How to support push OOP debugging when device is in production mode?

[tolauwae]

[carllocos]

I was also thinking about this since the out-of-things paper is about debugging in production if desired. If believe the problem of debugging in production is more than just about the pushing of events but also for instance about ensuring that the state between the local VM and the MCU remains in sync after performing proxy calls.

This is what I believe is important to keep in mind when working on this problem:

• After performing a proxy call the state between local VM and the MCU VM should be in sync. This is also mentioned in issue https://github.com/TOPLLab/WARDuino/issues/78
• After performing a proxy call the state of the MCU should be restored to what it was prior the call. So that the application can resume correctly.
• What about side-effects? we could restore the state after the call but turning off a led could not be undone at that moment.
• I think we need different ways on how to deal with events being generated on the MCU. One approach could be to simply redirect all the events so that the VM does not handle them, or to take a of the event and send the copy to the local VM.