Self-contained proof of malicious behavior for Round 3 of KeyGen

[fjarri]

Suppose we are running the KeyGen protocol (Fig. 5, or KeyInit in the library terms), and in Round 3, when verifying a message from another node, the Schnorr proof verification didn't pass.

The goal with generating evidence of malicious behavior (#43) is to create a self-contained data blob which would prove for anyone that the owner of the signing key corresponding to some verifying key did something wrong. Which means that it can only contain the publicly available data and the messages from that signer. But the Schnorr proof in question depends on rid, which is assembled from the data sent by all the other nodes. How do we make a self-contained evidence package in this case?

One possibility is that rid is not really necessary - which will eliminate the problem. What kind of an attack will be possible if it is removed for the proof's parameters?