Proposed 1.7.0 Release Schedule
Hi, SciPy 1.6.0 was released December 31/2020 (a bit late), a little more than 4 months ago, and I think we'd like to keep a roughly biannual release cadence. I'd like to proposed the following schedule for 1.7.0: - May 26/2021: branch 1.7.x - May 29/2021: rc1 - June 11/2021: rc2 (if needed) - June 20/2021: final release As always, it is a good idea to start tagging things that should be in the 1.7.0 release & please do help with reviewing PRs/issues that are tagged--current counts are: - PRs: 24 open with 1.7.0 milestone - issues: 8 open with 1.7.0 milestone Also great to update the release notes wiki as appropriate: https://github.com/scipy/scipy/wiki/Release-note-entries-for-SciPy-1.7.0 Thoughts/objections for the proposed schedule? The new build time dependency/transpiler, Pythran, would be a notable addition to this release, assuming that moves forward. Best wishes, Tyler
On Sun, May 2, 2021 at 7:38 PM Tyler Reddy <tyler.je.reddy@gmail.com> wrote:
Hi,
SciPy 1.6.0 was released December 31/2020 (a bit late), a little more than 4 months ago, and I think we'd like to keep a roughly biannual release cadence.
I'd like to proposed the following schedule for 1.7.0: - May 26/2021: branch 1.7.x - May 29/2021: rc1 - June 11/2021: rc2 (if needed) - June 20/2021: final release
Sounds good to me. As always, it is a good idea to start tagging things that should be in the
1.7.0 release & please do help with reviewing PRs/issues that are tagged--current counts are:
- PRs: 24 open with 1.7.0 milestone - issues: 8 open with 1.7.0 milestone
Also great to update the release notes wiki as appropriate: https://github.com/scipy/scipy/wiki/Release-note-entries-for-SciPy-1.7.0
Thoughts/objections for the proposed schedule? The new build time dependency/transpiler, Pythran, would be a notable addition to this release, assuming that moves forward.
There's a couple of nice PRs pending that depend on it with only a pure Python fallback. While I'm pretty happy with how the use of Pythran is shaping up, I think we should leave it as an optional build dependency only for this release, and try to make it non-optional afterwards. There's always the chance we'll find something unexpected in the release, and there's Windows build thing that needs sorting out - it works for all wheels, but Ilhan is having some trouble with a local build setup. Cheers, Ralf
I would also appreciate if other win users can give it a go. It might be the case that my system is the exception so we can ignore my issue. On Sun, 2 May 2021, 21:08 Ralf Gommers, <ralf.gommers@gmail.com> wrote:
On Sun, May 2, 2021 at 7:38 PM Tyler Reddy <tyler.je.reddy@gmail.com> wrote:
Hi,
SciPy 1.6.0 was released December 31/2020 (a bit late), a little more than 4 months ago, and I think we'd like to keep a roughly biannual release cadence.
I'd like to proposed the following schedule for 1.7.0: - May 26/2021: branch 1.7.x - May 29/2021: rc1 - June 11/2021: rc2 (if needed) - June 20/2021: final release
Sounds good to me.
As always, it is a good idea to start tagging things that should be in the
1.7.0 release & please do help with reviewing PRs/issues that are tagged--current counts are:
- PRs: 24 open with 1.7.0 milestone - issues: 8 open with 1.7.0 milestone
Also great to update the release notes wiki as appropriate: https://github.com/scipy/scipy/wiki/Release-note-entries-for-SciPy-1.7.0
Thoughts/objections for the proposed schedule? The new build time dependency/transpiler, Pythran, would be a notable addition to this release, assuming that moves forward.
There's a couple of nice PRs pending that depend on it with only a pure Python fallback. While I'm pretty happy with how the use of Pythran is shaping up, I think we should leave it as an optional build dependency only for this release, and try to make it non-optional afterwards. There's always the chance we'll find something unexpected in the release, and there's Windows build thing that needs sorting out - it works for all wheels, but Ilhan is having some trouble with a local build setup.
