In response to this excellent blog post by Dave Herman, Jay McCarthy pointed the reader to this interesting paper about creating a language construct that allows you to modularly redefine how your code (that you may or may not have written) gets evaluated (in other words, nonstandard interpretation).
Nonstandard interpretation
is a powerful and useful construct that supports such tasks as tracing, security logging, sandboxing, error checking, profiling, code instrumentation and metering, run-time code patching, and resource monitoring.
Obviously, these things aren’t new features. You can probably already perform nonstandard interpretation using your language of choice.
The interesting thing about this approach is that rather than performing compile time environment or code manipulation, a new language feature is applied to perform the non standard interpretation at runtime.
While both will get you to the place that you want to go; only the latter offers you an opportunity to build a conceptual framework for understand how you reached your destination. Neat!