Even known SAFE Network put many efforts of network takeover attack, I think that there is something that could secure the network, but haven’t find something like that in my searches.
Why more an more people think that Bitcoin is secure? Well, I can put up 3 main reasons:
They read the source and protocol description, did the Math and has been convinced. (Or that they just trust other people opinion and research)
They saw it work on practice
It has not been hacked yet! (Assuming that the cryptographic signatures/encryption/hashes is secure)
My problem with SAFE Network is that the third reason couldn’t be verified currently.
In Bitcoin, we can say that:
The block chain is valid, i.e. no money was forged “illegally”.
We had not detected a massive block chain reverse. (And we familiar the small one)
What I’m asking is a “proof of hack”, which means that if someone detect something is wrong, he could prove it to the users.
Wrong things for example:
Steal safecoins from account (very dangerous!)
Forge illegal safecoins to his acount (e.g. restore account previous amount)
Tamper with stored data
I don’t say it has to be completely transparent, but if someone was hacked or someone notice illegal activity, he can prove it. (I believe it can be done without block chain)
It may sound bad that users can show how insecure the network is, but I think if no report would be, it will prove the security of the network.
Because of the nature of the network, someone can claim that their safecoins were stolen but would have a very hard time proving it. Likewise it would be nearly impossible to prove that it wasn’t.
This leaves anecdotal data as the only “proof” either way, and open to FUD on the one side and “lalalalalalalalala it’s all good! I’m not listening to such” on the other.
So the only way to give a vaguely reliable proof of hack would be to (a) find a way to do it, (b) say you were going to do it and in exactly what way and what proof you’d present, (c) establish that the hacked condition did not already exist, (d) do the hack, and then (e) present the proof. Even this would have to be done in coordination with a trusted party to be able to rely on whether the conditions had been met or not.
Interesting. Hadn’t thought of it this way.
It would be rather hard to prove that one “had been hacked” by an unknown source, even if one were sure of it oneself.
A one example that I think is safecoin stilling proof.
If I understood correctly the protocol, the safecoin account state is stored somewhere in the network.
(This is my understanding of the protocol, I might not be 100% correct, but I believe that it’s similar)
One way of theoretical massive node attacker to steal(or even add) coins is to restore account to previous state(he can’t create a new account state, since the user sign them).
If the account state was signed timestamped, and he was attacked this way, he can prove he has a state that is newer than the current one, signed by the correct near addresses nodes. (BTW: I think it should be done automatically by the client application)
In this case there is two possibilities - either he was truly attacked, or he himself succeed in attacking the network. Either way it is an example for proof of hack.
An extension of this idea would be to have some close addresses nodes “spectators”.
Similar mechanism could be applied to a network storage.
You do realise that coins are actual SD objects and have owner. Unlike other digital currencies that use a ledger system to hold your balance.
Your “wallet” on SAFE will be sent a message when a coin is transferred to you and from there keeps a tally of the coins you own. But in any case the coins you own each have your ID as the owner of the safecoin SD object.
So the attack to “restore” coins to you involves an attack to change the owner of each and every coin back to your ID. It does not matter what any “State” says.
But you are right in that if your wallet says you have “X” coins but your actually have “Y” coins then something is wrong and could be evidence of a hack. But remember that each and every coin holds its own state independent of an account “state”
But in order to prove it, he must hurry up before this powerful attacker would mess up his coins too.
If you introduce timestamping in the signature, she can always show the previous timestamped state with current timestamped state(by asking close nodes to timestamp the response). This may allow a proof in other scenarios, like simple storage reversing like I mentioned before.
What do you thing about this timestamping in signature?
But the data chain is coming see new thread. It can make use of time apparently. It may provide some of the concreteness or trust that is sought. I love how SAFE as of late seems to (remains to be proven) have absorbed all of the btc funtionality, its like the first real world test of the concept and its passed it (tentative) in the womb.