Is a randomized default value primary key possible? - Mailing list pgsql-novice

From April L
Subject Is a randomized default value primary key possible?
Date
Msg-id 3.0.5.32.20020519121004.06db8ec8@mail.i-netco.com
Whole thread Raw
In response to Re: Some CSV file-import questions  (Andrew McMillan <andrew@catalyst.net.nz>)
Responses Re: Is a randomized default value primary key possible?  (Ron Johnson <ron.l.johnson@cox.net>)
Re: Is a randomized default value primary key possible?  (Josh Berkus <josh@agliodbs.com>)
List pgsql-novice
After reading about data types, create table, and other search results for
"default" and "random" in the postgres idocs, I haven't found a way to do
this:

Instead of a sequentially auto-incrementing primary key, I would like a
random 16 byte character value (assumedly hex) to be automatically created
as the default value for each new record.

I guess I could just use a large random number, created with the postgres
random() function - however I don't see any way of seeding it, or know if
that is necessary.

The purpose of using a random rather than sequential number is to prevent
people being able to access other's records by guessing.

Has anyone else encountered this challenge, and do you know of a way to
generate a random default value for the primary key?

Thank you,

- April

pgsql-novice by date:

Previous
From: Andrew McMillan
Date:
Subject: Re: Some CSV file-import questions
Next
From: Tom Lane
Date:
Subject: Re: Forgotten the master password of db