VDF clean ups #770
VDF clean ups #770
Conversation
jonas-lj
commented
Apr 23, 2024
•
jonas-lj
commented
- Derive Serialize/Deserilialize automatically and use bcs for all serialization and deserialization.
- Simplify: Get rid of unnecessary abstractions.
- Add more tests and update existing tests.
- Refactoring: Move unit tests to modules.
- Update benchmarks and cli.
jonas-lj
force-pushed
the
jonas/vdf_update
branch
from
c788ab8
to
4b03382
Compare
| #[test] | ||
| fn test_discriminant_from_seed() { | ||
| let seed = [1, 2, 3]; | ||
| let target_size = 512; |
There was a problem hiding this comment.
shall we do at least 1024?
There was a problem hiding this comment.
Also with the new paper we could theoretically support 2K+ sizes, shall we do or not? Do we actually have a default strong one, generated with some nothing up my sleeve properties?
There was a problem hiding this comment.
It does support larger discriminants, but it takes a long time. I'll update the test to generate a 1024 bit discriminant instead (takes ~9 seconds) and compare it to a test vector from chiavdf.
There was a problem hiding this comment.
Re a larger discriminant, generating a strong "nothing-up-my-sleeve" is straight-forward to do. 1) We demonstrate that the implementation follows chiavdf for 1024 bits, 2) decide on some future information, say a bitcoin block hash or similar, 3) generate a 3072 bit discriminant with that seed and use that as a fixed discriminant..
| @@ -54,3 +54,34 @@ impl QuadraticForm { | |||
| } | |||
| } | |||
| } | |||
|
|
|||
| #[cfg(test)] | |||
There was a problem hiding this comment.
after this is merged I'd expect some fuzzy test coverage here (even via LLM aid)
jonas-lj
force-pushed
the
jonas/vdf_update
branch
from
7689c88
to
28bd286
Compare
jonas-lj
force-pushed
the
jonas/vdf_update
branch
from
550a606
to
0ceb924
Compare
