On Wed, Dec 01, 2004 at 11:39:12PM +0000, Andy Powell wrote:
> A revised, and hopefully final, version of the Abstract Model working
> draft is at
>
>   http://www.ukoln.ac.uk/metadata/dcmi/abstract-model/

It's generally looking great!  Some mostly minor comments follow...

> - It would be better if we modelled 'syntax encoding scheme URI' and
> 'vocabulary encoding scheme URI' as separate entities in the model.

Figure 2 gets this wrong - presumably a typo
(s/Vocabulary/Syntax/ in lower left corner).

> - Does the model get the definitions of simple DC and qualified DC right?

I like the implicit acknowledgement that there is no firm
consensus on this.  However, the text says:

    in general terms, the phrase 'simple DC' is used to refer
    to DC metadata that does not make any use of encoding
    schemes and element refinements.

But then the table says that both types of Value Dumb-down
result in a "new value string".  If this is the case, then
the text above should perhaps say something like:

    in general terms, the phrase 'simple DC' is used to refer
    to DC metadata that does not make any use of encoding
    schemes and element refinements and has string values.

> Of these, there was quite a long discussion around the meanings of simple
> DC and qualified DC. No consensus was reached. We therefore agreed to
> remove definitions of these terms from the Abstract Model.

I think that the way you mention them without defining them
too precisely is fine -- and better than not mentioning them
at all.

A few other points:

-- It is not crystal clear that "descriptions" is shorthand for
   "DCMI metadata descriptions".  This could be fixed by adding
   "descriptions" in italics and parentheses when they are first
   mentioned:

        The abstract model of the _resources_ being described by
        DCMI metadata descriptions (_descriptions_) is as follows:

-- It was not clear to me why the following point is necessary:

        Each _property_URI_ may be repeated in multiple _statements_.

   It seems vaguely confusing... -- either requiring more
   explanation, or (I suspect) unnecessary.

-- I'm slightly bothered by the wording that a string or a
   rich value "is a representation of" the resource.  Maybe
   I'm reading "representation of" a bit too literally, but
   to me the words evoke something like a visual depiction.
   For example, a portrait photograph can be "a representation
   of" Andy Powell.  For me, a wording like the following
   would not have the same associations:

        Each... string stands for the resource...

        Each rich value... is some text... that stands for the
        resource...

-- I find myself stumbling over the sentence: "Note that
   where the _resource_ is the _value_ of a _property_, the
   _class_ is referred to as a _vocabulary_encoding_scheme_."
   Coming right after a sentence talking about resources that
   are subjects of DCMI metadata descriptions, one can read
   it as saying that "when the subject of a description is
   the value of a property...".  Maybe break into a separate
   point and clarify the wording.

-- I do not understand the reference to DC XML namespaces in:
   "Note that software should make use of the DCMI term
   declarations represented in RDF schema language and the
   DC XML namespaces to automate...".

-- I do not understand the sentence: "Note that the abstract
   model does not indicate that the combination of a DCTERMS
   URI syntax encoding scheme with a value string implies
   a value URI or resource URI."  I think you mean to say
   something like:

       Note that the abstract model does not indicate that a
       value string associated with the syntax encoding scheme
       http://purl.org/dc/terms/URI does not imply a value URI.

   Not sure about the reference to "resource URI" here... --
   do you mean a "value URI" that is then also a "resource URI"
   of a "related description"?

-- Instead of "1:1 principle", maybe "one-to-one principle"?

-- "The value of the DC Date Property is a point in time" -- can
   a range of dates be considered a point in time?

I may have one or two more comments in awhile.

Tom


--
Dr. Thomas Baker                        [log in to unmask]
Institutszentrum Schloss Birlinghoven         mobile +49-160-9664-2129
Fraunhofer-Gesellschaft                          work +49-30-8109-9027
53754 Sankt Augustin, Germany                    fax +49-2241-144-2352
Personal email: [log in to unmask]