Yeah, there are many people who have done a lot of work for Stan and the community without a lot of recognition (at least publicly). This is something that I think everyone agrees can and should be improved. There have been many discussions about how to better recognize the many different ways people contribute and I think we will be headed in the right direction there. Some of it may require the new governing body, but either way I think this is something that that is high on the priorities list to address and is an issue that is important to me personally (i.e., that everyone contributing feels properly recognized), and even if I’m not part of the new governing body I will personally make sure this issue is not left hanging in any way that I can.
Thank you @lauren and everyone else working on this!
Even if it ends up as the current electorate voting in the new SGB, I don’t think it will be the current electorate voting on the CoC, so I hope you don’t withdraw your proposal. I think the new SGB (more details on that coming soon), like you say, can decide on a proposed design in consultation with the people working on this and do an in-depth community consultation.
Thanks for tagging me, I’ve been keeping my work account separate and don’t check the personal one much during the day so I missed most of this discussion.
With regards to the CoC discussion: I agree with @lauren and @Daniel_Simpson that the effort of developing a CoC and reporting structure that works for the Stan project is a good use of time. If that ends up taking liberal amounts of text from the NumFocus CoC that’s fine, but in the end the modifications we make are going to be critical.
Going through this process will bring some issues to the surface that we could temporarily avoid. If they are not resolved now, they will definitely come up the first time a serious complaint arises. Based on how other projects have handled that scenario I am really hoping Stan can do something less messy. It’s unfair to the people involved in any reported issue to complicate their situation by simultaneously dealing with old organizational problems.
Somebody commented that sorting out a full proposal is a lot of effort (thus the suggestions for more borrowing). I think this implicitly assumes that the current pool of contributors is limited to the core of people who have obviously been working on Stan (PR’s, meetings, etc…) and find it overwhelming. I don’t think that’s a real limit because we are looking for people who understand a) the broader Stan community, b) the issues related to a CoC/reporting, and c) can write effectively. It’s not the same pool as the people who can write c++ template meta-programming code (although there is definitely overlap). As Dan Simpson pointed out, Lauren currently can’t vote but she’s already offered to draft text so this is one of the situations where there’s very competent labor offered with the project (presumably people would have to read the NumFocus CoC too if they were going to vote on it).
@seantalts I didn’t mean to silence your proposal by refusing to let the developers vote on it, I just felt that it was unfair to ask me to present a very time consuming proposal to allow the developers to vote when we have an SGB - not to mention confusing and time consuming to the community to consult on both. I am still happy to present an abstract to the SGB alongside any others who like to present alternative approaches. As you highlighted @mitzimorris and @stevebronder have both supported the “accept numfocus” idea and might like to do so.
I’ll wait for more details on the plan for the new SGB.
Edited to add:
This is not an issue I have with the project and I feel recognized for the small amount I do, even though it is not a CV line. I would like to see more skills in the Stan community recognized other than the devs, but I do not think I am a good example of this.
Thanks, I know :) Just realized I didn’t feel strongly enough and it seemed like you might be interpreting the NumFOCUS proposal as some kind of rebuke, so I want to step out.
I felt we’d reached a point where we were struggling to move forward, but I didn’t feel it was a rebuke. I’m glad we have a pathway forward and if you change your mind or if others agreed with you I would be happy to have both options presented to the SGB. Otherwise I hope you’ll contribute your thoughts on the full proposal! :)
+1 to that. I brought this up the first time around a few years ago and was shouted down because others asserted that any CoC was better than nothing and that it’d take too long to work out enforcement. (I come from a family of defense attorneys, so this was a disturbing position for me.)
As everyone can see from the exchanges above, this is a hot-button issue and tempers are already flaring. But please, let’s stop the ad hominem attacks or we’re going to be violating whatever CoC we come up with as we come up with it!
I haven’t read NumFOCUS’s code of conduct policy. Is there something specific people don’t like about it or is it just the not-invented-here issue that @alexpghayes says will undermine any code-of-conduct we adopt from another source?
A bit more independent than us doing it internally, but we’re a source of income for them, so I wouldn’t say “completely independent.”
If we had a way to add developers at the moment, we could solve that by adding you. But as is, with the new SGB, we’re in limbo on new devs. Not all of our devs write Stan code. Among our devs, I don’t think @andrewgelman or @Daniel_Simpson have written any code for Stan, and @avehtari only added code recently.