That “relevant member roster”, in this case, is a roster of people who we, as a community, consider the very best of us and the most emblematic of our values. I would rather that honor be hard to come by and more easily lost than the alternative.
Moreover, I expect our values to change, and I expect the board at the time to both embody those values and to be permanently in transition between the old and new values; I don’t need unanimity for them to represent me.
That’s fine, but the text of the actual proposed change (which I assume you haven’t read) doesn’t have anything specifically to do with Fellows (which I assume you’re talking about). Those were just named as a “motivating example” in the first post.
This applies to all classes of PSF member (Sponsor, Managing, Contributing, Fellow. and Basic). No special rules apply to any of them here (although Basic members didn’t have voting rights to begin with).
That assumption is unwarranted, inaccurate, and rude.
Then you have my sincere apology for that.
Jeepers! I don’t see anything impolite or unprofessional about David’s post. Some hypothetical person might be overzealous in the future and in general some people are sanctimonious. Whom is that rude to?
And I agree with David and Tim that it’s ironic that David’s post was silently canceled by the powers that be since we’re talking about people being silently canceled by the powers that be. It worked out this time, but that sort of control over the conversation (broadly meant) without outside scrutiny is precisely what I am against.
I find your interpretation of David’s comment utterly baffling. I think there’s a lot in this thread that underscores why we need higher rather than lower barriers to excluding people, or excluding their communication, for that matter.
I wish they had a variety of emoticons for reacting to a post. It’s clear as mud to me how flagging here works even after reading the docs:
https://meta.discourse.org/t/understanding-post-flags-in-discourse/275
From that, it appears unlikely to me that “the powers that be” squashed the post to begin with, only that David Lord had the moderator power needed to _un_squash it. Seemingly this kicked in instead:
Flags from well-trusted users might automatically hide the post based on their trust level and flagging history.
“Well-trusted” isn’t defined. But, regardless, in that case no human moderator was involved in squashing the post - it was a “hidden algorithm” decision driven by an ill-specified trigger.
Usenet remains a great improvement over its successors. Have a problem with some poster or topic? Add it to your killfile yourself. Instant end of problem. But then nobody else can impose their own view of what “should be” a problem to you. Utter chaos would surely result! 
.
This is kind of fascinating to me, because the situation I’m worried about – setting the requirements for consequences so high that offenders effectively know they will never face consequences – is something you’ve apparently seen happen many times.
I would think the realist position here would be to confront the common actually-occuring problem rather than focus all effort on a hypothetical problem we’re still missing an example of (since the NixOS example, as mentioned, turned out to be the community revolting againt the leadership, not the leadership trying to purge the community).
Hi Petr,
Thanks for bearing with us while we worked on a response to the rest of your question.
Change 1: What does “volunteering” mean?
If someone is paid to do open-source work, does it mean they’re not eligible?
I’m personally OK with losing voting privileges myself, but, for context: I’m lucky to be employed by the PSF. I’m paid by the month, without set working hours. As I read it, everything I do for the benefit of Python is part of my job. I can’t really “volunteer”.
That’s fine by me, for me. But if someone successfully sets up a Patreon or GitHub sponsorship, or chooses a lower-paying job to work on open source, it might work the same way. I don’t want to discourage that.
Contributing Members will continue to be subject to self-certification. The bar is that you work 5 hours per month on a project that furthers the foundation’s mission (often that works out to less in some months and more in others). If you work more than 5 hours per month, beyond what you feel you are paid for, then that is likely to be enough. We have not required a bar higher than that in the past. If you’re looking for a guideline in your case, you may wish to consider what constitutes a “full-time” role in your region (definitions in the US vary between 30 and 32 hours per week depending on which piece of legislation you’re looking at, for instance), but again, it’s down to self-certification. If you’ve done work beyond what you’re paid for, in furtherance of the PSF’s mission, you can use that in your self-certification.
Re your question about what constitutes an open source licence? Seconding what Deb wrote, the PSF’s mission has used the Open Source Definition by OSI, and that has applied to code contributions under the previous formulation of the bylaws. We’ll work on policy for non-code works that follow that same spirit once the bylaws give us scope to do so.
Note that section 4.7 of the bylaws allows the Board to add alternate eligibility requirements to the Contributing Member class, which could include clarifications to the existing language, should there be overwhelming concern that the language is inadequate. As mentioned, we received legal advice on the text of this proposed amendment.
