each layer inherits most of the previous layer's functionality and adds additional restrictions:
- sneakernet/implicit transport honest threat model
- bip draft dependencies:
- concurrent & extensions - basic data format
- scrubbing - enforcing boundary between privacy domains e.g. own signing devices and collaborative txn construction
- hpke - optional - for embedding of out-of-band messages into psbt files
- define net settlement / payment confirmation
- base definition assumes out-of-band
- can be made in-band with hpke
- all "messages" are just PSBTs with a partial order (can't guarantee a join semilattice in general)
- should have discussion of privacy
- interactive transaction construction w/ pluggable transports in honest threat model
- use cases
- multisig
- internally segregated wallets (e.g. operational vs. customer funds)
- collaborative batching
- net-settlement
- define message space to avoid in-band
- framing format for duplex stream transports
- broadcast
- optional compatibility with lightning interactive-tx by bridging
- semi-honest
- considerations for on chain structure
- restrictions on transport
- anonymous broadcast by relay w/ anonymous p2p transport
- implications for liveness
- bft
- threat model, incentives
- coalition formation
- ownership proofs & online key enrollment
- listen advertisements
- co-spend proposals
- coalition formation proposal
- role of aggregator in bootstrapping agreement
- causal log structure, bft crdts = liveness in async/eventual consistency communication model (this may make sense to split, and move the non-bft component to interactive honest)
- validity proofs
- bft agreement
- setchain/redebelly + weak coordinator (bootstrapped by aggregator)
- TEE leader without redbelly? weak coordinator?
- distributed (set union consensus)
Originally posted by @nothingmuch in #11 (comment)
each layer inherits most of the previous layer's functionality and adds additional restrictions:
Originally posted by @nothingmuch in #11 (comment)