From: | Alex Pilosov <alex(at)pilosoft(dot)com> |
---|---|
To: | Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> |
Cc: | pgsql-sql(at)postgresql(dot)org |
Subject: | Re: temporary views |
Date: | 2001-10-05 21:01:33 |
Message-ID: | Pine.BSO.4.10.10110051658420.6702-100000@spider.pilosoft.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-sql |
On Fri, 5 Oct 2001, Bruce Momjian wrote:
>
> What actually should happen is that the view should go away at the end
> of the session. However, unlike indexes, we can have several tables
> involved in a view so it is hard to know exactly how to handle this.
>
> Seems like a TODO item, at least. What we could do is to create views
> as TEMP if they use temp tables and drop the view as soon as the session
> ends . You of course would have to recreate the view each time but
> because it is a _temp_ view, it could be done reliably by multiple
> backends at the same time.
Didn't someone suggest dropping saving of parsed plans with OIDs
altogether, and saving the underlying query instead? The point was that
parser and planner are fast enough to make it unnecessary to save plans.
I don't remember what was disposition of that idea....
-alex
From | Date | Subject | |
---|---|---|---|
Next Message | Peter Eisentraut | 2001-10-05 21:12:19 | Re: SQL CONSTRAINTS - Constraining time values from two |
Previous Message | Bruce Momjian | 2001-10-05 20:42:58 | Re: temporary views |