> > ----- Forwarded message from Leslie Daigle > <leslie@thinkingcat.com> ----- > > > > I'd have some concern about: what happens when there is discrepancy > > between the digit and textual representations (i.e., it's an > > error). > > ITU-T Rec. X.660 series states that the number form of an OID > is mandatory, > but not the name form. Hence the number form takes precedence. It is important to recognize that the Name 'form' of an ASN.1 OID is not normative and is not transmitted in the ASN.1 encoding. As a result working groups defining OIDS do not in general take particular care or even any care whatsoever to ensure that the name form of the OID is unique and unambiguous. As with any spec half the discussion on any given ASN.1 schema is quite likely to involve discussion on the naming of variables and identifiers. I have participated in working groups where OIDs have been renamed from one draft to another on several occasions. I have done the same thing myself. The number form is generaly treaded with considerable respect and care however. I am absolutely opposed to attempts to provide 'clarity' that introduce ambiguity and are almost certain to introduce errors and complexity into programs. Using the name form of an OID for a purpose that was not intended falls into that category in my opinion. The name form is unnecessary and will introduce confusion complexity and error into applications. It should be rejected. Michael's original proposal is fine as is. Phill
Received on Wednesday, 29 November 2000 16:14:04 UTC