Re: Check constraints on non-immutable keys

On Wed, Jun 30, 2010 at 9:47 AM, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> We currently allow this:
>
> postgres=# create table t(a timestamptz not null primary key, check(a > now()));
> NOTICE:  CREATE TABLE / PRIMARY KEY will create implicit index
> "t_pkey" for table "t"
> CREATE TABLE
>
>
> Which seems very wrong. For one thing, a dump of this database can not
> be restored if now() has advanced enough into the future (which it
> will eventually). It also makes impossible to do things like SET a=a
> on the table.
>
> Yes, this is clearly a stupidly defined constraint, but why do we allow it?
>
> Shouldn't we disallow anything that's not IMMUTABLE in a check constraint?

suppose you did do this: shouldn't you then also recheck the
constraint if the function is create/replaced?

merlin