Conversation

This looks like a basic hit job on Jeremy 🧐
Quote Tweet
NumFOCUS found I violated their Code of Conduct (CoC) at JupyterCon because my talk was not “kind”, because I said @joelgrus was “wrong”. This sets a bad precedent. Joel was not involved in NumFOCUS’s action, was not told about it, and did not support it fast.ai/2020/10/28/cod
Show this thread
2
29
Looks like they walked it back as a procedural error, but the damage is done. Personally, I am onboard with the *objectives* of CoCs, but skeptical of them as a *mechanism*. They’re just too vulnerable to abuse and incompetence (latter appears to be the case here)
Quote Tweet
Public Apology to Jeremy Howard: numfocus.org/blog/jeremy-ho
2
20
People should use words like “incompetence” or “apathy” or “we’re stupid” in apologies like this. If you deny malice, you must accept stupidity. Otherwise it’s a get-out-of-jail-free card for unaccountable bureaucrats who were just spectators of miraculous errors without errants.
38
This is one of the secondary reasons I suggested winding down refactor camp in 2019 if we couldn’t find a way to decentralize/virtualize it for bureaucratic abuse resistance (weirdly Covid-friendly call as well).
5
While we had no problems in the years we ran it, the creeping expectations of CoC burdens etc made it increasingly taxing to run. I’m glad we quit on a high note without any serious issues in the 7-year run. It likely won’t be resurrected (though not for this reason).
6
In the experiment we’re trying to make decentralized structure and 1:1 trust do almost all the heavy lift work of ensuring a safe environment. There are no CoC compliance bureaucrat roles. The CoC is mostly “don’t make us make a rule for this.”
12
2x2: attacker org vs target org. Both ranging from centralized to decentralized. I think right-wing malice manifests like nerve gas in the subway (centralized attack on decentralized target), while wokewashed malice is more like a nuclear attack. Centralized on centralized.
2
4