First Release #1
Conversation
| /// Given domain as `<g>`, `CosetOfDomain` represents `h<g>` | ||
| /// | ||
| /// Constraint equivalent is in `r1cs_std::poly::domain`. | ||
| #[derive(Clone, Copy, Eq, PartialEq, Debug)] | ||
| pub struct Radix2CosetDomain<F: PrimeField> { | ||
| /// A non-coset radix 2 domain: `<g>` | ||
| pub base_domain: Radix2EvaluationDomain<F>, | ||
| /// offset `h` | ||
| pub offset: F, | ||
| } |
There was a problem hiding this comment.
This exists in r1cs-std now, right?
There was a problem hiding this comment.
Nope... r1cs-std contains the constraint code, but this one is native. (The reason I defined this is that coset domain is not in ark-poly yet. If coset domain is used elsewhere, I can move it to ark-poly)
There was a problem hiding this comment.
Oh completely missed that this is native. The difference from the coset_* operations on the existing EvaluationDomains is that this allows specifying the choice of offset for the coset, right? If so, I think it's useful to have in ark-poly. I think the approach in arkworks-rs/algebra#88 (comment) makes sense
There was a problem hiding this comment.
Exactly. It Sounds good! I may make a PR after I finished up BCS, but if anyone have time go ahead. (Until then I will keep this impl in ldt)
|
I think we should also have an issue detailing that we need to make an LDT trait, and LDT Gadget trait. I think its fine to not block making the repo public on this, as the exact trait API will likely be informed by your bcs work |
Description
Feature included in first release:
DirectLDTand its constraintsBefore we can merge this PR, please make sure that all the following items have been
checked off. If any of the checklist items are not applicable, please leave them but
write a little note why.
Pendingsection inCHANGELOG.mdFiles changedin the Github PR explorer