Account Options

  1. Sign in
The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Google Groups Home
« Groups Home
sci.crypt
+ new post
About this group
Subscribe to this group
This is a Usenet group - learn more
Message from discussion Baillie-PSW - Which variant is correct?
The group you are posting to is a Usenet group. Messages posted to this group will make your email address visible to anyone on the Internet.
Your reply message has not been sent.
Your post was successful
 
From:
To:
Cc:
Followup To:
Add Cc | Add Followup-to | Edit Subject
Subject:
Validation:
For verification purposes please type the characters you see in the picture below or the numbers you hear by clicking the accessibility icon. Listen and type the numbers you hear
 
Marcel Martin  
View profile  
 More options Jan 9 2004, 5:14 pm
Newsgroups: sci.crypt
From: Marcel Martin <m...@ellipsa.no.sp.am.net>
Date: Fri, 09 Jan 2004 23:12:44 +0100
Local: Fri, Jan 9 2004 5:12 pm
Subject: Re: Baillie-PSW - Which variant is correct?
Print | View thread | Show original | Report this message | Find messages by this author
Henrik a écrit :

> While searching for Baillie-PSW I found a few variants:

> A. n = 2 or 3 (mod 5); base-2 pseudoprime (Fermat); Fibonacci
> pseudoprime.

> B. strong pseudoprime to base 2; passes Lucas test A.

> C. base-2 strong pseudoprime; in the sequence 5, -7, 9, -11, 13,...
> find the first number D for which (D/n) = -1; Lucas pseudoprime test
> with discriminant D on n.

> D. base-2 pseudoprime; Lucas test; n = 2 (mod 5); Lucas  pseudoprime
> test.

> Now some questions:
> 1. All of the above are claimed to be Baillie-PSW. This confuses me.
> Which one is the right one?

Probably all of them.
Personally, for Primo I use the variant C (except the test is a Lucas
strong pseudoprime test, i.e., not just a Lucas pseudoprime test).

> 2. My sources are old. Are there counter-examples to the Baillie-PSW
> test today?

AFAIK, none. Not only none was found but nobody succeeded in building
a counterexample.
Primo is used since now more than 3 years and, for each primality
certificate it produces, all intermediate 'primes' are checked with
this test. If one was composite, the certification would necessarily
have failed. This never occurred. In fact, it is presumably that no
composite less than, say, 10000 digits can fool this test.

BTW, in my docs, I don't use "Baillie-PSW" to refer to this test but
simply "BSW" (Baillie-Selfridge-Wagstaff). I have nothing against
Pomerance. Just, I reported the name I found in François Arnault's
thesis (page 72).

mm
--
http://www.ellipsa.net/
m...@ellipsa.no.sp.am.net  ( suppress no.sp.am. )


 
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
©2013 Google