“plasma cash + verified state transactions”
I’m very curious about this type of construction and have a post about it @[0]. Namely, taking the property of Plasma where people are responsible for their own data availability and using validity proofs to limit the amount of data (history) a user would need to keep around to secure their coins to O(#coinslots).
From where I was at there’s some more thought needed in the direction of the worst-case exit procedure and ensuring a spend+exit of the same coin can’t happen in weird edge-case scenarios, but I like the idea in the OP of including on-chain exit proposals in the validity proof as a potential solution.
If you do it that way, how would you publish the full state to chain and authenticate it if the operator disappears?
If the construction is like what I’m referencing, then users themselves are responsible for the data availability that proves ownership of their own funds. It ends up looking like the familiar Plasma Cash setup, where, users can attempt to exit coins and other users can challenge the exits within the challenge period. You can exit at your leisure, unless you need to challenge someone invalidly attempting to exit your coins.
Similar synchrony assumption to Plasma Cash variants-- that is, a user needs to sync once per challenge period to challenge any potentially invalid exits on their coins.
[0] Validity Proofs + Plasma Cash = Simpler Exit Game/Coin History?