"Hello World" data logging - from the microcontroller or the orchestrator?

[sgbaird]

Module 5 is the data logging module, Module 6 is the "put it altogether" module, and module 7 is the "all-purpose lab sensor system".

For each of these, should the idea be to have learners use MongoDB's Data API with urequests directly on the microcontroller or to use PyMongo on the orchestrator after the data is received?

For Module 5, either is probably fine.

For Module 6, there's a higher likelihood of running into a memory error.

Module 7 kind of assumes no orchestrator (and no MQTT) - the idea was to upload directly to MongoDB via the Data API at regular intervals. Also, Module 7 is sort-of optional, but nice in that it leaves the user with something useful. Should every learner have a full light-mixing demo set up and running, or would it be more useful to most researchers to have their own sensor system set up?

[sgbaird]

Decided not to do module 7, and I tried to add in the companion notebooks for MongoDB and HiveMQ some instructions on how to do the two-way communication Python <==> Python (rather than MicroPython <==> Python). It's still worth considering whether to increase the visibility and amount of effort towards showing Python <==> Python.