Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

BLS-12381 London Inclusion #269

Closed
timbeiko opened this issue Mar 5, 2021 · 11 comments
Closed

BLS-12381 London Inclusion #269

timbeiko opened this issue Mar 5, 2021 · 11 comments

Comments

@timbeiko
Copy link
Collaborator

timbeiko commented Mar 5, 2021

Kelly Olson (EIP-2565 champion) would like to champion the inclusion of BLS-12381. Opening this issue to discuss.

@shamatar
Copy link

shamatar commented Mar 6, 2021

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

@jochem-brouwer
Copy link
Member

This is about EIP-2537?

@shamatar
Copy link

shamatar commented Mar 6, 2021

I’m only guessing that most likely it is unless there is some alternative idea how to introduce it

@timbeiko
Copy link
Collaborator Author

timbeiko commented Mar 7, 2021

@shamatar I think we probably should, yes. I just don't have his GH handle.

@ineffectualproperty
Copy link

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.

@shamatar
Copy link

shamatar commented Mar 8, 2021

I've added Kelly as an author in this PR

@timbeiko
Copy link
Collaborator Author

timbeiko commented Mar 14, 2021

From the ETH R&D discord:

Alexey, TurboGeth:

BLS Support as a precompile - I still think evm384 is a better solution to the underlying problem and needs to be pursued instead.

Martin, Geth:

not sure. I'm slightly in favor of it over 384 right now, but I'm on the fence.

@shamatar
Copy link

shamatar commented Mar 14, 2021

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 :)

@shamatar
Copy link

@timbeiko I’ve added Kelly to authors of the EIP for logistics, so can you add it to London formally?

@poemm
Copy link

poemm commented Apr 1, 2021

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 .

@timbeiko
Copy link
Collaborator Author

We agreed on ACD111 to not include this in London, so closing this.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants