> I think this is a typo? Some of the dc:identifier examples in
> the XML spec [1] that have URIs as values are written like this:
>
> <dc:identifier>
> http://www.ukoln.ac.uk/
> </dc:identifier>
>
> And some like this:
>
> <dc:identifier xsi:type="dcterms:URI">
> http://www.ukoln.ac.uk/
> </dc:identifier>
>
> Should it be consistent?
I think the intention was to distinguish between "Simple DC" in 4.3 and
6.1 and "Qualified DC" in 5.3? In the former, the "abstract data model"
doesn't include the idea of an "encoding scheme", so the use of the
xsi:type attribute would be limited to the latter.
The text sugggests that the example at 6.2 differs from that at 6.1 only
in the addition of the rights stuff, so, yes, maybe you're right there,
and the introduction of the xsi:type attribute is misleading. I'll
discuss it with Andy.
> Also I don't know if people are aware of RSS 2.0 (XML) [2]
> but I expect that a lot of people in the blogging community
> are going to look at the RSS 1.0 (RDF) modules for DC [3] and
> DCTERMS [4] and think that they can use them in RSS 2.0 --
> perhaps an example could be added to the XML guidelines to
> show how to properly include DC in RSS 2.0?
Hmmm.... I take your point about the potential for confusion over
modularisation issues for RSS 1.0/2.0 users, but I'm tempted to suggest
that is really for the RSS community/communities to address? Is there an
intention to define "modules" for RSS 2.0? Or to extend/revise the
description of RSS 1.0 modules to indicate how they might be used in RSS
2.0 as well?
Personally, I'm rather wary of taking [1] any further in the direction
of trying to describe/define support for modularisation/"application
profiles".
Pete
|