XML complexity, namespaces (was WG)

[Steven R. Newcomb]

[James Anderson:]

> I would take "used to do architectural forms" to mean that
> the information to be inferred from a namespace
> declaration would supplant (some portion of) that which
> would otherwise have been provided by existing
> architectural declarations. Otherwise namespaces are just
> "architecture-neutral".

> When I wondered about how this might work, it occurred
> that the identity between the URI in a namespace
> declaration and that in the system-id of a IS10744:arch PI
> might be used to infer mappings equivalent to those
> provided by the individual architectural mapping
> attributes.

> The problem with this is that it provides no means to map
> the local parts of universal names. Without such a means,
> either a given element type maps to exactly one
> architectural form (my original question), or the local
> parts of the respective type names must be identical in
> all architectures (which would seem an equally severe
> restriction).  That is, the standard architectural
> declarations are still necessary and, again namespaces are
> "architecture neutral".

Yes, I agree that you'd have to enhance the XML Namespaces
Recommendation, one way or another.  I personally feel such
enhancement would be a positive direction for the evolution
of XML Namespaces.  Since I believe that architectural forms
(or something that's functionally equivalent) are both vital
and inevitable, I believe there are really only 3 choices
here:

(1) Invent something completely new in order to support
    architectural inheritance at the document level.  If
    this means inventing a new schema language to replace
    DTDs, so be it.  (But that alone won't satisfy the
    requirement that architectural inheritance be possible
    without a DTD or something like one.  Namespaces already
    have the virtue of working without a DTD.)

(2) Use the ISO/IEC 10744:1997 way of doing it (which is
    already being done with XML in several quarters).  It
    works with or without a DTD, incidentally.  This would
    be just fine with me; I favor collaboration between ISO
    and W3C.  However, the ISO way of inheriting an
    information architecture isn't perfect.  I'd still like
    to see the future evolutions of XML and SGML remain in
    harmony with each other.  Several people have complained
    in this forum that Namespaces have various deficiencies.
    Frankly, the alleged deficiencies haven't bothered me at
    all [yet].  What bothers me is the stunning departure
    that Namespaces represent from a path of harmonious
    co-evolution with the ISO family of architected
    information interchange standards, and the fact that the
    distinct goal toward which Namespaces is driving us is
    not at all clear to me.  Was Namespaces designed as a
    merely temporary, expedient, under-high-pressure
    solution?  Or is it a step on workable path toward a
    sensible, efficient, reliable future?

(3) The answer to the latter question can be "Yes" if we

    * Recognize that the reasons for doing architectural
      inheritance are a compelling superset of the reasons
      for inventing XML Namespaces.

    * Accept that there is great need for orderly evolution
      in XML-land, and allow/cause XML Namespaces to evolve
      in the general direction of inheritable architectures.

    I like this third option better than choice #1 because I
    favor having a only few, general, powerful syntaxes and
    syntactical features, as opposed to an endless panoply
    of special-case add-ons.  The danger that XML (or,
    rather, the usefulness and efficiency of XML) will die
    of such obesity is very real.  There are numerous
    examples of standards, both de facto and de jure, that
    sooner or later became useless in exactly the same way.

-Steve

--
Steven R. Newcomb, President, TechnoTeacher, Inc.
srn at techno.com  http://www.techno.com  ftp.techno.com

voice: +1 972 231 4098 (at ISOGEN: +1 214 953 0004 x137)
fax    +1 972 994 0087 (at ISOGEN: +1 214 953 3152)

3615 Tanner Lane
Richardson, Texas 75082-2618 USA

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
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 at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)