BLS-12381 London Inclusion
Closed this issue · 11 comments
Kelly Olson (EIP-2565 champion) would like to champion the inclusion of BLS-12381. Opening this issue to discuss.
Would it be sufficient to add Kelly as an author of the 2537 to give him edition and championing powers? I'm working on the changes requested during moving from "Draft" to "Review" status, and can also include him
This is about EIP-2537?
I’m only guessing that most likely it is unless there is some alternative idea how to introduce it
Would it be sufficient to add Kelly as an author of the 2537 to give him edition and championing powers? I'm working on the changes requested during moving from "Draft" to "Review" status, and can also include him
This works great and my intent was to champion (and modify as needed) @shamatar 's existing proposal.
From the ETH R&D discord:
BLS Support as a precompile - I still think evm384 is a better solution to the underlying problem and needs to be pursued instead.
not sure. I'm slightly in favor of it over 384 right now, but I'm on the fence.
As usual, choice is between:
- native precompile, ready, tested, fast on every platform (EIP-2537)
- EVM384 PoC concert proposal, slower in any case, requires EVM repricing to be even reasonably not super expensive, and for now can only achieve this "reasonable" performance in evmone (Geth based fork is few times slower)
If someone makes a set of ADD_X/SUB_X/MONTMUL_X is would be a great feature, but it requires an EIP, proper plan, pricing, and may be few years of time looking at how things usually work :)
@timbeiko I’ve added Kelly to authors of the EIP for logistics, so can you add it to London formally?
Since EVM384 was mentioned, I will share the current status. Over the past few months, several breakthroughs in EVM384 have significantly decreased gas cost and runtime. At the request of core devs, I have been preparing the following update doc https://notes.ethereum.org/--JjliY8T_-qIdvAQKQlcg?view .
We agreed on ACD111 to not include this in London, so closing this.