Dear all,

Though the Collection Description Profile is not going through
a formal public comment period (not required for profiles),
there is some substantive discussion happening on the
DC-COLLECTIONS list.  

Andrew, as shepherd, could you please track and summarize
the discussion in Colima?  For example, some of Douglas's
comments are directly relevant to the questions we want to
address in the review.

Tom

On Mon, Aug 14, 2006 at 05:53:59PM +1200, Douglas Campbell wrote:
> X-Mailer: Novell GroupWise Internet Agent 6.5.4
> Date: Mon, 14 Aug 2006 17:53:59 +1200
> Sender: DCMI Collection Description Group <[log in to unmask]>
> From: Douglas Campbell <[log in to unmask]>
> Subject: Re: New draft of DC CD AP (2006-08-01) available
> To: [log in to unmask]
> 
> I'm hoping to have a more in-depth look in the next week, but in case I don't I just wanted to pass on a couple of things I noticed, more of a cosmetic nature...
> 
> AP document:
> - Introduction - Collections and collection-level descriptions: A diagram could help here, eg. showing the hierarchic relationship of: catalogue -> record -> object.  We use the word collection for both "group of objects" and "group of records about those objects", which may be easier to comprehend if you can see them.  This is an introductory diagram separate to the entity-relationship diagramin the Data Model section.
> - Introduction - The DC AP section (starts "A DC application profile specifies..."): It might be helpful if this and the similar part in the AP Summary document was identical.  Wearing an implementer hat, because I had read the summary document first I found myself checking word for word in case there was some extra meaning/specification in this version I was going to miss.
> - Introduction - Functional requirements: Could add to the end of the list "the discovery of the types of items in the collection"???
> - Vocabs - Property Usage: There are labels and definitions from two scopes (the original scheme source and for this AP), it might be helpful if they were labelled similarily - it took a while for me to realise that "source definition" and "usage in this DCAP" rows in the table were both definitions (sorry, I didn't read the property usage section first).  Maybe "Label from source", "Label in this DCAP", "Definition from source", "Definition in this DCAP" so the relationships are clearer???
> - Vocabs - Property Usage: It might be useful to insert a generic statement that any scheme may be used rather than "if no class is listed the AP does not specify a class from which values should be drawn".
> - Collection Properties: These don't follow the standard DCMES ordering - might be useful to indicate the ordering method used.
> 
> DCType Vocabulary
> - Catalogue type: Was the Catalogue type definition (as specified in your correction in your email below) intended to include indexing/abstracting databases (ie. that have descriptions of individual articles journals)?  It appears the "Index" type is purely for search engine type indexes because they are "typically generated automatically"
> - Definitions: I think the definitions should be cut down to one or two sentences, and any further clarification (such as the examples currently in the definitions) be placed in a separate "Comments" row.
> 
> Thanx,
> Douglas
> 
> >>> [log in to unmask] 9/08/06 00:53:21 >>>
> A new draft of the DC Collection Description Application Profile is
> available at
> 
> http://dublincore.org/groups/collections/collection-ap-summary/2006-08-0 
> 1/
> 
> http://dublincore.org/groups/collections/collection-application-profile/ 
> 2006-08-01/
> 
> A revised version of the type vocabulary, renamed the Collection
> Description Type Vocabulary as it deals with those classes of Collection
> which are Collection Descriptions, is also available:
> 
> http://dublincore.org/groups/collections/colldesc-type/2006-08-01/ 
> 
> (I've already noticed one error in my editing of that that document -
> the definition of the "Catalogue" type/class should be
> 
> A collection of individual records describing the items, and the
> intellectual content of those items, of a second collection. There may,
> in the individual records, be information about collections but that is
> not the focus of the catalogue. Catalogues are typically created with
> significant human input.
> 
> i.e. as it was in the previous version. I seem to have managed to
> overwrite it with the definition of the (new) type/class
> "CollectionDescription".)
> 
> This is intended as something of an "interim version": the profile
> includes the uses of the cld:itemFormat and cld:itemType properties, but
> it does not yet separate out the specification of the properties used to
> describe the Collection from the specification of the properties used to
> describe the Collection Description.
> 
> Pete
> ---
> Pete Johnston
> Technical Researcher, Eduserv Foundation
> Web: http://www.eduserv.org.uk/foundation/ 
> Email: [log in to unmask] 
> Tel: +44 (0)1225 474323

-- 
Dr. Thomas Baker <[log in to unmask]>
Director, Specifications and Documentation
Dublin Core Metadata Initiative