RE: it works! wasn't thought out
- here's how to catch a 2-pseudoprime... off the cuff...
if `I' is already prime, then the algorithm lets it slip through;
if `I' is a pseudo-prime, then it catches it before saying it's prime!
is my idea creative enough ??? can someone verify it ???
my computer is very, very primitive
Mr./Dr. Kosinar... give it a try now, reluctantly, or otherwise.
the method below is unproven, but not unfounded... and a bit much to
take in... I'd even take a comment from the fatphil Carmody, if he's
still in the Group
15 FOR I= 7 TO 2001
20 IF 2^(I-1) MOD I= 1 THEN //the usual 2-PRP test
25 Y= 2^(((I-1)/2)-3) MOD I
30 YY= I MOD 4
32 // check for pseudoprime before saying the number is prime!
35 IF YY= 1 AND ((Y> I/8 AND Y< I/2) OR (Y> I*3/4 AND Y< I*7/8))
40 PRINT I, "composite"
45 END IF
50 IF YY= 3 AND (Y< I/8 OR (Y> I/2 AND Y< I*3/4) OR Y> I*7/8) THEN
55 PRINT I, "composite"
60 ELSE PRINT I, "prime"
65 END IF
70 SLEEP 1
75 END IF
80 NEXT I