comments on Editor's Draft of UCR

I worry that the informality of the use cases encourages reading much too
much into the power of the rules.  Words and Phrases like "will be
rejected", "must", "never", "for anonymity reasons", "can", "desired",
"assume", "believes", "process", "will", "required", "considered",
"recommended",   I suggest that a disclaimer along the lines of

	However, this informality can lead readers to the conclusion that
	rules can perform arbitrary actions in the real world.  This is not
	the case - the RIF WG has not yet decided on the ultimate power
	that rules will have.

just before Section 2.1.


I believe that many of the motivates links are not correct.

Motivations that I believe are not supported:

Compliance Model

	RIF must define a compliance model that will identify
	required/optional features. 

	not motivated by Use Case 2.1, Use Case 2.3, 


Coverage

	RIF must cover the set of languages identified in the Rulesystem
	Arrangement Framework.

	not motivated by Use Case 2.1, Use Case 2.2, Use Case 2.4, Use Case
	2.5, Use Case 2.6, Use Case 2.7, Use Case 2.9

	This is more of a problem with the wording of the Coverage CSF than
	anything else.  The Coverage CSF should probably be rewritten to
	something like

	RIF must cover a reasonably wide selection of rule sets.


Semantic tagging

	RIF must have a standard way to specify the *intended* [emphasis
	added] semantics (or semantics style) of the interchanged rule set
	in a RIF document.

	not motivated by Use Case 2.2, Use Case 2.3, Use Case 2.4, Use Case
	2.10 

 	unless, of course, all that it means is that RIF rules have to have
	a formal semantics.


Default behaviour

	RIF must specify at the appropriate level of detail the default
	behavior that is expected from a RIF compliant application that
	does not have the capability to process all or part of the rules
	described in a RIF document, or it must provide a way to specify
	such default behavior.

	not motivated by Use Case 2.2, Use Case 2.4, Use Case 2.5, Use Case
	2.6 

Embedded metadata
	
	RIF must support metadata such as author and rule name.

	not motivated by Use Case 2.2

OWL data

	RIF must cover OWL knowledge bases as data where compatible with
	Phase 1 semantics.

	not motivated by Use Case 2.4, Use Case 2.6



The pointers for "Different Intended Semantics" and "Limited number of
dialiects" do not work.

Use Case 2.5 mentions that it depends on labelling rules with tags
providing meta-data.  However, there does not appear to be any use of or
need for this ability withing the use case.

The wording about the Widescale Adoption goal is missing the influence of
the Alignment CSF.

The Alignment CSF is very strange in that it only mentions XML.  Is not
alignment of RIF with RDF and OWL also appropriate here?  Similarly, why
does not Aligment with the Semantic Web support Widescale adoption?

There appears to be some confusion between the CSF "Coverage" and the
requirement sometimes called "RAF coverage" and sometimes just called
"Coverage". 



The bottom line:

The connections between the Use Cases and the requirements is too vague.  I
suggest pulling the suspect ones.  If that leaves some requirements
unmotivated, then either this should be noted (with a comment something
like "This requirement is expected to be motivated by a revised version of
Use Case 2.x.") or removed for now.


Peter F. Patel-Schneider

Received on Monday, 26 June 2006 17:37:57 UTC