*read-default-float-format*?



While testing the converted SLATEC routines for Bessel functions, I've
come across a small problem.  *READ-DEFAULT-FLOAT-FORMAT* is
unconditionally set to 'double-float in clmacs.lisp.  This file is
loaded early in the compilation process so it causes 0.1 to be read as
a double-float when, in fact, f2cl really wanted that to be read as a
single-float. 

f2cl can't portably force a single-float 0.1 to be printed as 0.1f0,
so f2cl can't do this.  However, since this only happens in a couple
of places, I can easily manually change 0.1 to 0.1f0 so that it is
read correctly, but I prefer not to do this because I think the
conversion from Fortran to Lisp should be untouched.

Is there really a need to set this?  Couldn't this be bound in the
top-level loop?  Or do we really need to ensure that all numbers read
are double-floats while compiling?

Ray