W3C home > Mailing lists > Public > semantic-web@w3.org > April 2015

Re: RIF a failure? (WAS: Question on dereferenceable RIF rules)

From: Paul Gearon <gearon@gmail.com>
Date: Thu, 9 Apr 2015 13:57:48 -0400
Message-ID: <CAGZNPFk6KE3xVpsoKtw5m+QWAH+2j-oagtp6fGTvBB+BnkfhvQ@mail.gmail.com>
To: Paul Tyson <phtyson@sbcglobal.net>
Cc: Dave Reynolds <dave.e.reynolds@gmail.com>, Semantic Web <semantic-web@w3.org>
On Thu, Apr 9, 2015 at 9:59 AM, Paul Tyson <phtyson@sbcglobal.net> wrote:

> What am I missing that hundreds of other system architects get who
> aren't making this choice?
>

In principle, I agree with you, but in practice I can see why people don't
use it.

The first thing that comes to mind is that much of the time rules are
written for a given system, and not for interchange. Rule descriptions for
one system often won't make sense in the context of another system. So that
removes one incentive to use RIF.

Also, the syntax of RIF entails a lot of overhead. The XML and RDF
representations are rather verbose, and not easily written. The
presentation syntax is supposedly for display purposes only, though given
the shortcomings of the formal syntaxes we implemented a parser for this,
as it was the easiest syntax to read and write. Even then, there are
elements that are more verbose than needed.

Technically, the semantics of rule systems are usually the same, regardless
of syntax, so using RIF shouldn't matter, and the other benefits it brings
would seem to make it a good choice. But simple rules come out being more
verbose than other rule languages that I've used.

I really can't say, but I'm guessing that for others who have looked to
implement a rule system have had similar thoughts to the above.

Regards,
Paul Gearon
Received on Thursday, 9 April 2015 17:58:16 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:42:59 UTC