instead of "Pure Prolog"

When I originally proposed the requirement "RIF Core must cover Pure
Prolog" [1] I thought the term "Pure Prolog" was better defined and
somewhat less inclusive than it is [2].

>From WG discussions it seems like the right term is probably "Sequential
Horn Clauses with Prolog Syntax", which I propose we abbreviate to
"SH-Prolog" (SH for Sequential Horn, of course).   So the requirements is
now

   RIF Core must cover SH-Prolog

or 

   RIF must cover SH-Prolog

depending on how we're handling phasing.  In general, I'd like to see
requirement titles stay short, using created terms like this if needed.
Any verbose requirement title is likely to contain phrases we'll be
saying and writing a lot.


How does SH-Prolog differ from ISO Prolog (for which none of us have the
spec), or SWI-Prolog or XSB, etc?  Basically it's the subset without any
built ins.  I'm not exactly sure how to define that.  That is, I don't
know if this rule:

    foo :-
       halt.

should be syntactically forbidden from SH-Prolog, or just have different
semantics from those in real prologs.  So there's more work to do in
making a spec, some test cases, and maybe a validator/interpretor or
something.  This work that would be a starting point for people trying
to really use RIF with real Prologs some day.

      -- Sandro

(I think this e-mail completes actions 3 and 4.)

[1] http://www.w3.org/2005/rules/wg/wiki/RIF_Core_must_cover_pure_Prolog
[2] http://www.w3.org/2005/rules/wg/wiki/Pure_Prolog

Received on Monday, 8 May 2006 22:46:51 UTC