Andreas Fackler

[In hbbft](https://github.com/poanetwork/hbbft/blob/6d5689fb97fe5d058964181c77ca0fc46d5467be/src/broadcast/merkle.rs) I tried representing the proof just as its index and the vector of sibling digests. For us, that reduced the size considerably, and the proof generation (`MerkleTree::from_vec`) and...

Awesome, that `contract!` macro would be perfect! And I wasn't aware of https://github.com/paritytech/ethabi/pull/85; I guess that also makes my suggestion obsolete. Or do you think it would make sense to...

Thank you, we weren't aware of this! (At least I wasn't.) The [GMP benchmarks](https://gmplib.org/gmpbench.html) look very different. Were your GMP libraries compiled with similar optimization settings on both systems?

`Hydrabadger` is already generic in the type of contribution `C`. The idea is that `C` is a list of transactions, but it can be anything! So I'm not sure I...

We should also add some benchmarks, similar to [the ones in threshold_crypto](https://github.com/poanetwork/threshold_crypto/blob/master/benches/bench.rs), but maybe more deterministic ones.

Sounds good! Yes, I guess perf isn't that useful for now then. I'd add benchmarks and then look at potential higher-level optimizations. (Questions like: Do we need to reduce in...

(I'm not sure anyone is currently working on this crate, sorry.)

Also, _verification_ might be parallelizable? (Not sure right now whether this applies to our two VDF schemes.)

I'm probably missing something, but [I was wondering](https://github.com/poanetwork/hbbft/issues/55) whether some kind of special v*/termination message is necessary anyway, to guarantee that all instances terminate (not for output, of course, but...

I think this is for the case of a faulty sender: If the sender uses shards that aren't actually part of an erasure coding to create an otherwise valid Merkle...