Falko Strenzke <[email protected]> writes:

>Strongly unforgeable signatures do not lead to unique X.509 certificates

The solution to this problem is to think in terms of allowlists, not
blocklists, even though this goes against decades of established X.509 bad
practice, in turn based on older bad practice with credit-card blacklists [0],
which goes back to even older bad practice with cheque blacklists (hey, it's
gotta start working some time if we just keep doing it again and again).

If you encode a certificate non-canonically it's no longer valid in an
allowlist environment, while with a blocklist it is.  So you don't need to
Rube-Goldberg your signatures, just change the way you think about working
with certs.

Peter.

[0] They were still called blacklists back then.
_______________________________________________
TLS mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to