Randy, all our objects are now defined in terms of CMS.
CMS defines a multiple sign model. Its integral to the definition of CMS.
If we defined a multiple sign model for our data objects now, then we effectively future proof the design against any emergent need to have multiple sign over objects to validate them for RFC3779.
This is a remarkably low barrier right now, in a world of single sign. it is designing a data model to encompass a potential future need.
I do not believe one implementors view on optimal processing defines the scalability of multiple signer checking: for instance, a DB centric model would have no particular problem locating the set of direct signers, and in a pre-validated certificate hierarchy, (ie the covering trees for each certificate already being known) the edge- validation of the object by the EE certificates which signed it, which can be validated against the DB seems to me to be scaleable.
-George _______________________________________________ sidr mailing list [email protected] https://www.ietf.org/mailman/listinfo/sidr
