Re: The TransRelational Model: Performance Concerns

On Fri, 12 Nov 2004 12:53:35 +0100, "Rene de Visser" <Rene_de_Visser_at_hotmail.de> wrote:

>Maybe you could list some of the many other mistakes and wrong assumptions?

Of course.

For instance he is assuming that the page cache does not exist, that all the DBMSs are based on disk storage with little page sizes, that the implementation of the TRM tables don't use BTrees or hashing techniques, that a relation or a significant part of a 100000 tuples relation don't fit in main memory, he only considers the disk page loading costs, etc.

He also assumes that we need to make room for the inserts and the room is not already there.

Etc, etc.

To only analize single relation queries is more than enough to discard the analisys.

>The analysis did not seem ridiculous to me.

I thought that it should be for most of the people here.

>However for instance when programming in AP5 most of my relations have two
>columns, some with three, and occasionaly four. So I guess it does not
>reflect that when programming in a truly relational language, that the
>relations don't tend of have many columns.

When you program in a truly relational language the relations tend to have a few attributes, and the use of the join operation tends to be extensive.

This is one of the many reasons because the TRM is an important advance. Joins are almost free with the TRM.

Regards. Received on Fri Nov 12 2004 - 14:21:22 CET