My day job, @ Wanelo, requires cranking on all cylinders, and so I’ve been pretty busy with life, the universe, and everything, although it mostly means Work, the Wife, and the Cat). In all this constructive chaos, I totally forgot to mention, and properly introduce to the world something I’ve been working on during many sleepless nights this September :)

So without further ado, LaserCutter — Ruby Gem (a library and a CLI) for making PDF designs of laser-cut boxes, which fit/snap in together at all edges using tabs that go in and out. The output of the library is a PDF document. Typically next step would import that PDF into Adobe Illustrator for additions and touch ups, and then sent off to a laser cutter for the actual, well, cutting.

In addition to the ruby gem, there is also simple web front-end, called MakeABox.IO, which can get you started creating boxes without installing any gem, or running anything on the command line.

If you’ve made any enclosures of your own, like either for your electronics gear, or a lighting show, or your cat auto-flushable toilet, you probably used one of the existing and free tools out there. I did not find them all when I started making boxes, but over time the choice seemed to come down to three options:

It came down to a simple fact that I was not at all happy with BoxMaker, or the alternatives. Boxmaker had lots of features, but also some bugs; it’s source code was difficult to read and modify, it produced asymmetric box designs ---- including corner pieces that would sometimes be disconnected from the main shape ---- and it had no automated tests. The last part made it difficult to think about contributing fixes.

To top it off it was written in Java, which was the right choice back in 2001.. but we are not there anymore. I think.

So I embarked on writing my own, and as of a few weeks ago, the beast is now feature complete, as far as basic box making is concerned. It’s written in pure ruby with the Prawn library, it has a beautiful command line interface, it’s box cut out designs are handsomly symmetric, and all of my recent boxes have been made using it. And yet, there is so much more that I want to do more with it in the future.

To the coders reading this, I will shamefully admit that the current incarnation of this library (version 1.0.3) contains a number of hacks, and there is a consistent lack of clean intentional software design. I’ve sort of taken a brute force approach in a couple o places, and consequently been awarded a pretty low Code Clime score. (Although who are they to judge! Some computers, dammit :P) But on the bright side, I am planning a refactor before I do more with it. The algorithmic part of this ended up being a bit tricky (surprise!), but the kinks are ironed out and the tests are there to prove it.

My process typically revolves around the following steps:

I’ve gotten pretty good at Illustrator, and these days it takes me anything from 10 minutes, to a couple of hours (for a complex project) to design an enclosure. Then of course I haul to Techshop and print it on their Epolog Laser Cutter, while swearing profusely.

Perfect Thickness

Having tried nearly all thicknesses at the thinner range of the imperial scale, such as 1/8, 1/4, 3/16 of an inch, I decided that for plastic 3/16 is really the sweet spot: it does not bend as much as 1/8, but is not as heavy as 1/4.