Code of Conduct for Stan

I also think that, while everyone should be held to a high standard, people who are being paid to work on Stan by Stan should be held to very stringent standards. And that any deviations from the absolute best should have a decent chance in ending their paid position.

With volunteers we get what we get. But we don’t pay everyone and we should not pay people who are not good member of and ambassadors for the community


No. We just have to adopt one. We default to NumFOCUS because we don’t have our own.

1 Like

Maybe someone from the SGB can weigh in here. My sense is that this is where the experience of NumFOCUS provides a pathway that we should take advantage of.

My thought is that the various “forms” of interaction with the Stan community would take that as a basis and provide examples of what it would look like to demonstrate the standards of behavior advocated for in the CoC.

For example:

Be empathetic, welcoming, friendly, and patient.** We remember that every NumFOCUS project and program is crafted by human beings who deserve to be treated with kindness and empathy. We work together to resolve conflict and assume good intentions. We may all experience some frustration from time to time, but we do not allow frustration to turn into a personal attack. A community where people feel uncomfortable or threatened is not a productive one.

What does that look like for a maths PR?

I also thought it might be possible to highlight that there might be people below the NumFocus board who you might want to contact (i.e., you might want to contact your meetup organizer because you know them) for support with the process. That might be something NumFocus can guide us on though - they might have decided against that for good reasons and imho I think we should accept their guidance here.

Hi all. Thanks for all the feedback! Just to give some quick context here: this came up because we’re working on fleshing out the descriptions of the new Stan community positions that we’re setting up (Discourse manager, webpage manager, Stancon manager, etc), and we wanted codes of conduct for all these positions. So this was a good motivation for us to think about code of conduct more generally.

I was tagged but have no preferences. Most familiar with the NumFOCUS CoC and it seems like a good default - I like @lauren’s idea of adopting it and adding some Stan-specific interpretations and reporting strategy.

1 Like

There needs to be something in the CoC to manage weird conflicts that are specific to this project. For instance, problems with a Columbia Or Aalto employee shouldn’t be decided by other Columbia or Aalto employees. (I think they’re the only two places with more than one person right now. But in general a conflict of interest should be declared if you share an employer)

it sounds like the current (provisional) SGB is trying to add more hierarchy to the project - adding to inequality, adding to more ways for people to feel excluded, especially since these positions will now be paid positions “positions with tips” - so all managers have an incentive not to do anything that’s not SGB approved in order to collect.

the SGB asked for a two month extension to sort out elections. how’s that coming along?

@mitzimorris I think that’s a different question for the SGB - I’m actually unsure whether that’s been decided upon or not.

Regardless of whether those positions go ahead or not there are multiple efforts to produce a code of conduct for multiple reasons. To me it makes sense to decide on a proposal together and then submit it to the SGB for discussion.

My proposal is to adopt the numFOCUS proposal with two adaptions reflecting @anon75146577’s comments. The first is to distinguish between those paid to work on Stan and those who volunteer on Stan and how that works in terms of violations and consequences. The second is to adjust the reporting listed in the numFOCUS code of conduct to also include people within the Stan community who might also be able to receive complaints (for example the StanCon committee for StanCon related violations).

The last part of my proposal is that it is interpreted with specific examples of how to resolve conflict, potential consequences and reporting strategies for each expression of the Stan community.

@steven01 has put a lot of work in an alternative proposal. Steve do you think that there are ideas in your proposal that my proposal doesn’t capture? Can we merge them together in some way?

Edited for pronouns/clarity


that sounds good.

my point is that we need to work on preventing situations that lead to complaints or violations - at the point where there’s something to report, it’s too late.


I completely agree there!

My hope is that clear, well publicized examples of what the standards of behavior do and don’t look like will help!

Is your proposal the numfocus one or do you have a version with the two revisions you mentioned above? I read the numfocus one last night and it seemed good to me! I should re-read it thinking about Dan’s comments. If you have a version with the revisions would be nice to look over.

ftr I have no tying binds to the one I put up, I don’t have a lot of experience with these sorts of things so was mostly winging it. I didn’t know anyone else was working on a CoC. My intent was mostly to trying to have something that said

  • Don’t be a jerk
  • Here’s what’s bad
  • Here’s how to report things that are bad

I think the numfocus one covers that and hits more of Dan’s comments

Glad the ball is rolling!


Yes. This is an important thing. The SGB as it currently stands is very white, very male, and very Columbia (or ex-Columbia) heavy. That may not be the right balance for a reporting group. Perhaps it would be a good idea to let people self-nominate into a small EDI and Conduct Committee that has an ex officio seat on the SGB but is run independently of it.

Edit because I checked the numbers: The provisional SGB is:

  • 4 Current Columbia Employees
  • 2 Previous Columbia Employees
  • 2 Women (who haven’t, to my knowledge, been employees of Columbia)

I don’t think that’s a problem necessarily for an SGB (half-and-half Columbia) but I think it’s less great for dealing with CoC issues.

I spoke too soon! I am in favor of adopting the NumFOCUS CoC completely and essentially as-written with the amendments @lauren originally proposed, I believe just switching over the reporting form to our own reporting form and the points-of-contact to our own points-of-contact. I think it should be evident by now that our community’s value-add to the universe is not coming up with new governance schemes :P

The NumFOCUS CoC already has a section on conflicts of interest that I quite like. Regarding the sharing-an-employer conflict of interest, I’d say the same logic applies to anyone working together on Stan in general. For that reason, maybe it’s a good idea to have half of the folks on the committee be from totally outside the project. We might talk to some other projects at NumFOCUS to see if we can trade some of our folks time on their CoC committees for some of their folks time on our committee. Then we could get a fair number of reasonable outsiders familiar with the general scientific computing space involved as disinterested third parties.

We can also just ask NumFOCUS how they deal with conflicts of interest given that I’m sure their CoC Enforcement Team only has people employed by NumFOCUS on it.


This CoC applies to people who do not work on Stan as well as people who do. Sharing an employer is an issue for a number of reasons. For example, I have certain responsibilities if one of the people involved is related to UofT that may be in conflict with the CoC. For example, I may have a duty to break the confidentiality and inform someone of the accusations. This is pretty common in north american universities. Typically if I have enough information to voluntarily recuse myself due to a conflict of interest, I have enough information that I would be unable to keep it confidential. That’s why automatic exclusion rules are important.


Ah, that’s interesting. That makes sense. What other automatic rules would we have? And how do we fill empty spots due to these rules?

Rust CoC has some nice things on moderation.

@lauren do you want me to make a PR on the project proposals github with the NumFocus CoC? I can give you write permissions and we can move the discussion and changes over there for visibility

1 Like

Yeah that would be great, thank you.

Discourse people who are interested in the code of conduct but are not github users please send me a message and we’ll work out a platform that suits us all. :)

Thanks for the time you and @stevebronder are putting into this issue and converging on one proposal.


Oh wait, I don’t think automated exclusion rules would do anything w.r.t. mandatory employer reporting because either:

  1. Both people involved in the CoC complaint are employees at X company and both subject to the same reporting obligations, or
  2. At least one of the people involved is not an employee at X and therefore the situation is not subject to reporting requirements.

I just went and did Columbia’s mandatory sexual harassment training recently and it was pretty clear about what you have to report. I imagine other institutions are similar.

I’ve just come across this on StackOverflow, and it seems to have some nice bits that could be helpful in case we are writing our own version of a CoC: