You might be interested in the old Plan 9 experiments sam & samterm. https://en.wikipedia.org/wiki/Sam_(text_editor)
Sam itself was an experiment in evolving the ed/ex/vi model of editing to a more capable structured regular expression basis, which doesn't apply here. Plan 9 itself was a (giant bundle of) experiment(s) in moving the unix model fully across the network via the filesystem, which is slightly more relevant here. The part that I think might interest you is samterm, which is basically "separate the display process from the editing process, and see if you can end up with all of efficiency, capability, and network transparency". From what I recall, it was relatively successful on those fronts, but it was supplanted pretty quickly by a different experiment in "editor plus" called acme, which is still used (although fairly niche) today. The key approach for samterm that might make it interesting and/or totally irrelevant is this: sam drives samterm not via a graphics or display protocol, but by an on-the-line protocol that describes the editing changes themselves (because it was largely built around that concept of structural regular expressions). Likely, it'll mostly be a sideline of academic interest (so I forked the thread over to emacs-tangents), but there's nothing wrong with those! ~Chad
--- via emacs-tangents mailing list (https://lists.gnu.org/mailman/listinfo/emacs-tangents)