If you disagree w/ some mod decision, please either bring it up w/ the admins or report it to the CoC WG.
(FWIW this was actually not a mod decision; it was an automatic action taken by Discourse in response to a user report, and was contradicted by explicit mod action, reinstating the post’s visibility. But in general, yes.)
And, as I said, I see it every frickin’ day in US politics. The PSF bylaws probably can’t do much about that, though.
I was on the PSF Board for over a dozen years. The idea that Board unanimity is hard to achieve is just rhetorical fantasy to me. Looking at the record, abstentions have become far more common, as the Board is increasingly asked to vote on things various Directors know nothing about. But dissenting votes remain conspicuous by absence. While a Director may well feel unqualified to vote on giving a $500 grant to fund a pizza party for some claimed-to-be Python-related user group in a country they couldn’t find on a map, they vote on “weighty” matters. Which kicking a member out should be.
Sexual harassment (the example of mine you quoted)? The PSF is already acutely sensitive to that. If you know of cases where a PSF member is getting away with that, report it to the CoC WG. I have every confidence they’ll pursue it as a priority. Heck, Steve Holden (who was as highly placed in the PSF at the time as it’s possible to get) once got slapped down for making a mildly off-color joke. His position afforded him no immunity, nor should it have, nor did he think it should either. He took his lumps gracefully, and didn’t do it again.
I saw Thomas Wouters here say “there are, in fact, Fellows who repeatedly flaunt the CoC”, but he left it at that. If it’s so, the question isn’t really bow many Board votes are needed to kick them out, but why the CoC WG and Steering Council have apparently decided to do nothing about it. Or why nobody reported it. Or … it’s hard to address such vague intimations.
I don’t feel obligated to address strawman arguments. I explicitly said, more than once, that a hostile takeoover doesn’t need to compromise the Board - the most effective way is to gain control of the groups empowered to cut members out of community life. Chris (and also, on psf-vote, Tres) were exceedingly clear about that this happened in the NixOS mess. Because they effectively have no functional Board at all, the cabal of “moderators” doing the purging is accountable to nobody.
You’re free to poo-poo it, but I won’t. I worked with Chris and Tres (more in the Zope than in the Python worlds), and know them for a fact to be honorable and fair-minded. They are not themselves jockeying for power in the NixOS community. That they both concluded (as interested fans of the project’s code) that there’s gross abuse of power in play here is strong evidence to me.
Thanks Brett, but let’s just let that small question drop. I wrote a few words. Some person or some bot flagged it as inappropriate; I think that was misguided, as do other commentators. In the scheme of thing, it’s trivial and unimportant whether my one paragraph in a long thread is available.
Python is an absolutely wonderful and inclusive community. Catch my forthcoming keynotes at PyCon Nigeria and PyCon Africa that are about exactly that (with a whole lot of complicated social theory piled on top… and a little bit of Python).
Like every group of people, it’s often easy for us to get too emotionally invested in outcomes and in personalities. I’ve done that. I’ve done that as a PSF Director. It’s much better for a few hot heads during some stressful month not to find it too easy to take bad actions.
The contentious proposal here is what gives the Board the power to do something about Fellows (or any member) violating the CoC. I’m honestly confused to the point where I had to read this multiple times, as this quote seems to be asking why the Conduct WG or the Steering Council haven’t done anything if these cases exist.
To which I will say, because as a former Conduct WG member I can’t comment on specific cases: look at the transparency reports. Look back through the python core mailing list archives. To say the WG and the Steering Council have done nothing about it reads to me as someone who’s unaware of all those groups do.
The WG does make recommendations, but can’t take action, because the Board and the SC are who take action. And the SC has taken actions against Fellows who behaved in bad faith. This proposal gives the Board similar powers to the Steering Council. I don’t understand why that’s contentious.
Actually, there’s been almost no contention here over whether the Board should be empowered to revoke PSF membership. The arguments have been more over how a high a bar the Board should have to jump.
I’m quite aware of that recommendations of the CoC WG have been accepted many times by, e.g., the Steering Council. That’s not a bone of contention either.
That’s why I’m kind of baffled by the quote from Thomas. If there’s a significant body of CoC WG recommendations that have been rejected, that’s the first I heard of it. Are there? (I know: it’s secret ).
