New video: Why the blocksize limit keeps Bitcoin free and decentralized

[benjamindees]

I think you're mistaking the kind of anonymity being talked about here.  Say some random authority wants to force miners to only mine transactions that the authority approves of and not mine any that match some kind of blacklist, and to not extend any chains that contain violations of these rules.  If successful this would substantially undermine the purpose and goals of Bitcoin.

I see.  So you think such a blacklist is technically feasible?  How often would it need to be updated, in order to be effective?  How could it be enforced, unless the authority in question controls 51% of the network?

And, in the other case, all it takes to evade such regulation is access to a single miner somewhere in a region that is friendly to Bitcoin.  We can pretty much assume that will always exist, in one form or another, for the few who need it.

[1714520369]

1714520369

[1714520369]

1714520369

[1714520369]

1714520369

[1714520369]

1714520369

[1714520369]

1714520369

[edmundedgar]

I think you're mistaking the kind of anonymity being talked about here.  Say some random authority wants to force miners to only mine transactions that the authority approves of and not mine any that match some kind of blacklist, and to not extend any chains that contain violations of these rules.  If successful this would substantially undermine the purpose and goals of Bitcoin.

Yes, it's technically feasible. The thought is that the attacker (presumably a large government, or group of governments) would make laws forcing miners in their jurisdiction to follow their rules, not Satoshi's. The idea is then that if miners are able to operate anonymously we'd have a defence, because they'd all just hide behind Tor and the government couldn't get them because it wouldn't know who they were.

The problems with the defence are:

1) It's not enough for the miners to be _able_ to duck behind Tor and hide under the radar rather than comply they also need to be _incentivized_ to go off the radar. There would still be a chance that they'd get found out. ("With an electricity bill like that you're either mining bitcoins or growing weed, let's search your house and find out which...") Better to comply rather than risk going to prison.

2) Small-scale anonymous miners still have to out-compete miners who are working out in the open, which they can't do any more because Bitcoin mining is too capital-intensive, regardless of what you do with the block size. (See my previous post.)

3) The attacker wouldn't have to use coercion. They could use bribery instead. Anonymity doesn't help with that, because thanks to Bitcoin, you can bribe people anonymously. Decentralization doesn't really help either; If anything it makes it worse, because smaller miners have less to lose if their actions reduce the value of Bitcoin compared to the amount they stand to gain.

4) If all that failed, the attacker could just set up the mining infrastructure themselves. It's expensive to set up, but you make up for it in mining revenue. They have access to cheaper capital than the private sector, so they could probably just out-compete the non-compliant miners and drive them out of business. If they weren't able to operate as efficiently they'd need a subsidy, but only enough to cover the difference between their inefficient operation and the efficient private-sector operations, plus a little bit extra to make sure they wiped out their competitors' profit margins.

All this makes me think that the chances of the Tor defence being both necessary and sufficient are pretty minimal. On the other side of the equation, throttling the network and forcing transaction costs up takes away one really effective defence Bitcoin has against hostile government activity, which is the ability to take the world's payment systems hostage so that the government can't shoot it without hitting something it cares about. This is done by getting loads of businesses and non-profits using Bitcoin, right on the block-chain, so that everybody depends on Bitcoin and will stop their donations to any politician who messes with it.

In fact, if you were a representative of a hostile government or somebody trying to protect the legacy banking system, I think the most effective move you could make right now would be to make a video trying to persuade the Bitcoin community to cripple their own currency by throttling it at 7 transactions per second...

[Impaler]

Videos argument is irrelevant because ASICS and the ever growing size of the block-chain, indeed the very nature of PoW will inevitably centralize all mining regardless of block-size.  Fundamental nature of BTC implementation lead to centralization and simply demanding that nothing change will not prevent that, the rules of the protocol constitute an unstoppable force or 'invisible hand' so to speak.

