Re: Conformance Labels

On Mon, 2006-09-04 at 21:07 -0400, Chimezie Ogbuji wrote:
> Below are my attempts to document some conformance labels for GRDDL (thinking out loud). 
> Initially, I was going stratify the terms *completely* along the 
> mechanisms themselves, but it didn't seem very useful.  For instance, I ended up with terms (for 
> the source document) such as:
> 
> - GRDDL XHTML Document (either associated with the data-view metadata profile or with a specific HTML profile document - with it's own GRDDL hooks)
> - GRDDL XML Document (bleh..)
> - GRDDL-by-Namespace-Document (couldn't think of a concise label even though)
> 
> In the end, it didn't seem very helpful to define labels for the source 
> document (from which the RDF is gleaned) at more granularity than:
> 
> # GRDDL (XML?) Document #
> 
> An XML document which registers specific hooks for a GRDDL client to use 
> to extract Resource Descriptions from it.

Could you elaborate on what you mean? What does it mean for a document
to register a hook? Documents aren't people/agents; they
don't _do_ things (except perhaps to say something).

Is this a GRDDL XML Document?
	<z/>

How about this?
	<z xmlns="http://example.org/terms" />

How about this?
	<z xmlns="http://www.w3.org/2001/sw/grddl-wg/pg23" />

Does it depend on the representations available from .../pg23 ?
i.e. does the question of whether an XML document is a GRDDL
document depend on the state of the web as well as the text
of the document?

> As for the client (or processor), It did seem useful to stratify 
> the labels this way, since the mechanisms are very specific (metadata profiles 
> and namespace documents) and I could imagine GRDDL implementations that 
> supported a subset (rather than all) of the mechanisms.

I can imagine them, but I don't want to encourage them by giving
conformance labels to them.

> It seems sufficient to stratify the labels for implementation conformance 
> along the XML / XHTML divide:
> 
> # GRDDL XHTML User Agent (or Client?) #
> 
> A software agent which implements both of the XHTML-specific GRDDL 
> mechanisms for extracting Resource Descriptions from XHTML.  In 
> particular, the interpretation of the http://www.w3.org/2003/g/data-view 
> meta data profile and it's implications for transformation links [1].  In 
> addition, the interpretation [2] of the data-view:profileTransformation 
> property in Resource Descriptions extracted from an arbitrary metadata 
> profile.
> 
> # GRDDL XML Processor (or GRDDL-aware XML Processor?) #
> 
> An XML processor which implements the mechanisms specific to the 
> extraction of Resource Descriptions from XML documents.  In particular, 
> the data-view:transformation attribute [3] and attempting to dereference 
> and interpret [2] namespace documents.
> 
> I prefer GRDDL XHTML User Agent (only because that's the language [4] used 
> in the section regarding metadata profiles) and GRDDL XML Processor.
> 
> It's worth noting that there is a recursive element to the level of 
> conformance.  For instance, a GRDDL XHTML User Agent would not be able to 
> extract Resource Descriptions from a metadata profile which consisted of 
> an XML document that had a data-view:transformation attribute.

Indeed, so let's not encourage that.

> So, the conformance labels are:
> 
> - GRDDL Document
> - GRDDL XHTML User Agent
> - GRDDL XML Processor
> 
> [1] http://www.w3.org/2004/01/rdxh/spec#grddl-xhtml
> [2] http://www.w3.org/2004/01/rdxh/spec#ns-bind
> [3] http://www.w3.org/2004/01/rdxh/spec#grddl-xml
> [4] http://www.w3.org/TR/1999/REC-html401-19991224/struct/global.html#h-7.4.4.3
> 
> Chimezie Ogbuji
> Lead Systems Analyst
> Thoracic and Cardiovascular Surgery
> Cleveland Clinic Foundation
> 9500 Euclid Avenue/ W26
> Cleveland, Ohio 44195
> Office: (216)444-8593
> ogbujic@ccf.org
> 
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Received on Tuesday, 5 September 2006 13:52:58 UTC