I agree with you that that first bit is a problem. Though I don't know how to balance that with the desire for privacy / the desire to avoid publishing what amo

I agree with you that that first bit is a problem. Though I don't know how to balance that with the desire for privacy / the desire to avoid publishing what amounts to a personal attack on the person, which can follow them into other communities, which is a penalty significantly beyond banning from a single community. In the case where the person was working in good faith but just had a spell of poor judgment, it seems ethical to allow them the "right to be forgotten," which implies not putting information anywhere that a search engine can index. And in the case where they're actually working in bad faith, they probably are going to be more than happy to make legal threats against everyone involved, which is a chilling effect on the submitter and the moderators. Even if they're correct and confident in everything they say, proving it in court is an unpleasant process. Maybe a compromise is to say that complaints are private provided the person on the receiving end also agrees to keep the details private, and if they start to litigate it in public, it's permissible to publish the complaint. I'm not sure that helps you with the bad-faith participant, though, but in this case it seems like it's a good-faith contributor who just exercised a whole lot of bad judgment (and perhaps the implicit encouragement to not have this saga Googleable by future employers would have been good for everyone involved). The other option is to have a strong community norm of contributing with pseudonyms, as e.g. Wikipedia does, and then your "right to be forgotten" is your right to pick a new pseudonym, and the whole moderation process can happen in public. But for something like Go, it's a norm to contribute with your real name as known to employers. To your second point, the way I read it is not that they made an ex-post-facto change to the code of conduct for this incident, it's that this was already in the works (especially given his comment that he was involved in writing the CoC). If I'm misinterpreting that, then yes, I agree that it's a problem. I'm of the opinion that communities also need to state who can change the code of conduct and what process they'll use, and changing it in a way that surprises the community is bad. (That said, as long as you specify it up front, I'm totally okay with the BDFL model for that, and I think it's the right model in most cases. A CoC is only as good or bad as the people who enforce it; its value is in setting expectations clearly.)