Now, for those other comments...

[John Kunze wrote:]
> You are replying to a brand new thread, so I take it you think that the
> message itself wandered.  It was concerned entirely with DC audiences

Hmm... well, my perspective was that this was a new e-mail-conversation-thread 
on an ageing topic of "what the User Guide writing team wants to 'fix' in 
Dublin Core", and it was the topic that I felt was wandering out of the
space of this group.

> I agree completely.  The four phases identify the broad activity that will
> be occupying the different kinds of DC "users".  Unless I'm very mistaken,
> we imagine that this usage is limited to general purpose or DC-aware tools,
> such as word processors, text editors, robots, search engines, and browsers.
> 
> While designing these tools is clearly out of scope for meta2, designing
> the DC with these phases and tools in mind is entirely appropriate.

Yes -- but it is a delicate balance to achieve the right proportioning
of "understanding the needs of the space" and "attempting to completely
define the space".

> > The issues you are getting at, I believe, have more to do with ensuring
> > that people will provide _"good"_ metadata -- data that will be accurate
> > and useful in a number of contexts.  This may be more of a problem with
> > trying to write a User Guide than with the DC itself.
> 
> The Guide is concerned with "good" metadata, and I'm interested in your
> suggestion that the DC is not.  But that's a topic for a separate thread
> since my message was meant to be only about foundation issues.

I am questioning your foundation issues.

As I believe I said in my earlier message, a User Guide is no doubt
concerned with detailing how to provide "good metadata".  And, as you
have pointed out, this is application- and user- specific.  

My perspective, as a person involved with building indexers and services,
is that I want a _structure_ into which I can put/have people put 
resource descriptions -- according to some agreed-upon conventions.   Those
conventions may be agreed upon by Bunyip and some enterprise, and we may create
metadata that you consider _bad_ -- but it works for us.  Or, those
conventions may be according to rules proposed by some librarian community
dedicated to getting the Web properly indexed.    Or, a given resource
may have both, for different purposes (which is why I object to embedding
metadata in an object, although it is often the most convenient mechanism).

So my point was that "good" metadata is in the eye of the beholder, and
is therefore beyond the scope of the work of this group.  

How I have understood your position is that you want to use DublinCore
to ensure that people can tag their resources _Correctly_, and the 
global resource index will be built.  

_That_ is what I object to, because it seems to me that many of your proposals
for how Dublin Core should be "fixed" stem from having one perspective on
how it should be used.  I understand that you are attempting to make that
perspective as general as possible, but it is still one perspective.

> Perhaps this is what you meant by wandering, that DC discussion should
> focus on putting on the finishing touches.  I happen to believe that's
> an overly optimistic assessment.

I don't suggest that DC is perfect or beyond improvement.  I think some
good proposals have been made for declaring the status of what exists
today (calling it 1.0) and moving discussions to future versions.    

Leslie.

-- 

----------------------------------------------------------------------------

  "The light at the end of the tunnel             Leslie Daigle
     is often the blindingly obvious..."          Vice President, Research
                                                  Bunyip Information Systems
                      -- ThinkingCat              (514) 875-8611
                                                  [log in to unmask]
----------------------------------------------------------------------------