Don’t get me wrong, I’m all for RLM when it makes a language more readable.
But the author of that article adds a lot of boilerplate just to rearrange some code that he admits in the introduction is a clearly understood pattern, and the result is less than clean, IMNSHO.
(Coined, AFAIK, by Richard Suchenwirth over at the Tcler’s Wiki. So yeah, sorry, it’s a term used when discussing code that uses certain special features of Tcl.)
Lots of "look what we can do!" code ideas should be seen as "what NOT to do."
I am guilty of this. Sometimes you just play with stuff to play with it. As long as no one is recommending doing it in real code, its entertaining and even has some capacity to help understand by digging into how things really work.
There is nothing wrong with mixing your “fun” and “work” boxes.
Hmmm, now that I think of it, yes; there is no problem mixing them.
Though, I'm sure that me spending almost a month learning WIN32 API definitely counts more as a "fun" thing, because:
A) Modern developers making C++ GUIs are certainly not going to use outdated, obselete tools from 300 million years ago, and
B) Spending a month making a program with ~7200 lines of code is probably considered too slow for a work environment.