Re: Introduce VariableRepresentation class (was: variable representations edited for spec)

> You shouldn't have omitted the sentence that followed this:
> 
>    About the only thing I can think of is to query for
> VariableRepresentation.
> 
> :-)

Oh yeah, sure :-)

>> Suppose we have a GUI to build Hydra templates.
>> The knowledge that MyXyz is a VariableRepresentation
>> would allow the GUI to offer it as possible value for variableRepresentation.
> 
> Anything else?

The scenarios I can think of ar variations of this;
basically, the possibility to say: “this thing is like those two others”.

>>> Maybe... but much more interesting would be
>>> to know how to use those (newly found) representation formats. We can't
>>> describe that at the moment. Thus I see limited value in introducing such a
>>> superclass and a range.
>> 
>> How about "I don't understand representation X, so I'm trying Y instead".
> 
> Not sure I follow...

Nah, never mind, it's a variation of the above use case.

> Nothing.. as you said, machines are dumb. Humans might be able to infer
> something from the name (URL) or by reading the natural language definition
> they get by dereferencing the URL...

Another benefit then: that somebody else can create a new property,
and say that the range is Hydra's VariableRepresentations.
We allow reuse that way.

> which reminds me that we should add a
> pointer from the concepts definition to the relevant section in the spec
> (rdfs:seeAlso?).

Yes, great idea.

Ruben

Received on Wednesday, 31 December 2014 16:35:11 UTC