The recent change in https://github.com/Dygmalab/Kaleidoscope-Bundle-Dygma/pull/2 converted data0 and data1 to fixed-length arrays, but the free calls in the destructor remained, causing undefined behavior. I noticed that when calling unbind and bind on the USB port I would get a failed to set config error on my PC and the keyboard wouldn't work at all (except for the LEDs), but unplugging and replugging the USB cable would fix it. I also sometimes had to unplug and replug after starting up the PC.
Oddly I hadn't seen this issue before, while I was testing PR#2, but I got a new motherboard today, and after I had installed it the problems started appearing.
With this change I am able to unbind and bind without the error above, and it always works after turning on the PC.
I probably should have caught this when doing the original PR, but I assumed the upstream change was fine, so I didn't look through it super closely.
The recent change in https://github.com/Dygmalab/Kaleidoscope-Bundle-Dygma/pull/2 converted data0 and data1 to fixed-length arrays, but the free calls in the destructor remained, causing undefined behavior. I noticed that when calling unbind and bind on the USB port I would get a failed to set config error on my PC and the keyboard wouldn't work at all (except for the LEDs), but unplugging and replugging the USB cable would fix it. I also sometimes had to unplug and replug after starting up the PC. Oddly I hadn't seen this issue before, while I was testing PR#2, but I got a new motherboard today, and after I had installed it the problems started appearing.
With this change I am able to unbind and bind without the error above, and it always works after turning on the PC.
I probably should have caught this when doing the original PR, but I assumed the upstream change was fine, so I didn't look through it super closely.