You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@mathause I'm personally really uncomfortable pushing out a release when there are multiple known breakages against dependencies - see Must have list. Help is welcome!
Activity
dcherian commentedon Oct 22, 2019
#3406 needs a review
mathause commentedon Oct 28, 2019
Would be nice to get this out - just hit #3426. (I now updated to master but this is besides the point).
crusaderky commentedon Oct 29, 2019
@mathause I'm personally really uncomfortable pushing out a release when there are multiple known breakages against dependencies - see Must have list. Help is welcome!
crusaderky commentedon Oct 29, 2019
@barronh any update re pseudonetcdf 3.1?
barronh commentedon Oct 31, 2019
I believe there are two issues here.
First, the 3.1 release was not 2.7 compliant. It mixed named keyword arguments with
**kwds. That could be fixed easily.Second, the camx reader changed subclasses, which affects the xarray test for CF variables. I could update the test that would fix xarray test-case.
crusaderky commentedon Oct 31, 2019
@barronh, we don't care about Python 2.7 compliance. We support >=3.6.
barronh commentedon Nov 4, 2019
Got it. This is a test update to a test that is backward and forward compatible. I'll get something checked in.
mathause commentedon Nov 5, 2019
xref for numpy 1.18 numpy/numpy#14831
dcherian commentedon Nov 5, 2019
@mathause Thanks for filing these upstream issues!
dcherian commentedon Nov 7, 2019
Once #3491 is merged, we should be ready to release since numpy will fix itself (numpy/numpy#14841)
mathause commentedon Nov 7, 2019
Don't we have to figure out whether to use
np.meanornp.nanmeanfor datetime vectors?dcherian commentedon Nov 11, 2019
I'm in favour of releasing even if we are currently broken with numpy master.
Our current release is also broken with numpy master, so the situation is no worse...
crusaderky commentedon Nov 15, 2019
If someone could review and merge #3537 and #3531, we can close this!
crusaderky commentedon Nov 19, 2019
I think we're good to go!
dcherian commentedon Nov 19, 2019
I agree! @crusaderky, @max-sixty: do either of you have the time to make a release?
max-sixty commentedon Nov 19, 2019
Yup I can do it later this afternoon!
dcherian commentedon Nov 19, 2019
👍
dcherian commentedon Nov 19, 2019
Thanks @max-sixty I activated the RTD build.