[
Lists Home |
Date Index |
Thread Index
]
Seairth Jacobs <seairth@seairth.com> wrote:
>
> From: "Hunsberger, Peter" <Peter.Hunsberger@stjude.org>
> >
> > All of which is a long winded way of arriving at the
> conclusion that:
> > you don't need namespaces if you precisely understand the
> context of
> > the combined vocabularies, but until you combine the
> vocabularies you
> > probably want the name spaces if only to help in combining the
> > vocabularies.
>
> I have been working on a new vocabulary (RNA) [1]. The
> specification allows additional vocabularies (Vocabulary A)
> to be combined, as long as the following limitations are met:
>
> 1) The document must always be of RNA, but may contain
> Vocabulary A. In other words, RNA cannot be used inside of a
> document of Vocabulary A.
> 2) Vocabulary A elements and attributes must always be
> namespace-qualified.
>
> As a result, RNA is never namespace-qualified. In an earlier
> incarnation of the spec, I had placed RNA within a namespace
> since I knew there was potential for combination with other
> vocabularies. In some conversations on namespaces, it was
> pointed out to me that this wasn't necessary given the above
> limitations and that keeping namespaces out allowed for
> easier processing of the document using standard tools. This
> made sense to me, so I removed the namespace stuff (with the
> exception of "extensions" [2]). Now, this vocabulary has no
> namespace but has specific rules for namespaced vocabularies
> to be used with it.
>
> It seems to me that this is the correct way to go about
> handling namespaces, at least in this case.
The assumption that RNA cannot be used inside of some other vocabulary seems
perhaps dangerous to me: on person's metadata is another persons data.
Sooner or later someone will want to gobble up RNA and encapsulate it in
some strange way. Perhaps if only to present a live RNA example on a Web
page? Sure, in that case you can probably arrange for no collisions, but
maybe someone will find a reason to do some kind of indirect/dynamic or
recursive RNA (ack!) type thing and then what?
|
