Posted by olkurbatov
Oct 27, 2026/16:48 UTC
The notation in the formulas for $C$, $R_1$, $R_2$, and $s$ required correction to maintain consistency with established standards like BIP-327 and BIP-445. Specifically, the coefficient $c_i$ should be adjusted to $\hat c_i = g \cdot c_i$, reflecting the convention of applying accumulated parity to each signer's secret. This adjustment ensures that the equation $s = r + e \cdot a$ holds, where $a$ is defined as $\sum_{i \in T} \hat c_i k_i + \tau$.
If the generator $g$ takes the value $-1$, the original equations inadvertently prove knowledge of an incorrect discrete logarithm. The calculations then yield $C = [x + \tau]B$, which is interpreted as a share of $X + [\tau]G$ rather than the intended $A$. This discrepancy causes the proof to fail verification according to BIP-374 when checked against $A$. As a result, the session properly aborts instead of erroneously accepting an invalid share.
Despite these changes, the responses $z_i$ and the per-signer verification checks remain unaffected because they are computed using $K_i$. This retention further safeguards the protocol's correctness. An update to the corresponding documentation will ensure clarity and alignment with this corrected approach.
TLDR
We’ll email you summaries of the latest discussions from high signal bitcoin sources, like bitcoin-dev, lightning-dev, and Delving Bitcoin.
We'd love to hear your feedback on this project.
Give Feedback