Re: Private key usage period extension

Peter Gutmann wrote:
> Erik Andersen <[email protected]> writes:
> 
>>In my mind, the validity of the private key should not spread outside the
>>validity period of the certificate.


As Russ wrote, a relying party doing signature verification deals with
_public_ key validity and with certificate validity only. So private
key validity might go ignored by many/most verifiers of digital signatures.

Renewal of certificates with same key also seems not that unusual.


> 
> It's not meant for that, in fact it's the exact opposite, it's an extremely
> useful extension for when you want to say that, for example, a signing key is
> valid for one year but the certificate used to verify its signatures is valid
> for ten years.  The lack of a capability for doing this has been plaguing
> cert-based signatures for years, leading to all manner of workaround hacks to
> deal with verifying signatures after the cert has expired.

For "online" signatures (for the purpose of online authentication),
the reference time will should be the real time of the verifier, not a
time asserted by the signer himself in the signature blob.

In contrat to that, "offline" signature verification, in particular
for archived or batch-processed material often uses the reference time
from the signature blob -- and unless that signature is countersigned
by a trusted third-party timestamping authority, a signer-asserted
signing time is pretty fluffy, and could be made to fit within a
"private key validity period".


-Martin

_______________________________________________
pkix mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/pkix