[Shootout-list] Many tests cheat

[Isaac Gouy]

--- skaller <skaller@users.sourceforge.net> wrote:
> Hmm, well purely functional systems can't solve
> certain problems efficiently. Arrays are a well 
> known example, the best performance is 
> around O(log n), most people believe it can be 
> proven functional arrays can't exist although it
> hasn't be done yet. 

Isn't Clean a pure functional language?
Referential transparency and destructive update.

-snip-
> The intent of my comment is not, however, to suggest
> the test is inadequate, although that may be the case.
> Rather to point out that defining what an array is,
> and requiring some vague 'do it the same way' constraint
> is not necessary nor sufficient to achieve fairness,
> and is most likely to be an ill formed requirement,
> unless it actually refers to an observable quantity
> that can be checked.
> 
> AFAICS there are only 4 observables: time, memory,
> lines of code, and the actual required output.

You are going to move beyond simple criticism and provide ideal test
definitions, aren't you?



		
__________________________________ 
Do you Yahoo!? 
Check out the new Yahoo! Front Page. 
www.yahoo.com