Pieter Wuille

@Rspigler If there are up to 2 substitution errors, it is 100% accurate. If there are more or other errors, it is certainly going to be wrong (if it locates...

@Rspigler I don't see what these statistics have to do with that. If a user types an incorrect address, and it is more than 2 substitution errors, or a different...

> How often are there 3 or more substitution errors, that the algorithm thinks is 2 or less? Depends on the model of actual errors, but likely close to 1...

> why not make it so users can only paste and clear this field and not allow character edits at all? But there should still be a way to enter...

FWIW, bitcoin core has an AppVeyor MSVC/Windows CI environment. Maybe it's worth looking into adding one for libsecp256k1 directly.

With #831 + the hddivsteps improvement, I think the difference between vartime and consttime inverses is much smaller (around 1.8 us vs 1.2 us on my Zen+ CPU, where signing...

Concept ACK, I still need to go over the logic changes.

Wowwow, calm down!

See https://github.com/sipa/secp256k1/commits/pr1031 for two commits on top of this. One renames your delta to theta (as the variable operated on is really defined as delta-1/2 compared to delta as defined...

Updated my branch with additional assertions and a few more improvements to the documentation. Code review ACK. I'm happy to do my changes on top in a separate PR, though...