xml-dev - Re: Namespaces and Schemes

["Thomas B. Passin" <tpassin@home.com>]

Sean B. Palmer said -

>
> According to 6.3.2(2):-
> "The author of a document uses namespace declarations to
indicate the
> intended interpretation of names appearing therein; there
may or may not be
> a schema retrievable via the namespace URI."
>
> Can we simply declare a namespace for a document then, and
validate it using
> that namespace as a Schema URI? Is that legal? There seems
to be some
> abiguity in that statement "may or may not be".
> BTW: I am one of the more outspoken defenders of XML
Schemas, and their
> usage. I also advocate Namespaces for validation purposes,
but I would like
> to see clarity...at the moment, I'm not quite sure what is
wrong and what is
> right.
>
 A namespace is just a string that has a uri: format.  It
has no inherent significance so far as the namespace or
xml-schema is concerned.  And the prefix is just an
arbitrary string that, it is hoped, helps a human read the
document and reduces the character count.

Of course you can use any word (string) any way you want.
You can assign all sorts of meanings to it.  For example,
xslt assigns to its namespace the meaning "this element is
an xslt instruction".  Just be aware that it is ***you***
doing it on top of some standard base like xml-schemas.

Much of this confusion would never have arisen if all the
examples using namespaces had used uri's that did not look
like url-style retrievable locations.  Then it would have
been clear that no actual document had been intended.

Cheers,

Tom Passin