Hash: SHA512

On 2015-06-19 15:37, Eric Lombrozo wrote:
> OK, a few things here:
> The Bitcoin network was designed (or should be designed) with the
> requirement that it can withstand deliberate double-spend attacks that
> can come from anywhere at any time…and relaxing this assumption
> without adequately assessing the risk (i.e. I’ve never been hacked
> before so I can assume it’s safe) is extremely dangerous at best and
> just horrid security practice at worst. Your users might not thank you
> for not getting hacked - but they surely will not like it when you DO
> get hacked…and lack a proper recovery plan.
> Furthermore, the protocol itself makes no assumptions regarding the
> intentions behind someone signing two conflicting transactions. There
> are many potential use cases where doing so could make a lot of sense.
> Had the protocol been designed along the lines of, say,
> tendermint…where signing multiple conflicting blocks results in loss
> of one’s funds…then the protocol itself disincentivizes the behavior
> without requiring any sort of altruistic, moralistic assumptions. That
> would also mean we’d need a different mechanism for the use cases that
> things like RBF address.
> Thirdly, taken to the extreme, the viewpoint of “signing a conflicting
> transaction is fraud and vandalism” means that if for whatever reason
> you attempt to propagate a transaction and nobody mines it for a very
> long time, you’re not entitled to immediately reclaim those funds…they
> must remain in limbo forever.

I'm not talking about changing the protocol - I'm talking about the 
business relationships between users of Bitcoin.

I would expect a payment processor to inform the merchants of relevant 
double spends that it observes on the network, even if the payment is 
actually successful, so that the merchant can decide for themselves 
whether or not to pursue it out of band.

Mining is a kind of technical fallback that allows the network to 
resolve human misbehavior without human intervention. If nobody ever 
attempted to make a fraudulent payment, we wouldn't need mining at all 
because the signed transaction itself is proof of intention to pay. That 
it exists doesn't suddenly make fraud less fraudulent and mean that 
users who are in a position to pursue out of band recourse shouldn't do 

I agree that there are valid reasons for replacing transactions in the 
mempool, I just think they should be implemented in a way that doesn't 
facilitate fraud.

I'd also like to note that "prima facie" doesn't mean "always", it means 
that "the default assumption, unless proven otherwise."

Version: GnuPG v2


Bitcoin-development mailing list

Reply via email to