FWIW, I was unable to find any relevant transparency reports online after making best search efforts. For example, Google search turned up nothing even remotely relevant for the query “psf coc transparency report” - although it did give links to CoC reports from other organizations.
And I know about PyCon transparency reports (which are easy to find), but those are about CoC actions taken by PyCon staff at PyCons. To my eyes, those aren’t plausibly related to the quoted claim.
I don’t know of any CoC WG recommendations that have been rejected. I don’t know if the CoC WG has ever recommended a member be ejected from the membership. The fact that it’s not a thing that’s currently possible without a very messy, public spectacle that harms victims of abuse very certainly has something to do with that. The Board is unable to deal adequately with any CoC WG recommendations that someone be ejected from membership under the current rules. I’m certain this also influences the interactions the CoC WG has when dealing with incident reports, in particular with the victims of misconduct. It doesn’t help the perception that the CoC is without teeth, and thus without merit. (The PyCon US CoC has teeth. The PSF one, very much less so. This proposal doesn’t add significant teeth. It adds a single, fairly minimal one.)
The PSF Board is given the responsibility and duty to mange the PSF. This is a very concrete thing. They have a lot of power over the PSF, and the membership. They have, among other things, a duty of care, making sure the Foundation is handled well, and a duty of obedience, making sure the Foundation and the Board follows all the laws, rules and regulations. We already entrust them with all that, and they are already accountable for it. (In fact, to some extent they are personally liable for it.)
The PSF Board needs the tools to do that work. They have the authority to do quite a lot with the Foundation, except evict members. This has come up as a real, concrete problem. It’s come up as an actual legal risk. It’s something that needs addressing sooner rather than later. I would say it’s already inhibiting some of the handling of CoC violations, because of the lack of a way to enforce rulings on offenders. It’s very likely worse situations are to come, just because of the continued growth of our community. Even if – and I don’t agree that that’s the case, but for the sake of argument – even if we were to say the ability to evict a PSF member hasn’t been necessary so far, this is not something that can be held back and only be put in place once there’s a “real need” for it.
This is not, as was mentioned before, an “easy way” to get rid of a PSF member. This isn’t easy, and it shouldn’t be easy. This is a last resort situation, as many people have attested to. A unanimous vote (of the whole Board, remember, not just quorum at a meeting) about a CoC issue is not an easy thing. The Board takes these things seriously – is required to take these things seriously. The Bylaws as drafted don’t allow for the Board to decide to kick out a member willy nilly, the member has to have “failed a condition of membership”, which means standards that apply to all members (or all members of that membership class). And there is a process for the member to be heard by the Board, to object or clarify or try and understand the situation. The membership as a whole also has a remedy here, which is a special meeting, possibly with a vote of no confidence.
It’s already been explained – several ways! – why the vote shouldn’t necessarily be unanimous. Requiring the majority of the Board to vote for it is better than a unanimous quorum, right? Nobody on the Board can get sidelined just to get this vote through. Getting a unanimous Board vote is an unrealistic standard for a member eviction both because of logistics, as well as giving a single Board member the ability to block a measure the majority considers absolutely necessary. Requiring a supermajority sounds like a nice idea (and as I mentioned before I’d probably vote for it as a separate bylaws change) but given that the Board currently has 12 voting members, it only makes one vote difference. I do not believe it is a good idea to delay this measure by a year because of such a minor difference, and I encourage everyone to vote for this proposal the way it stands.
Hello! Some of you may not know me, but I am an active member of the PSF community, and have been an events organizer for about a decade now. I am co-founder of PuPPy, the Seattle Python User Group, and PyCascades, a regional Python Conference in the Pacific Northwest. I have a lot of experience dealing with CoC complaints, both as a member of CoC committees, and as an organizer who has to handle the fallout of decisions made by CoC committees.
I agree 100% with what Thomas is saying. This is a real concrete problem and allowing the board to remove Fellowship status is an important tool that is critical to our community’s safety.
I’ve seen multiple people say something along the lines of “The CoC WG and the Steering Council can already remove people from their positions, so allowing them to remove PSF Fellow status seems useless”. Friends, it is not useless.
Here’s a hypothetical situation:
There is a PSF Fellow deeply embedded and active the community. Behind the scenes they are sexually harassing and threatening other Python community members. This behavior goes on for a while. A series of CoC complaints are filed and they’re handled one by one over a period of months or years. Eventually, after the Nth CoC report, they are deemed too much of a risk to the community, and they are removed from their leadership positions within the Python community.
