Re: Database Independence: Approaches and Achievability

From: Todd Gillespie <toddg_at_linux128.ma.utexas.edu>
Date: Thu, 6 Sep 2001 20:50:39 +0000 (UTC)
Message-ID: <9n8niv$k7v$1_at_geraldo.cc.utexas.edu>


Jackie Burhans <jackie.burhans_at_usa.net> wrote:
: I have been trying to do some research on "database independence" and
: have not had much luck. I've seen in this forum topics about "data
: independence" logical vs. physical and even distribution independence
: (not to mention one spam about financial independence. But nothing on
: what I hear termed "Db independence". To the folks I talk to "DB
: independence" seems to mean that I have an application that talks to
: DBx and I can swap it out for DBy at any time? Seems to me that's a
: bit simplistic and, curiously, I don't see many people in this forum,
: on the web or anywhere else talking about how to do this and the
: extent to which its possible. Any ideas?

Probably b/c it's not really a theoretic domain. It's a one-off hack project. 99% percent of the people who describe DB independence are Java folks who are a bit uneasy with databases. Their approach with DB independence is to write lowest-common denominator SQL that does as little as possible, and then do the bulk of the JOINing and other working inside the client application. The performance of this approach is as slow as you expect. I recommend smacking these people. The last 1% will write their SQL into a hash table and key it on which DB is being addressed ATM. So they can still write real SQL, and tune it to each platform.

This really isn't a DB problem, this is an application design problem. Maybe you could clarify what you're looking for here? Received on Thu Sep 06 2001 - 22:50:39 CEST

Original text of this message