Greg Werner wrote:
>From a technical perspective once the user receives an EOL notification for a particular TSA he/she would have to manually refresh all roots to establish new and updated trust points. Or, another preventive option is to check revocation status for outer layer trust anchors on a periodic basis (once per day or week) and if the TSA root is no longer valid (revoked) then time stamp records from that particular authority should no longer be accepted.I have a little problem understand this. When should "time stamp records" no longer be accepted? I assume first, time stamp records means 'time stamps' according to RFC 3161 or ISO stuff.
Accepting a time stamp occurs: - by the requesting party, at this time revocation checking can be done as usual. - by a relying party that shortly after the event, i.e. in the normal course of the distributed workflow. Also I don't see a problem using whatever normal checking is done. I am not considering the third case: - by whoever trying to 'reverify' a time stamp LONG time after the event.
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature