the bcd20 library is a self contained implementation of binary coded decimal floating point arithmetic together with trig functions. most of the functions found on ordinary scientific calculators are implemented. the library delivers a precision of 20 decimal digits and has an internal working 28 digit mantissa. exponents range +/- 9999.
the library compiles under windows using visual c++ and also for windows/cygwin and linux using gcc.
bcd20 is used in Thomas Okken's excellent Free42 hp42s simulator.
download either bcd20v1.2.tgz or bcd20v1.2.zip
comments, suggestions and bug fixes welcome on the forum.