W3C home > Mailing lists > Public > public-sws-ig@w3.org > March 2004

Re: strawman concise syntax (was Re: [OWL-S] new IOPE example #1)

From: Bijan Parsia <bparsia@isr.umd.edu>
Date: Wed, 31 Mar 2004 11:09:50 -0500
Message-Id: <D66FDD91-832D-11D8-AEC8-0003939E0B44@isr.umd.edu>
Cc: public-sws-ig <public-sws-ig@w3.org>, David Martin <martin@AI.SRI.COM>
To: danny666@virgilio.it

On Mar 31, 2004, at 11:02 AM, Danny Ayers wrote:

> Bijan Parsia wrote:
>> I'm interesting to know if you think mixing RDF and some other syntax 
>> a la your proposal is better than having a complete drop in 
>> replacement. I'm inclined to the latter, myself.
> Hmm, must confess I'm in two minds. For true human-legibility a total 
> replacement certainly is better, but something along the lines of  my 
> suggestion would keep most of the key parts of the RDF 
> machine-readable in the interchange syntax, while the tasty part (the 
> interface definition itself) is human readable directly, machine 
> readable given a little translation. Of course pretty much any syntax 
> can be used locally, and translated at the end points, so it's not a 
> big deal in any case...

Yeah. One set of experiences to look at is XSLT & XQuery. Your 
suggestion is a bit like having the nice syntax for XPath. I tend to 
think that the mixes don't work out so great.

> The choice of a Lisp-like syntax is interesting.

This was just for developement, as Drew already had a few languages in 
this flavor.

> Who the target audience for this? If it's 'old school' krep folks, 
> then it's bound to do well. However if the aim is to get web 
> developers to use it, then I'd anticipate resistance, and would have 
> thought swapping the braces for angle brackets (a la OWL Presentation 
> Syntax) would be good politics ;-)

A future direction is develop a more Web Developer friendly syntax. 
I'll also note that the current syntax was developed with an eye to 
being directed mapped to the ontology, rather than have tons of 
convenience features. The latter is also a goal.

> But on both points, given that work that has already been done on the 
> surface syntax I'd lean towards the current approach.


I'll be interested in your reaction to the PE expression solution, 
since that allows for embedding non-xml based formula languages (like 
KIF). But there, the goal isn't readabilty, but accomodating code for 
which there isn't an already defined XML syntax.

Bijan Parsia.
Received on Wednesday, 31 March 2004 11:10:57 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:54:12 UTC