[r-t] Big search

Mark Davies mark at snowtiger.net
Tue Jan 22 08:49:54 UTC 2008

My Big Search has finished, uncovering a bug in my code in the process. I 
must have used a 32-bit register for the "time taken" calculation, which for 
some reason is performed in milliseconds. 2^32 milliseconds is a mere 1193 
hours, so the count wrapped over at that point.

Total running time was in fact 1271 hours, with 287.9 trillion nodes visited 
and 458 billion compositions found and proved - that's just over 100,000 per 
second throughout the search.

Now to see if there's anything decent in the results... ;-)


More information about the ringing-theory mailing list