I was researching about zero-knowledge proofs and in this link http://web.mit.edu/~ezyang/Public/graph/svg.html I've seen the exercise question:
- Currently, you can only select adjacent pairs of nodes to check. Would the proof still be zero knowledge if you could pick arbitrary pairs of nodes to check?
I think it would still be "zero knowledge" because it shouldn't leak information to verifier, but there won't be a proof if you select non-adjacent pairs since the reveal phase doesn't prove anything. Is this a trick question?