Better EIP-7702 ergonomics#325
Conversation
RomarQ
left a comment
There was a problem hiding this comment.
Not sure it matches with the following description. Will try to have a better look a it later today
If any step above fails, immediately stop processing the tuple and continue to the next tuple in the list. When multiple tuples from the same authority are present, set the code using the address in the last valid occurrence.
sorpaas
left a comment
There was a problem hiding this comment.
Looks right. It's also just equivalent if we just remove the item from authorization_list if signature verification fails.
|
Ye, I think it is better to skip the item when building the |
You're both right, that would be equivalent. However, removing items at the I think the current approach is as efficient and better follows EIP-7702 specs given these constraints. |
* feat: ✨ make authorizing_address in authorization list optional * test: ✅ test None authorizing_address * style: 🎨 fmt
Description
Using a
Option<H160>for the authorization address allows to send aNonevalue when the signature (r,s,y_parity) is invalid and the authorization needs to be skipped as described in step 3 of EIP-7702.