"Richard Heyes" <richardh@xxxxxxxxxxx> wrote in message news:481F096C.9040107@xxxxxxxxxxxxxx >> You obviously do not understand what "technically perfect" means when it >> comes to data normalisation. > > Obviously. > >> That's why solutions which are thrown together are often incapable of >> being expanded to include new requirements. > > I've never adovocated "throwing together" a solution. Merely that the > solutions proposed by programmers or database designers aren't always > necessary. I never use databases which have been designed by others, especially DBAs who have fancy but often unrealistic ideas on how databases should be designed. I design the databases myself, then write the code to access them. If I later find that I need to change the database design then I can do so without having to ask anyone's permission. > > If you create a "throw away" solution >> then never try to expand it, simply throw it away and start again. > > For what reason? If you create a working solution and never try to expand > it, chances are you don't need to, so what's the advantage in throwing it > away and generating work for yourself in reproducing it? Your experience in the real world must be very limited as it is often the case where a customer starts off with a "simple" requirement then keeps expanding it as time goes by as he dreams up more things that the system should do for him. If at day #1 you say "these reqirements are very simple, therefore require nothing more than a cheap and chearful solution which can be thrown together in a few minutes" you end up with a system that is not designed to be expanded. As soon as the first enhancement request comes in you will find yourself on the tail of a snake. -- Tony Marston http://www.tonymarston.net http://www.radicore.org > -- > Richard Heyes > > +----------------------------------------+ > | Access SSH with a Windows mapped drive | > | http://www.phpguru.org/sftpdrive | > +----------------------------------------+ -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
