Re: [HACKERS] pg_dump problem? - Mailing list pgsql-hackers

From Chris Bitmead
Subject Re: [HACKERS] pg_dump problem?
Date
Msg-id 3732E240.A347F511@bigfoot.com
Whole thread Raw
In response to Re: [HACKERS] pg_dump problem?  ("D'Arcy" "J.M." Cain <darcy@druid.net>)
Responses Re: [HACKERS] pg_dump problem?
List pgsql-hackers
> Ultimately I think you need to get away from using OIDs in your 
> top level applications.

I don't give a rip about standard SQL. What I care about is real object
databases. A fundamental principle of object theory is that objects have
a unique identity. In C++ it is a pointer. In other languages it is a
reference. In an object database it is an oid. In the NSHO of a fellow
called Stonebraker, you should be using oids for everything.

BTW, I was looking through the original 4.2 docs, and I noted that in
Postgres 4.2 every class had not only an oid, but an implicit classoid,
allowing you to identify the type of an object. What happened to this?
It would solve just a ton of problems I have, because I'm using a very
OO data model. It sounds like Postgres used to be a real object
database. Now everybody seems to want to use it as yet another sucky rdb
and a lot of essential OO features have undergone bit-rot. What happened
to building a better mouse trap? 

Have a read of shared_object_hierarchy.ps in the original postgres doco
to see how things should be done. Sorry for the flames, but I used to
work for an ODBMS company and I'm passionate about the benefits of
properly supporting objects.
 Depending on them causes these kinds of problems
> and moves you farther from standard SQL in your app.  Use of the OID
> (IMNSHO) should be limited to temporary tracking of rows and even then
> it should be in middle level code, not the top level application.  I
> offer the use of OIDs in pg.py in the Python interface as an example
> of middle code.
> 
> I suggest that you replace the use of OID in your database with a serial
> type primary key.  That allows you to dump and reload without losing
> the information and it performs the same function as OID in your code.
> 
> --
> D'Arcy J.M. Cain <darcy@{druid|vex}.net>   |  Democracy is three wolves
> http://www.druid.net/darcy/                |  and a sheep voting on
> +1 416 424 2871     (DoD#0082)    (eNTP)   |  what's for dinner.

-- 
Chris Bitmead
http://www.bigfoot.com/~chris.bitmead
mailto:chris.bitmead@bigfoot.com


pgsql-hackers by date:

Previous
From: "D'Arcy" "J.M." Cain
Date:
Subject: Re: [HACKERS] pg_dump problem?
Next
From: Tom Lane
Date:
Subject: Re: [HACKERS] cvs 6.5 regression tests on Linux x86