Variable Precision Floating Point?

[Rana Ian]

Christoph & Joannes,

Hi. Are you still interested in offering a variable precision floating point classes? If so, and an open source c# implmentation already exists, would you be interested in using it?

Harry Smith is offering his C# VPCalc as freeware with source. I've asked him if he would be interested in contributing some of his classes to Math.Net.

I have an interest in being able to use a robust, open source math library in C#. I remember a recent Math.Net wishlist including variable precision floating point. Having recently found Harry's variable precision floating point source I thought, "Wouldn't this be a great help to Math.Net?"

You can download Harry's variable precision floating point code at http://pw1.netcom.com/~hjsmith/Calc/XPCalc.html.

Best,

Rana Ian

[Rana Ian]

All,

I got a very nice email from Harry saying he is open to contributing his code to Math.Net. XPCalc, which has the c# variable precision floating point classes, is already licensed under GPL. But it was nice to hear from Harry all the same. A perusal of XPCalc code shows the GPL license.

So, are these classes of interest to Math.Net? And if so, who wants to add the classes? Shall I?

=)

Rana

    http://home.netcom.com/~hjsmith/Calc.html

You will find the following notice in its source code

/// <summary>
///         XPCalc - Extra Precision Floating-Point Calculator
///         Developed in TURBO Pascal 5.5 and 6.0, converted to Visual Basic 6
///         and now in Visual C# .Net
///
/// ---------------- *** LEGAL NOTICE: *** ------------------
///
/// This program is free software; you can redistribute it and/or modify
/// it under the terms of the GNU General Public License (GPL) as published
/// by the Free Software Foundation. cf. the file COPYING.txt.
///
/// ------------ *** end of legal notice *** ------------------
/// </summary>

[Joannes Vermorel]

Dear Rana,

Christoph and I are interested by an arbitrary precision numerical library in Math.Net. We have been planning to put those elements in a namespace called MathNet.PreciseNumerics. I am in charge of the MathNet.Numerics part of MathNet. For now, nobody is working on MathNet.PreciseNumerics.

I think (and Christoph will certainly confirm) that we will welcome any help for the MathNet project. If you want to contribute to the MathNet.PreciseNumerics, you are welcome.

Please note that MathNet aims to provide a very clean C# framework". This is, in my view, the most important thing. The objective is not to make
quick & dirty port of everything to C#, the objective is to have very clean and very documented libraries.

Please contact me directly at joannes AT_NOSPAM_ vermorel DOT KOM.

Best regards,
Joanns Vermorel