Should we adopt the new NumFOCUS Code of Conduct?

45 views
Skip to first unread message

Hans Moritz Guenther

unread,
Oct 15, 2024, 10:20:38 AM10/15/24
to astro...@googlegroups.com
Hi,

NumFOCUS is adopting a new code of conduct  (CoC) in Jan 2025 - see here: https://numfocus.org/2025-code-of-conduct
Sponsored and affiliated projects such as Astropy are free to adopt that Code of Conduct or continue to use their own (new projects accepted after Jan 2025 will be assumed to be under the new CoC, but can opt-out if they choose to and nothing changes for old projects like us, unless we actively opt-in to use the NumFOCUS CoC).

I suggest that Astropy opts-in to use the new NumFOCUS CoC. 

What’s different?
Essentially, all CoCs just say “be nice to each other” in a more formal way. The new NumFOCUS CoC ( https://numfocus.org/2025-code-of-conduct) is a lot more detailed than our current Astropy CoC (https://www.astropy.org/code_of_conduct.html) and a lot more specific. The goal is to help people to understand the intention better. Being specific matters because cultural norms about what exactly it means to e.g. “treat people with respect” differ throughout the world, so in my opinion it is better to add more details on what we do and do not want to see in this community. (I’ve been part of the group that wrote the new NumFOCUS CuC. That doesn’t mean that each word is exactly what I would have liked it to be, but I think that we all made reasonable compromises to end up with the current document.)

Also, the current Astropy CoC says a lot about “We pledge to …” but it doesn’t say what happens when someone does not follow those pledges, what the process is to report and incident, and what will happen after. Again, being specific and spelling this out brings everyone on the same page about who needs to do what and who has the power to impose sanctions, remove people etc.

What is the role of the NumFOCUS CoC working group?
NumFOCUS will form a CoC working group to handle incidents at the NumFOCUS level (e.g. on the NumFOCUS slack or at NumFOCUS meetings). This group will offer support to projects, IF THE PROJECTS WANT THAT. For example, it could help with cases where the astropy-internal ombudsperson themselves are conflicted or simple would like some confidential outside advice or resource. NumFOCUS CoC working group members will receive some form of formal training, the details of which are still to be worked out. If Astropy adopts the NumFOCUS CoC, we can choose to work with this group as needed or on a case-by-case basis, but we don’t have to. (If we stick with our own CoC, we cannot use this group as a resource, because they won’t have the capacity to handle > 100 difference CoCs if every project has their own).
There are currently > 40 nominations for this group (incl me), which will be formed later this year.


Of course, we all hope that we’ll never need to use a CoC because everyone just acts appropriately in the first place, but if something happens, we need this in place to react. In the last 14 years of astropy or so, there have been very few incidences, but it is better to be prepared.

Moritz

Pey Lian Lim

unread,
Oct 16, 2024, 8:53:08 AM10/16/24
to astropy-dev
Hi Moritz,

Thanks! Personally, I am in support of this, as I see the benefit of sharing a process with a larger cross-project org with more resources on such things.

I have a question: How does this affect the current Ombudsperson role in Astropy Project?

I will also add these here just so we remember to revisit as this topic progresses:

https://github.com/astropy/astropy-project/issues/21 (Add a process section to code of conduct)
https://github.com/astropy/astropy-project/issues/10 (Make a new "Unacceptable Behaviors" Document)
https://github.com/astropy/astropy.github.com/blob/main/code_of_conduct.html
https://github.com/astropy/.github/blob/main/CODE_OF_CONDUCT.md
https://github.com/astropy/astropy-APEs/blob/main/APE8.rst (Astropy Community Code of Conduct)
https://github.com/astropy/astropy-project/blob/main/roadmap/roadmap.md#government-management-and-personnel -- "Adopt the NumFOCUS Code of Conduct (CoC) and build a community ombuds/CoC team to curate the guidelines and enforce them."
* And then there is a matter of how to propagate this to all the other repositories in astropy GitHub org and how to recommend/enforce for Affiliated packages.

Thanks again for moving this forward,
Pey Lian

John Swinbank

unread,
Oct 16, 2024, 11:46:04 AM10/16/24
to astro...@googlegroups.com
Hi Moritz,

The contents of the NumFOCUS CoC are clearly appropriate and well-thought-through, and I agree that it would be smart for Astropy to benefit from this effort.

However, I wonder a little bit about what I'm going to loosely refer to as “branding” issues. Assuming the CoC is adopted word for word, it means that participants at events organized by Astropy will be agreeing to a CoC referring to “NumFOCUS events”. I'm betting that most members of our community don't know what NumFOCUS is, and will be confused by this wording (and the other similar wording in the CoC).

