|
Hello, Nicolas.
The changes you noted were made after a review of the previous draft
revealed that the unintentional changes had been introduced which made
it inconsistent with RFC 3335 ("as1"). That, as I believe you
will agree, would not be a good thing; hence the updated text again
brings them into synchrony.
Note that the signed-receipt-micalg parameter is still honored for signed
messages: the "algorithm used to calculate the MIC" refers to the
digest returned inside the signed MDN as the "received-content-MIC", which
may be different from the algorithm used to sign the MDN itself.
The signed-receipt-micalg parameter refers to the algorithm used to sign
the MDN.
For an unsigned message that requested a signed MDN, a digest will be
computed using SHA-1 and returned as the "received-content-MIC".
In either of the above cases where a signed MDN is requested, the MDN
is signed using the algorithm specified in the signed-receipt-micalg parameter
in the disposition-notification-options header (left-to-right precedence
if more than one algorithm specified) , or SHA1 if not specified.
Hope that clarification resolves the issue to your satisfaction.
Regards,
Richard
Nicolas Bielza wrote:
in draft-ietf-ediint-as2-14.txt,
section 7.4.3: For
signed messages, the algorith used to calculate the
MIC MUST
be the same as the algorithm that was used on the
message that
was signed. Then
if the message is signed, the signed-receipt-micalg parameter is not taken
into account;But
the draft continues with:If
the message is not signed, then
the SHA-1
algorithm should be used. So
the signed-receipt-micalg parameter is never taken into account ? (And
there's no way of requesting a MD5 MIC for an unsigned message).
|
