Quote from: avivz78 on December 07, 2013, 09:40:52 PM

Quote from: Peter Todd on December 07, 2013, 08:46:52 PM

Going back to your original post:

Quote

We note that block propagation times are the primary obstacle for scalability.

The obstacle to scalability is keeping Bitcoin decentralized while scaling up; we know that Bitcoin can scale if we sacrifice decentralization - Visa and Mastercard are doing just fine. Ultimately you're proposing something that solves one problem - the high granularity of confirmations and the long wait associated with them - at the expense of scalability with decentralization. So don't claim you've done anything other than presented an interesting academic analysis of a specific trade-off possible in the system.

Peter,

I keep trying to agree with you.

One last try before I give up:

Surely you must agree that if there were no block propagation delays, there would be no centralization associated with scaling up. Blocks go instantly to everyone, and there are no problems with miners getting more than their share of the blocks.

No, I bet he won't agree on that.
Let's say we want the network to validate 1 million tps. Even if nodes communicate with each other using the ansible (zero block propagation delays) you need all full nodes to validate 1 million tps. How many full nodes do you think we would have?

If people stupidly complain about ASIC miners "causing centralization" or "SHA256 being less democratic than scrypt", wait to see what they have to tell about having a room-sized super-computer (I heard NASDAQ market execution computer is that big) alongside their ASICs machines for Pow. Well, they will need that computer even for full nodes that don't mine.

Quote from: avivz78 on December 07, 2013, 09:40:52 PM

I guess the difference in our perspectives is that I don't think these are optional tradeoffs. If at some point bitcoin becomes mainstream, there will be an increased pressure to allow more transactions through (or alternatively, fees will soar higher than the money transmitters Bitcoin is trying to replace). You then run into several problems. One is a security problem against attackers. The other is a pull towards centralization. All I'm saying is that we claim to solve the first problem. The second clearly remains.

Bitcoin will never be able to process the tps the full dollar system (with all their banks hierarchy) on its own.
We need off-chain transactions, micro-transactions between two parties with transaction replacement, etc.
p2p currencies are a trust-less extreme, but there's other monies that beautifully leverage trust, we need them to complement each other (I'm talking about LETS and other local currencies, ripple-like credit networks, simple IOUs, etc).
Technologies like Freimarkets (the private chains part), two-phase-commit Ripple or Open Transactions are not as trust-less as bitcoin, but they scale much better.
So, yes, this is a tradeoff and we know bitcoin can't scale to be a global near-monopoly money like the usd is today.
Luckily there's monetary innovations out there that will complement bitcoin much better than national currencies do today.

EDIT: I still find the proposal interesting, just wanted to explain that propagation times are not the only obstacle for scalability.