>I see from the "primes added" mail that

Actually this is correct. The page will take the floor and report 50896

>"10^50897-5*10^25448-1" has got 50897.999999999 decimal digits! I eagerly

>await the first prime with exactly e^(pi*sqrt(163)) decimal digits...

digits, but will

use the whole expression (roughly log(n)+1) to sort the list. The problem

is solely that

you got to see something you should not have--an internal format.

CC