RE: Comment on * DRAFT * Rules Working Group Charter 1.60

Adrian,
 
This makes sense. I have only one question. To what extent to we need to
make decisions in this first phase to arrive at a place where we can do
this in the second phase. For example, I believe someone brought up the
question of whether this should be defined as a FOL. Does this need to
be decided in phase 1 or can it safely be deferred until later?
 
Just a thought.
 
James

________________________________

From: Adrian Walker [mailto:adrianw@snet.net] 
Sent: Tuesday, August 23, 2005 10:56 AM
To: Lynn, James (Software Escalations)
Cc: public-rule-workshop-discuss@w3.org; www-rdf-rules@w3.org; Dieter
Fensel; edbark@nist.gov; Michael Kifer; Jim Hendler
Subject: RE: Comment on * DRAFT * Rules Working Group Charter 1.60 


James --

At 10:00 AM 8/23/2005 -0400, you wrote:


	Adrian,
	
	Certainly your points are valid. But I wonder if it wouldn't be
	beneficial to come up with a core language which would serve as
an
	easily translatable subset of any vendor's language. There would
be a
	number of ways this could be used, e.g., creating modules of
rules which
	conform to the core restrictions, or even just grouping or
tagging the
	rules which conform to allow automatic translation of some part
of a
	vendor's rules. 


I think we only differ on timing.  

The proposal in [1] is to start with a simple recommendation that works.


That (or some other form of input-output level interoperation) should
get the various rule vendors and research systems all interoperating,
without getting into "engine wars" about what should be derivable from a
set, or ordered list, of rules and facts.  

The scenario is then that we gain valuable practical experience from
rules systems, and OWL inferencing, that interoperate cleanly and
accountably on the SW.  One of the issues that can be clarified
practically in this way is monotonic negation vs NAF & SNAF.  Hopefully
this approach also meets Dieter's concerns about the current draft of
the charter.

A later recommendation could be along the lines that you suggest,
informed by the practical experience enabled by the earlier one.  It
would probably also have to specify abstractly the behavior of a few
carefully selected engines.

                                Cheers, 

                                        -- Adrian
 
[1]  http://www.w3.org/2004/12/rules-ws/paper/19/



INTERNET BUSINESS LOGIC (R)
Online at www <http://www.reengineeringllc.com/> .reengineeringllc.com
<http://www.reengineeringllc.com/>  

Adrian Walker
Reengineering LLC
PO Box 1412
Bristol
CT 06011-1412 USA

Phone: USA 860 583 9677
Cell:    USA  860 830 2085
Fax:    USA  860 314 1029




	-----Original Message-----
	From: www-rdf-rules-request@w3.org
[mailto:www-rdf-rules-request@w3.org]
	On Behalf Of Adrian Walker
	Sent: Tuesday, August 23, 2005 7:49 AM
	To: Sandro Hawke
	Cc: public-rule-workshop-discuss@w3.org; www-rdf-rules@w3.org;
Dieter
	Fensel; edbark@nist.gov; Michael Kifer
	Subject: Re: Comment on * DRAFT * Rules Working Group Charter
1.60 
	
	
	Sandro --
	
	At 01:39 AM 8/23/2005 -0400, you wrote:
	>We're talking about one language which is a superset of many of
the 
	>common languages, so it can be used as an interlingua.  You
translate 
	>your ruleset into it, and if you can translate it back out into
another
	
	>vendor's language (because it has enough features), your rules
will 
	>mean the same thing.
	
	Unfortunately, it makes no sense to translate rulesets from one
vendor's
	language into another [1].
	
	This is because the results you get from running rules depend on
which
	variation of which kind of engine is used.  So, your rules will
	regrettably
	*not* "mean the same thing" in another vendor's system.
	
	One can visualize acres of list discussions of the form "were
you using
	Jena 1.5.4.3.2a with the 5.2.1.3.4 fix to backward chaining
engine with
	mysql semi-persistence to get the result that George Washington
is the
	current president of the US?"
	
	Why not start small, with something that works?  Recommend a way
in
	which diverse rule systems can interoperate at the input-output
level,
	e.g. as in [1].
	
	Once you get that going, there will be practical reasons for a
few
	engine behaviors (aka "semantics") to emerge as ones that can
usefully
	be the subject of the next recommendation.
	
	HTH,          Cheers,  -- Adrian
	
	[1]  http://www.w3.org/2004/12/rules-ws/paper/19/
	
	
	
	
	INTERNET BUSINESS LOGIC (R)
	Online at www.reengineeringllc.com
<http://www.reengineeringllc.com/> 
	
	Adrian Walker
	Reengineering LLC
	PO Box 1412
	Bristol
	CT 06011-1412 USA
	
	Phone: USA 860 583 9677
	Cell:    USA  860 830 2085
	Fax:    USA  860 314 1029