Hexapod Robot Kinematics - part 4 (First meatspace demo - warts and all)

A couple of years ago, I built this Dynamixel powered Hexapod. It sports 18 AX-12A servos and a custom controller board of my own design. It looks impressive, but it lacks proper kinematics firmware.
I have to admit that I underestimated the complexity of robotics. Building a robot is easy. Making it behave the way you want is a bit more difficult.
In order to properly learn the fundamentals of kinematics for this system, I decided to create a simple hexapod simulator. This allows me to play around with topology, forward and inverse kinematics, various gaits and gait transitions.
I'm currently running the simulator under Windows and MacOS, but I am planning to also compile it for Raspberry Pi, which I in turn can use as a more generic robot controller.
This is the second part of the series, where I let the simulator act as a controller and stream servo parameters to a robot arm/leg.
I have also designed a new ESP32- based controller board for my hexapods. This should be able to either run the kinematics part directly, or act as a receiver for a remote controller.
In this episode I show you the first test of the hexapod actually running in meat space. The kinematics calculations are done in the simulator that is written in golang. The controller board runs FreeRTOS and is using an espressif port of the Dynamixel2Arduino library.
In the video, you can see that hexapod struggles to move forward. After having watched the footage a couple of times in slow motion, i'm beginning to suspect that I may have wired up the coxa joints in reverse, so that the legs are fighting against each other.
00:00 - Intro.
00:08 - Excuses, excuses
01:23 - Show & tell
02:39 - Demo time
05:12 - Someone has inverted their coxas...
08:12 - The SloMo guy