Re: Relationals vs. Objects Databases I

Richard Ronteltap wrote:

> >Mabel TrigĂ˜is wrote in message <01bd230c$6c367590$ec0bbac3_at_m3bj1>...
> >>I am not trying to make the 'polemique'. I am preparing a paper and I
 would
> >>need to clarify some points:
> >>
> >>Comparing relational, object oriented and objects databases.
> >>
> >>1) Differences.
> >>2) Common points.
> >>3) Future of relational, object oriented and objects databases.
> >>
>
> Take a look a this, the Cetus object link:
> http://www.rhein-neckar.de/~cetus/software.html
>
> Because this was the subject of the finishing project on my Computer
> Science study, I'll insert my abbreviated $0.02 here:
>

May be you should study some more. Like most of the studentsyou are jumping to conclusion based on incomplete data and without having any practical experience.

> There is no technical reason when ODBMS should not replace
> RDMBS. An ODBMS can easily be made a superset of RDBMS
> functionality. Some problems of RDBMS are:
>
> - No (object) identity for referencing

You sure can have one if you define one explicitely.

> - No duplicate handling (In RDBMS, what looks the same, is the same)

May be it is good. Less confusion.

> - Bad complex data handling. Everything must be broken down into
> rectangular tables.

With the object model you also have a list of attributes which is what? One dimentional?Relational data model does require more elaboration, but it does your mind good.

> The relations between these smashed bits are not part of the model.

If you'll define these relations you will have them. And they will be handled better then by ODBMS.

> - Bad programming language integration (impedance mismatch)

Not proven statement.

> SQL and relational algebra are computationally incomplete and
> there is no fitting alternative that is.

The same for the most OO languages and OO approach in general.

> - No extensibility. You can't add new basic data types.

Yes, you can.

> Ok, ok, so I'm not very objective. the reason for this is, that I believe
> that relational databases are frustrating the computer evolution towards
> global distributed computing.
>

That's what it is. You are just adept of the new religion called ODBMS.

> Richard
Received on Mon Jan 19 1998 - 00:00:00 CST