Re: Use of the term "hierarchy" (and table/class)

From: Alexandr Savinov <>
Date: Sat, 27 Aug 2005 18:36:35 +0000
Message-ID: <>

Marshall Spight schrieb:
> Alexandr Savinov wrote:

>>But actually what are
>>tables if not a physical container for records?

> A logical container for records.

Then I see the following ways to resolve the conflict: 1. it is a conflict of terms only (the problem is solved) 2. we need to define more precisely what we mean by physical container and logical container (my definition see below) with an oppurtunity to go to item 1 above or to some other more interesting conclusions.

Here is my treatment:

  • physical container is responsible for its internal objects representation (identification) and access (given those identifiers). This is why an object cannot change its physical container so any object is deemed to live where it was born.
  • logical container is flexible because it is based on object properties according to the following definition (part of COM): a value is a logical container for all object that have it. Thus changing an attribute value we can change a container this object belongs to.

In this sense a table is physical container for its records. Note that these definitions are parts of a data model rather then informal interpretations and associations.

What is your treatment of physical and logical containers which you mean when say that table is a logical container for records? And what is then an example physical container?

Received on Sat Aug 27 2005 - 20:36:35 CEST

Original text of this message