On 1/04/19 11:09, Martin Landa wrote:
Hi,
po 1. 4. 2019 v 10:10 odesílatel Moritz Lennert
<mlenn...@club.worldonline.be> napsal:
How good is Python 3 support in trunk currently, especially for wxGui.
As would be expected, most errors I've seen in the past seem to be
linked either to wxGUI, or to lib/python.
Do you think it would be a good idea to create a 7.8 release branch
before having solved most of the Python 3 problems, or wouldn't it be
better to focus on that in trunk and create the realease branch only
once this has reached a reasonable state.
I would say that we are not expecting that 7.8.0 will be perfect and
all problems/bugs will be solved. If we plan a perfect release that we
will never release :-)
I obviously agree, but there is a difference between releasing a tech
preview or a real release, i.e. something that we officially consider
stable (but not perfect).
Moreover Python3 support will be widely tested
when there will be packages available to install. I would expect that
many problems will be discovered after release. Then we can come with
a more frequent point releases.
So, IIUC, the idea is to create an official release to get more testers
? I'm a bit afraid that if we create an official release and many people
hit upon errors that we will actually scare away potential users...
But all depends on the actual current usability for a normal user (i.e.
someone relying on the GUI).
I'm sorry, I have been fairly absent from grass-dev in the last months,
and the situation will not get much better in the near future, and so I
don't want to block if others are for releasing quickly. Also, I'm not
aware of the testing efforts that have already happened (except for
Mundialis). Are the automatic tests run systematically in a Python 3
environment ?
Moritz
_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/grass-dev