[amincd]

On the other side of the equation, throttling the network and forcing transaction costs up takes away one really effective defence Bitcoin has against hostile government activity, which is the ability to take the world's payment systems hostage so that the government can't shoot it without hitting something it cares about. This is done by getting loads of businesses and non-profits using Bitcoin, right on the block-chain, so that everybody depends on Bitcoin and will stop their donations to any politician who messes with it.

In fact, if you were a representative of a hostile government or somebody trying to protect the legacy banking system, I think the most effective move you could make right now would be to make a video trying to persuade the Bitcoin community to cripple their own currency by throttling it at 7 transactions per second...

+1

[benjamindees]

All this makes me think that the chances of the Tor defence being both necessary and sufficient are pretty minimal.

...

In fact, if you were a representative of a hostile government or somebody trying to protect the legacy banking system, I think the most effective move you could make right now would be to make a video trying to persuade the Bitcoin community to cripple their own currency by throttling it at 7 transactions per second...

Ding, ding, ding.  Hobbling Bitcoin transactions and mining over Tor is such a ridiculous solution that only those working for the "good guys" could suggest it with a straight face.

[tvbcof]

...
On the other side of the equation, throttling the network and forcing transaction costs up takes away one really effective defence Bitcoin has against hostile government activity, which is the ability to take the world's payment systems hostage so that the government can't shoot it without hitting something it cares about. This is done by getting loads of businesses and non-profits using Bitcoin, right on the block-chain, so that everybody depends on Bitcoin and will stop their donations to any politician who messes with it.

Like I've said before, a healthy and active solution of off-chain activity would provide the necessary user numbers to give the protection you (supposedly) seek.  Any regulatory attacks are going to be a broad enough brushes to impact all such economic activity.  You seemed to have no answer for that.

I also said back in 2011 that a viable way to attack Bitcoin might actually be to either leave it alone or foster it's growth and let it burn itself out (or mutate away it's peer2peer nature) due to uncontrolled growth.  The relative lack of molestation and surprisingly positive media attention that allowed/facilitated growth over the last few years makes me wonder if that was not a course of action chosen.  The likes of Chuck Schumer have been interestingly mum since their initial burst of verbal diarrhea.

Again, I see a light and highly distributed core value system with a myriad of off-chain solutions of different postures as being a pretty hard whack-a-mole problem to attack on a sustained basis.  Which is why it appeals to me.

In fact, if you were a representative of a hostile government or somebody trying to protect the legacy banking system, I think the most effective move you could make right now would be to make a video trying to persuade the Bitcoin community to cripple their own currency by throttling it at 7 transactions per second...

I don't see it as crippling in the least.  Nor do I see a lack of room to grow at our current settings.  On the other hand, the fantasies of a one-world currency that are everything to everyone are almost certain to end in tears as I see it.

[johnyj]

Very nice video, and another point did not mentioned in the video, if you change the block size limit, you will split the blockchain into two fork and cause lots of confusion

I always prefer speed over functionality when it comes to software, the functionality can always be done at user level with many local adaptation, good for scalability, good for a robust infrastructure

[Prattler]

The video is too extreme for my taste. Obviously, a 2MB or 4MB block size limit will not kill bitcoin. I still applaud the author for pointing out the problem!

The basic idea is extremely sound: the block size limit needs to always be kept relatively small and increased only when there is actual transaction demand (not SatoshiDice). Running a full network node should not be too expensive and should stay available to hobbyists. This would keep bitcoin free and open, which is what we all want!

[amincd]

Like I've said before, a healthy and active solution of off-chain activity would provide the necessary user numbers to give the protection you (supposedly) seek.  Any regulatory attacks are going to be a broad enough brushes to impact all such economic activity.  You seemed to have no answer for that.

What "healthy and active solution of off-chain activity"? All off-chain activity occurs on centralized third party payment processors that you have to trust your coins with.

People seem to be convincing themselves that there is some awesome off-chain solution that doesn't exist. The only secure decentralized off-the-chain solution I know of is payment channels between two parties using lock-time that allow for fee-free micropayments, but this is only useful in a small percentage of transactions, where you have an ongoing financial relationship with someone, and since it still requires the creation of one on-chain transaction to set up the secure payment channel, with $20 fees, it would be almost useless.

