Clarify the SPEC process: scope, procedure

[stefanv]

This PR aims to clarify the SPEC process to make it very clear to new contributors what the procedures are for submitting a SPEC, as well is what is, and is not, in scope for a SPEC.

/cc @scientific-python/spec-steering-committee

[bsipocz]

Thanks, these feels like great improvements. I have some minor comments.

[ivanov]

This indeed reads better, thank you!

[jarrodmillman]

Thanks!! I am on vacation until Monday, so I just took a quick scan. I will do a more careful review then. It would be worth trying to get as many steering committee members approving this as possible. Just to make sure we are all on the same page.

You may want to review

• https://scientific-python.org/specs/steering-committee/#how-are-specs-accepted
• https://scientific-python.org/specs/steering-committee/#fn:1

to make sure those sections are in sync with these changes. Also may be worth making sure the information isn't duplicated by linking from here to this page or vice versa.

[jarrodmillman]

You may also want to make a PR template and update the quickstart script to make sure people are looking at (and following) the process information.

Probably best to work with #393 and update the quickstart script to address the limitations of the PR template.

[lucascolley]

LGTM!

[tupui]

LGTM 👍

[pllim]

Overall LGTM. Thanks!

Question: This document does not say what happens if a SPEC proposal is rejected; did I miss it elsewhere?

There is also issue with the flowchart in dark mode but I will open separate issue for that (#396) as dark mode is out of scope here.

[stefanv]

Also see inline comments from SPEC meeting notes

[bsipocz]

The build failures looks relevant this time, the links don't resolve as expected. Maybe use {ref}`ref-name` syntax instead?

[stefanv]

You may want to review

* https://scientific-python.org/specs/steering-committee/#how-are-specs-accepted

* https://scientific-python.org/specs/steering-committee/#fn:1

to make sure those sections are in sync with these changes. Also may be worth making sure the information isn't duplicated by linking from here to this page or vice versa.

There's still some duplication, but it is now clearly marked.

See 051ad80

[stefanv]

Question: This document does not say what happens if a SPEC proposal is rejected; did I miss it elsewhere?

I don't know if we have anything specific to say on this situation. Two reasons for rejection are:

1. The idea is not applicable to the wider ecosystem. That could change over time.
2. There aren't two authors willing to work on the SPEC. Something that can be addressed.

Perhaps the SPEC committee here uses their best judgment.

[stefanv]

OK, should be ready for another round of review.

[jarrodmillman]

The build failures looks relevant this time, the links don't resolve as expected. Maybe use {ref}`ref-name` syntax instead?

This looks to be resolved now.

[bsipocz]

Thanks Stefan!