What is SuperCollider?

SuperCollider is a synthesis and signal analysis tool used all over the world by artists, academics, and professional engineers. I know of at least one college course that uses it to teach coding, math, and signal processing, and I've used it in my live audio production/synthesis experiments with New Orleans musicians since 2017. It's an amazing, one-of-a-kind, powerful tool that's completely free to anyone in the world under the GPL, which means it has to stay free. I have a lot of plans with it, like an educational program for kids in New Orleans, some interactive art installations, lots of cultural interchange, and continued joy & personal development.

I had developed this sweet touchscreen theremin and was just starting to perform with it when the HID client in SuperCollider stopped working due to lack of maintenance since it was originally crowd-funded in 2013. SuperCollider relies on the community for maintenance. Most of the people qualified to perform that maintenance are busy with much more lucrative (and unfortunately often destructive) work for the corporate tech industry.

What is HID?

Remember when keyboards and mice had their own port, and everything else needed a custom driver to work? All the gamepads, flight yokes, drawing tablets, braille displays, and MIDI controllers needed the manufacturer to write custom code for every operating system, and the user had to go through the extra hassle of installing it from the CD-ROM. Then, early this millennium, the industry came out with "Plug N Play". Hooray!

The underlying standard that makes this work is called HID -- Human Interface Device. No more custom driver code! When you plug it in, the firmware on a device just emits a report in a standard format that says, "Hey I'm a peripheral, I have these LED's you can blink to indicate the state of these settings, I have this rumble motor, and I have these buttons with these names and usages" and the OS or application goes "alright, I know all about that from birth, let's play".

It took about 20 years for all the usage tables to get fleshed out, but nowadays thanks to HID you can plug in just about any kind of device over USB, Bluetooth, various IoT links, and there's a specification for it. This includes things like VR headsets and even more esoteric things in the field of robotics and manufacturing.

What work are we funding?

SuperCollider's original HID client written in 2013 was based on a fork of the HID library from an early stage of its development, and thus includes limited functionality compared to what's available now. It lacks the ability to rumble motors and blink LED's (or anything else implemented in HID Feature Reports). It also can't talk over Bluetooth or some of the newer IoT hardware protocols.

It worked well enough anyway for a long time, but subtle differences between SuperCollider's outdated client and the latest HID library shipped with the operating system led to the malfunction that broke my touchscreen theremin and anything else that uses HID. The entire SuperCollider programming environment crashes when you try to use HID on Linux with SuperCollider version 3.13.

I have already made the changes to use the latest official HID library, and now there's just a bunch of bugs to chase down in the SuperCollider client code. There are a couple of memory leaks, and the client code to do things like calibrate joysticks and blink LED's is unfinished. Finding and patching each of those differences is time-consuming and painstaking work. I expect this work to take up the bulk of my day-to-day time for several weeks, and may involve help from others in the developer community.

To begin with, I am preparing a draft of software unit tests for each HID component in SuperCollider so that this maintenance can be performed sustainably in the future without such a big push. When the upstream library changes in the future, we'll catch it every time someone compiles the code.

You can follow the development and discussion on github.

How much do we need?

It's hard to tell how much time the bugfix phase after the tests are done will take, but I am guessing that with $5,000 I can get this done or at least spearheaded with enough inertia to finish. This covers my living expenses and allows me to focus exclusively on this work, and affords a small budget to compensate other developers for their assistance as needed.

I wish that I could afford to donate my time, but my ethical/professional/personal path has led to a divestment from most of the lucrative ways to monetize my skills. I've spent the last 7 years becoming deeply familiar with SuperCollider and its internals instead of working in the industry. So I am seeking community support.

The commercial, educational, and artistic value that we can produce all over the world with SuperCollider is hard to account for -- not easy to sell to a venture capitalist -- but still a crucial part of the ecology, especially pertaining to accessibility for those with lack of access to formal education or other resources. To build the world we're building, we need all hands on deck, and I feel strongly that SuperCollider is a powerful tool for making that happen.

Thank you for your time and support!