SMT proofs with ics23 v0.10.0 or later

[yito88]

We're using arse-merkle-tree (our modified Sparse Merkle Tree) because the old ics23 crate didn't support SMT proofs.
ics23 v0.10.x supports them.
We can use SMT for IBC-related data.

• Require ibc-rs v0.42.0 or later Upgrade ibc-rs to 0.47.0 with mainstream tendermint-rs #2084

• Gaia, Osmosis, and Noble should support SMT proofs, provided by ibc-go v7 or later

  • Gaia v0.15.1 or later supported
  • Osmosis v23 or later supported
  • [Blocked] Noble doesn't support because it is still using ibc-go v4

[yito88]

Counterparty chains should be able to verify SMT proofs. They have to use:

• ics23 v0.9.1 or later
• ibc-go v7 or later

But the latest Gaia v13 (also the next v14) is still using ibc-go v4 (ics23 v0.9.0) which has been already EOL...

[cwgoes]

@yito88 do you know if there have been any updates here Gaia-side?

[yito88]

In Gaia's main branch, ibc-go has been updated to v7 supporting SMT proofs.
However, it's not included in the latest release v14.1.0.

[yito88]

Gaia v15.1.0 has been released, which version has ibc-go v7.
Started replacing arse-merkle-tree.

[yito88]

Osmosis v23 has ibc-go v7 already.
Noble is still using ibc-go v4 according to https://github.com/circlefin/noble-fiattokenfactory

[cwgoes]

@yito88 Is this issue still relevant?

[yito88]

The latest Noble uses ibc-go v8. To reduce the Merkle tree size, we could go back to SMT for IBC-related data. I'll test it.

[yito88]

hmm, the current ics23 can't support our SMT. Our hack to replace the key and value in the proof doesn't work by the changes which forces to apply the given spec instead of the leaf spec.

[yito88]

We need to keep using arse-merkle-tree. We could open it again if ics23 would support other SMTs