[Qgis-developer] Moving towards QGIS3?

[Nyall Dawson]

On 24 August 2016 at 08:19, Paolo Cavallini <cavallini at faunalia.it> wrote:
> Il 23/08/2016 23:14, Nyall Dawson ha scritto:
>
>> We've been closing as they are fixed, so everything still open is
>> outstanding.
>>
>> I don't think anyone's specifically working through that list yet - we
>> are very much still in the "cleanup" phase and removing years of legacy
>> cruft.
>>
>> There's still a lot of work to do to remove all the remaining
>> deprecations, but at least with every removal the code gets a bit
>> cleaner and future tasks are easier.
>>
>>> Should we prioritize the remaining ite, or redefine them as
>>> necessary|useful|nice to have, so we know the minimum set to complete
>>> for 3.0 release?
>>> All the best.
>>
>> So far everything has been done by just a handful of devs in their spare
>> time. I'm not sure how useful it is for someone to specifically
>> prioritize them. I think it's being driven more by how annoying each
>> individual dev finds that part of the API!
>>
>> I'd say we are still very early days in the work towards 3.0 and release
>> checklists could come later.
>
> Hi Nyall,
> thanks for your reply. Here at the HF we are discussing about this, and
> it seems to me that the work can be split in two major areas:
> * removal of cruft and general cleanup
> * implementation of new, better APIs.
> It seems that the first phase is being done, partly automatically, in
> large part manually, in small steps, whereas a good part of it could be
> made automatic, with some perl magic etc.

I disagree - yes, some of this was done using scripts in the past (the
removal of V2 from names), but I'm sure Jürgen would agree that there
was still a lot of manual grunt work required.

Personally, I'm ULTRA cautious whenever removing things marked as
deprecated, because they are still sometimes used in the codebase or
core python plugins. It takes a lot of careful grepping through the
code to make sure there's no unpleasant side effects (especially since
only processing has unit tests which would inform us of breakage).

Is there any concerns about the current approach? I've been thinking
progress is moving along nicely...

> So my proposal: set aside some of our budget for QGIS grant to allow key
> people (I'm thinking especially to Juergen, possibly also Nyall,
> Matthias and others) to do the first part as they feel it more efficient
> and appropriate. This should pave the way for the second part, which
> could therefore be faster.
> How does it sound?
> All the best.

I'm sure none of us would object to being paid to do work we currently
voluntarily do ;)

Nyall

> --
> Paolo Cavallini - www.faunalia.eu
> QGIS & PostGIS courses: http://www.faunalia.eu/training.html
> _______________________________________________
> Qgis-developer mailing list
> Qgis-developer at lists.osgeo.org
> List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer