RE: Namespaces

At 09:07 22/5/97 -0700, Andrew Layman wrote:
>Sean Mc Grath points out that the namespace problem recurrs in many
>modern programming languages, and that these may give clues to a
>solution.
>
>1	The URI-at-the-front method:
>

>2.	The "Java" method:
>
>3.		The "Modula-2" method:


The problem with all these methods is scope. 

XML is concerned with documents that are longer than those on the WWW today.
XML needs to be able to deal with compound documents, where different
fragments have been taken from different sources, created with different DTDs.
In reusing fragments we don't want to have to change the tagging.

Therefore what is needed is a way of saying that from this element till you
find the matching end-tag, consider that a separate name space has been set
up whereby all tags within this scope can be considered to have
automagically been qualified by a specified name space qualifier.

The fact that this name space qualifier has the effect of ABC:DEF is
irrelevant. What is important is that users don't have to remember that they
need to qualfify DEF by ABC and can continue to call DEF by the name they
have been used to using when creating DEFs in their normal source documents.

What my proposal sought to do was to provide a scopable statement of how far
name spaces applied, while still giving a mechanism for referencing from one
scope to another with ID references. This feature seems to be ignored by
most other proposals. One question they must all ask themselves is " how can
I be sure that an ID in a reused fragment conforming to DTD X will not
conflict with an ID in a DTD for fragment Y". I would be very grateful to
hear how this can be acheived wth any of the three proposals made by Andrew
Layman..
----
Martin Bryan, The SGML Centre, Churchdown, Glos. GL3 2PU, UK 
Phone/Fax: +44 1452 714029   WWW home page: http://www.sgml.u-net.com/

Received on Friday, 23 May 1997 06:31:00 UTC