Weikeng Chen

Ale mentioned that this paper is related: https://eprint.iacr.org/2021/1359.pdf And yes, using BLS-24 gives us the advantage that the prime can be smaller (due to the higher embedding degree), and it...

Pratyush added this as a documentation improvement goal. I guess there is a reason to separate it from the existing serde. A probable reason is that we want different types...

I think I will need read into this a little more. For me, the flexibility to deserialize G_1/G_2 encodings with or without subgroup checks, and with or without compressions would...

Then the question becomes this: what if I want to be able to serialize a transaction in two different ways, one compressed+checked, one uncompressed+unchecked? Basically, the passing of the "wrapper"....

Ah, I think I am thinking too much---in that case, we just don't use "derive".

The wrapper also implies a cost though: now accessing `tx_proof` requires you to use `.0` to access the element inside the wrapper. I feel maybe a good idea is to...

I feel that there is a consensus to have some support for `serde` as long as it is doable, and we just need a plan to merge the API as...

update: currently debugging on an unsatisfied PC check using tracing.

Note: This branch uses the `constraints` branch of `poly-commit`. Once `poly-commit`'s PR is merged, this part needs to be changed.

(Help needed. The no_std error seems weird, though ci.yml for this branch has been updated.)