Re: circular relationships ok?

From: Volker Hetzer <volker.hetzer_at_ieee.org>
Date: Thu, 02 Mar 2006 15:23:58 +0100
Message-ID: <du6v5u$41i$2@nntp.fujitsu-siemens.com>

Bob Hairgrove schrieb:

> On Wed, 01 Mar 2006 18:05:33 +0100, Volker Hetzer
> <volker.hetzer_at_ieee.org> wrote:
> 

>> Hi!
>> Just in general, are circular relationships something that
>> can always be avoided?
>> Or, given a model with a circular relationship, possibly
>> spanning several tables, is there a way to get rid of them?
>
> Do you have an example?
Yes, for a PCB (printyed circuit board) design tool: There are nets on a pcb. Each net can have constraints, like length or impedance.
Now, each constraint has a structure, namely an expression. This expression contains several tokens, like constants, literals and *nets*.
So, for instance, I can have a constraint that net A has to be 5mm longer than net b.

The way I have modeled this until now is: Net(NetName)
FreeNetConstraint(NetName,ConstraintName,...) Token(TokenName,TokenValue)
ExprElement(NetName,ConstraintName,TokenName)

Now, ExprElement can be a Token *or a net*. So I modeled Token and Net as subtypes of ExprElement.

I'd love to have a place where I could upload an image but I don't.

Lots of Greetings and Thanks for bothering! Volker Received on Thu Mar 02 2006 - 08:23:58 CST