Re: [Bitcoin-development] Coinbase reallocation to discourage Finney attacks

[Gareth Williams <gacrux@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 1932 bytes --]

On 30/04/14 00:13, Mike Hearn wrote:
> Every time miners and nodes ignore a block that creates >formula() coins
> that's a majority vote on a controversial political matter

Actually, there's one more thing I'd like to add. Apologies to the list,
but it bears repeating:

* rejecting a block at validation

Is /very different/ from:

* reinterpreting a block that has already passed validation

Nodes ignoring a block that creates >formula() coins is rejection at
block *validation*. That's the protocol working as it is supposed to.
It's deliberately an all-or-nothing mechanism; you can't pick and choose
which blocks you like. If 51% of the network say your block is invalid,
they're now on a different fork. The consequences are this drastic on
purpose, for stability.

Nodes reinterpreting a block that has already passed validation is
almost the polar opposite of this. They're /ignoring the protocol/ and
making up their own meaning for stuff. Sidestepping the mechanism in the
paragraph above. I would hope it'd be self evident that this is dangerous.

Adam Back is arguing practicality in this thread. I'm arguing
fundamental principle. (And, er, someone else is randomly throwing
around ad hominems, which we'll politely ignore; Mike could work for
Lucifer himself and his good ideas would still be good, and his bad
ideas would still be bad, so let's just stick to the ideas eh.)

So, fundamental principle: don't reinterpret history!

We have validation for a very good reason. Undermine it and you might as
well have an unvalidated system like Counterparty, which I wouldn't
ever trust with more than the value of a small hamburger.

If the economic majority starts reinterpreting history (through whatever
voting mechanism / side-channel you like) that completely undermines
Bitcoin's validation, and its PoW. It's worse than 51% of miners
deciding to rewrite history.





[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 555 bytes --]