Re: pointers on representing tree in db?
Date: 21 Apr 2001 01:41:27 GMT
Message-ID: <9bqog7$8tr$1_at_geraldo.cc.utexas.edu>
Steve Long <steven.long_at_erols.com> wrote:
: normal forms are more academic than practical. often, data is denormalized
: in systems where selects are more common than updates, inserts, and deletes.
: normalization makes retrieval slower but I, U, and D are faster.
: show me a database that is 3NF (or higher) and i'll show you a database that
: is as slow as molasses in january at the equator.
I'll believe you on 5NF, but I won't give you the time of day dissing 3NF. Show me a system where the designers eschew normalization, and I'll show you a system where no one knows where data lives, which table is authoritative, where applications don't update all redundant columns, where queries are just plain wrong.
Normal forms are not ivory tower wanking -- they are essential to good relational design. Received on Sat Apr 21 2001 - 03:41:27 CEST