Complex pattern creation

@adambrower It's been very gratifying for me to see a tool like Sandify being used by Sisyphus owners (as well as DIY'ers). Great thanks to @jeffeb3 for adding polar output compatible with Sisyphus' native format (thr files). I began making Sis tracks over 20 years ago for the first prototype, but actually got started ten years earlier, working on a different drawing machine that also had no pen-up ability - the first Eggbot. My goal then -  creating single, continuous paths that look cool both while being drawn and when completed - remains the same now. Since I was not formally trained as an artist, I never learned to draw by hand skillfully. But I took to early 2D CAD software quickly, and soon was able to control the "pen" using simple algorithms (a.k.a. turtle graphics). Eventually, AutoCAD became my main tool. Most of the routines I used to create the tracks that ship in the "default" playlist, I wrote in AutoLISP, (a(cumbersome(language))) which at the time, was the only algorithmic tool inside AutoCAD. That changed in 2012, when I joined a makerspace in Mpls. Surrounded by younger DIY CNC enthusiasts, I was introduced to Grasshopper (an amazing plugin for Rhino). It took me a while to get used to GH's GUI-based programming - but I eventually ported most of my LISP routines into GH "definitions." The main advantage with regard to algorithmic track hunting, is that using AutoLISP, the process was one of typing numbers (values of equation variables), hitting enter, and seeing the result drawn on the screen - over and over again,  while tryig to "dial in" on something of interest. With Grasshopper, I do the same thing, but with sliders.  And if I keep the number of points along the path to a reasonable number, the drawing morphs in real-time as I move a slider. A definite game changer.

How to find track "gems?" - I suspect it's a lot like gold: spending a whole lot of time panning in the right place using capable tools helps. Of course, unlike actual gold, which has an accepted value, deciding which tracks are "worthy" is a process I believe to be outside the domain of  analytic tools. For a long while, my analytic side was both suspicious of and a bit uncomfortable with this. But I've come to cherish this "tension" and the seeming paradox of spending so much of my time building and collecting analytic tools, solely in pursuit of non-analysis-driven goals.

Foxwarp2 is a beautiful pattern. It has a bit of a 7 sides shape, but it is rotated perfectly to overlap. I would guess you can get close with sandify by making a 7 sides polygon and picking the right spin parameter. But there's more to it. The organic pattern of the sides can't be recreated in sandify yet. Bob and I have been working on some warp techniques to add "texture" or "waves" to the output shape, but it is actually very hard to script something that is a bit random and is actually pleasing (just like Bruce said, the hard part is deciding what is good). As a tool maker, I would like to make it easy to find pleasing patterns without limiting someone's ability to be truely creative. That's the hardest and most interesting challenge.

I would also like to add in the ability to really just draw a shape, and then send it through the loops, spins, and other effects. But that's just in the design stage right now.