A New Class of Unsafe Primes

[Peter Gutmann]

The IACR ePrint archive contains a paper from 2002 titled "A New Class of
Unsafe Primes", http://eprint.iacr.org/2002/109, which proposes a fast way to
find a prime factor p of an integer n when 4p - 1 has the form db^2, where d =
{ 3, 11, 19, 43, 67, 163 }.  I haven't been able to find any references to
this anywhere, is this something like the p +/- 1 factoring methods where the
values they're effective against are so unlikely that they can be safely
ignored, or is it just that no-one's ever noticed this paper?

Peter.