Floating point numbers used here are really char strings dimensioned to six
places. ex: char float1[6],float2[6]. I admit it is very cumbersome changing
from strings to floats and back again but if you want floats with CC65 as it
stands this seems to be the only alternative.