Recently received my 3' table, setup was easy, its been running great. I went through a variety of tracks in the beginning to see what it can do, and noticed that it was only using one ball. I figured it must be select tracks that use two.
I ended up playing the default playlist, and at some point it began using two balls, which then appeared to continue doing so on tracks that previously only ran with one...perhaps my memory is bad.
Had a power outage the other day due to some electrical work being done, so it had to reboot, and its now been back to a single ball for all tracks. The small ball is in the corral, so it isn't in some random spot.
Is there a trick to getting the two balls moving?
Yes, there is a trick, and I apologize for that. But we really wanted to start shipping the wood tables, and we weren't quite ready with the app. We are just about to push an update that includes a better demo playlist (that will actually use two balls). But, if you want to explore:
There should be two tracks at the bottom of your track library, named "attach" and "detach". If you run attach (either as a single track or within a playlist), EVERY track after that should have two balls. Counter-wise - after detach, all tracks should have one ball. This is still kinda raw. But no matter what happens - it will not harm your table - that I'm sure of!
Somewhat off topic, so apologies in advance. Would there eventually be an option to upgrade the metal tables with two-ball support? The lack of a "garage" for the second ball is certainly an issue, but I can think of some ways this could potentially be addressed...
In somewhat related news, out of curiosity, I wrote a primitive program that would convert existing tracks into a track that would show what the path of the ball would look like if it were attached to the opposite end of the arm--that is to say, what the second ball would draw in isolation. This is actually a very easy operation on a track file; just replace every rho value with 1 minus rho (technically, theta should also be replaced by pi + theta but that only serves to rotate the path). Although this doesn't replicate the unique result of having two balls drawing the original and "inverse" path simultaneously, it did result in additional tracks that I found interesting. In fact, this notion of applying a transformation to a track could be generalized.
No apologies accepted (since never needed :). Off topic is OK. There's no theoretical reason 2-ball can't work for any version of Sisyphus. But implementation and support are the sticking points. That being said, we will be rolling this out soon. Your "primitive" program is spot on: (1-rho) and (theta -Pi). I call this "Tantalus mode" (sticking with the wickedly cruel Greek punishment theme). It only applies when both balls are attached. Don't have a name yet for if only the tail ball is running.
"Can anyone point me to some reading material that might help me progress into a more complex track making game?"
Welcome to my world! First off - although I'm pretty sure when it comes to sand plotting, I have more area under the curve than most - I'm no authority. But with regard to compexity - I would divide it, for now, into those tracks that are generated algorithmically, and those not. So the pumpkin track, and any tracks that you recognize (including text), are non-algorithmic. All the others -'crop circles', spikey or swirly/loopy spirals, and fractals - are created via multilayered recipes (instructions) -aka algorithms. Both branches are infinite design spaces. Markyland created Sisyphus For The Rest Of Us so that people could more easily explore non-alg space. If you're interested in algorithmic pattern space - start googling. All you need to do is place the connect-the-dots in the right order :).