Response to DC2

From: Dan Connolly <connolly@w3.org>
Date: Mon, 05 May 2008 14:35:07 -0500
To: public-rif-comments@w3.org
Message-Id: <1210016107.4651.363.camel@pav.lan>


I noted these two bits of the RIF syntax:

"Each predicate and function symbol has precisely one arity"

"A well-formed term is one that occurs in a well-formed set of fomulas."

  -- http://www.w3.org/TR/rif-bld/

Those seem to be not web-wide definitions, but definitions
that just apply to one file or something. Otherwise,
to take an arbitrary example, the function symbol ABC:
what is its arity?

The context-sensitivity of those definitions seems
to conflict with the requirement to be able to
merge rule sets:

"4.2.12 Merge Rule Sets 
RIF should support the ability to merge rule sets. "
http://www.w3.org/2005/rules/wiki/UCR#Merge_Rule_Sets

If ABC has arity 2 in one rule set and arity 3 in
another, what happens when those rule sets are merged?

Is it worthwhile making the requirement more precise as follows?

  any collection of well-formed RIF formulas is itself well-formed

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
gpg D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Dan,

Thanks for your comment. No, this case does not violate the requirement but supports it. Predicates and functions have one arity, and if URI's are used for them then their arity must hold across documents. Merging them would correctly result in an error.

-RIFWG