Ed/merge sig agg

[elliedavidson]

This PR is a wrapper around this PR: #1568

[shenkeyao]

LGTM!

[philippecamacho]

I get this failure when I run the tests on my machine.

    at src/traits/networking/libp2p_network.rs:549

The application panicked (crashed).
  2023-08-21T22:45:30.303380Z ERROR libp2p_networking::network::behaviours::gossip: error publishing gossip message Err(Duplicate)
    at libp2p-networking/src/network/behaviours/gossip.rs:218

Message:  TEST FAILED! Results: [("Test Overall Safety Task", MismatchedLeaf)]
Location: /home/leloup/repositories/espressosys/HotShot/testing/src/test_runner.rs:168

Backtrace omitted. Run with RUST_BACKTRACE=1 environment variable to display it.
Run with RUST_BACKTRACE=full to include source snippets.
FAILED

failures:

failures:
    webserver_libp2p_network

test result: FAILED. 0 passed; 1 failed; 1 ignored; 0 measured; 0 filtered out; finished in 11.11s

[alxiong]

1. I thought @dailinsubjam is going to move SignatureKey into its own crate under /crates/ and out of /types? does that affect this PR as well? (or is this PR supposed to eclipse Sishan's?)
2. I got lost in the many abstractions of crypto-team-defined QuorumCertificate, jellyfish's AggregatableSignature, and here HotShot's SignatureKey, AssembledSignature, SignedCertificate, and HotShot's own QuorumCertificate, we should clarify them and remove unnecessary layers of abstraction. (maybe in a separate PR)
   in fact, I'm lost because I don't know which trait is for what purposes and thus cannot comment on what API I expect it to have.
3. would be great if @elliedavidson could provide more details in the PR description, at least highlight the main changes, and what changes deserve special attention (if any)

[alxiong]

it felt underconstrained w/o specifying the equivalence between the key in ENTRY and K.

I wonder if you want to use:

use hotshot-types::traits::stake_table::StakeTableScheme;

pub struct HotShotConfig<K,ST: StakeTableScheme, ELECTIONCONFIG> 
where ST::Key: K, {
  pub known_nodes: Vec<K>,
  pub stake_table: ST,
}

[elliedavidson]

1. Yes, that will happen in a separate PR!
2. I agree. In general HotShot traits have gotten messy. Here is an issue for this, but we unfortunately have other priorities to work on right now: Remove unnecessary layers of abstraction between crypto primitives and HotShot traits #1595
3. Normally a PR would have more details! Sishan's original PR did have more details. I only made this PR since Sishan can't make it while she is off.

[elliedavidson]

Re this particular comment - I agree - good catch. I think we should probably just remove the known_nodes since it effectively the same thing as known_nodes_with_stake. Issue here: #1596

[elliedavidson]

@philippecamacho
This is fine; the fix for that test is being upstreamed in another PR.

[philippecamacho]

LGTM