Hi, I agree with Nyall. As this is a major release this requires more time, have a look at how long it took Python 3 to finally become used for an idea on how people treat major breaks in API.
As we already have 2.14, 1.6 and 2.18 out the door as a really good base I don't see a need to rush this out. I have plans for 3.0 but don't have a Windows build setup yet. Regards, Nathan On Wed, Oct 26, 2016 at 4:10 PM, Neumann, Andreas <[email protected]> wrote: > Hi, > > It would be fine for me to have a double dev cycle for 3.0. Devs should > have enough time to make proper decisions and work on the API. And python > devs probably also need more time to get their most important plugins in > shape for the new API, qt5 and Python 3. > > Perhaps it would be good to also have more than one month for testing and > bug fixing - extending this to two months. We could simply shift the one > month bug fixing/testing of the first cycle towards the end of the second > cycle. > > So this modified proposal would mean: > > Feature freeze at the end of May, release at the end of July. Right? > > Fine with me, if we still care about the 2.x branch where necessary - > preferably investing more in 2.18x than in 2.14x (my personal opinion). > > Andreas > > On 2016-10-26 07:53, Nyall Dawson wrote: > > Hi all, > > I'd like to start the discussion around this early so that we can plan > ahead and not have to make a last-minute decision. > > What are everyone's thoughts on extending the timeline for 3.0? In my > opinion things are currently going really well, we have Qt5/python3 > builds which are stable enough for daily use and there's been a ton of > cleanups to the code. > > There's a lot of changes still coming in, and I think there's SO much > room for making things better that I don't like the idea of the early > 2017 deadline for the final release. I'd much rather extend this out > by another cycle and really getting the platform ready for the next > series of QGIS releases. > > We could always put out a "preview" release in March, without frozen > API, if desired. > > So, what's everyone's thoughts? Good idea? Bad idea? > > Nyall > _______________________________________________ > Qgis-developer mailing list > [email protected] > List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer > Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer > > > > > _______________________________________________ > Qgis-developer mailing list > [email protected] > List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer > Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer >
_______________________________________________ Qgis-developer mailing list [email protected] List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer
