Jim,
Thoughts in line.
Jim Schaad wrote:
>The following do not appear to be well address in the present ESS
>document.
>
>1. For the message S1(S2(M)) - If S2 validates and contains a receipt
>request but S1 fails to validate. Should the receipt be generated?
>
>
Yes I think it should. The layer with the signed content identifier and
the receipt request is verifiable so one should be returned.
>2. For the message S1(S2(M)) - If S2 validates and contains a receipt
>request, S1 contains an MLExpansionHistory attribute, but cannot be
>validate due to either a) missing certificate or b) unknown signing
>algorithm. Should a) the receipt be generated and b) the
>MLExpansionHistory attribute be obeyed?
>
>
For a - yes
For b - no
>3. The following is a "new" case S1(E1(?)) - S1 contains a receipt
>request, E1 cannot be decrypted due to the lack of a lock box for the
>receipient. Should a receipt be generated?
>
>
Yes I think it should. It's only telling the originator that the
message was received not that they could open it. If the recipient
can't open it they should send a message indicating as much.
>Jim
>
>
>