[squeak-dev] Integer>>isPrime why reverting Knuth's algorithm P (probabilistic primality test) with iteration over division with even numbers?

Andres Valloud avalloud at smalltalk.comcastbiz.net
Thu Dec 24 20:53:34 UTC 2009


The problem I see is one of confusion.  If you ask an integer isPrime, 
then it would be reasonable to expect a deterministic answer, not a 
probabilistic one.

On 12/20/2009 4:19 AM, Enrico Spinielli wrote:
> Again my point is:
>
>     was there a problem to solve?
>     Was it about performance?
>
> If the answer is YES, then a fix (your proposal for example) is needed
> if NOT leave what was there originally...double implementation is just
> making maintenance more difficult.
> Optimizing for the sake of it has never proved correct...
>
> My 2 cents...
> Bye
> Enrico
>
> On Sat, Dec 19, 2009 at 11:18 PM, Cédrick Béler <cdrick65 at gmail.com
> <mailto:cdrick65 at gmail.com>> wrote:
>
>     or simply
>
>     SmallInteger>>isPrime
>     the current version
>
>     LargePositiveInteger>>isPrime
>     Knuth's version
>
>
>     --
>     Cédrick
>
>
>
>
>
>
> --
> Enrico Spinielli
> "Do Androids dream of electric sheep?"— Philip K. Dick
> "Hear and forget; see and remember;do and understand."—Mitchel Resnick



More information about the Squeak-dev mailing list