On Sunday 20 November 2011, 13:53:20, Sebastian Hungerecker wrote:
On 20.11.2011 07:50, Haisheng Wu wrote:
Hmm... I think I made a little confusion so I put my finding here: http://haisgwu.info/posts/2011-11-20-euler-problem-14.html
I do got stack overflow thus need several compile opts to fix it. Not sure if it is what you mean by "You get overflow using 32-bit types here."
No, what he meant was that your n does not always fit into a 32-bit integer, so the integer will overflow (i.e. once it exceeds its maximum value it will wrap around and become negative).
That. With Word32 you won't get negative values of course, but the wrapped-around values are completely bogus nevertheless. With signed 32-bit integers, the overflow puts you into an infinite loop of negative values for several starting values, the smallest is something around 113000 iirc. With Word32, I never tested what the smallest starting value landing you in an infinite loop is, but (2^32-1)/3 surely does (~> 0 ~> 0).