Pete,
Floating point precision is a bit confusing.
(1) The original and still the primary use of fpprec is to control
precision in bigfloat computations. (Bigfloat = Maxima's
arbitrary precision float.)
(2) fpprec is also used in EXPLODEN (src/commac.lisp)
to tell how to format ordinary floats for output.
(3) There is also a global variable fpprintprec which controls
formatting for bigfloats.
I believe (2) is suboptimal; either it should be changed to fpprintprec
(and the documentation for fpprintprec should say it applies to
both floats and bigfloats) or maybe we can invent a separate
variable to control float format only. At the moment I'm inclined
to just use fpprintprec.
Probably tex needs to be updated to respect the precision flag,
whatever we decide for it to be.
FWIW
Robert Dodier