@braid Thanks for giving the talk. I just watched the recording of it. You covered a lot in a short amount of time. One clarification that I'd like to add: I wouldn't say that the approach taken by MSC2883 was outright rejected by the IETF people. Some people did raise concerns, but some people also see some potential in it. In the mean time, there are a some recent IETF drafts that try to address the issue of MLS's linear ordering requirement. One is
datatracker.ietf.org/doc/draft… which I believe should be compatible with the MSC2883 approach and would fix one of the big concerns with the MSC2883 approach. And another one is
datatracker.ietf.org/doc/draft… which is a very different approach.