Hi All, While waiting for Christoph to drop the other shoe on 1.10.0b1, I thought I'd try again to start a discussion on the 1.11 release. If we want to get the next release out in a timely manner there should be some advance planning to cover at least the following three items - Release manager: I can do that again, but would be more than happy to let someone else give it a shot. - Goals to meet: - __numpy_ufunc__ - masked array and recarray fixups for astropy - ? - Release date to shoot for, I'd suggest Feb 1, 2016 Thoughts? Chuck
On So, 2015-08-16 at 14:04 -0600, Charles R Harris wrote:
Hi All,
While waiting for Christoph to drop the other shoe on 1.10.0b1, I thought I'd try again to start a discussion on the 1.11 release. If we want to get the next release out in a timely manner there should be some advance planning to cover at least the following three items
* Release manager: I can do that again, but would be more than happy to let someone else give it a shot. * Goals to meet: * __numpy_ufunc__ * masked array and recarray fixups for astropy * ? * Release date to shoot for, I'd suggest Feb 1, 2016
Sounds good to me. Feb. 1 probably would mean a feature freeze in the beginning of January? I guess we could try to push a bit harder to a stricter 4 months cycle, but with the common and of year stress and holiday season it sounds reasonable. Can't think of any goal to explicitly meet [1], but I guess that is mostly downstream wishes anyway ;). Personally, I would like to see Pauli's overlap detection and the proposed indexing changes by the next release. That should be doable, but they are not that explicit goals. Our Austin discussions might have some smaller stuff/wishes that should have some priority. - Sebastian [1] However, I would very much like to see the organizational stuff to be finalized this year. But I know it is probably quite a bit of work which can drag on.
Thoughts?
Chuck
_______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On Mon, Aug 17, 2015 at 10:53 AM, Sebastian Berg <sebastian@sipsolutions.net
wrote:
On So, 2015-08-16 at 14:04 -0600, Charles R Harris wrote:
Hi All,
While waiting for Christoph to drop the other shoe on 1.10.0b1, I thought I'd try again to start a discussion on the 1.11 release. If we want to get the next release out in a timely manner there should be some advance planning to cover at least the following three items
* Release manager: I can do that again, but would be more than happy to let someone else give it a shot.
While I don't want to do the job, I will volunteer to help fix the documentation on how to do this job where needed.
* Goals to meet: * __numpy_ufunc__ * masked array and recarray fixups for astropy * ? * Release date to shoot for, I'd suggest Feb 1, 2016
Sounds good to me. Feb. 1 probably would mean a feature freeze in the beginning of January? I guess we could try to push a bit harder to a stricter 4 months cycle, but with the common and of year stress and holiday season it sounds reasonable.
That means branching early to mid December, sounds fine to me.
Can't think of any goal to explicitly meet [1], but I guess that is mostly downstream wishes anyway ;). Personally, I would like to see Pauli's overlap detection and the proposed indexing changes by the next release. That should be doable, but they are not that explicit goals. Our Austin discussions might have some smaller stuff/wishes that should have some priority.
- Sebastian
[1] However, I would very much like to see the organizational stuff to be finalized this year. But I know it is probably quite a bit of work which can drag on.
not related to a release, but +10 for this goal Ralf
participants (3)
-
Charles R Harris -
Ralf Gommers -
Sebastian Berg