in reply to Weird precision issues in float addition
The usual rounding errors.
if you just need 1 decimal place, why not scale it all up by 10?