Phantom device (change BROADCAST ID)

[thomsna]

Two Arduino pro mini ATmega328.
A master set at address 45 and a slave set at 44.
Bus on pin 12.

While sending to device 124 even though it didn't exist on the network, I get ACK values back.

Can someone reproduce this?

Here, I wrote a script to check for a response on all addresses.
Code for master:

#include <PJON.h>

PJON network(12, 45);

char content[] = "ajtyenflhiehfnvk";

void setup() {
    Serial.begin(115200);
    delay(250);
    uint8_t test = 0;
    long scan_time;

    for (int i=0; i<256; i++) {
        scan_time = millis();
        while (millis() - scan_time < 20) {
        int response = network.send_string(i, content, 16);
            if (response == ACK) {
                test++;
            }
        }
    if (test > 0) Serial.println(i);        
    test = 0;
    }
}

void loop() {};

Code for slave:

#include <PJON.h>

PJON network(12, 44);

void setup() {
  Serial.begin(115200);
};

void loop() {
  network.receive(); 
};

[aperepel]

@thomsna if I read the code correctly, address 124 is the broadcast. In your teat, try counting how many acks you get and from what addresses to confirm.

[thomsna]

@aperepel Thanks for clearing that up. 124 is indeed the broadcast address.
The code above will check for the ACK from a device on all addresses. An ACK is received on address 124, even though no device is set to that address.

[aperepel]

Great, this is not an issue, it looks. I would maybe prefer to have the broadcast on a more 'classic' address like 0 or 255, than in a middle of the range. @gioblu any comment on that? 124 doesn't even look cool in binary :)

[thomsna]

+1 for setting broadcast address to 0, like I2C.
It would also make the code clearer in my situation to count from 1 to 255, instead of taking care of excluding address 124.

[gioblu]

Hi @aperepel @thomsna to be consistent and not subjected to versioning changes (as the one you are proposing) bugs, always use constants if you can: for(uint8_t i = 0; i < 256; i++) if(i == BROADCAST) continue; so if BROADCAST value will change during development this will not affect your sketches.

Could be more clean I agree, we should run extended tests, but if all is ok, I don't see limits in changing BROADCAST to 0. Thank you again for your support and passion in this project 👍

[gioblu]

Ciao @aperepel, @thomsna.
With the release of the 2.0 version this issue has been fixed and BROADCAST id is now 0, as correctly proposed 👍 . Thank you again for your support and passion in this project.