From a versioning control point of view WebDAV seems to be exactly the thing that you're trying to achieve, especially in regards to Delta-V?
I might be wrong, did I miss something? I guess what I'm trying to find out is what the ultimate goals or features of the WebFS are, given that WebDAV exists and is supported by git as a remote filesystem. WebDAV seems to have a specification and lots of support by other server implementations, so I'm wondering what kind of featureset was missing in there that WebFS now includes.
The overall internal cell/branch structure of got seems to be similar to how git uses blobs and trees as a list of references to blobs.
You seem to have also built a p2p library which I found pretty neat and might also be interesting to other HN readers (go-p2p) [3] that you use now as a transport layer. What I specifically liked about it is that it's based on QUIC, so it has no troubles staying under the radar of traffic classification, while still being able to punch holes. The generic abstraction by different protocols is really nice, and similar to what I've tried to do with my p2p networks. Seeing network protocols just as a way to transport (or encapsulate) things usually makes the logic behind swarm interactions more robust.
[1] https://github.com/brendoncarroll/webfs/blob/master/pkg/webf...