Hi Mikael,
The document is looking good!
In the section "Value Strings, Value String Languages and Syntax
Encoding Scheme URIs":
s/occuring/occurring/
I confess I'm still a little bit hazey about whether/when we need to say
"RDF URI reference" rather than "URI", but given that RDF Concepts uses
the first form, I guess it's fine to stick with that when talking about
things in the RDF graph.
On that basis, I wondered (but I'm not sure about this!) whether
- an RDF predicate that is the URI rdf:type
should read
- an RDF predicate that is the RDF URI reference rdf:type
[and should we expand the QName to a URI here?]
And similarly for
- an RDF predicate that is the URI dcrdf:valueString
Again I'm not completely sure whether the following is a problem or not,
but I wondered whether some of the current phrasing in the section "RDF
shorthand for RDF plain literals and RDF typed literals" was a little
bit ambiguous about whether it was referring to the URI or to the class
identified by the URI. I was thinking particularly of the use of "...
equals rdfs:Literal".
So I wondered whether maybe
- The DCAM Vocabulary Encoding Scheme of the value is an instance of
rdfs:Datatype or equals rdfs:Literal.
might be better phrased as something like
- The DCAM Vocabulary Encoding Scheme of the value is an instance of
rdfs:Datatype or is the class rdfs:Literal.
just to make it clear that we are referring to the class not the URI?
Similarly, change
- The DCAM Syntax Encoding Scheme of the DCAM value string is a
sub-class of or equals the DCAM Vocabulary Encoding Scheme of the DCAM
value
to
- The DCAM Syntax Encoding Scheme of the DCAM value string is a
sub-class of or is the same class as the DCAM Vocabulary Encoding Scheme
of the DCAM value
And
- If the DCAM Vocabulary Encoding Scheme is not equal to rdfs:Literal,
the DCAM value string may not have an associated DCAM value string
language.
to
- If the DCAM Vocabulary Encoding Scheme is not the class rdfs:Literal,
the DCAM value string may not have an associated DCAM value string
language.
And
- If the DCAM Vocabulary Encoding Scheme of the DCAM value equals
rdfs:Literal, an RDF plain literal is used.
to
- If the DCAM Vocabulary Encoding Scheme of the DCAM value is the class
rdfs:Literal, an RDF plain literal is used.
(An alternative would be to rephrase the sentences to talk about
"Vocabulary Encoding Scheme URIs" and "Syntax Encoding Scheme URIs", but
that might be more cumbersome!)
> I have now added some figures and examples to the draft.
>
> Thanks for the suggestions regarding syntax to use in the examples.
> After some thought, I decided to go with RDF/XML, mostly
> because it it most easily imported directly into tools, such
> as the W3C RDF Validator service.
Should the sentence in the intro
- It does not refer to any specific RDF syntax encoding (such as the one
defined by the RDF/XML Syntax Specification [RDFXML]).
be qualified slightly? It's still correct that the mapping is to the RDF
graph and isn't specific to RDF/XML, but the doc does include some
RDF/XML examples now.
> However, I'm still open to arguments that we should include
> graph/N-TRIPLES/... versions as well.
>
> Please consider the pictures and examples and provide
> suggestions for improvements/additions.
In the two examples in the section "RDF shorthand for RDF plain literals
and RDF typed literals":
s/ResoruceURI/ResourceURI/
Also I think the constructs labelled "ValueURI" should be labelled
"ValueString", in both cases (for "Learning Biology" and "2005-06-10").
> I've chosen to not illustrate every section, instead relying
> on a few carefully (?) chosen diagrams that illustrate the
> more difficult parts.
> Does this work or not?
I think it works well!
Cheers
Pete
---
Pete Johnston
Technical Researcher, Eduserv Foundation
Web: http://www.eduserv.org.uk/foundation/
Email: [log in to unmask]
Tel: +44 (0)1225 474323
|