Jim,
I'm sorry that my PEM WG meeting minutes didn't make the
intent of this field clear. The short form is as follows:
PEM, as defined by 1113bis, consists of three processing
steps, of which the first and third are specific to the application.
The first step is SMTP canonicalization (clearly application-specific)
and the last step involves representing the PEM header fields in
a fashion suitable for transport in a given mail system and in
encoding the PEM-processed message in a form suiatble for transport
in a given mail system (MIC-CLEAR being an expcetion to the second
part of this step). Thus onlt the middle step, i.e., MIC calculation
and optional encipherment, are pure PEM. This leads to the suggestion
that we parameterize PEM for use in different environments, allowing
both the initial step (canonicalization) and the third step (header
representation and content encoding) be different for different
message system environments. The Content-Domain field is intended
to specify the processing applied to the plaintext message which is
PEM processed. The specification of the third step will be different
for each message system and is currently specified only for vanilla
822 messages transport.
Steve