Re: pgsql/src/backend catalog/heap.c utils/adt/tid.c

From: Hiroshi Inoue <Inoue(at)tpf(dot)co(dot)jp>
To: Neil Conway <nconway(at)klamath(dot)dyndns(dot)org>
Cc: Hiroshi Inoue <inoue(at)postgresql(dot)org>, pgsql-committers(at)postgresql(dot)org
Subject: Re: pgsql/src/backend catalog/heap.c utils/adt/tid.c
Date: 2002-05-23 04:00:16
Message-ID: 3CEC6950.33A48395@tpf.co.jp
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers

Neil Conway wrote:
>
> On Wed, 22 May 2002 03:46:58 -0400 (EDT)
> "Hiroshi Inoue" <inoue(at)postgresql(dot)org> wrote:
> > Log message:
> > Allow
> > CREATE VIEW as SELECT CTID, ....
> > SELECT currtid( a view, ..).
>
> Hiroshi,
>
> Can you elaborate on exactly what your modification allows?

Before the change, views have meaningless system columns.
After the change, views have no system columns basically
and users could add columns with system column name.

My main intention is to allow scan by TIDs on views.
Scan by TIDs is the fastest random access to PG tables
and I'm using the scan to implement updatable cursors
in ODBC driver. I'd not like to discourage users to
use views.

> Also, would it be appropriate to add regression tests for this?

Not sure.
AFAIK few people are interested in system columns
other than OID.

regards,
Hiroshi Inoue

In response to

Browse pgsql-committers by date

  From Date Subject
Next Message Hiroshi Inoue 2002-05-23 07:49:39 Re: pgsql/src/backend/catalog heap.c
Previous Message Tom Lane 2002-05-22 21:46:40 pgsql/src/backend Makefile