Cheers, Ralf
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
Hi, A few PRs that might benefit from some extra attention: - a stats PR with a tricky debate related to the "nan" policy: https://github.com/scipy/scipy/pull/13572#discussion_r625334934 - a set of alternate PRs that have been described as one of the bigger pain points for optimize: https://github.com/scipy/scipy/pull/13143#issuecomment-841737552 Best wishes, Tyler On Mon, 3 May 2021 at 04:43, Ilhan Polat <ilhanpolat@gmail.com> wrote:
I would also appreciate if other win users can give it a go. It might be the case that my system is the exception so we can ignore my issue.
On Sun, 2 May 2021, 21:08 Ralf Gommers, <ralf.gommers@gmail.com> wrote:
On Sun, May 2, 2021 at 7:38 PM Tyler Reddy <tyler.je.reddy@gmail.com> wrote:
Hi,
SciPy 1.6.0 was released December 31/2020 (a bit late), a little more than 4 months ago, and I think we'd like to keep a roughly biannual release cadence.
I'd like to proposed the following schedule for 1.7.0: - May 26/2021: branch 1.7.x - May 29/2021: rc1 - June 11/2021: rc2 (if needed) - June 20/2021: final release
Sounds good to me.
As always, it is a good idea to start tagging things that should be in
the 1.7.0 release & please do help with reviewing PRs/issues that are tagged--current counts are:
- PRs: 24 open with 1.7.0 milestone - issues: 8 open with 1.7.0 milestone
Also great to update the release notes wiki as appropriate: https://github.com/scipy/scipy/wiki/Release-note-entries-for-SciPy-1.7.0
Thoughts/objections for the proposed schedule? The new build time dependency/transpiler, Pythran, would be a notable addition to this release, assuming that moves forward.
There's a couple of nice PRs pending that depend on it with only a pure Python fallback. While I'm pretty happy with how the use of Pythran is shaping up, I think we should leave it as an optional build dependency only for this release, and try to make it non-optional afterwards. There's always the chance we'll find something unexpected in the release, and there's Windows build thing that needs sorting out - it works for all wheels, but Ilhan is having some trouble with a local build setup.
Cheers, Ralf
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
There's also a request for core devs to "tie break" a decision here: https://github.com/scipy/scipy/pull/11263#issuecomment-840265287 On Sat, 15 May 2021 at 18:26, Tyler Reddy <tyler.je.reddy@gmail.com> wrote:
Hi,
A few PRs that might benefit from some extra attention:
- a stats PR with a tricky debate related to the "nan" policy: https://github.com/scipy/scipy/pull/13572#discussion_r625334934 - a set of alternate PRs that have been described as one of the bigger pain points for optimize: https://github.com/scipy/scipy/pull/13143#issuecomment-841737552
Best wishes, Tyler
On Mon, 3 May 2021 at 04:43, Ilhan Polat <ilhanpolat@gmail.com> wrote:
I would also appreciate if other win users can give it a go. It might be the case that my system is the exception so we can ignore my issue.
On Sun, 2 May 2021, 21:08 Ralf Gommers, <ralf.gommers@gmail.com> wrote:
On Sun, May 2, 2021 at 7:38 PM Tyler Reddy <tyler.je.reddy@gmail.com> wrote:
Hi,
SciPy 1.6.0 was released December 31/2020 (a bit late), a little more than 4 months ago, and I think we'd like to keep a roughly biannual release cadence.
I'd like to proposed the following schedule for 1.7.0: - May 26/2021: branch 1.7.x - May 29/2021: rc1 - June 11/2021: rc2 (if needed) - June 20/2021: final release
Sounds good to me.
As always, it is a good idea to start tagging things that should be in
the 1.7.0 release & please do help with reviewing PRs/issues that are tagged--current counts are:
- PRs: 24 open with 1.7.0 milestone - issues: 8 open with 1.7.0 milestone
Also great to update the release notes wiki as appropriate: https://github.com/scipy/scipy/wiki/Release-note-entries-for-SciPy-1.7.0
Thoughts/objections for the proposed schedule? The new build time dependency/transpiler, Pythran, would be a notable addition to this release, assuming that moves forward.
