[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Namespace Registry? And what about Architectures?

• From: Clark Evans <clark.evans@m...>
• To: xml-dev@i...
• Date: Tue, 09 Feb 1999 23:47:55 +0000

I just finished reading the RDF specifications on the web.
It's nice that they are putting the classification scheme
that they have come up with in their own namespace. This
is good, and un-assuming.  I approve.

However, now I have two concerns:

a) I feel that the RDF specification should show
how architectures would be used.  This to me
is very critical.  I don't see RDF being used
directly, I only see it as a documentation mechinism
that I can include in my DTDs.  Does this make sence?

In this way, if someone wants to define a transformation
from my documents to theirs, they have two choices, 
they can define the transformation directly from mine,
or they can define a transformation in terms of an 
RDF syntax, and thus, transitively handle my documents.
Correct?

b) And, much less importantly, to me a namespace 
refers to a global way to say who has defined a name.

I know the current name space mechinism creates an 
"alias" for a URI, via the xmlns:alias="... URI ...".
I would think, that it might be prettier to use 
full-fledged-reverse-url-like-thingys similar to Java.

Thus, the "rdf" alias becomes "org.w3.rdf".  I reason
that a DTD, enabled with 10+ archetecture cross-references
it would quickly become a nightmare figuring out who
is "rdf".  Is there a way to specify a DTD wide alias?

<!NAMESPACE
  org.w3.rdf
  "http://w3.org/TR/1999/PR-rdf-syntax-19990105#" 
>

<!NAMESPACE
  com.manhattanproject.pdb
  "http://manhattanproject.om/xml/pdb-rdf#"
>

Thus, you can then do fun stuff like

<!ELEMENT x ... >

<!ATTLIST x org.w3.rdf                 : type   #FIXED "object" 
            com.manhattanproject.pdb.1 : person #FIXED "true"
>

Anyway, I'm just playing... but when you want to name 50+ 
equivalences for your object (which may be common practice
if you have about 100 suppliers... each with their own
product definitions).  I can't see how the plain old aliases
will scale all that wonderfully.

Of course, you could probably do something like this anyway,
the namespace spec dosn't prevent a "." in an alias does it?

Best,

Clark

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@i...
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo@i... the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo@i... the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@i...)

• References:
  • CORBA's not boring yet. / XML in an OS?
    • From: Robb Shecter <shecter@d...>

• Prev by Date: Re: Storing Lots of Fiddly Bits (was Re: What is XML for?)
• Next by Date: ANN: IBM XML Parser for Java (XML4J) v2.0.0
• Previous by thread: Re: CORBA's not boring yet. / XML in an OS?
• Next by thread: RE: CORBA's not boring yet. / XML in an OS?
• Index(es):
  • Date
  • Thread