FAQ

Informative.

Isn't a signed, anchored record already enough? It proves the record was not changed. It does not prove the result would recur. A record can be perfectly sealed and perfectly wrong.

Why not just re-run with the same lab? That is a Repeat Run. It shows the lab is consistent. It cannot show the lab is right.

Why fresh canaries instead of the same ones? The same canaries would be replay. Reproof reproduces the outcome against the same control, not the same packets.

Who can commission a Check Assay? Anyone, against any Registered proof, including a competitor of the Provider. The Commissioner's interest is declared, and the same independence and umpire rules bind it.

What if the check disagrees because the original was honest but unlucky? Divergence records a direction, not an accusation. The umpire decides which result stands, not why.

Can a Provider keep commissioning checks until one agrees? It can commission them, but every one is published, and no status can be shown without the full count of agreeing, diverging, insufficient, abandoned, and lapsed checks.

Does reproof make a Unit of Proof worth more? Nothing in this specification assigns value. A reader may weight by Reproof Status and must say so (PP-SPEC-080).

Can a claim from another framework be reproven? No. Only a committed Proof Protocol run can be reproduced. Anything else is verification evidence: it can be proven by a new run, not reproven.