Walter Bright <[email protected]> wrote:
bearophile wrote:
Walter Bright:
Are you sure? What's the mistake in the code?
This is the code in the Overview, it prints 1899:
http://codepad.org/lzRtggEL
This is the code I have suggested in bugzilla, it prints 1027:
http://ideone.com/D9ZqQ
Wolfram Alpha says they are 1027 (I have left out the last number
because Alpha uses a closed interval)
http://www.wolframalpha.com/input/?i=primes+in+2+..+8190
It's been printing 1899 primes since the 80's. Just for fun, google
[sieve of eratosthenes 1899 primes]
I wonder if that has to with there being 1899 primes in 3..16384.
http://www.wolframalpha.com/input/?i=primes+in+2+..+16384
An implementation of just that is what I found most of the time,
owing back to "A High-Level Language Benchmark" by Jim Gilbreath,
in BYTE Magazine, september 1981, pages 180-198.
Sadly, I cannot seem to find a version of the original article
online. Anyone know of one, that we may get to the bottom of this?
--
Simen
