A Stupid XML Project is the result of combining XML with some existing, ancient idea, and spinning it as a brand new invention. The concepts in the ancient idea are often given new names that conform to Xml Speak.
For example, a number of programming languages are cropping up from the realization that a printed notation for structured data can be used as the syntax for encoding a computer program, and that within that program, the notation can also specify literal data. Wow, a code-as-data paradigm is born out of XML! Sound familiar? Lisp with angle brackets instead of parens? [⇒ Parens Dialect]
Except that the idea is old, and this particular incarnation of it is so atrocious, nobody in their right mind would want to use it.
Referring to Xslt Language, for example? Or worse: Xexpr language [www.w3.org 
; reinventing a tiny subset of Lisp with XML read syntax], Water language [the vacated website www.waterlang.org ; Lisp, coated in XML syntax].
This is part of the larger problem that people want to use a markup language as a language for everything [⇒ More About Algorithmic Markup]. There are other languages that are easier to parse and read, because markup languages are specialized for embedding information in raw text documents. I don't think markup languages make sense for any other use.
I came to the slightly kinder conclusion that XSL is merely Lisp By The Back Door. The (incomplete) mapping seemed very obvious to me as a Lisp and XSL newbie. It seemed most likely that someone needed something Lisp flavoured but needed to avoid the problems of Ai Winter. -- Matthew Astley
Do we smell Band Wagon Syndrome?
The issue is what to do when you want a human readable, easily parseable expression syntax. You look at something like XEXPR based on XML syntax. Unfortunately, it is verbose and appears to have comparatively little support; a Google search for XEXPR Evaluator doesn't turn up a bunch of code.
See original on c2.com