Closed KtorZ closed 4 years ago
@KtorZ, @dcoutts, so we no longer store the stake pool ID in the registry?
Shouldn't a single owner keypair be usable for several stake pools, in principle?
I agree this is probably a can of worms we might not necessarily want to open..
Shouldn't a single owner keypair be usable for several stake pools, in principle?
Also, cc @disassembler, as he's got stake in this question (pun, most likely not intended, probably, maybe).
@deepfire also note that these validation only applies to Jörmungandr and the incentivized testnet. Haskell nodes are working based on a different model and do not require all this.
@KtorZ, @dcoutts, so we no longer store the stake pool ID in the registry?
Indeed. We store the owner public key.
Shouldn't a single owner keypair be usable for several stake pools, in principle?
I am not sure about the answer here for it depends on Jörmungandr's design. Both options seem plausible (allowing it or, rejecting an already known public key). The registry as updated with this PR would work with both option.
Merging, as per Matthias' suggestion.
Although I wonder if we might have another change because of @disassembler's requirements.
See also: https://github.com/input-output-hk/testnet-stake-pool-registry-validator/issues/2
I highly suggest reviewing this commit-by-commit, the git history is somewhat meaningful and should help make the review little easier.
I did some smoke tests, fiddling a bit with the CLI, looks okay: