Post
by torwag » Thu Aug 17, 2017 7:14 pm
Hi, there is no need to recompile and reflash after doing it initially unlike :
1. one want to update micropython and the precompiled modules coming with it
2. want to add modules to the image, so called frozen bytecode, to have a customised version of mp or for memory safeing reasons.
Micropython comes with a little file system pretty much like a usb memory stick but much smaller, created in the memory of the chip. This can be used to store modules beside of self written files e. g. boot.py and main.py. Simply copy them over as you would copy a file to somewhere else. There are tools to do this.
Actually it is just one of the major advantages of micropython that you do not have to recompile and reflash after every single change.
Furthermore, as soon as you have micropython running, all the stuff you find for other platforms could in principle work on the esp32. However,
1. ESP32 is still experimental and rather new, some stuff might not be yet included or is buggy.
That was the reason I mentioned the esp8266 port which is much more stable and mature. As both ports are very similar in many aspects, one can assume that if something works under the esp8266 it should eventually work under the esp32, however for now it might take a while to implement all what is needed and to fix bugs. Devs are happy for every helping hand.
2. Sure nobody expects something to work what is physically not available. If one chip has a high speed ADC, but another chip has not, nobody can add this (in software).
As for MQTT (one Q two T) it is a lean protocol to send and receive data between different units. It requires however a server, called broker., I guess looking for MQTT and broker will give you a good idea...
Enjoy