Re: double linked list

From: Ed Prochak <edprochak_at_adelphia.net>
Date: Wed, 12 Feb 2003 19:08:22 GMT
Message-ID: <3E4A9F2B.8050608@adelphia.net>

--CELKO-- wrote:

>>>[book on TREES & HIERARCHIES IN SQL] Are we rediscovering the past?
>>

> Network model databases have their place and can out perform a RDBMS
> in those cases. It's just that those cases are few and far between. In
> my experience, it's when the data is static, ie, load data once and
> read many times very fast. Dynamic data just bogs done on link
> updates. <<
>
> Actually, I spend most of the time showing how to code hierarchies
> without using links, but by taking a set-oriented approach, like the
> one I just did a "cut & paste" on.
>
> Actually, the old network and IMS databases are really fast to load
> and access. They just have no flexibility. It's like a Polar Bear --
> great in one environment and no place else.
>
> If you put parts subordinate to their suppliers, and want to get all
> the sources for a "Metric #5 Machine Screw", you will be chasing
> pointer chains from the supplier, thru his catalog, back to the next
> supplier, thru his catalog, etc.

Yes. I expected you knew this but it didn't hurt to ask. (I've seen others not know the past and proclaim their network model to the the "next big thing".)

Now what about the UPDATE issue I asked about. I was really more concerned with that. It reminds me somewhat of the old line numbering issue in BASIC. After so many inserts, you have top renumber a large number of records. How do you avoid that?

-- 
Ed Prochak
running: http://www.faqs.org/faqs/running-faq/
family:  http://web.magicinterface.com/~collins
--
"Two roads diverged in a wood and I
I took the one less travelled by
and that has made all the difference."
robert frost

Received on Wed Feb 12 2003 - 13:08:22 CST