Further, participants are invited to report violations to the NumFOCUS CoC WG, via a NumFOCUS branded form. This seems like it's ceding a significant part of Astropy's project stewardship to NumFOCUS, which I'm a little sceptical about given that I have a number of concerns about the competence they've historically shown dealing with our finances. I do note that your message gives us a (block caps!) disclaimer of “IF THE PROJECTS WANT THAT”, but the text on the page seems to imply that this form of reporting is intrinsic to opting in to using the CoC.

In short, I like this idea in principle, but I wonder if we'd benefit from doing an s/NumFOCUS/Astropy on the CoC text.

Cheers,

John
> --
> You received this message because you are subscribed to the Google Groups "astropy-dev" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to astropy-dev...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/astropy-dev/C8087DFB-DBB1-4F54-AD99-0AEA9195E3B3%40gmx.de.

Hans Moritz Guenther

unread,
Oct 17, 2024, 8:36:08 AM10/17/24
to astro...@googlegroups.com
In short, I like this idea in principle, but I wonder if we'd benefit from doing an s/NumFOCUS/Astropy on the CoC text.

Yes, we should search&replace “NumFOCUS" with “Astropy”. I’m not sure if a simple search&replace will do, there might be a few cases where we want it to say NumFOCUS (e.g. in acknowledgements), but in principle, we should do that.

As for the role of our astropy ombudsperson: That’s up to us: We already voted on changes to APE0 to allow the Ombudsperson to actually be a “group” and also to allow to have that group shared with projects. So, our APE0 would allow us to use that NuMFOCUS group as “our” group, as long as at least one astropy representative (our “ombudsperson”) is on that group.

However, as NumFOCUS envisions it, projects who adopt this new CoC are free to have their own ombudsperson/COC working group / whatever it's called in that project and only push up to NumFOCUS CoC working group on request or use them as a resource (e.g. use their training materials or to provide a reporting email address that’s independent of the project leadership).

So, in that sense, the question if we want to replace or augment our own ombudsperson with that NumFOCUS group is a separate issue.
I would suggest that, initially, we just adopt the NumFOCUS CoC, but keep our own Ombudsperson. We can re-evaluate that at some point in the future.

Moritz

Pey Lian Lim

unread,
Jan 21, 2025, 6:53:29 PMJan 21
to astropy-dev

Dear Projects,

This is a detailed email with important information, so please read it carefully to the end.

Starting February 1, 2025, the new NumFOCUS Code of Conduct (CoC) will come into effect. Projects that entered into collaboration agreements with NumFOCUS before this date can opt-in to adopting the new NF Code of Conduct, though it is not mandatory.

A key update in the new CoC is the establishment of the independent and trained NumFOCUS Code of Conduct Working Group (NF CoC WG). This group will handle reports of CoC violations, among other responsibilities. You can learn more about the NF CoC WG members, their training, as well as the charter on which they operate, here.

To help your project make an informed decision, we invite you to attend one of the following information sessions. These sessions will cover details about the new CoC, the NF CoC WG, and related procedures.

Session Details:

  • February 13: 9:00–10:00 AM PST | Zoom link (Passcode: 174844)

  • February 19: 21:00–22:00 PM PST | Zoom link (Passcode: 720434)

  • March 4: 8:00–9:00 AM PST | Zoom link (Passcode: 361979) 

Time Zone Converter

Additional Notes:

  • No registration is needed to attend.

  • Sessions will be recorded and shared via the NF Projects mailing group, the Projects Slack channel,and NF GitHub.

  • We encourage live attendance to participate in the Q&A portion.

Opt-In Process for Projects Collaborating with NumFOCUS Before February 1, 2025:
Projects collaborating with NumFOCUS before February 1, 2025, are encouraged to consider adopting the new Code of Conduct. We will contact all projects in March 2025 with a form to formally indicate their decision.

Thank you for your attention to this important update.

Best,
--

Kamila Stępniowska

Projects DEI Coordinator

numfocus.org


hamogu

unread,
Jan 22, 2025, 9:45:57 AMJan 22
to astropy-dev
As s short note: I am going to be a member of the new NumFOCUS Code of Conduct working group. We have not yet met, many details on how this works in practice will have to be worked out, but please feel free to reach out to me with any questions.

Moritz

Pey Lian Lim

unread,
Jun 20, 2025, 1:17:02 PMJun 20
to astropy-dev
Astropy has opted in with "recommendations". As such, here is the issue for discussions:


and here is the accompanying PR to update the actual code of conduct for the whole Project:


I plan to propose a breakout session to discuss this at Astropy Coordination Meeting 2025 next week, if there is enough interest.

Thank you.

Reply all
Reply to author
Forward
0 new messages