There's a couple of nice PRs pending that depend on it with only a pure Python fallback. While I'm pretty happy with how the use of Pythran is shaping up, I think we should leave it as an optional build dependency only for this release, and try to make it non-optional afterwards. There's always the chance we'll find something unexpected in the release, and there's Windows build thing that needs sorting out - it works for all wheels, but Ilhan is having some trouble with a local build setup.
Cheers, Ralf
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
Thanks Tyler. Regarding Optimize. It seems that lots of PRs and issues are pointing towards an overhaul of the module. My position would be to solve what can be solved for the release with open PRs and for the next release we should lay out a plan to evolve the module. What do you think Andrew? Could we plan a specific meeting for that maybe? Cheers, Pamphile
On 16 May 2021, at 02:48, Tyler Reddy <tyler.je.reddy@gmail.com> wrote:
There's also a request for core devs to "tie break" a decision here: https://github.com/scipy/scipy/pull/11263#issuecomment-840265287
On Sat, 15 May 2021 at 18:26, Tyler Reddy <tyler.je.reddy@gmail.com> wrote: Hi,
A few PRs that might benefit from some extra attention:
- a stats PR with a tricky debate related to the "nan" policy: https://github.com/scipy/scipy/pull/13572#discussion_r625334934 - a set of alternate PRs that have been described as one of the bigger pain points for optimize: https://github.com/scipy/scipy/pull/13143#issuecomment-841737552
Best wishes, Tyler
On Mon, 3 May 2021 at 04:43, Ilhan Polat <ilhanpolat@gmail.com> wrote: I would also appreciate if other win users can give it a go. It might be the case that my system is the exception so we can ignore my issue.
On Sun, 2 May 2021, 21:08 Ralf Gommers, <ralf.gommers@gmail.com> wrote:
On Sun, May 2, 2021 at 7:38 PM Tyler Reddy <tyler.je.reddy@gmail.com> wrote: Hi,
SciPy 1.6.0 was released December 31/2020 (a bit late), a little more than 4 months ago, and I think we'd like to keep a roughly biannual release cadence.
I'd like to proposed the following schedule for 1.7.0: - May 26/2021: branch 1.7.x - May 29/2021: rc1 - June 11/2021: rc2 (if needed) - June 20/2021: final release
Sounds good to me.
As always, it is a good idea to start tagging things that should be in the 1.7.0 release & please do help with reviewing PRs/issues that are tagged--current counts are:
- PRs: 24 open with 1.7.0 milestone - issues: 8 open with 1.7.0 milestone
Also great to update the release notes wiki as appropriate: https://github.com/scipy/scipy/wiki/Release-note-entries-for-SciPy-1.7.0
Thoughts/objections for the proposed schedule? The new build time dependency/transpiler, Pythran, would be a notable addition to this release, assuming that moves forward.
There's a couple of nice PRs pending that depend on it with only a pure Python fallback. While I'm pretty happy with how the use of Pythran is shaping up, I think we should leave it as an optional build dependency only for this release, and try to make it non-optional afterwards. There's always the chance we'll find something unexpected in the release, and there's Windows build thing that needs sorting out - it works for all wheels, but Ilhan is having some trouble with a local build setup.
Cheers, Ralf
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
On Sun, 16 May 2021 at 22:14, Pamphile Roy <roy.pamphile@gmail.com> wrote:
Regarding Optimize. It seems that lots of PRs and issues are pointing towards an overhaul of the module. My position would be to solve what can be solved for the release with open PRs and for the next release we should lay out a plan to evolve the module.
It'd certainly be good for the number of active optimize maintainers to increase, to tackle PRs and issues, etc. optimize is probably one of the most heavily used parts of scipy, that's why it gets a lot of traffic on github. I think the module is probably in not too bad a good shape. A review of outstanding issues would give an idea of what types of issues are commonly experienced, it's not clear that a large scale redesign would necessarily change anything. One thing I'm mainly interested in minimisations of scalar functions. I had a desire to implement a class-based minimization system for those, but I'm fairly convinced it would have to be developed outside scipy in a spin-off project first.
