begin  quoting [EMAIL PROTECTED] as of Sun, Mar 16, 2008 at 11:07:19PM -0700:
> On Sun, Mar 16, 2008 at 10:15:29PM -0700, David Brown wrote:
> > The distinction I'm trying to make is writing the documentation before
> > writing the code.  Non trivial code is usually not well defined until it is
> > being written or has been written.  Thinking otherwise is mostly an
> > illusion.
> 
> I would agree with that.  In the signal processing consulting work I've done I
> always seem to put it in words before I try to code it.

I tend to fill up a LOT of scratch paper.

Eventually, I get a picture that I'm not totally unhappy with.

Not really something that fits into words...

> Some test driven development fan (TDD!) at Pycon writes his *tests* before
> writing any code.  I haven't gone that far yet.

I've tried that. It's difficult.

It works a lot better with pair-programming.

-- 
The best design approach for me appears to get in to an argument.
Stewart Stremler

-- 
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-lpsg

Reply via email to