Aaron Craelius

> @aaronc assigning you for review for this: > > - is the function we are giving simple and extendable enough? > > - does this introduce any bugs such...

here are my thoughts: * Issues with coin types and HD derivation paths are purely a wallet issue and completely opaque to state machines. As long as the desired Ethereum...

How would you model initial token distributions? As a sequence of mint messages? (Which by the way doesn't exist yet). I think supporting generalized message execution makes sense but can't...

Makes sense regarding the create denom message. Practically it's hard to deprecate old features though. Usually we need to create something new while keeping existing stuff at least in maintenance...

We should have a discussion to align on encoding. I suggest having a deterministic output (no trailing zeros) and flexible input. We must explicitly reject NaN and Infinity though. Can...

I want us to distinguish two types of module accounts: - used internally and shouldn't receive coins. These probably shouldn't have an address at all and should be stored separately...

> > I suggest we call them module escrows > > This suggests that it will receive coins. Users send money to escrow. > > The use case is to...

The alternative is on receive hooks in accounts to reject sends.

Looks like CI is failing here because I'm using a replace for schema in collections, but many places that have collections referenced don't have the replace for schema, but do...

> api looks good, the registration would be in app.go correct? Which registration do you mean? The way it's setup here, app.go would need to call `EnableIndexer` if there's app.toml...