Representation of float and smallfloat in ISQL-SE - DOS

I have a DOS application which uses C-ISAM to store "C" record structures. Most of the C record structures contain C "double" types. I wish to access this C-ISAM data through ISQL-SE but, when I create and run the corresponding SQL schemas, I notice that ISQL is converting the "float" definitions to "DECIMAL(16)". The format of DECIMALs is not IEEE and therefore I cannot access the numeric data in the files. WHY DOES INFORMIX CONVERT FLOATS TO DECIMAL UNDER DOS ? Short of re-working all my applications code to use DECIMALs instead of doubles, does anyone know of a solution or are Informix going to release a version of ISQL-SE on DOS that supports true IEEE floats/doubles? If it matters, my current version of ISQL is 2.10.06E. Thanks in advance. Nick Rumble