W3C home > Mailing lists > Public > public-esw-thes@w3.org > February 2005

Re: [PORT] progress update

From: Charles McCathieNevile <charles@sidar.org>
Date: Fri, 04 Feb 2005 14:10:02 +0100
To: "Miles, AJ (Alistair)" <A.J.Miles@rl.ac.uk>, "Dan Brickley" <danbri@w3.org>
Cc: public-esw-thes@w3.org
Message-ID: <opslob60ujw5l938@saturne>

On Fri, 4 Feb 2005 12:22:38 -0000, Miles, AJ (Alistair)  
<A.J.Miles@rl.ac.uk> wrote:

> 1. What to do about the SKOS Core 'Collections' vocab
...
> I was thinking to change the names of these properties, do you think  
> this is a good idea?  What could we call them?

Please don't change the URIs. Give them clearer labels and descriptions if  
you think the human-readable stuff currently doesn't help much.

(I have yet to meet a change of URI that struck me as anything but a bad  
idea :-(

> Stella has suggested 'skos:Array' instead of 'skos:Collection', because  
> the new BS thesaurus standard  calls these things 'arrays of concepts',  
> and it would be nice to be in line with BS as much as poss.  But I'm  
> concerned that 'array' has a specific meaning to compscis which is a bit  
> different.  For 'arrays of concepts' sometimes the ordering is  
> meaningful, and sometimes it isn't, whereas to compscis an 'array' is  
> always an ordered set.  Anyway, what do you think?

I think coining new URIs for something is a bad move, since it focusses  
attention on the wrong aspects of RDF (the syntax, rather than the model).  
Plus it means people need to keep collecting information about new names  
for old rope.

> 2. URIs in guide examples.  Currently all example concept URIs are of  
> the form http://www.example.com/something#concept - because this URI  
> form is a compromise between the folks who say you must use hash URIs,  
> and the folks that want each concept URI to be able to de-reference to a  
> different representation.  Tom B is concerned that this type of URI  
> looks exotic, and we should explain the choice.  I want to avoid  
> discussion of URI form in the SKOS Core guide as much as poss.

If you want to avoid discussion of URIs, just don't talk about them :-)

Seriously, URIs in RDF are effectively opaque unless someone decides to  
make their own URIs something else. Having seen a lot of the discussion,  
and thought fairly hard about it, I don't see much point anymore in  
worrying - just use a URI and keep using it.

> Also I wrote a new introduction, see  
> http://www.w3.org/2004/02/skos/core/guide/2005-01-25.html ... what do  
> you think?

I reckon I should have a look... :-)

cheers

Chaals


-- 
Charles McCathieNevile - Vice Presidente - Fundacion Sidar
charles@sidar.org                      http://www.sidar.org
     (chaals is available for consulting at the moment)
Received on Friday, 4 February 2005 13:17:11 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:38:53 GMT