Model #357 Update

One of my favorite ways of falling asleep is to think about puzzles I’ve been working on and the design challenges they pose. Well one evening last November as I was drifting off to sleep I turned over in my mind – for the umpteen millionth time – the sequence of manipulations that solve Model #357 as well as other manipulations that a puzzler might try in their search for the solution. And to my horror I realized I had found not just one unintended solution but a whole family of them.

So the next morning I tried out my theory and sure enough it solved #357. Trivially. “Golly, quite the bummer,” is not what I shouted at the top of my lungs but it captures the general idea. By this time I had already sent my design drawings off to my machine shop and he had already produced the first article (that’s what they call the first item of a production run). So I asked him to pause production while I figured out how to fix the design – and that ended up requiring a lot more work and a lot more time than I expected. For most of November through most of February I experimented with and prototyped different approaches until I found one that worked reliably. The version I had given my machinist was version 6; the final version (well I hope it’s final!) is version 13.

In case you’re curious, here’s what version 6 looked like post bead-blasting and pre-anodizing:

Anyhow, I’m hugely grateful that I found the problem rather than one of my customers. And also I’m pretty lucky for having caught it before my machine shop had fabricated a full production run’s worth of defective puzzles. If all goes well, Model #357 will be available for sale in a month or so. And yes, I’m turning the version 13 design over in my head most nights – so far so good!