All,

I guess you could add to this list:

Events lasting varying periods
Works of art picked up, reworked and put down again by their creators

And so on.

DWG should perhaps propose a start/end specification using something like
DCMI Period as the framework?

Ian Williams

-----Original Message-----
From: DCMI Date Working Group [mailto:[log in to unmask]] On Behalf Of
Childress,Eric
Sent: 20 July 2004 15:19
To: [log in to unmask]
Subject: Re: date and date ranges

Pete,
        Great questions.

1. My memory of the discussions of the original Date WG are fading, but if
memory serves there was little argument about "date" being read as either an
instant or a range.  I vaguely recall some discussion about "created"
perhaps presumptively being assumed to be a single instant, but this was not
universally held nor ultimately expresed as special case in the definition.
I would advance the argument that an "event" by definition has two terminal
points.  So to my mind date created can in fact be expressed as a range.
It's not difficult to think of cases where the creation event elapses more
than 24 hours, or even several years (e.g., a collection of correspondence)
or possibly quite long periods (e.g., fossilization).

2. The W3CDTF standard does or does not allow for the expression of date
ranges.  Determining which is the case is on the workplan of the current
Date WG.  The examples offered seem to be entirely for expression of a
single instance, but the thrust of the document seems to be the adoption of
a particular syntax (among several options in ISO 8601) to assure that
century and year are always explicitedly recorded.

Other insights?

Eric

> Eric Childress
> Consulting Project Specialist
> OCLC Research {http://www.oclc.org/research}
> OCLC Online Computer Library Center, Inc.
> 6565 Frantz Rd., Dublin, OH 43017 USA
> US: (800) 848-5878 or (614) 764-6000
> Fax: (614) 718-7361 email: [log in to unmask]
>
>
>
>
-----Original Message-----
From: Pete Johnston [mailto:[log in to unmask]]
Sent: Tuesday, July 20, 2004 8:09 AM
To: [log in to unmask]
Subject: dc:date and date ranges


Hi all,

I should have thought this through earlier, but a discussion on
dc-collections focused my mind....

Looking at the definition of dc:date, it seems to me that, strictly
speaking, a value of dc:date ("A date associated with an event in the
life cycle of the resource.") _is_ supposed to be a single date, _not_ a
date range. That definition seems pretty unambiguous.

On those grounds, it seems to me that a date range should not be used as
the value of a property which is declared to be a subproperty of
dc:date.

But the definitions of some existing refinements of dc:date in the
dcterms vocaulary do explicitly endorse the use of date ranges as values
e.g.

dcterms:available = Date (often a range) that the resource will become
or did become available.
dcterms:valid = Date (often a range) of validity of a resource.

And since

dcterms:valid rdfs:subPropertyOf dc:date .
my:a dcterms:valid "1999/2000" .

implies

my:a dc:date "1999/2000" .

that implies that date ranges are acceptable values for dc:date.

Is there not a contradiction in these existing definitions here?

Cheers

Pete
-------
Pete Johnston
Research Officer (Interoperability)
UKOLN, University of Bath, Bath BA2 7AY, UK
tel: +44 (0)1225 383619    fax: +44 (0)1225 386838
mailto:[log in to unmask]
http://www.ukoln.ac.uk/ukoln/staff/p.johnston/