text from naming breakout

NAMING-BREAKOUT-PROPOSAL-1: RIF will use URIs (IRIs) in the style of RDF
and OWL, using them to identify at least: predicates, functions,
datatypes, constants (OWL individuals, not literal values), rules, and
rulesets.  Translators to and from languages which do not use URIs as
names will need use namespace prefixes or other name-mapping system. 

NAMING-BREAKOUT-PROPOSAL-2: RIF does not mandate any particular mapping
scheme. It will provide a metadata vocabulary to allow translators to
record the mapping if they choose to.  For example, in rule systems
where the predicates refer to fields of java objects the metadata
annotations might be used to carry the java fully-qualified classname
and field name.  For RIF these are purely annotations, any use of them
for carrying semantic information steps outside of RIF.  This metadata
should be extensible in the usual RDF way. 

ROUNDTRIP ISSUE: When you round-trip a ruleset through a RIF dialect in
which it is covered, you are guaranteed to get back a ruleset with
identical semantics. But what about the aspects which are not part of
the semantics of the ruleset, such as variable names, whether a variable
is considered "anonymous", rule names, and ordering in languages where
ordering affects performance but not semantics?

Received on Saturday, 4 November 2006 22:53:51 UTC