Readers of this article might also be interested in a piece I originally did for Doctor Dobb's Journal (www.ddj.com), which I have revised and put on line at http://www.third-bit.com/~gvwilson/xmlprog.html. In brief, I believe that the shift to representing code in XML is almost inevitable at this point, but (a) it isn't going to take off until we can give programmers WYSIWYG editors capable of handling rich content (like in-line SVG and MathML), and (b) the real benefit will be that this shift will make high-level program manipulation feasible.
It's a very good article and the argument goes along similar lines of processability and extensibility of XML languages.
I personally think that 'the future is here' - using a language such as o:XML gives immediate benefits that cannot be neglected. Simple things like extending a type definition with test cases is _really_ easy. Producing high quality documentation is _really_ easy. The source code has real value as structured, accessible and unambiguous data. And when using a schema-aware editor (such as Emacs), o:XML is not very difficult to code in either.
Previous Message
