Re: double linked list

From: Galen Boyer <>
Date: 11 Feb 2003 23:19:25 -0600
Message-ID: <>

On 3 Feb 2003, wrote:

>>> When you are ready to demonstrate a superior implementation in

> another SQL
> RDBMS please let us know. And I don't mean superior in terms of some
> theoretical concept like building a database in fifth normal form. I
> mean superior in terms of performance and scalability in the real
> world. <<
> Funny you should ask :)!! I have a book coming out later this year on
> TREES & HIERARCHIES IN SQL (Morgan-Kaufmann), which gives several
> different models in Standard SQL-92 for tree structures which are
> superior in terms of performance and scalability compared to the
> A few years back, we ran a test on a huge tree with the following
> method versus cursors and CONNECT BY; it was 20 to 100 times faster
> for 75,000 nodes and 12 levels deep. It could also do things like
> compare the structures of sub-trees in one SELECT statement.

You have done nothing, whatsoever to prove that Oracle is a horrible product. Nothing, zilch, nada!

What you have done is present a design that many have, I know I have, implemented to speed up access to hierarchies. But, you want to know what, we implemented these designs in Oracle. I deliberately try not to use connect by very often, but it comes in handy, especially for small sets and quick hits. What the flip is wrong with offering that up? When someone wants to really implement hierarchies and fast access to them, they design a home-grown one, like yours.

So, please tell us how Oracle is a "horrible product". This time, stop selling your book and show where another database vendor's solution to hierarchical queries is superior to Oracle's "connect by". Even then, you will still have done, nothing, zilch, nada to prove that Oracle is a "horrible product", but you can't even start to defend the statement because you haven't even shown one vendor's superior solution to Oracle's "connect by", not one.

Galen deForest Boyer
Sweet dreams and flying machines in pieces on the ground.
Received on Wed Feb 12 2003 - 06:19:25 CET

Original text of this message