Quote from: jonald_fyookball on September 07, 2015, 07:29:53 PM

Quote from: brg444 on September 07, 2015, 07:26:08 PM

Quote from: Peter R on September 07, 2015, 07:09:09 PM

Quote from: brg444 on September 07, 2015, 07:01:43 PM

Quote from: Peter R on September 07, 2015, 06:51:05 PM

Quote from: jonald_fyookball on September 07, 2015, 06:37:25 PM

Quote from: brg444 on September 07, 2015, 06:23:47 PM

I'm just gonna drop this here Grin

https://botbot.me/freenode/bitcoin-wizards/2015-08-30/?msg=48477664&page=1

Quote

It's infuriating when someone makes an intellectually weak argument, but slateers it in so much (very nicely constructed) pretext and trappings that people who aren't interested or lack the context to evaluate it on its merits are too busy being mesmorized with the leather binding.

Which paper of Peter's is Greg referring to?

I thought Peter is now arguing we don't need any block limit for a "healthy fee market".

Is this a different paper?

Greg claimed the paper is fundamentally flawed and asked me several times to publicly retract it. Instead, I have examined his objections, and in addressed them I have been able to remove assumptions and strengthen the claims of the paper.

The fee market paper is interdisciplinary and really outside the realm of Greg's expertise: it is a mixture of physics and economics. My opinion of Greg is that he is an professional cryptographer, an amateur physicist, and a lousy economist and game theorist. Remember, he already "proved" that decentralized consensus was impossible...

And it is, Bitcoin uses carefully aligned incentives to work around the problem.

Your paper is fundamentally flawed because it addresses nothing resembling the current dynamics at stake in Bitcoin. More precisely it ignores the incentives for miners to centralize (as they have shown to have) to mitigate propagation times. In effect your paper clearly demonstrates it is more profitable to do so under free-floating blocks and you essentially rely on their altruism to maintain the validity of your model to make decisions going forward. In short, your work might be sound from a technical standpoint, in a vacuum, but can not be used to construct security models that depend on worst-behaviours assumptions.

I cannot accept that you could comment on Greg's game theory intelligence when you continue to wave away the clear negative externality present in your models. You chose to ignore the obvious tragedy of commons at stake and that is why your opinion can never be considered as long as you don't recognize the cost externalized to node and the overall centralization pressure suggested by YOUR alignment of the incentives.

My paper implies that the most "efficient" network configuration (at least from a superficial perspective) is a single miner in a large data center. No one is disputing this fact. No one is also disputing the fact that a monopoly miner like this has the ability to censor transactions and to double-spend (and earn a greater profit this way too).

Will mining centralize around one single miner? No one knows. This is why Bitcoin is still a risk.

But if this doesn't happen--if there remains more than a single miner--then the fee market does exist.

Great, but that's completely irrelevant to the decisions we have to make. In fact it strongly suggest against removing entirely the block size cap so you should really stop pulling up this material as it absolutely does not support your position.

What you essentially propose is to see "how far" the miners will be willing to centralize toward "the most efficient network configuration"

The fees SHOULD be irrelevant right now, but some people (in this very thread) are pointing to arguments that
we have to have small blocks because of fees, so I think Peter's paper actually is relevant if it dispells those fears.

You absolutely don't understand the implications of what Peter's paper suggest .

I couldn't careless if fees are irrelevant right now we need a model that takes them into account for the future.