Re: Tables and Views
Date: Sat, 02 Oct 2004 02:42:50 +0100
Message-ID: <sq1sl09bikeu7ovova677j67hqi073u8de_at_4ax.com>
On Sat, 25 Sep 2004 11:09:20 -0400, "Laconic2" <laconic2_at_comcast.net> wrote:
>In another discussion, the comparison between l-values and r-values was
>being made. Somebody said, "congratulations, you've invented the virtual
>expression"
>
>That started me thinking down a different path:
>
>What if we've been thinking backwards all these years. We've been saying,
>one way or another, that a view is an expression pretending to be a store.
Eh? A view is many things, but not that.
It's a way of providing a consistent interface to the actual data, even after changes to the underlying tables.
It's a way of making queries on the data more concise (put the complex joins and where clauses in the view rather than the queries).
It's a security feature, allowing access only to certain subsets of the data to different classes of user.
It's a window onto the data. It isn't, nor does it pretend to be, a store.
>How about looking at it the other way? How about viewing a table as a store
>pretending to be an expression?
>Where does this lead?
Madness. Utter madness.
Lemming
-- Curiosity *may* have killed Schrodinger's cat.Received on Sat Oct 02 2004 - 03:42:50 CEST