W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > July to September 1999

Schema Part 1 Comments (High Level)

From: Miller, Robert (GEIS) <Robert.Miller@geis.ge.com>
Date: Thu, 8 Jul 1999 16:11:27 -0400
Message-ID: <F2A665AC0054D1119D0F00805FFECA12032BBBA0@roc01bxgeisge>
To: www-xml-schema-comments@w3.org
Hi All,

I've been studying all manner of XML technical documents, both approved and
in development, so please understand that my brain may have overheated.
These initial comments on the XML Schema Part 1: Structures are from a high
level viewpoint.  They represent my personal thoughts on Schemas, and do not
represent specific needs of the X12C Subcommittee I chair.  I will be
reviewing all of the comments sent in thus far, and plan to get down to a
more earthly view in future comments.

In (near?) the beginning, man created the index sequential database.  He saw
that it was lacking, and so advanced to the hierarchical sequential
database.  Life was good for many years. As life became ever more complex,
the network database was invented to better deal with life's complexity.
But the NDB was found to be too complex for human or computer minds, and so
man came upon the mathematical revelation that all information could be
represented in a relational database, and life was made both good and
simple.  Now the RDB provided man spare time to be creative, and upon the
RDB foundation was invented the object oriented database..

In another beginning, man invented SGML, which begot XML.  And with some
knowledge of the past, man skipped the index sequential representation of
information, jumping directly to the hierarchical representation of data!?? 

Whoops, what has happened to the rest of man's data representation knowledge
learned from experience?

Should not the efforts to design a schema language for XML draw upon man's
full prior experience?  Yet I don't see that full experience reflected in
the first draft of XML Schema Part 1: Structures.  Where is the basic talk
of sets, members, and relationships, and the XML Schema framework to support
them?  Where is the separation of relationship definitions from data
definitions?  

Might it be beneficial to reflect upon established principles of database
design, and define XML structures which reflect and directly support these
principles.  I believe such an approach will yield a better end product than
would result from attempts to improve upon the prior attempts to define XML
data structures.

Cheers,      

  Bob Miller
    615-371-6037
Received on Thursday, 8 July 1999 16:13:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 18:12:46 GMT