Actually doing precisely what you're describing here is something I've
wanted to implement for a long time. I've thought through a few times
how it might work, and it would be pretty cool. Unfortunately, I
still haven't implemented it.
For anybody else who is reading this, the idea is that all of the
TimeTravel / realtime sync information would get syncd between a local
cocalc-docker (or electron, etc.,) instance and the main
cocalc.com
servers. This would make it possible to run cocalc on your own
computer, but fully participate in realtime sync with other users.
A good way to think about this is basically just replacing the web
browser (as a client to
cocalc.com) with web browser + local compute
(e.g, via docker). That's all -- it would just extend what a
cocalc.com client can do. It might also make sense to have a vscode
plugin that connect to
cocalc.com and provides a similar link.
Implementing something like this is inevitably going to happen at some
point, since it's so powerful and useful. However, it hasn't happened
yet. A tricky part is figuring out where the filesystem for a project
lives or how it is sync'd, and Harald's message above has some good
ideas about how to do that.
This is somewhat related:
https://github.com/sagemathinc/cocalc-compute
--
William (
http://wstein.org)