[Nut-upsdev] support for newer mustek powermust model 06da:0003

[Arjen de Korte]

> The machine shuts itself down upon "upsmon -c fsd", but the UPS
> doesn't turn the load off.

I kind of expected that already... :-(

> More, after doing an "upsdrvctl shutdown", the driver cannot regain
> control of the UPS anymore (every command during detection times out).

Most likely because we send it some command it doesn't understand which
locks the whole thing solid (like the I\r command). Instead of masking out
some commands we know the device *doesn't* understand, we'd probably
better let through the commands we know it *does* understand (to be safe).
To allow for testing, we might let them go through only when debugging is
enabled. One thing that could be a problem is the fact that the shutdown
command takes up too much space for the 8 byte buffer we have now. I'll
try to fix that later this evening.

> The same happens if one sends any command that isn't supported by the
> UPS (like doing a "test.battery.start.deep" instant command). The only
> way to restart the driver after this is by unplugging/replugging the
> USB cable.

Ditto. We have some wrinkles to iron out in the 'phoenix' subdriver,
before we can consider this stable enough for production.

Best regards, Arjen