From: Tom Lane <tgl@sss.pgh.pa.us>
To: Shigeru Hanada <shigeru.hanada@gmail.com>
cc: PostgreSQL-development <pgsql-hackers@postgresql.org>
Subject: Re: inherit support for foreign tables
In-reply-to: 
 <CAEZqfEfVJNptoTSEY3QrQ90K9WD2nxfM+6fsk3YTc1ZXZ1Y5-g@mail.gmail.com>
References: 
 <CAEZqfEfVJNptoTSEY3QrQ90K9WD2nxfM+6fsk3YTc1ZXZ1Y5-g@mail.gmail.com>
Comments: In-reply-to Shigeru Hanada <shigeru.hanada@gmail.com>
 message dated "Thu, 14 Nov 2013 13:51:40 +0900"
Date: Thu, 14 Nov 2013 00:28:26 -0500
Message-ID: <12432.1384406906@sss.pgh.pa.us>
Precedence: bulk
Sender: pgsql-hackers-owner@postgresql.org

Shigeru Hanada <shigeru.hanada@gmail.com> writes:
> I'd like to propose adding inheritance support for foriegn tables.
> David Fetter mentioned this feature last July, but it seems stalled.
>   http://www.postgresql.org/message-id/20130719005601.GA5760@fetter.org

The discussion there pointed out that not enough consideration had been
given to interactions with other commands.  I'm not really satisfied
with your analysis here.  In particular:

> 2) Allow foreign tables to have CHECK constraints
> Like NOT NULL, I think we don't need to enforce the check duroing
> INSERT/UPDATE against foreign table.

Really?  It's one thing to say that somebody who adds a CHECK constraint
to a foreign table is responsible to make sure that the foreign data will
satisfy the constraint.  It feels like a different thing to say that ALTER
TABLE ADD CONSTRAINT applied to a parent table will silently assume that
some child table that happens to be foreign doesn't need any enforcement.

Perhaps more to the point, inheritance trees are the main place where the
planner depends on the assumption that CHECK constraints represent
reality.  Are we really prepared to say that it's the user's fault if the
planner generates an incorrect plan on the strength of a CHECK constraint
that's not actually satisfied by the foreign data?  If so, that had better
be documented by this patch.  But for a project that refuses to let people
create a local CHECK or FOREIGN KEY constraint without mechanically
checking it, it seems pretty darn weird to be so laissez-faire about
constraints on foreign data.

			regards, tom lane


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers