Chain TryFrom errors rather than calling unwrap on them
#91
Conversation
| @@ -1,6 +1,6 @@ | |||
| [package] | |||
| name = "wsts" | |||
| version = "9.1.0" | |||
| version = "10.0.0" | |||
There was a problem hiding this comment.
If we're going to bump the version, should we remove that unused error variant?
There was a problem hiding this comment.
Yeah we can go wild here.
| /// threshold, where the degree of each poly is (t-1) | ||
| t: u32, | ||
| t: usize, |
There was a problem hiding this comment.
Curious, usize instead of u64 or u32? Same question for neg_shares.
There was a problem hiding this comment.
The goal was to get rid of the conversions inside the C callback, since we can’t return errors from there.
There was a problem hiding this comment.
Oh wow get_point still has an as cast 😮
There was a problem hiding this comment.
Ahhh I see. Got it.
We use usize because we often feed things into functions in the p256k1 library. That library wraps the libsecp256k1 C library via the rust-bindgen crate. And bindgen assumes size_t is usize (which is the default). We can change that default but keeping things as is allows p256k1 to be used on rust targets where the pointer size is not 64-bits.
There was a problem hiding this comment.
Yeah we need to do another pass on usize vs explicitly sized types. Anything that doesn’t go over the network is probably better s as usize.
… BadPolyCommitmentLen
|
Can we also replace the |
Yeah. I had aspirations of making this |
Ahh I see. Yeah, I was wondering why that was used. Well, if that ends up being a goal again we can change it back or something. |
Fixes #89