Documentation for failure modes — heartwood

Hey! It would be quite useful to have documentation for failure modes of Radicle.

I think the failure mode that is most likely to occur is two collaborators of a project independently advancing the repo’s HEAD while offline. Once they reconnect, Radicle fails to sync saying that no quorum could be reached.

What’s the correct way to recover from this situation?

If there’s already some documentation for this, sorry for missing it, would appreciate a link!)

Another failure mode I found (I’m just experimenting with a few scenarios):

init/import a new private repo
add a delegate, but do not sync the repo with the delegate yet, but do accept the change
realise that there’s only one node which can link you with this delegate at the moment!
try to allow access for that node
notice that the other delegate (which doesn’t have the repo yet) now has to accept this access change

The only ways to recover from this seem to be out-of-bounds (transfer the repo to the delegate some other way, e.g. by establishing a network connection somehow, or transferring the repo without using radicle at all, …)

This is another sharp edge people might run into quickly (I ran into this while trying to set up a test repo for another scenario).