Re: [Tails-dev] Update Electrum documentation for Tails 1.8 …

s7r,

Please see my replies that follow.

>> The main debate is over the DoS documentation. This is a good
>> summary by anonym of a worst case scenario: “Thanks to SPV, the
>> server can spoof the wallet balance. Hence the server operator can
>> scam Tails users, e.g. s/he can buy stuff from a Tails user, and
>> then bump their balance with that exact amount so it looks like
>> they've received payment.”
>>
>
> I don't exactly understand what you mean when you say DoS and not sure
> what would you like to include in the documentation. Obviously an user
> shouldn't trust an unconfirmed transaction, but this recommendation
> usually goes for full wallets as well not only SPV. This is already
> written everywhere and that's why Electrum shows the unconfirmed
> balance separately.
Full wallets do not suffer from the same vulnerabilities of SPV. I am
used to using Bitcoin in the most decentralized way, so, when I see an
SPV client using centralized servers run by strangers, I become nervous
especially when it is over Tor. That is a bad combination shown in the
example of “Bitcoin over Tor is not a good idea.” Security is not black
and white. There is a probability of risk that is assessed based on the
environment that the software is in. Perhaps I am too paranoid and you
are too confident. I hope that we can find a middle ground.
>
>> I strongly disagree. DoS should be mentioned as it has a
>> possibility (although unlikely) to have a devastating effect on
>> Tails users.
>
> How exactly? Can you explain me detailed where you think the DoS risk
> is? Again, the linked research paper has nothing to do with Electrum.
> The fact that an electrum server runs on top of bitcoin core which is
> mentioned in that research paper cannot be taken into consideration
> (how do we even know if the bitcoin core running on the electrum
> server we are connected to uses Tor for its peer to peer connections
> with other nodes).
>
> The problem here is that I don't know how you define DoS in this
> context. From my point of view an Electrum Server lying about an
> unconfirmed balance until first block is mined cannot be called DoS.
> (Also, in this case, the server has to OWN the coins apparently spent
> and target a certain user which is behind Tor (so anonymous) which is
> highly unlikely.).
The first mined block could never reach the client essentially putting
the user offline. Yes, it is unlikely, but this is Tails where we take
security seriously.
>>> - An Electrum server could not broadcast an outgoing transaction
>>> (payment) sent by you;
>> I'm not sure what you mean by this.
>
> When you send a transaction from Electrum, it's sent do the Electrum
> server to which you are connected. The server automatically feeds it
> to bitcoin core via cli command which broadcasts it to the peers (and
> into the network). The Electrum server could skip this step and drop
> your transaction, never send it to the network.
Wouldn't this be proof of DoS?