[Cryptography] ghash.io hits 50% of the Bitcoin compute power

[Perry E. Metzger]

On Fri, 13 Jun 2014 20:52:11 -0400 grarpamp <grarpamp at gmail.com>
wrote:
> >> This is only a risk if you want to spend or transfer your coins.
> >> If you are a hoarder it doesn't matter.
> >
> > That is untrue. The binding between your private key and a set of
> > coins depends on the integrity of the blockchain itself.
> 
> Wuh, false, last I checked unlike other networks, there are no
> serialized "coins" being passed around. You're awarded
> ownership of a balance in a public ledger by signatories that check
> your initial work of solving a hard math problem... the digging
> your own gold part of "mining". You spend off parts of that balance
> by signing it away to others

And how do these others know that the public key in question has any
connection to a particular set of coins in the ledger? How do they
know that some other key, in fact, isn't the correct one?

> If everybody agrees that todays height is legit,

Who is everybody?

> nobody can go back in time to alter that, because the
> millions of people with that legit historical ledger will call
> bullshit. 51+% only matters to future proposed ledger entries.

Do millions of people actually store and check the complete block
chain on a routine basis? Where did your little cellphone wallet
app get its blockchain from? When was the last time you verified the
custody chain of all coins? Does your cellphone wallet app do that
all the time? Where do people get this historical ledger from?

I think you're trusting quite a bit here.

Bitcoin isn't primarily protected by mathematics, it is primarily
protected by a social process that can be gamed. Gaming it is
nontrivial because the social process is protected by cryptography,
but there seems like a great deal of religion behind people's claims
of how well it all works in the face of attack, perhaps because a lot
of people want it to work very, very badly.

Perry
-- 
Perry E. Metzger		perry at piermont.com