Commentary by Mark Wahl, CISA
Organizing principles for identity systems:
Identity Schema Value Syntax Restrictions at Data Sharing Summit Day 2 (20070908)
During the second day of the DataSharingSummit the discussion of the initial set of identity schema metadata properties was continued with a focus on the metadata for an attribute to express restrictions on the value syntax.
Use cases for value syntax restrictions include
- validating attribute values being entered by an end user in a form,
- validating of identity attribute values exchanged between services,
- transformation of identity attribute values exchanged between services, and
- display control: improving the formatting of the display of values based on known constraints of the values.
For an initial specification of the minimal set of identity schema metadata properties, the low-hanging fruit is providing human and machine readable descriptions of the contents of simple, string-valued attributes in order to support validation of input based on patterns, and information for developers to learn more about how to format values of the attribute. There is no one format for encoding this description. Instead, there are several options which might be viable: regular expressions, XBNF and XML schema formats (of which there are several).
XBNF is a proposal by Marty Schlieff for encouraging reuse between formal language specifications of formal languages derived from BNF, which uses XRI "dollar notation" to provide unique names to symbols. A specification for this is still under development.
The discussion also highlighted the value in adding to the set of metadata test case values as specialized forms of example values which contain edge cases.