Hi all, This is just to let everybody on this list know that I have updated my arbitrary precision decimal arithmetic package for Calc.
http://kohei.us/ooo/decimal/ Since my last announcement, I have fixed numerous calculation bugs and added more functions. It now supports the following functions: DEC_ADD( left operand; right operand; precision ) DEC_SUBTRACT( left operand; right operand; precision ) DEC_MULTIPLY( left operand; right operand; precision ) DEC_DIVIDE( left operand; right operand; precision ) DEC_EXP( number; precision ) DEC_PI( precision ) DEC_SIN( number; precision ) DEC_COS( number; precision ) DEC_TAN( number; precision ) Again, I have provided a ready-to-use binary package for Linux for those interested in testing out this package with minimal effort. It is compiled on FC3, so YMMV. This UNO package is a good demonstration of what the arbitrary precision decimal arithmetic can do. Ultimately, I would like to see the underlying Decimal data type, which is implemented in C++, integrated into the main code base so that it can be used internally for types of calculations that require extra precision, with no rounding errors inherent in binary double-precision. However, the current implementation is terribly (and somewhat embarrassingly) slow to be useful at the moment, so the performance issue needs to be addressed before we can talk about possible integration. But I would like to know, if there is any interest among the core developers for its potential integration into the main code base if/when the performance issue gets resolved. Niklas? Eike? :) For those who know how to make my code run faster, please, feel free to poke at the code and give me feedback (or patch, JCA signed). I will greatly appreciate it. You all have a good day, Kohei -- Kohei Yoshida OpenOffice.org Calc contributor http://kohei.us/ooo/ --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
