
.BASIC/S II does not directly support floating point
manipulations, but there are three files on the disk
that will allow a limited form of floating point calculations
from within BASIC/S II compiled code.  The file FLOAT/BAS
is a BASIC/S I compilable program (BASIC/S II will NOT
compile it!) which compiles into FLOAT/CIM (also on the
disk).  You can write code to be compiled by BASIC/S II
which can do floating point calculations via USR calls
to FLOAT/CIM.  Such a program should be compiled at
the usual (default) starting address, 5200H, and should
do a DEFUSR=-9536 at the start (to set up the USR call
to FLOAT/CIM).  What this does is give you access to
two floating point variables, A! and B! (sometimes referred
to by just A and B).  Use the BASIC/S II integer variable
C to set up the function code that tells FLOAT/CIM what
to do; the various codes are listed in FLOAT/BAS.
SQR/BAS is a BASIC/S II compilable program that is heavily
commented and is an example of how to use FLOAT/CIM.
For example, to compute the LOG of 2.5 and print it out
on the video, you could write a BASIC/S II program as
follows :

10 DEFUSR=-9536:A$="2.5":C=19:'function code for VAL
20 X=USR(0):' this sets A!=VAL(A$), ie 2.5
30 C=14:'code for LOG
40 X=USR(0):'set A!=LOG(A!)
50 C=18:'code for STR$
60 X=USR(0):'set A$=STR$(A!)
70 PRINT A$:'print out the result!
80 END

See SQR/BAS for another such example.
To execute such a compiled program, FLOAT/CIM needs to be
in memory; just LOAD it from DOS READY (it is around 820
bytes long, and starts at DAC0 hex (56000)).  To make
a stand alone /CMD file, you can append FLOAT/CIM onto
the end of your /CMD file using Roy Soltoff's CMDFILE.
