XML.orgXML.org
FOCUS AREAS |XML-DEV |XML.org DAILY NEWSLINK |REGISTRY |RESOURCES |ABOUT
OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index]
Re: [xml-dev] XML Schema as a data modeling tool

Thank you. But that's amazing. A person who does not think that he or she has a phone number, rather than the other way around, would be exceptional. So there is a difference in "commonness" as to the two alternatives of parent and child.

I would not hesitate one moment to construct a conceptual model of an enterprise in which a person has zero, one or more phone numbers. I can transform this information content at any moment into a mapping of phone numbers to people, if the need arises. The model is sound and more readily understood by the majority or people than a set of "edges" between phone numbers and people. I do not understand that it "does not reflect the management of phone numbers, devices and people." I do not understand that I "haven't done data modeling". I think I have, all parts have been accommodated, no information is lost. The fact that I cannot assert that this is the only possible way to construct the model does not mean that it is not a sensible, useful, satisfactory way to do it.

Perhaps we have different notions of what a conceptual model is.

Hans-Jürgen



Von: Peter Hunsberger <peter.hunsberger@gmail.com>
An: Hans-Juergen Rennau <hrennau@yahoo.de>
CC: Michael Kay <mike@saxonica.com>; "xml-dev@lists.xml.org" <xml-dev@lists.xml.org>
Gesendet: 0:05 Dienstag, 1.Oktober 2013
Betreff: Re: [xml-dev] XML Schema as a data modeling tool

Here's a real simple real world example:

People have phones 
Phones have phone numbers

Except, sometimes the same phone number goes to a fax as well as a desk phone, or a single phone number is shared between multiple people, or people have more than one phone number.  So what sits at the top of the tree? Phone number or Person?  If your answer is that it depends on the problem at hand then you haven't done data modelling, you've modelled a single instance within our problem domain. Or, in-other-words, a tree can model a document that gives a single way to handle phones numbers, devices and people, but that single tree model does not reflect the management of phones numbers, devices and people  within the enterprise.

More generally, you cannot represent any many to many relationship as a single tree. 

So, the answer to your question is that these models are not ever best represented as trees.




Peter Hunsberger


On Mon, Sep 30, 2013 at 4:43 PM, Hans-Juergen Rennau <hrennau@yahoo.de> wrote:
I do not understand. Here comes my view.

(a) A system is a set of entities, which are concepts.
(b) In isolation, these concepts (say, customer, bookings, accommodations) are best modelled as trees, as all stakeholders understand them immediately and comprehensively, regardless of their background
(c) The fact that these trees reference each other suddenly implies that we should cancel our tree models which served us so well.

Why? And what shall replace point (b)?




Von: Peter Hunsberger <peter.hunsberger@gmail.com>
An: Hans-Juergen Rennau <hrennau@yahoo.de>
CC: Michael Kay <mike@saxonica.com>; "xml-dev@lists.xml.org" <xml-dev@lists.xml.org>
Gesendet: 23:14 Montag, 30.September 2013
Betreff: Re: [xml-dev] XML Schema as a data modeling tool

We may very well be aiming at different scales.  However, in my experience any set of entities that can be described a single concise hierarchy isn't going to require much management  buy in and as such using XSD for domain management isn't going to revolutionize the world of software development any time soon.  At any reasonable level of complexity -- one that does require management buy in-- hierarchies in the shape of trees become forests and binding them all together requires the modelling equivalent of ID/IDREF.  At that point you've got models that have cross references that must be manually looked up and traversed and you've reinvented the worst possible form of a graph for managing the complexity.

Peter Hunsberger







[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index]


News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 1993-2007 XML.org. This site is hosted by OASIS