Re: David C. Hay's Universal Data Model:

I was looking to use a variant of the universal data model in a CASE tool. I remember, some years ago, at a 'Systems Architect' class, that tool used only 2 tables (entities and relations) to capture all it needed to know. At that time, they told you the schema because you had to use SQL to generate custom reports.

For a case tool, it would be fine, I'm just wondering if it would be better to have direct support for other common things, like inheritence and versioning.

"kdc" <kcameron_at_navicsys.com> wrote in message news:r%IN4.167677$AT6.239887_at_dfw-read.news.verio.net...
> I have the book and enjoy it alot (don't care much for the Oracle Case
> Notation good old IE works for me)
>
> I would never try to implement universal model. I find it intersting in
 an
> academic sense but it would not perform well in a production environment.
>
> "js" <jeeves_97_at_yahoo.com> wrote in message
> news:nJPL4.289$IY6.33943_at_nntp0.chicago.il.ameritech.net...
> > Say, has anyone out there given much thought to the design of a
 Universal
> > Data Model? Something along the lines of the one found in David Hay's
 book
> > "Data Model Patterns"
> > (pg 256 for those who have it)
> >
> > The data model had the following tables:
> > Attribute
> > Attribute Assignment
> > Value
> > Relationship
> > Relationship Type
> >
> >
> > Value
> >
> > | |
> > Relationship-------- Thing ----------------------| |
> > | | --------- |
> > |

           |                        |


> > |

> >              |                        |       Attribute Assignment    |

> >     Relationship             |          |                         |

> |

> >         Type                     |          |                         |

> > |

> >                                  Thing Class