Content-Type: text/plain;
  charset="iso-8859-1"
From: Josh Berkus <josh@agliodbs.com>
Reply-To: josh@agliodbs.com
Organization: Aglio Database Solutions
To: <mallah@trade-india.com>
Subject: Re: Getting NEW and OLD in ordinary functions.
Date: Wed, 9 Apr 2003 12:13:57 -0700
User-Agent: KMail/1.4.3
Cc: <pgsql-sql@postgresql.org>
References: <200304091551.02277.mallah@trade-india.com>
	<200304091123.16123.josh@agliodbs.com>
	<1461.219.65.233.94.1049915353.squirrel@mail.trade-india.com>
In-Reply-To: <1461.219.65.233.94.1049915353.squirrel@mail.trade-india.com>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Message-Id: <200304091213.57184.josh@agliodbs.com>

Rajesh,

> This was for simulating CHECK (column in ('A','B' , 'C')) ;
>=20
> I already have 5 FKEYS  on that table dont' u feel fkeys  for this
> purpose is an overkill?
> eg here the valid values can range from say 3 to 10 (small number)
> whereas FKEYS are capable to check in a very large set residing in
> another table.
> also fkeys makes updates and inserts slower  (i feel)

The problem with CHECK constraints is:
1) If you forget what the range of valid values for the column is, you have=
 to=20
dump the constraint definition to find out;
2) If you want to add a new valid value, you have to:
	a. Freeze all data entry on the table
        b. delete the constraint
        c. replace the constraint
        d. test that the new constraint does not bug out
        e. restart database access
adding a new value to a reference list is *much* easier.

> another question i have seen CHECK traslating IN to a set of ORs
> does the same happen in pl/pgsql internally?  and does it really matter
> for here for small target sets?

No idea, and probably not.

--=20
-Josh Berkus
 Aglio Database Solutions
 San Francisco