[Shootout-list] Binary trees (ruleS)

[Bengt Kleberg]

On 2005-06-19 18:33, Jon Harrop wrote:
...deleted
> For me, the propect of having Isaac throw a submission back in my face because 
> of a rule he just imagined is the strongest reason to not join in development 
> of the shootout. 

if we replace the name with ''somebody'' i agree.  this is (imho) a good 
reason to have explicit rules for what tests that are acceptable, and 
what is not acceptable. presumably those that have this power currently 
prefers things as they are.
(http://lists.alioth.debian.org/pipermail/shootout-list/2005-May/002012.html)

> Also, these problems with binarytrees will disappear if we follow Skaller's 
> advise for designing the tests. This test shouldn't state that you must use a 
> binary tree, it should simply require you to solve a problem that is best 
> attacked with a binary tree.

this is ok for those that are good enough at specifying tests. i have 
not been able to do this, and would still like to design tests. i think 
this is why we have ''same way'' and ''same thing''. if one is really 
good at test design it is possible to do ''same thing''. i have to be 
content with ''same way''.


bengt