- How does state commitment compare to recursive zkp, like Mina ? They both achieve similar effect of ‘fast sync’ / light client. How fast is Mel’s light client sync, and why not ZK compress the process ?
- Would there be a layer two like https://aztec.network , or something as capable as that.
- Does it make sense to use Locutus (freenet2) with Mel light client ? Further, tightly integrate them ?
Both of 1 provide a hash of current chain state. And with this hash we can upgrade networks like Locutus to have consensus. I see two ways
- Use Mel with Locutus in the app, which doesn’t require official integration.
- Use Mel in Locutus contracts. This supplies Mel chain state to contract verification process.
I believe Locutus is superior to Fediverse/Matrix, see this.
- Could Melodeon Lang provide features like formal verification ?
afaik, Mel is even better for L2s like aztec
nullchinchilla has talked with lunarpunks about this (they are implementing their own sets of blockchain & crypto)
ZKPs add compute costs and complexity
no, zkp is definitely a bonus.
the compute cost is negligible.
ZKP’s will be a very important part of Mel’s eventual ecosystem.
Part of this will of course be the usual on-chain suspects — zkrollups etc. A big future milestone is to optimize the MelVM interpreter for implementing novel cryptography on-chain, including ZKP functionality.
But a big underexplored field is off-chain ZKP within off-chain peer-to-peer networks that use Mel for something else, like antisybil or PKI. e.g. how Fulgor/Rayo uses off-chain ZKPs to anonymize Lightning-Network-style payment channels, while the on-chain part requires no ZKP.
(And Fulgor/Rayo uses old ZKP from years ago; modern techniques would yield much better performance)