I think this is an excellent approach.
But I do think there are *three* kinds of fax under consideration:
(a) Fax data over-e-mail (Ifax, EIfax, ITU T.37):
Some aspects of the traditional fax model cannot be fully realized (IMO).
(b) Fax protocols over IP (ITU T.38):
This is very specific to carriage of traditional fax over the Internet.
Does not integrate well with Internet messaging.
(c) Fax service over IP (IPP/FAX, "session mode" I-fax):
I see this as the best current approach to provide a truly fax like
service
that can integrate well with Internet messaging.
I think that (a) and (c) can be developed in the way you suggest, adding
further semantics in a way that is common to both.
But I find it difficult to see how (b) fits into this larger picture. One
advantage I do see with approach (b) within this framework is T.30 fax
onramp/offramp flexibility: it allows us to separate the Ifax/T.30
interface from the IP/GSTN interface.
#g
------------
Graham Klyne
(GK@ACM.ORG)