Henry de Valence

@n4ss Cool, if you'd like to write a PR that adds endianness testing to CI (could be Github Actions instead of Travis, whichever is easier), I'd be happy to merge...

Removing the dependency on `tendermint-rs` wouldn't make our life easier, because we'd still need to use the Tendermint light client to speak IBC to Tendermint chains. Although Tendermint is only...

I think I'm missing something -- why does #284 need to be a hotfix to a previously released version, rather than just using the `main` branch? It seems like much...

As an onlooker, I'd recommend against using feature flags. They're quite a lot of work to build and maintain, and they bring a high risk of undetected breakage, because cargo's...

It seems like there are two possible approaches here: 1. Make IBC relaying more consensus-aware: provide mechanisms to tie IBC relaying and block construction together. 2. Make IBC relaying less...

> One option is not to remove, them, but replace them with a length prefix. Varint, like protobuf does > > Also, I don't see the use in prefixing something...

Probably Event's serialize impls should run through the proto type (using Serde's try_from/into attributes) and the proto type should have a ProtoJSON compatible Serde impl via pbjson_build.

Example of how this works for a random Penumbra data structure: the `CompactBlock` rust type serializes through the proto type: https://github.com/penumbra-zone/penumbra/blob/main/crates/core/component/compact-block/src/compact_block.rs#L20-L21 The corresponding proto type has a generated Serde impl...

> The proto file at https://github.com/cometbft/cometbft/blob/main/proto/cometbft/abci/v1/types.proto#L467 shows the type field isn't optional No, it doesn't, it says ``` string type = 1; ``` All proto3 fields are optional. A message...

The ABCI domain types should be on the list of PRs that need to be ported to main; not having those blocks us from rolling back to 0.34.