This could easily be the end of the story. However, this person, rightfully shunned from our community, wants to continue participating somehow. What are their options? Well, they can easily go to Python Adjacent communities. They can quickly establish themselves as an experienced professional, and can even point to their Fellowship status, as we all know it’s no small feat to be a PSF Fellow.
Python adjacent communities can, and will, make assumptions based on PSF Fellowship status. They must be an upstanding community member if they’re a PSF fellow, right? Come on in! We could use your experience, here, help us run an event.
Alternatively, maybe they don’t try to help or volunteer at other communities, but just go to their events and conferences. They can put on their conference badge “PSF Fellow”. An immediate indicator of community participation and excellence. A status symbol they can continue to use to perpetuate further abuse.
By allowing the PSF board to remove Fellowship status we are giving them the tools they need to help reduce harm caused by people who have Fellowship status. We’re removing a tool that abusers can use to perpetuate more abuse in and outside of our community. This is a critical tool for the safety of our community, and those communities adjacent to ours.
While the situation I outlined above is hypothetical, it’s also grounded in real events. A situation like this has actually already happened. This is already a known problem in our community, and we need to take action now.
I think it probably is very easy to get unanimity on things like giving a grant to a Python workshop somewhere, or most other routine business of the Board. But we’re talking about a very non-routine type of business, and we have enormous amounts of evidence from other entities (for-profit, non-profit, and otherwise) that unanimity is extremely difficult to achieve with this type of business.
Becoming a PSF Fellow requires that one be well-established and respected and connected within the Python community. What are the odds that, if a Fellow were to egregiously misbehave, we’d be able to find someone who still would want to stand by them no matter what, would insist that we have to weigh the loss of their long-time contributions, would say “well they’ve always been perfectly nice and professional to me”, would say that they’ve already paid enough of a price via lesser actions being taken, “why are we ending this person’s career over this”, etc. etc.? Because in a unanimous-vote-required world, all it takes is that one person.
This is why I pointed out the incongruity with your seeming complaint about having to sit through anti-harassment training – the reason why the executives get away with it is because they have just enough supporters who will go through that litany of reasons to withhold consequences.
So let’s do the realist comparison of the two failure cases under consideration:
So to me it seems very clear that (2) is the realistic threat and the one we should spend the majority of our time threat-modeling and dealing with. Instead of which, objections to the bylaw change have been focused almost exclusively on (1). I do not think that is a logical or productive way to spend this much of our time and effort.
That matches my guess. Is it unclear why I find that hard to reconcile with your earlier ““there are, in fact, Fellows who repeatedly flaunt the CoC”? How does that come to be? Nobody is filing reports against them? They are, but the CoC WG declines to recommend action against them (e.g., perhaps because they don’t agree with your opinion that they’re flaunting the CoC? not trying to be snarky, just trying to make sense of it all)? You pretty much just ruled out it’s the case that the Steering Council is rejecting recommendations to act against them.
I don’t really care if you respond to that. @ubernostrum says they’re concerned about “prominent/high-placed/long-time” people getting away with breaking the rules, and “therefore” the threshold for the Board kicking them out must be as low as possible.
I’d be very surprised if they did. They don’t report to anyone with the power to act on that.
Have to say that one surprises me! To the contrary, more than a few people are terrified of the power the CoC WG is perceived to have. For example, it makes no real difference to my retired life if I can put “PSF Fellow” next to my name, but I would be deeply pained if I lost my “commit bit”. There’s nothing I deeply care about that isn’t already subject to revocation.
But, no, I’m not typical here.
Certainly a comparatively minimal one to me, but a tooth the Board should nevertheless have.
I’ve asked before under what legal theory a membership organization can be held legally liable for anything a member does on their own. I’ve never heard of that before.
Depends on how it’s worded. By “unanimous” I had in mind all those Directors attending a meeting. I want to live in the real world here - it appears there’s almost never a meeting anymore that all Directors attend. I’m looking to raise the bar, not make it impossible.
See “real world” above .
I fully expect it to pass. I’ll have to content myself with cackling “told ya so!” when the catastrophic consequences become undeniable.
Seriously, while I very much do prefer a bar just as high as realistically possible, I’ll probably end up voting for it too.
Then stop feeding it?
I stopped getting value out of this a few replies back, so I’ll set an example by stopping right after I type the next period.