Andy Powell wrote:

> The proposed label and definition of the adaptability property are as
> follows:
> 
> Label: Adaptability Statement
> Definition:	Characteristics of the resource that affect how it can
> be modified for users or agents.
> 
> The following thoughts strike me...
> 
> If the property value is supposed to be a 'statement' (as the label
> suggests, and as 2 of the 4 example values in the proposal show) then it
> might be more consistent for the definition to start "A statement of the
> characteristics of the resource ..."?  
> 
> If the anticipated values are intended to simply be characteristics (as
> all but the last value above are), then it might be better if the
> proposed label is changed to 'Adaptability characteristics'?

I guess this will sound like the most awful semantic hair-splitting, but 
OTOH, I do feel that clarifying the "value space" of this property mas 
proposed might be helpful in highlighting some of the problems 
associated with it.

So picking up this point about "characteristics" and "statements"....

According to the DCMI Abstract Model, properties are used to make simple 
statements about resources, where each statement says

resource:X is-related-in-some-way-to resource:Y

i.e. a value is a resource.

I must admit I don't really grasp what sort of resource would/could be 
an "adaptability characteristic".

I can grasp the notion that I make statements about a resource in order 
to assert/record/disclose characteristics of a resource that are 
relevant to its adaptability, so e.g. I make statements that

resource:R some:hasAccessMode accessmode:Textual
resource:R some:deploysControlMechanism control:mouseOnlyControl

But e.g. control:mouseOnlyControl as a resource exists independently of 
the statement: it is not a "characteristic" of the subject resource, 
it's some abstract concept, the notion of "needing a mouse to control 
the resource". And of course that notion could be the subject of a 
statement too

control:mouseOnlyControl some:requiresDevice device:mouse .

I think we describe a "characteristic of a resource" only when we bring 
that concept into combination with a property in a statement; e.g. when 
we say:

resource:R deploys control:mouseOnlyControl

then I describe a characteristic of a resource.

For any property, I think it's helpful (neccessary even) to ask: "What 
set of things forms the value space for this property?", and if it's 
hard to answer that question or the answer is some rather odd mixed bag 
of things, then I think that's often a sign that the property needs 
rethinking. From the definition "Characteristics of the resource that 
affect how it can be modified for users or agents", I find it very 
difficult to define what the class of values is.

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/