Re: unique index on fields with possible null values

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Dmitry Ruban <dmitry(at)ruban(dot)biz>
Cc: pgsql-sql(at)postgresql(dot)org
Subject: Re: unique index on fields with possible null values
Date: 2007-07-27 02:49:31
Message-ID: 6521.1185504571@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-sql

Dmitry Ruban <dmitry(at)ruban(dot)biz> writes:
> I'm trying to find a solution for unique index on fields with possible
> null values.

You appear to be hoping that a unique index would constrain a column to
contain at most one null entry. It doesn't work like that, and I
strongly urge you to reconsider what you're using null for. I think you
are in for a world of hurt, well beyond this one particular point,
because the SQL spec's semantics for null are not designed around the
idea that it represents a single distinguishable value. What it
represents is "unknown".

regards, tom lane

In response to

Browse pgsql-sql by date

  From Date Subject
Next Message Dmitry Turin 2007-07-27 06:40:13 Re: calling webservice through postgresql function
Previous Message Paul Lambert 2007-07-27 02:18:36 Re: Join question