W3C home > Mailing lists > Public > public-rif-wg@w3.org > May 2008

Re: Any proposals for a Concise RIF XML Syntax?

From: Sandro Hawke <sandro@w3.org>
Date: Tue, 20 May 2008 10:25:12 -0400
To: Gary Hallmark <gary.hallmark@oracle.com>
Cc: public-rif-wg@w3.org
Message-ID: <22965.1211293512@ubuhebe>


> I also had a suggestion, which boiled down to this:
> 
> 1. start with fully striped
> 2. when the Class stripe is uniquely determined by its enclosing 
> property stripe, then omit the Class stripe.
> 
> XML Schema provides the "complexType" construct to implement #2
> This is similar in strongly typed OO languages like java -- One declares 
> the class information, e.g
> 
> Class C1 { C2 c2; }
> Class C2 { C3 c3; }
> 
> Then one constructs references such as c1.c2.c3, not c1.C1.c2.C2.c3.C3. 

The problem I see with this (and it's probably a problem with all stripe
skipping) is how it interacts with extensibility.  I imagine people
adding extensions by creating "subclasses".  If the name of that
subclass isn't looked at -- or isn't even serialized! -- then there's no
way to know the meaning.

For a rough example, an extension of NAF (as Axel once talked about)
into StableModelsNAF and WellFoundedNAF, would likely just change the
syntax by replacing a <NAF> class tag with <SMNAF> or something.
Obviously, if we ignore/omit the class stripe, you'd get the wrong
semantics.    Of course, one could force the class name into the
property name, but...    Hmmm.

    -- Sandro

> Sandro Hawke wrote:
> > Occasionally people talk about making the XML syntax for RIF more terse
> > and easy to read by humans.   I remember Jos and Hassan saying things in
> > this direction fairly recently.
> >
> > A long time ago we went through a suggestion I had for this -- I
> > proposed some rules for when you could skip a stripe as redundant -- but
> > we decided against that (with me concurring).
> >
> > I wonder if there are any other proposals for a concise RIF XML syntax?
> > If so, they'd need to come forward very very soon.  (Some would say it's
> > too late already, but...  *shrug*)
> >
> > I started a table where one could do an ad hoc version of this:
> >     http://www.w3.org/2005/rules/wiki/BLD_Syntax_Table
> >
> > Please edit at will, with comments here, if this is something you're
> > interested in.
> >
> >      -- Sandro
> >
> > [1] http://www.w3.org/2007/01/30-rif-minutes.html
> >     http://www.w3.org/2005/rules/wg/wiki/Arch/XML_Syntax_Issues_1
> >
> >   
Received on Tuesday, 20 May 2008 14:27:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:33:49 GMT