[r-t] Big searches

Ian Broster ian at broster.co.uk
Sun Dec 16 23:15:00 UTC 2007

> machine search, measured by number of nodes searched?

Wouldn't a better metric be along the lines: "...an algorithm that can  
find [all of?] the compositions by searching the fewest nodes"

It is clear that most exhaustive searches are very large - it is important  
before starting a search to appreciate how long it will run for - if you  
are waiting thousands of years to search compositions that don't start  
HHWHH then your time is better spent on improving the pruning alorithm  
than waiting for it to finish.


More information about the ringing-theory mailing list