qgis-psc/2015-April/002966.html
>>>      and following mails) it seems most people are happy with skipping the
>>>      2.10 release and having 3.0 in autumn (feature freeze September,
>>>      release
>>>      october).
>>>
>>>      It will be good to have an official PSC decision with voting about it.
>>>      Please give your votes if you agree with skipping 2.10 and go for
>>>      3.0 in
>>>      autumn.
>>>
>>>      (obviously +1 for me).
>>>
>>>      Regards,
>>>      Marco
>>>
>>>      --
>>>      Dr. Marco Hugentobler
>>>      Sourcepole -  Linux & Open Source Solutions
>>>      Weberstrasse 5, CH-8004 Zürich, Switzerland
>>>      marco.hugentobler at sourcepole.ch
>>>      <mailto:marco.hugentobler at sourcepole.ch> http://www.sourcepole.ch
>>>      Technical Advisor QGIS Project Steering Committee
>>>
>>>      _______________________________________________
>>>      Qgis-psc mailing list
>>>      Qgis-psc at lists.osgeo.org <mailto:Qgis-psc at lists.osgeo.org>
>>>      http://lists.osgeo.org/mailman/listinfo/qgis-psc
>>>
>>
>> -- 
>> Dr. Marco Hugentobler
>> Sourcepole -  Linux & Open Source Solutions
>> Weberstrasse 5, CH-8004 Zürich, Switzerland
>> marco.hugentobler at sourcepole.ch http://www.sourcepole.ch
>> Technical Advisor QGIS Project Steering Committee
>>
>>
>>
>>
-- 
Dr. Marco Hugentobler
Sourcepole -  Linux & Open Source Solutions
Weberstrasse 5, CH-8004 Zürich, Switzerland
marco.hugentobler at sourcepole.ch http://www.sourcepole.ch
Technical Advisor QGIS Project Steering Committee