From: Andre Krause

just a beginners question, not want to start any flames: why are you using C++/CLI mode, if you are using portable libraries like boost anyway? why not build a native C++ application? what are pro's and con's? i must admit that i never got my head up into .net stuff, because i thought: "no way, i completely will loose cross platform." but now with mono, thats not true that much anymore. i heard that bytecode might be even faster than native code, because these bytecode JIT compilers can "magically" optimise speed at runtime!! ( see http://psyco.sourceforge.net/, http://www.kano.net/javabench/ , http://www.idiom.com/~zilla/Computer/javaCbenchmark.html )

Valid question. I'm using C++/CLI because I want to use the new Windows Presentation Foundation ( WPF ) libries which only have a managed API. So my choices are C++/CLI, C#, VB or XAML. Using these libraries are the only way to get 'next-gen' hardware accelerated GUI applications as far as I know. Compared to the Win32 API, WPF is a huge win - it's saved me months since I didn't have to write my own windowing framework. If our application was Windows only I'd probably not be using boost but we also have a bunch of shared portable code that needs to run on Windows, x86 Linux and PPC Linux, so I need to figure out how to integrate boost properly into a C++/CLI application.