Talk:Pernicious numbers

From Rosetta Code

C solution too cryptic?

Hi, the C solution is a bit mysterious to me. The for loop seems to "remove the zeros" if I understand correctly, but then what's so special about 2693408940 ?? How can a simple bitwise & tell if the number of 1 (which is now the length since all zeros have been removed) is prime? A bit of explanation in the introduction of the section or in comment would be welcome--Grondilu (talk) 14:43, 12 March 2014 (UTC)

Edit, hum after some analysis it appears that 2693408940 encodes the prime numbers in its binary form: 10100000100010100010100010101100. From right to left, starting from 0, the one marks the primes. Sill this could use some explanations.--Grondilu (talk) 14:51, 12 March 2014 (UTC)
OK, I understand now. I added some comments in the code. Hope it's ok.--Grondilu (talk) 15:02, 12 March 2014 (UTC)
Er yeah, it was pretty cryptic, and your analysis is absolutely correct. But I changed the routine again for something faster, though I did put in some comments this time. I didn't retain the "0b010011..." literal format since it's a GCC exteinsion, but if you want it back in I'm not really troubled by it. --Ledrug (talk) 19:40, 12 March 2014 (UTC)

Wiki link removal

Why was the Wikipedia link removed (for Pernicious number) ? -- Gerard Schildberger (talk) 20:00, 12 March 2014 (UTC)

The "hidden link" to Wiki's link within the definition of pernicious number isn't obvious unless one hovers over it and then Rosetta Code then underlines/underscores it, indicating that it's a link (either that, or my eyes are getting really bad).   I can't see the harm on having an explicit link listed with the others. -- Gerard Schildberger (talk) 20:16, 12 March 2014 (UTC)