FW: Request for Comments: DSpace History System Descriptive Note *dra ft*

On behalf of Eric Miller, forwarded to the list with permission.

-----Original Message-----
From: Eric Miller [mailto:em@w3.org] 
Sent: Friday, May 09, 2003 12:50 PM
To: Bass, Mick
Cc: jason_kinner@dynamicdigitalmedia.com
Subject: RE: Request for Comments: DSpace History System Descriptive Note *dra ft*


Comments on http://web.mit.edu/simile/www/documents/historySystem/descriptiveNote/descriptiveNote.pdf

Section 2.1 Namespaces

- Use Dublin Core namespace http://purl.org/dc/elements/1.1/

- Use Harmony namespace (err.. again which one I'm not sure) *or* simply reuse the pieces of the event model that make sense to dspace and call it something else. Again, I don't think we should hold the Harmony namespace in such reverence.

- Please plan namespaces that resolve to more RDF. I'd suggest for example therefore *not* using namespace such as http://www.dspace.org/

- Provide RDF schemas formalizing Section 2.2

Section 3.1.2 is slightly confusing to me; are you saying with use
*both* abc:hasPart and dcterms:hasPart in the metadata? Further to then say they are 'equivalent' in a dpsace schema?

If thats the case, I suggest simply using *one* hasPart and not worry about drawing eq relationships.

Section 3.1.3

i think you mean to say in your example - s/rdf:Collection/dspace:Collection

I very much support adding Type information to the dpsace history data making this information more explicit and ultimately far more useful.

Section 3.2.1

see http://purl.org/dc/elements/terms/

Section 3.3.1 argues for simply creating a 'new' (simplified) event model vocabulary based on harmony work. I would support this.


Section 3.3.2

'typeless' resources could be inferred by declaring the rdfs:range of certain event properties (input, output) to be of dspace:Event.

Received on Friday, 9 May 2003 14:12:09 UTC