Francisco Giordano
Francisco Giordano
Yes I'd say a library for delayed updates is not necessary.
The goal of the delay is to ensure there is time to cancel a malicious transfer. A malicious transfer would always use the minimum delay possible, and the option to...
IMO it should not be Ownable.
Thanks for bringing this up @wighawag. Super interesting. > Unfortunately this computation will be dependent on gas pricing. And as such an overestimation is required. To be clear, by "gas...
I'm looking at the EIP and it seems rather clear that a reverting `supportsInterface` should only be interpreted to mean "false" when querying about supporting ERC165 itself. > **How to...
The second bug is easy to fix. The first one is the tricky one. > the "after-the-call" check does not work if the implementer of `supportsInterface` is doing something like...
Pasting here @wighawag's suggestion from https://github.com/ethereum/EIPs/pull/881#issuecomment-491677748: > check for gasleft() after the call > >``` >// execute STATIC_CALL with 30000 gas >require(gasleft() > 30000/63, "not enough gas left"); >``` >...
> i.e. if the call failed and it possibly run out of gas (there's less than 30k/63 remaining), then we revert. Yes. Why "possibly" though?
Technically we can assume that the implementer will use at most 30000 gas. So we could send like 30001 (or a similarly small value), and then we can be sure...
>That's true, but this is not very clear since it does not mention the 30,000 rules neither here That requirement is mentioned in the general description for `supportsInterface: > This...