W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > August 1997

Re: XML Namespaces: Results of Montreal Meeting 22-Aug-97

From: Matthew Fuchs <matt@wdi.disney.com>
Date: Mon, 25 Aug 97 12:02:26 PDT
Message-Id: <199708251902.MAA02653@scrumpox.rd.wdi.disney.com>
To: andrewl@microsoft.com (Andrew Layman)
Cc: w3c-sgml-wg@w3.org, connolly@w3.org

You might want to post a pure ASCII version of your doc to make it
easier to critique by mail.

To make the namespace mechanism more extensible, the href attribute of
the namespace PI should be able to point to an entity as well as a
URI.  It would also be very nice if you could declare a namespace as an
entity, and use the entity name as the prefix.  This would:
1)  allow namespaces to interact with Catalogs when they are
2)  let those who gag at the use of PIs use namespaces with their
consciences clear.


<!ENTITY zoo SYSTEM "http://zoo.org/schema.dtd">


<!ENTITY zoo SYSTEM "zooSchema" -- eventually with catalogs -->


I don't think this changes anything fundamental, but is perhaps a
little more "orthodox."  The entity declarations could also be changed
to PIs when the document is transmitted somewhere.

It is interesting to note that without validation, this is all just
syntactic sugar and could be accomplished entirely with attributes,
however if we include validation, then GIs with different namespace
prefixes would have different content models, and namespaces would not
be just sugar.


> Several important design decisions regarding XML namespaces were
> concluded at the XML meeting in Montreal this past Friday. The attached
> document contains my understanding of the resulting definition of XML
> namespaces and their application to document instances. It is
> essentially the same as the "Layman-Bray" namespace proposal, but has
> been refined in a number of particulars, so I recommend a careful
> reading of the document.
> Dan: Can you post this, superceding the similar document I sent a week
> ago?
> --Andrew Layman, Microsoft Corporation
Matthew Fuchs
Received on Monday, 25 August 1997 14:59:49 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:25:11 UTC