RE: What to do with CryptoBinary?

Brian LaMacchia writes:

>> CryptoBinary and base64Binary are not exactly 
>> equivalent -- there are further restrictions 
>> on a CryptoBinary because it is a representation of
>> a single bignum

This is exactly the sort of thing I had in mind when I suggested that 
higher level application or library software could key on the 
"CryptoBinary" type name to do the additional validation. 

Indeed, I think that some have failed to notice that the schemas design in 
this respect anticipates layered families of  type libraries and 
associated schema processors.  Core compatibility is what we describe in 
the schema specification, and everyone must implement it compatibly.  On 
the other hand, someone might want to propose a higher level standard for 
math-oriented processors that will additionally recognize and validate a 
library of math types including perhaps a "PrimeNumber" type (which might 
be declared in the schema as merely a nonnegative integer).  When 
presented with the same schema and document, an ordinary XML schema 
processor would validate the content as an integer, and would correctly 
report its type name as PrimeNumber.

Similarly, a processor, library, or application with knowledge of digital 
signatures could recognize and do the additional validation of 
"CryptoBinary".  Processors which merely conform to the XML schemas 
recommendation would correctly validate the content as base64Binary and 
would correctly report the type of any such element as "CryptoBinary".

So, I think this further supports the suggestion that the additional type 
be retained. 

------------------------------------------------------------------------
Noah Mendelsohn                                    Voice: 1-617-693-4036
Lotus Development Corp.                            Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------------

Received on Monday, 26 March 2001 23:36:30 UTC