But while there might be moments when somebode "shows off" in a short routine like those in the "8052.com code challenge", it's impossible on a long run for extensive projects.
I view the coding challenge mas an excellent opportunity for someone to "show off" without hurting anyone.

So if you find something "fancy" in a common program it is almost certainly because of "the exception" (trying to meet some resource limit - space, speed, money).
I can not count the number of times I have seen totally unnecessary "fancy" code (plenty of unused resources - time and space) causing someone to ask me to fix "that little last bug". Invariably this turns out to be "those pesky bugs we can not get rid of" and the only reasonable cure is a major effort to "unfancy" the code.
That being said, I have seen ONE case where "fancy" code was justified.
In a new development, if you get to the point where "fancy" code is even considered STOP, find another chip (more memory/faster) and then continue. You will find that you have saved tons of money and lots of agrrevation.

just calculate:
a reasonable estimate of one hour of programming (salary - insurance - tools - space - training - vacation - sickleave - supervision) will be $100.
So, of you use mundane code and spend 2 hours including debugging the cost is $200. If you do the same fancy it will take, say 10 hours to code and 5 hours to debug at a cost of $1500. The difference is $1300, so if the "fancy" stuff is in order to use a RC2 instead of a RD2 you need sell 1300 units to make the "fancy" wirth it. WHEN, because of the "fancy" stuff, a little bug show up someone must go to the customer to fix it. It typically cost us $500 to service a unit. NOW where are the "savings".

Yes, I will try to make a unit as cheaply as possible, but risking "a little pesky bug" that show up after 1000 units are shipped to save $1 - no way.

Now I say: Isn't all of said absolutely obvious?

Erik

I recall once a novice proudly told me "I have saved $0.75 of hardware". I asked "how long did it take you and how many units will you build" The answer was "about 5 hours, we hope to sell 10 units". I quickly calculated and asked "so they pay you $1.50 per hour?"