As I said yesterday:

Or just post "interesting" solutions to problems.
Forum readers can decide for themselves which are 
the interesting attributes.

- number of characters
- number of tokens
- time
- space
- innovative use of the language
- elegance
- etc.

That is, the "rules" as they have been all along.



----- Original Message -----
From: greg heil <[EMAIL PROTECTED]>
Date: Thursday, August 31, 2006 9:29 am
Subject: Re: [Jprogramming] Re: Sierpinski triangle

> On 8/31/06, Roger Hui <[EMAIL PROTECTED]> wrote:
> > However, comb takes time and space linear in the size of the 
> result whereas comb1 takes exponential.  Yet comb1 would have been 
> preferred according to the "number of characters" criterion.
> 
> Switching to a words criteria would not affect that. Any one
> dimensional valuation will have the same kind of deficits. ... unless
> there is a robust "market" of intelligent "bidders" which backs a
> "price" valuation. Even so some "bidders" will have variations in
> "preferences". And most "markets" have "externalities" which are not
> fully "costed".
> 
> A measure based on actual storage costs and easily evaluated has some
> constancy which supports a broad valuation of solution classes. In a
> sense the character metric is one of the broadest and most convenient.
> Metrics based on performance are either inexact (eg big O) or have
> constricting architectural assumptions (counting assembly language
> ops), and can be hard to evaluate. Characters are simple and broadly
> understood.


----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to