Yup, I can confirm that.
As it seems the device OS fork for mesh devices was taken before the extension of the publish buffer.
@rickkas7, can you forward that?
Test code
const char txt[] =
"123456789112345678921234567893123456789412345678951234567896123456789712345678981234567899123456789A"
"123456789112345678921234567893123456789412345678951234567896123456789712345678981234567899123456789B"
"123456789112345678921234567893123456789412345678951234567896123456789712345678981234567899123456789C"
"123456789112345678921234567893123456789412345678951234567896123456789712345678981234567899123456789D"
"123456789112345678921234567893123456789412345678951234567896123456789712345678981234567899123456789E"
"123456789112345678921234567893123456789412345678951234567896123456789712345678981234567899123456789F"
"123456789112345678921234567893123456789412345678951234567896123456789712345678981234567899123456789G";
bool done = false;
void setup() {
Particle.subscribe("send", pubLen, MY_DEVICES);
}
void loop() {
if (done) return;
Particle.publish("send", txt, PRIVATE);
done = true;
delay(5000);
}
void pubLen(const char *event, const char *data) {
char len[6];
snprintf(len, sizeof(len), "%5d", strlen(data));
Particle.publish("len", len, PRIVATE);
done = false;
}
(Photon 1.0.0 reports len == 622
, Argon 0.8.0-rc.27 reports len == 255
)