improve examples' use of mesh.begin() and renewAddress()

[2bndy5]

This solves #203

While reviewing all the examples, I added some comments to the Linux C++ examples and ported the work from the mesh examples in pyrf24. I can revert the changes to the python examples if people are still using python2.7 for some reason, but I'm pretty sure boost.python has phased out maintenance support for python2.

[TMRh20]

This commit just brought to light another issue with the examples.

        if(!mesh.renewAddress()){
          mesh.begin();
        }

I think should be?

        if(mesh.renewAddress() == MESH_DEFAULT_ADDRESS){
          mesh.begin();
        }

I guess when I wrote the examples, I was thinking it would return something else.

[2bndy5]

I don't recall seeing that snippet anywhere. Why would you call begin() after renewAddress() fails?

But I would agree. Failure from renewAddress() is identified when it returns 04444.

[2bndy5]

Wait, the docs don't mention what is returned when renewAddress() fails.

I suppose it is a good sign that users haven't noticed this oversight (or that no one reported it).

[2bndy5]

Should I also add a while loop in the child node examples that demonstrate how to renewAddress() on write() failures?

        if (!mesh.write(&displayTimer, 'M', sizeof(displayTimer))) {
            // If a write fails, check connectivity to the mesh network
            if (!mesh.checkConnection()) {
                // The address could be refreshed per a specified timeframe or only when sequential writes fail, etc.
                printf("Renewing Address\n");
                mesh.renewAddress();
            }
            else {
                printf("Send fail, Test OK\n");
            }
        }
        else {
            printf("Send OK: %u\n", displayTimer);
        }

Or should we add a var isConnected that triggers a while (mesh.renewAddress() == MESH_DEFAULT_ADDRESS) in the loop()? This idea (to me) seems overly complex for a simple example.

[TMRh20]

I don't recall seeing that snippet anywhere. Why would you call begin() after renewAddress() fails?

Its in all the client/child examples:

      // If a write fails, check connectivity to the mesh network
      if ( ! mesh.checkConnection() ) {
        //refresh the network address
        Serial.println("Renewing Address");
        if (!mesh.renewAddress()) {
          //If address renewal fails, reconfigure the radio and restart the mesh
          //This allows recovery from most if not all radio errors
          mesh.begin();
        }
      } else {
        Serial.println("Send fail, Test OK");
      }
    } else {
      Serial.print("Send OK: "); Serial.println(displayTimer);
    }
  }

It allows recovery from hardware failures etc. and makes the radios hot-swappable. I've swapped out faulty radios without having to reset the MCU...

Should I also add a while loop in the child node examples that demonstrate how to renewAddress() on write() failures?

Sure.

[2bndy5]

I found that the RF24::isChipConnected() wasn't exposed in the old py wrapper. I fixed this on RF24 master branch.