It hath occurred to me that there are several very effective contraptions that would demonstrate Second Life's advancing technology in a compact package, while educating and entertaining at the same time. Clearly, one of these is my designated example of shufflepuck. What else is there?

WHO WILL MAKE A ROCK TUMBLER IN SECOND LIFE?

I've never owned one, natch. But, I love the results they produce. In case you aren't acquainted, a rock tumbler is something you dump rough stones and some special chemicals into, and they bounce and jiggle and sway, in sort of a boulderiffic take on urban music from the granite ghetto, and rough becomes smooooth. Takes time, oyes, but that patience rewards itself with pretty pebbles.

1. Rock tumblers depend on physical collisions. In SL, the Havok 1 physics engine doesn't play well when lots of objects are smashing together. All those fancy computer calculations add up, and enough will not not only bring a a simulator (aka region) to its Newtonian knees, but possibly CRASH it. Having 100 rocks and even more tiny grains all banging each other is my idea of a good time, but clearly not feasible at present.
    
2. If you've ever seen most rocks in SL, i.e. the Linden landscaping kit, they are jagged and decidedly unappealing, but, like the best we can do at present. Making them smoother would either require turning them into some variation of a sphere or a torus without a hole (WTH?), which would look weird too. More faces on a prim, or perhaps more elegant, new shapes to accomodate things like this, might be the way to go.
    
3. Tumble-polished rocks are shiny. We only have 3 degrees of shininess, not including NONE. I see this could pose a problem if you wanted to make a translucent (transparency in SL does have a smoother variable percentage, from 0-100%) tumbler and watch for gradual changes in the presumably scripted rocks. Ever seen a floating cube of water? Yes, I feel the same way. Also, the shininess is oriented towards metal. Convincing donut glaze is hard to do.
    
4. Sound limitations. We aren't able to smoothly crossfade between loops, and for anyone who's really listened to a rock tumbler knows, it sounds repetitious, but deep within, if you lend your ear, is a credible granular groove. A stone never rolls the same way twice, after. If I put one stone in, that'd be different from ten stones. A more dynamic audio environment should be able to account for these differences.
   

Solving each of one of these steps is a prerequisite to coming up with a rock tumbler in SL. I believe it will be done—just not yet. Kinda like lightning is a sign of a rainstorm, rock tumblers in SL will show far we've come in a lengthy, finely-polished strike.

And there you have it.