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

Re: SD5 - Namespaces [fmt]

From: altheim <altheim@mehitabel.Eng.Sun.COM>
Date: Fri, 16 May 1997 17:39:10 -0700 (PDT)
To: w3c-sgml-wg@w3.org
Message-ID: <libSDtMail.9705161739.19412.altheim@mehitabel>
From: Jean Paoli <jeanpa@microsoft.com>
[...]
>  SD5 - Namespaces:
>  -----------------------------
> 
>  Proposal: Every element comes from some schema, which becomes the
>  default schema for that element. The names within the default schema
>  are usable within the element without qualification. Names from other
>  schemata are also usable (provided they are compatible with the
>  elements definition) but must always be qualified.
> 
>  Every element can have one or more foreign schemata introduced to it
>  by a well-known attribute which refers to a schema by URI, and gives
>  it a short name used for qualification.
> 
>  <XML>
> 	 <*xml-schema xml-ref="http://www.bigbookstore.com/schema"
>  xml-as="bk" />

I think the idea of namespace inheritance is a great idea, but I wish
you hadn't tied the example to your SD2-Structured Attributes proposal,
which I really dislike (New, incompatible syntax? What does it purport to do
that can't be done better with existing markup? Mystified...). This was
(and may still be) an issue for WAI/ICADD, and requires a completely
SGML compatible solution.

I don't claim to understand architectural forms in all their myriad
complexity, but I was under the impression that they would serve this
purpose. If that is true, then is this proposal in competition with AFs?
If so, a lot of thought has gone into architectural forms, unless I'm
misreading this and AFs are orthogonal to the need that SD5 professes to
handle. Is this possibly another case of the WWW not-invented-here syndrome?

If architectural forms cannot provide the necessary ties into an external
namespace for your needs, then perhaps a modification and simplification
of AFs would better suit everyone than starting over from scratch.

Rick's alternative also seems preferable: (as he states) clearer, less
confusing, and compatible with SGML (which at least until recently ;-)
was one of our stated goals).

Murray

...........................................................................
Murray Altheim, SGML Grease Monkey                    <altheim@eng.sun.com>
Member of Technical Staff, Tools Development & Support
Sun Microsystems, 2550 Garcia Ave., MS UMPK17-102, Menlo Park, CA 94043 USA
         "Give a monkey the tools and he'll build a typewriter."
Received on Friday, 16 May 1997 20:39:41 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 10:04:33 EDT