Gemini 2 serial alignment bug

[bgottula]

While testing the Gemini 2 serial interface in a control loop I encountered a problem where I would hit a timeout while waiting for a response to a get position command. After this occurred, I found using minicom that I would only get a response to the command being sent after sending the first character of the next command. The only way to recover seemed to be unplugging the USB cable or power cycling Gemini 2. The control loop was also sending some native commands that do not produce any response, so one hypothesis is that commands were being sent to Gemini 2 too quickly. I no longer encountered the issue after inserting echo commands in between each of the native commands that had no response, the thought being that the code would wait for responses to the echos before proceeding to the next important command.

[bgottula]

Wireshark capture of the USB serial interface. The issue begins near the end of the capture (filter on usb.addr).
serial_busted.pcapng.zip