> From [log in to unmask] Fri Nov 29 04:14 PST 1996
> > Two element qualifiers that are almost completely undefined in the original
> > DC specification are "Type" and "Identifier".  There's one lousy sentence:
> > 
> >     "In the example below, it is assumed that the sub-elements Type
> >     and Identifier have been defined." ... [ 4 examples appear also ]
> 
> Took me a while to work out what was going on here, as it seemed to 
> be referring to a fairly 'old' implementation, and I had to try hard 
> to remember that far back...! [never did find John's guilty line in 
> the original workshop report, either...  :-(  )

I found the quote in

http://www.oclc.org:5046/oclc/research/conferences/metadata/dublin_core_report.html

I don't know what you mean by old "implementation".  Perhaps you mean the
original DC workshop report, the DC spec that several groups have been
planning projects around?  We could use some names for the various
DC-derived metadata sets being talked about:  the Newcastle set,
the User Guide set, the original set, etc.

> > This is simply not enough to work with.  The User Guide Group was unable
> > to describe them without making too many second guesses as to how they
> > interacted with the Role qualifier and other elements in general, and so
> > it dropped them.  Pending meta2's formulation of clearer term definitions
> > and goals, these qualifiers needlessly complicate the DC and are too hard
> > to explain, so I propose to merge Type's function into Role and omit
> > any mention of Type and Identifier qualifiers from the DC spec.
> 
> 
> If I understand this correctly, I DEFINITELY disagree.
> 
> A lot of the current use of DC (certainly in the UK anyway) is in 
> terms of headers slapped into HTML documents by tools like 
> http://www.ncl.ac.uk/~napm1/dublin_core/sp_meta_creator.html.
> 
> These tools, and people doing it by hand, are making a HUGE use of 
> the TYPE qualifier;
> 
> <META NAME="DC.author" CONTENT="(TYPE=phone) +44 (0)1904 43 3954">
> <META NAME="DC.author" CONTENT="(TYPE=fax) +44 (0)1904 43 3939">
> etc...
> 
> The suggestion to lump this in with ROLE simply doesn't make any 
> sense, as how is being a telephone a role for an author? However, a 
> telephone number is a TYPE of detail about an author...

Indeed, how is being a telephone a TYPE for an author? (your phraseology)

That aside, I would never suggest that Role is an appropriate place to
indicate a phone number.  I do agree with you that the original DC spec
has no useful place to put an author's phone number.  This is a failing
of that spec.  I think a qualifier name more appropriate than Type for
this purpose (namely, identifying subelements), would be something like
"Sub".  There are, however, a number of other strategies to consider
for dealing with subelement names that don't involve a Type or other
qualifier at all, but that is definitely not a subject for this thread.

> Indeed, _I_ don't use ROLE at all, and haven't seen it on many recent 
> pages anywhere I've looked. Perhaps ROLE is the one to be ditched?

I'd guess that ROLE has too many adherents to be ditched.

> IDENTIFIER can, as far as I can see, certainly go, but we do need a 
> sensible use of the SCHEMA and TYPE qualifiers such as (for example);

There is no SCHEMA qualifier.  I take it you mean SCHEME qualifier.

> <META NAME="DC.date" CONTENT="(TYPE=current) (SCHEME=ISO31) 1996-11-28">
> <LINK REL=SCHEMA.dc HREF="http://purl.org/metadata/dublin_core/elements#date">
> <LINK REL=SCHEMA.iso31 REFERENCE="ISO 31-1:1992 Quantities & units -- 
> Part 1: space &  time">

Here's my previous message's point:

   "... so I propose to merge Type's function into Role and omit
   any mention of Type and Identifier qualifiers from the DC spec."

> Sorry if I've misunderstood, but the message certainly seemed to ME 
> to play up the importance of ROLE and IDENTIFIER, while denegrating 
> (the very useful) TYPE and ignoring SCHEMA altogether...

I'm glad you think I ignored SCHEMA and SCHEME, because they're unrelated
to any point I was making, but I'm baffled that you interpret "omit any
mention of ... Identifier" as playing up the importance of IDENTIFIER.

I think your real issue is that I ignored the work your group is doing
with the TYPE qualifier.  This was not meant in any way to trivialize it;
in fact, the functionalty you're trying to achieve is important to me,
even if I question the method you've chosen.  The reason I ignored your
Type specification is that the closest thing the DC has to an official
specification is, in the absence of any blessing from Stuart, the original
DC workshop report.

-John