From: MikkelFJ <mikkelfj-anti-spam@...>
Date: 2002-11-29T08:55:28+09:00
Subject: Re: A lesson learned


"Gavin Sinclair" <gsinclair@soyabean.com.au> wrote in message
news:140701c296ef$8538af70$436032d2@nosedog...
> Folks,
>
> Today I rediscovered the truth of that very important saying in software
> developement: "Make it work, make it right, then make it fast."

and D. Knuth: "early optimization is the root of all evil".

I see your point, but I disagree to some extend. The approach is right, but
using a plain stupid algorithm that doesn't scale is generally bad. Things
get abused and used in ways you can't imagine.
It can take a looong time to get code sorted out once in production and it's
difficult to make fundamental changes. Today IBM developer works weekly
letter had a point about "no amount of work can compensate for an intially
bad design", drawing parallels to an exploding turkey (the editor is
obviously insane).
Intelligent design is good. Overoptimization and unnecessary complexity is
bad.
To put this back to Ruby - it's fine to start out in Ruby, make it work and
work right, probably plenty fast, otherwise you already have the right,
intelligent approach that can be optimized in a faster language.

Mikkel