OK Dave,
The commands work!
Wow!
Right now my initial tests are quick and dirty; holding with a couple of the GPIO inputs to 3.3V with a jumper wire, but I wanted to report back ASAP,
I will do further testing tonight and tomorrow morning with a Ryde and a Pi3 Stream Receiver to confirm all the switching logic works and that the GPIO logic is correct (not inverted etc.). So far it looks very promising indeed. I saw one issue, but I think this might be 'key bounce' with me manually holding the GPIO wire to 3.3V, so I need to do the proper test. My concern here is if the Ryde input signal flaps and hence the GPIO pin flaps what the affect would be on the ATEM macro. A quick flap seems to halt the ATEM macro, but I should be able to work-round this in the macro design. It might not be an issue in practice with a real Ryde GPIO output connected to the GPIO input. I mention this FYI purely in case it could affect the IR mode under some signal conditions too.
I see it seems to have the 15-minute ID feature. I have yet to customize my macro to support this, I've not done it yet, but I'll do that as part of my testing too.
I've spotted a couple general bugs, but I think I know what the cause is:
First I updated the software from existing to new version, but kept the old config file, but after initial (failed) tests I then decided to update again, but not to keep my config file, but to use your default file.
One mistake I made in my file was adding "outputswitchcontrol=html" in addition to "outputswitchcontrol=ir". This might have caused the issues I describe below.
1. When updating the software from 202202260 to 202202260 the server did not reboot, but left it in a state where ssh was refused by my client (generic Ubuntu terminal window). See comment in 2. below.
- Impact: Would potentially require a site visit for a power cycle.
Note: have also seen this behavior on some previous Ryde versions when upgrading.
2. When performing reboot from the menu when running 202202260, the ssh connection was closed and further attempts to reconnect were refused, but the graphics carousel continued on the HDMI output.
- Impact: Would potentially require a site visit for a power cycle.
This condition I repeated twice, but not again, it might be that it was not happy with my original config file, it could also explain the previous Ryde issue I had too. I suspect that an illegal item in the config file could cause this behavior.
I have been thinking about how I can remote power cycle the Pis, my favored option right now it to connect the reboot pin of all the Pis to a relay contact on the ModBus controller to send a remote reset command to all the Pis. Might need a diode "and" gate.