Re: UniBasic code that should work

Hi
Yes and no. Always store in Internal format - ICONV and always calculate in
external format - OCONV or just do a division as Jeremy did. But get the
precision right for comparisons by getting rid of any trailing bits.
According to Symeonb the command in Unjdata is
FLOAT.PRECISION is the command you need in unidata to configure your wide 0
Regards
Peter McMurray

<symeonb@xxxxxxxxx> wrote in message
news:1143635195.763568.171740@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Personnally I would always make sure my numbers are iconv'd correctly
otherwise you can get in a right old mix !



.