What do you think Andrew? Could we plan a specific meeting for that maybe?
Cheers, Pamphile
The draft release notes are ready for review/suggestions: https://github.com/scipy/scipy/pull/14118 We have about 20 open PRs with milestone, which is a bit high this close to branching: https://github.com/scipy/scipy/pulls?q=is%3Aopen+is%3Apr+milestone%3A1.7.0 Quite a few are marked "approved." Some are likely to get bumped--I'm probably going to get a bit stricter tomorrow with bumping PRs that aren't on a clear trajectory to merging by Wednesday based on the comments or available expert reviewer bandwidth, CI issues that haven't been fixed, etc. Best wishes, Tyler On Sun, 16 May 2021 at 19:02, Andrew Nelson <andyfaff@gmail.com> wrote:
On Sun, 16 May 2021 at 22:14, Pamphile Roy <roy.pamphile@gmail.com> wrote:
Regarding Optimize. It seems that lots of PRs and issues are pointing towards an overhaul of the module. My position would be to solve what can be solved for the release with open PRs and for the next release we should lay out a plan to evolve the module.
It'd certainly be good for the number of active optimize maintainers to increase, to tackle PRs and issues, etc.
optimize is probably one of the most heavily used parts of scipy, that's why it gets a lot of traffic on github. I think the module is probably in not too bad a good shape. A review of outstanding issues would give an idea of what types of issues are commonly experienced, it's not clear that a large scale redesign would necessarily change anything.
One thing I'm mainly interested in minimisations of scalar functions. I had a desire to implement a class-based minimization system for those, but I'm fairly convinced it would have to be developed outside scipy in a spin-off project first.
What do you think Andrew? Could we plan a specific meeting for that maybe?
Cheers, Pamphile
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
I'll probably try to bump rc1 forward into early-mid next week because of the long weekend here in the US. Maybe a few more backport/bug fixes will trickle in as well. Tyler On Sun, 23 May 2021 at 18:17, Tyler Reddy <tyler.je.reddy@gmail.com> wrote:
The draft release notes are ready for review/suggestions: https://github.com/scipy/scipy/pull/14118
We have about 20 open PRs with milestone, which is a bit high this close to branching: https://github.com/scipy/scipy/pulls?q=is%3Aopen+is%3Apr+milestone%3A1.7.0
Quite a few are marked "approved." Some are likely to get bumped--I'm probably going to get a bit stricter tomorrow with bumping PRs that aren't on a clear trajectory to merging by Wednesday based on the comments or available expert reviewer bandwidth, CI issues that haven't been fixed, etc.
Best wishes, Tyler
On Sun, 16 May 2021 at 19:02, Andrew Nelson <andyfaff@gmail.com> wrote:
On Sun, 16 May 2021 at 22:14, Pamphile Roy <roy.pamphile@gmail.com> wrote:
Regarding Optimize. It seems that lots of PRs and issues are pointing towards an overhaul of the module. My position would be to solve what can be solved for the release with open PRs and for the next release we should lay out a plan to evolve the module.
It'd certainly be good for the number of active optimize maintainers to increase, to tackle PRs and issues, etc.
optimize is probably one of the most heavily used parts of scipy, that's why it gets a lot of traffic on github. I think the module is probably in not too bad a good shape. A review of outstanding issues would give an idea of what types of issues are commonly experienced, it's not clear that a large scale redesign would necessarily change anything.
One thing I'm mainly interested in minimisations of scalar functions. I had a desire to implement a class-based minimization system for those, but I'm fairly convinced it would have to be developed outside scipy in a spin-off project first.
What do you think Andrew? Could we plan a specific meeting for that maybe?
Cheers, Pamphile
_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev
participants (5)
-
Andrew Nelson -
Ilhan Polat -
Pamphile Roy -
Ralf Gommers -
Tyler Reddy