broadcast/youtube.go: live monitoring routine continuously power cycling hardware when problem occurs

(https://bitbucket.org/ausocean/vidgrind/src/8fda3b349afa854898b6c8cf8950fa67e19e932f/broadcast/youtube.go#lines-263:295)

This is happening due to two compounding issues. Firstly, the nIssues counter is not reset after the hardware is reset, and secondly, because we do not provide enough time after hardware restart for data to be sent again, i.e. we get no data error and the cycle continues.

Comments (1)