Here’s how it works: you write a program in a programming language of your choice, say .NET. Once compiled you send it through the first step of XMLVM which analyzes the produced CIL and creates an XML document out of it. It would end up looking like something similar to this:

<clr:ldc type=”int” value=”2″/>
<clr:rem/>

Next this XML document is fed through what XMLVM calls the data-flow analysis (DFA). Basically you can think of DFA as a pseudo-language that simply describes the operations that the program is trying to perform. Once in this form the code is considered portable. XMLVM then lets you pick a target, for example the Java JVM, and automates the conversion of the DFA to an XML representation of the java byte code. From there it’s an easy conversion back to true java byte code.

Now think about this in practical terms for a second. That means that you can write a program in a .NET language (C#), and have it automatically ported and compiled to Java. Expand on this a bit and consider that you can write the same program in any language and have it converted to any other language. Currently the XMLVM offers a lot of other cool options as well and has actually been designed a lot with mobile devices in mind. Now you can write a program once and have it automatically converted to Objective-C, to run on the iPhone, and to Java to run on Android.

I really hope that this project continues to improve and I will certainly be watching it closely. It is still very early in development but from what I have seen it is simply brilliant.