Re: Binary Relational Modeling call
Date: Sun, 11 Sep 2011 15:23:55 -0700
Message-ID: <d_WdnaJfZdBiqfDTnZ2dnUVZ5sOdnZ2d_at_giganews.com>
Ivan wrote:
>>If Teacher is a set, what's it a set of?
>
> elements are contents of sets, but they are not shown by BRM graph. they should be available for browsing only in user application.
>
> binary relations have their name, source set and destination set. they state possible bindings between elements of source and destination sets. Basically, a set can be seen as a table with its fields represented by relations that source from that set. elements are records of that table. here is example:
> https://sites.google.com/site/binaryrelationalmodeling/compare-invoice%286%29.png
>
>
>>So, your data model has at least two structural elements: sets and >>binary relations. Does it have any other structural elements?
>
> those 2 are crutial structural components of BRM. however, notation of uniqueness of elements is welcome to take place in graphs (one-to-one relationship is done in the same way many-to-may relationship is, with intermediate set, but with uniqueness constraint). although i'm aware of importance of uniqueness, uniqueness notation is not specified and programmers have to implement one in the way they find appropriate.
So, you are saying BRM doesn't have logical identity?
>>Note that at 2 structural elements, BRM is already structurally more >>complicated than the relational model, which has only relations. What >>compensating benefit does one get for the added complexity?
>
> the real value of BRM graphs is in its simpleness.
Whoa! Hold on there! We have already established BRM is structurally more complex than the relational model not simpler. It's real value cannot come from something it doesn't even have. Received on Mon Sep 12 2011 - 00:23:55 CEST