Re: Odd/bad sentence in 5.4.1

Thomas Roessler wrote:
> On 2008-04-04 15:29:00 +0100, Stephen Farrell wrote:
> 
>> - If an RP is doing relaxed path validation, then it can ignore
>> the current time when considering notBefore and notAfter fields.
>>
>> I'd be open to allowing non-overlapping validity periods in cert
>> paths when doing relaxed path validation, but there's probably no
>> point if the underlying crypto APIs already insist on some overlap.
>> (Which I think is the case, can't recall really.)
> 
> The additional question here is whether relaxed path validation
> should be permissible for any validated certificate, or maybe
> prohibited for augmented assurance?

Relaxed path validation should be prohibited for AACs. I can't
think of a sensible augmentation (of the authentication) that
would be true for all time.

S.

Received on Friday, 4 April 2008 14:41:39 UTC