Is the author of the spec here?

I notice in section 6.1 reliability, the word “must” in clock going backwards mitigation is not MUST.

Intentional?