Thanks. Didn't notice that, because... you have to place your mouse exactly in the right spot, and only then an arrow icon will pop up under it, and only clicking in that small spot will make the hopper dispense a ball.
Ridiculous we are still waiting on a Turing Tumble llvm backend that will emit decent marble netlistings. I swear if it doesn’t have proper tail recursion, you will have an actual mess when the stack overflows near the mere thought of fib(n=10)
The wikipedia page on Fredkin & Toffoli's billiard ball computer[0] mentions a 2011 paper Robust Soldier Crab Ball Gate[1] co-authored by Andrew Adamatzky of the Unconventional Computing Centre:
"Soldier crabs Mictyris guinotae exhibit pronounced swarming behavior. Swarms of the crabs are tolerant of perturbations. In computer models and laboratory experiments we demonstrate that swarms of soldier crabs can implement logical gates when placed in a geometrically constrained environment."
I think it would be fair to call the system Turing complete if it is theoretically capable of addressing an arbitrary amount of memory. Of course, no physical computer or Turing machine actually has unbounded memory.
> They also make the computer Turing-complete, which means that if the board were big enough, it could do anything a regular computer can do (in theory!).
They do account for limited resources when they make the claim, thankfully.
I think feeding marbles to it repeatedly constitutes “looping”. Think of them as clock pulses going through a CPU. If you want to signal that the computation is finished, you just direct the marble to a bucket (interceptor) instead of to the bottom.
If you built a conveyor to automatically push marbles from the bottom back to the top, then you would have a slow but functional automatic computer.
When I was a kid a teacher donated an old game to the class that was a rudimentary version of this, Avalanche[1]. None of the kids had seen this game before but the satisfying mechanical movement and thinking it took to get all the marbles to fall all at once made it became very popular. I had these fond memories of Avalanche, and did actually think of it as a visualization when studying hardware in college, but over time my recall had made it more elaborate. Surprising to see how simple the old game actually was. If me and my classmates got so engaged with a primitive ancestor of Turing Tumble this new game could really catch on with kids, as long as it isn't overly complicated.
Got one in the Kickstarter and my kids love it. Great for the 7-10 age range with adult partnership. The comic story is even decent! I imagine 12+ could get pretty far on their own.
Agreed. I was surprised by how much my 6 year old became engaged in the story. He frequently asks to play/solve more so he can find out what happens next in the story.
Check out the Turing Tumble Education site [1] for PDFs of the teaching resources (which includes the puzzles included in the actual set) and links to the in-browser simulations of the Turing Tumble, including one in VR.
When I was a kid I made contraptions similar to this using hotwheels tracks with differently sized holes cut into them and slips of paper. Wasn't turing complete, but could do some things. I would have had a blast with this thing. I might buy one now so I can gift it to my nephew when he's a little older.
https://jessecrossen.github.io/ttsim