After the first fail, the first bug. I like bugs in a hardware project, they usually occur in real word. For this bug, is during a night under the stars in my garden. When the CMV is at the hight speed, and without other sound, I clearly listen many speed switches. After a good night, I looked this graph:

Ok, there is a bug. Two bugs in fact.

A first bug, the root of bugs, is my rewrite in rust of open2300. There is many fails (~90% of reads). I don’t (yet) find where is the problem.

But reseted the speed when the read fails is a second bug. Using ws2300 as subcommand in the cronjob caused sending (null). I simply handle this case in the vmc script via this commit.