From: | Chris Bitmead <chris(at)bitmead(dot)com> |
---|---|
To: | "Robert B(dot) Easter" <reaster(at)comptechnews(dot)com> |
Cc: | pgsql-hackers-oo(at)postgresql(dot)org, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: [HACKERS-OO] UNDER and INHERITS |
Date: | 2000-05-27 02:18:02 |
Message-ID: | 392F305A.8BB67EA3@bitmead.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
"Robert B. Easter" wrote:
> The official single
> inheritance UNDER, is designed to support inheritance of constraints
> and the sharing of indices from the maximal supertable down into its
> subtables. The
> maximal supertable is required to have some UNIQUE NOT NULL attribute
> for this purpose (SQL-1999 Foundation, Section 11.3, Syntax Rule 7.g).
> I feel that maybe standard single-inheritance UNDER and the current
> PostgreSQL extension, INHERITS, can be used together to complement each
> other. INHERITS provides a simple multiple inherit ability. UNDER
> provides a feature-rich single inheritance container where subtables
> are extensions the maximal supertable. One change I think is not
> unreasonable, is that INHERITS allow parent tables to
> be dropped. I'd like to know the reason why its not allowed now.
The reason dropping parent tables, and inherited indexes and constraints
don't work is that no-one has bothered to implement them.
In so far as creating an index on only one table might be useful (as is
the case in postgres now), the extension "create index on only table"
would seem appropriate. No sense on making blanket rules that under must
inherit them and inherits can't.
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2000-05-27 02:22:46 | Re: SPI & file locations |
Previous Message | The Hermit Hacker | 2000-05-27 02:04:49 | PostgreSQL v7.0 branched ... |