On Wed, Mar 20, 2002 at 05:08:33PM +0000, Paul Millar wrote:
>
> The way I think of this is if you're looking at the overall picture, then
> you are choosing which algorithm is best for a particular "job".
That's all down to design. I'm gathering that I misundestood what you
were getting at in the first place. Or decided I wanted to rant. :)
With a nice sound design, it's often easy to implement a noddy algorithm
to get the basic system running, then go back and improve it later.
The example that springs to mind is a graphics program I was working on
a couple of years ago. The idea was to draw a spline path[1]. First of
all, I took the noddy approach of drawing straight lines between the
control points and having the ball follow them. Once I had that (and
the rest of the program) working, I modified it to use bezier curves to
draw the path.
And since I was really getting into it, I got bored and replaced the
bezier curve code with Catmull-Rom which produced the smoothest path
through the control points. :)
But yeah, I get your point.
[1] Through a dog's stomach, duodenum and oesophagus (sp?) though not
necessarily in that order. Don't ask. :)
--
[EMAIL PROTECTED] http://www.wossname.org.uk/~mathie/
--------------------------------------------------------------------
http://www.lug.org.uk http://www.linuxportal.co.uk
http://www.linuxjob.co.uk http://www.linuxshop.co.uk
--------------------------------------------------------------------
