>What's the problem that you're trying to solve asking validators not to use >PQ-downgrade protection?
DNSSEC is already a pretty complex landscape. Algorithm rolls are complex, we don't have a lot of experience signing with multiple algorithms. Adding extra restrictions to all use of a new algorithm brings us in unchartered teritory. Operational problems with DNSSEC tend to show up long after the IETF finished writing a specification. There is also a bit of misuse of the term 'downgrade protection'. A preference of ML-DSA-44 over ED25519 suggests that ML-DSA-44 is strictly more secure than ED25519. Which is, as far as I know, today just not the case. If we go one step further and also throw hybrid into the mix, then without signalling the situation becomes a lot more complex. _______________________________________________ DNSOP mailing list -- [email protected] To unsubscribe send an email to [email protected]
