Table Inheritance Discussion

Поиск
Список
Период
Сортировка
От Curt Sampson
Тема Table Inheritance Discussion
Дата
Msg-id Pine.NEB.4.44.0208120819490.517-100000@angelic.cynic.net
обсуждение исходный текст
Ответ на Re: Why is MySQL more chosen over PostgreSQL?  ("Jordan Henderson" <jordan_henders@yahoo.com>)
Список pgsql-hackers
On Thu, 8 Aug 2002, Jordan Henderson wrote:

> I think what would be useful is to discuss the theory part.

As do I.

> - Date has 3, however his most current work is dated 2000, The Third
> Manifesto SECOND EDITION.

This is actually Date and Darwen.

I think we should also add Date's _An Introduction to Database Systems,
7th Edition_, as it covers some relational stuff in more detail than
than _The Third Manifesto_. For example, it investigates the details of
automatic view updatability, which came up during this discussion, and
which most books just completely cop out on. (For example, _Database
System Concepts_ just points out a couple of problems with view
updatability and says, "Because of problems such as these, modifications
are generally not permitted on view relations, except in limited
cases.")

> - Silberschatz, Korth, Sudarshan, A book I am sure we have all read,
> Database System Concepts - Third Edition.
> ...
> In any case, we should use the current editions of these books, not
> something the author has reconsidered, re-written, and published again.

In that case we ought to use the fourth edition of this book.

Here are some questions I'd like to see people answer or propose
answers to:
   1. What models of table inheritance have been proposed, and how   do they differ?
   2. What models of table inheritance are actually implemented in   currently available database systems?
   3. What are the advantages of describing something using table   inheritance rather than an equivalant relational
description?
   4. If you think table inheritance is "object oriented," why do   you think so.
   5. How ought we to fix the table inheritance in postgres?

The last question comes up because, during the conversation up to this
point, we seem to have implicitly accepted that table inheritance is
an "object-oriented" way of doing things. Thinking further on this,
however, I've decided that it's not in fact object-oriented at all.

cjs
-- 
Curt Sampson  <cjs@cynic.net>   +81 90 7737 2974   http://www.netbsd.org   Don't you know, in this new Dark Age, we're
alllight.  --XTC
 



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Alvaro Herrera
Дата:
Сообщение: Re: [COMMITTERS] pgsql-server/ oc/src/sgml/release.sgml rc/back ...
Следующее
От: Don Baccus
Дата:
Сообщение: Re: Table Inheritance Discussion