comments on FLD, BLD, and Core

Dear all,

An elegant and solid framework for specifying RIF dialects in FLD, a
first general logical rule dialect in BLD, and a subset of BLD and PRD
(Production rules) constitute a set of documents that can lead the
rule designed/implementor in coming up with own dialects in a clear
and concise manner. It's only a matter of spreading the word now and
convincing communities like the LP community to use it.

Some minor comments on the different documents:

on "FLD: RIF Framework for Logic Dialects".

Section 2.2: It says "However, dialects cannot redefine the semantics
of the predefined quantifiers".  Does this mean that one cannot change
the domain of quantification in a dialect? For example, in LP instead
of a closed domain assumption (with the constants of the program) and
open domain assumption as in FOL.

Section 2.4: item 4: It is not immediately clear from this explanation
what the difference (besides a syntactical) between closed lists and
open lists is. All its arguments t1, ..., tm, t are terms (both in
closed lists and open lists).

Section 2.5: Def. Schema for an external term: Why is the order of the
names of the variables important. In other words, why are (?X ?Y; bla;
loc) and (?Y ?X; bla; loc) different?

on "BLD: RIF Basic Logic Dialect"

Section 2.4: Is there a reason why in an annotation (* id phi *) phi
is a frame formula (or a conjunction thereof) and not also allowed to
be an atomic formula?

On "Core: RIF Core Dialect"

Example 1 in "Core" and "BLD" are the same (i.e., both Core). It would
be better to have in BLD an example that is BLD but not Core.

--
best regards,
Stijn

-- 
http://www.kr.tuwien.ac.at/staff/heymans/priv/

Received on Thursday, 20 August 2009 08:32:48 UTC