Post
Topic
Board Development & Technical Discussion
Re: Is someone monitoring large parts of the network? (evidence+firwall rules)
by
libcoin
on 16/03/2015, 09:47:35 UTC
Also, the approach used here to link IP and TX is complete nonsense in my opinion.
Even if you are connected to all nodes out there, it is still possible to receive a transaction from an IP first even though someone else is the originator.

1. Imagine there is a small node with a bad internet connection sending a transaction. He has a list of nodes he is connected to and has to broadcast to all of them.
2. User X (having a huge internet connection) is the first one to broadcast and Eavesdropper Z is the last one in the peer list getting his broadcast at the very end.
3. Due to the slow internet connection the time between broadcasting to User X and broadcasting to Eavesdropper Z is sufficiently large.
4. User X relays the TX immediately after receving the TX from the originating node to all of his peers ... including the Eavesdropper Z which is also connected to him.

Now the problem kicks in. Eavesdropper receives the TX from User X before receiving it from the slowly connected originator, and therefore thinks the connection was actually created by X.

How the heck such approach could ever be decided to be implemented?

E.K. totally agree - but it is still quite useful for statistically monitoring traffic btw countries.

/M