Re: Proposed response to Graham Klyne

On Thu, 19 Jun 2003, Jim Hendler wrote:

> I'm okay with this one except for one thing -- you have
>
> At 5:57 PM +0100 6/19/03, Ian Horrocks wrote:
> >This document only provides definitions of various parts of OWL.  Turning
> >these definitions into effective procedures is a task for implementors.
>
> which is begging for someone to say "you need a long CR for this" --
> however, our web page has a pointer to Sean's validator which proves
> this can be implemented - so why don't we say
>
> "This document only provides definitions of various parts of OWL.  Turning
> these definitions into effective procedures is a task for
> implementors (c.f.the OWL Species Validator, available from the WG
> web page or at [5]), which is such an implementation).
>
>
> [5] http://phoebus.cs.man.ac.uk:9999/OWL/Validator"

As anyone who knows me will confirm, I'm always the first to diss my own
implementations. Having said that, In this case I am a little wary of
using this as a justification here (if that is in fact what is happening).
I'm *pretty sure* I'm getting most of my implementation "right", but it
would not at all surprise me if there are places where it's a little
flaky, (for example in areas like imports or data ranges, should anyone
wish to probe it :-).

I would agree that it shows that one can make a good stab at implementing
a parser (where I mean here something that turns the RDF into some other
structure and tries to do some validation on the way), but I wouldn't
claim that it shows I know how to tackle the whole language.

This is perhaps nit-picking, but I don't want it to appear like *I'm*
claiming I've built a 100% correct OWL validator. Because I don't think I
have (yet :-)

Cheers,

	Sean

-- 
Sean Bechhofer
seanb@cs.man.ac.uk
http://www.cs.man.ac.uk/~seanb

Received on Friday, 20 June 2003 05:15:23 UTC