[
Lists Home |
Date Index |
Thread Index
]
- From: Sparkman Bill <Sparkman_Bill@ems.prc.com>
- To: "'KenNorth'" <KenNorth@email.msn.com>, Hien Luu <Hien@Icarian.com>, "Eve L. Maler" <elm@east.sun.com>
- Date: Fri, 3 Mar 2000 10:32:36 -0500
KenNorth wrote:
Perhaps Hien was asking about "referential integrity" in the database
context, which means more than asserting a linking relationship. The
relationship, and rules about the relationship, are defined to a container
(database) that blocks actions that would subvert data integrity.
I approached this subject last summer. Basically my solution is this:
1) XML is a data format. A transport tool so to speak.
2) Applications may dynamically create XML files which contain content.
3) This content, when created, may contain relationships which reflect
exactly the condition of the source data. This includes RDB relationships.
4) It is up to the Application to create it correctly.
5) XML COTS tools or homegrown tools can be used to maintain the data
6) If a COTS tool is used it will at best verify that the current version of
the data is valid and well formed according to XML rules.
7) The Application or homegrown tools must verify the unique conventions
created to reflect the RDB...
Hien - I hope this sheds some light on how you can create an environment
that works now. This list and others are constantly working on ideas to put
more controls or layers on XML. Sometimes I wonder if this is really the
right approach. I like the KISS method. XML does a great job of formatting
data. It is readable. Anyone can then process it. Should they try to
provide behavior that a RDBMS or ORDBMS does? An example is your request to
verify referential integrity. A program can be created to verify how you
implemented referential integrity. But this currently needs to be
customized based on how you implement it.
Bill Sparkman
***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************
|
