perhaps inefficient implementation

[itnygr2]

hi, i'm fairly new to the sieve of atkin. but i think i found a python project that runs much faster than this one. so i think there is some problem that slows down the implementation perhaps.

the python project: https://github.com/mccricardo/sieve_of_atkin

that project used around 15 seconds to generate primes up to 10^7, but this project used around 15 seconds to generate primes up to 10^5.

i used the c++ implementation (which should be faster than anything python), as i do not run maple on my computer.