So far this has means orchestrating a set of processes just to talk to bluetoothd, automatically connect to boards, guess which /dev/input/event* device each one is associated with, and start reading those and sending data somewhere over OSC. Every step of the process is about 10 times as difficult as I initially expected it would be, which is why, after months of work, I'm still putzing around with process orchestration instead of actually doing anything with music.
The bit of hair on this yak that I'm currently shaving is building a TUI library for Deno so that I can build a nice dashboard for keeping tabs on the state of all the different pieces. I prototyped a bit of the orchestration process in Node-RED, which was neat, but not exactly what I'm going for. There is a TUI library for Deno already, but it's kind of buggy and non-ergonomic for my brain.
So far this has means orchestrating a set of processes just to talk to bluetoothd, automatically connect to boards, guess which /dev/input/event* device each one is associated with, and start reading those and sending data somewhere over OSC. Every step of the process is about 10 times as difficult as I initially expected it would be, which is why, after months of work, I'm still putzing around with process orchestration instead of actually doing anything with music.
The bit of hair on this yak that I'm currently shaving is building a TUI library for Deno so that I can build a nice dashboard for keeping tabs on the state of all the different pieces. I prototyped a bit of the orchestration process in Node-RED, which was neat, but not exactly what I'm going for. There is a TUI library for Deno already, but it's kind of buggy and non-ergonomic for my brain.