W3C home > Mailing lists > Public > www-webont-wg@w3.org > January 2003

Re: apologies and TEST update

From: Jeremy Carroll <jjc@hpl.hp.com>
Date: Mon, 6 Jan 2003 12:24:36 +0100
To: www-webont-wg@w3.org
Message-Id: <200301061224.36376.jjc@hpl.hp.com>


DanC wrote:
> Please remove all the stuff about "OWL systems" "OWL reasoners"
> and whatnot. Please let's just specify the owl vocabulary
> and its semantics, and leave specification of software systems
> out of our spec.

> Rationale:

> Our charter doesn't require us to do it;
>   http://www.w3.org/2001/sw/WebOnt/charter#L778
> we have little experience doing it. Based on my experience
> with HTML, XML, and XML Schema specifications,
> I think it takes a lot of time and provides
> relatively little value.

I would like to hear other people on this.
I had previously (mis?)heard WG members wanting some clear statement of what 
it means to be OWL compliant.

I think the text I am proposing [1] tries to be minimalist, and is *not* 
setting us up to do conformance testing, or even to have a conformance test 
suite.

> We need enough implementation experience to convince
> ourselves (and our reviewers and The Director) that
> the design of the vocabulary/language is useful
> and correct. We don't need to do software conformance
> testing.

I agree with the last sentence.

I note that one of our external reviews [2] dsicusses:
[[
This has a more pragmatic impact in the development of tools for
    managing such ontologies; the current state of tool
    interoperability (for DAML+OIL) is lamentable.  Tool manufacturers
    need to have an easily implementable, stable spec to be able to
    make tools that will interoperate.  Can OWL/Full provide this,
    while allowing such full expressivity of metaclasses, arbitrary
    set operations, etc.?
]]
I believe that without a clear statement of what an OWL/DL and OWL/Lite tools 
are meant to do then we will in practice end up with OWL Full diluted to 
different strengths and in different, non-interoperable, ways. While I 
believe that the developers of Jena, cwm, and euler may find this more 
attractive than having to implement stuff they currently don't have a handle 
on, I do not believe that that was WG consensus.

i.e. as HP rep I would be happy to not have a conformance section, since the 
only sort of conformance statement that I believe the WG might agree to is 
one that HP will face some non-trivial cost in meeting. As editor, I am not 
sure that I would be doing my job if I simply strike the conformance section.

Also as HP rep, I have had a clear message from the HP community that I should 
be prioritising the overall success of OWL over any specific HP concerns.

Jeremy

[1] 
http://lists.w3.org/Archives/Public/www-archive/2002Dec/att-0091/01-main.jsp.html#conformance

[2]
http://lists.w3.org/Archives/Public/public-webont-comments/2002Dec/0009.html
Received on Monday, 6 January 2003 06:27:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:56 GMT