first stab at eip155:0 lang in eip155/caip2.md #153
Conversation
|
|
||
| ### Special Case of Chain-Independence | ||
|
|
||
| The `chainId` value `0` is valid per the original [specification for the chainId system][EIP-155] but unused in (and prescribed by the tooling for) the canonical registry mentioned in that same specification. |
There was a problem hiding this comment.
This makes sense in the context of EOAs or contracts that canonically deploy to the same address in EVMs. IIRC though there's some "EVMlike" networks where the addresses don't match and those edge cases should be explicitly mentioned here.
The one that stands out in my mind is Starknet I think it was or maybe Aztec where the contract addresses wouldn't match for 4337. I remember talking with @yoavw at WalletCon in Denver a few years ago where this was a struggle he was encountering for smart accounts, but not the exact details. So the takeaway here is probably the following:
-
We should check whether those edge cases still exist
-
Explicitly define how they should be handled when the same address doesn't match in that case
Discussions coming out of the CAIP-25 syntax overhaul conversation made me realize this should've been specified a long time ago!