Nullifying pointer actually not good?

[JG-Adams]

This was confusing for my C++ brain. The addition I made here where you nullify pointer after freeing memory does absolutely nothing because the function takes a copy of a pointer. I'm used to thinking reference hence the mistake.

However, there are other confusion to this. I realized that actually nullifying pointer can cause program to crash rather than be stable. Take this example:

case ENET_EVENT_TYPE_DISCONNECT:
if constexpr (Server) {
deactivatePlayer(event.peer->incomingPeerID+1);
printf("Client%u has disconnected\n", event.peer->incomingPeerID+1);
if (event.peer) {
enet_peer_reset(event.peer); // Reset peer safely
//event.peer = nullptr; // Ensure we don't use it again. But this cause crashes???
}
BitWriter ms(packet_buffer, MESSAGE_SYSTEM_SERVERMESSAGE);
ms & "Player has disconnected!";
message_relay(event, ms, NET_CHANNEL_SYSTEM_SECURE);
}

One would want to send out message of a player leaving. So you would reset the peer thinking it should render it destroyed. My message_relay() would go through a loop and acquire the peer from the host. Not the event. and check with currentpeer->state == ENET_PEER_STATE_CONNECTED.

If you uncomment the nullification of event's pointer to peer this would crash. I don't understand how, but event's peer and host's peer somehow shared the same peer? So if you nullify one from event it applies to host too! This is very counter intuitive I would say. Are we not supposed to nullify them at all? I'm guessing that to be the case. However, you can nullify host and peer that exist outside of event.

Edit: I think it's shared probably because the service function for event may actually be updating the host with its own peer pointer. Shrug