This leaves relying on third party payment processors. With $20 fees for on-chain transaction, it would be expensive to get your coins out once you have transferred them to a third party off-chain processor. It's not like it is now, where you can easily get your coins out once you've deposited them at some website.

To avoid high fees, the coins will have to stay within a network of payment processors that have trust-based relationship with each other. They might settle the debts they have with each other at the end of each business day with large on-chain transfers, much like modern banks that use CHIPS.

It would be MUCH EASIER for a government to attack a large payment processor than a high bandwidth/hash-rate node. A node is mobile: it simply needs to connect to the internet to be active. A payment processor needs to rebuild those trust-based relationships with other major payment processors to be useful, so once it's shut down, it would take BTC a long time to repair.

The shut-down would be especially damaging to BTC because it would cost the many account holders $20 each to get their coins out. It's more likely that payment processors will transform into banks, in which case BTC will be just another type of bank credit, that's expensive to transfer and is closed in by all types of bureaucratic walls.

A solution that leads to a limit of 7 peer-to-peer transactions per second and $20 transaction fees is not a good one.

[benjamindees]

I think you're mistaking the kind of anonymity being talked about here.  Say some random authority wants to force miners to only mine transactions that the authority approves of and not mine any that match some kind of blacklist, and to not extend any chains that contain violations of these rules.  If successful this would substantially undermine the purpose and goals of Bitcoin.

I see.  So you think such a blacklist is technically feasible?  How often would it need to be updated, in order to be effective?  How could it be enforced, unless the authority in question controls 51% of the network?

And, in the other case, all it takes to evade such regulation is access to a single miner somewhere in a region that is friendly to Bitcoin.  We can pretty much assume that will always exist, in one form or another, for the few who need it.

Yes, it's technically feasible.

Just saying this doesn't make it true.  Can you actually answer any of my questions?

This regulation FUD is a complete canard, being pushed by investors with obvious agendas.  Quite frankly, I'm shocked that some of you are naive enough to fall for it.  Entangling Bitcoin with the existing financial system, and all the baggage that entails, would be the beginning of the end.

Perhaps Bitcoin can be shutdown within a country.  If so, it's their loss.  Perhaps Bitcoin could be completely destroyed by a motivated attacker that doesn't care about the costs.  This is widely recognized.  But regulated and micromanaged at such a level?  Give me a break.  They can't even keep banks from laundering billions in drug money.

[jl2012]

Video: http://www.youtube.com/watch?v=cZp7UGgBR0I

Website: http://keepbitcoinfree.org/

Pretty simple right now, but this is the beginning. For those of you at the 2013 conference, I'll be giving a presentation about off-chain transactions on Saturday as part of the tech stream.

Could the government shutdown your "off-chain solution"?

If yes, forget it.

If no, why?

[Peter Todd]

Could the government shutdown your "off-chain solution"?

If yes, forget it.

If no, why?

There do exist such off-chain solutions that the government can't shutdown; keep in mind the Silk Road is a very simple off-chain transaction system and it has high resistance to government shutdown. With good auditing and punishment for fraud you can make off-chain systems where the operators can be totally anonymous, yet you can still trust them to hold some Bitcoins on your behalf. It's similar to how miners are anonymous to you, yet the incentives in the Bitcoin system are such that a 51% majority of them will not attack Bitcoin.

Don't forget that moving a lot of tx volume to off-chain systems makes it much harder for governments, indeed anyone, to track where money is flowing in the blockchain simply because that information is only available from a whole host of third-parties, or doesn't exist at all in the case of provably anonymous chuam tokens. Unless the people you are up against are arguing that Bitcoin needs to be shutdown because they can't track transactions, having less public information about where money is flowing gives fewer examples of dodgy transactions to point too. For instance right now due to how SatoshiDice bets are completely public it's easy for anyone to say the majority of Bitcoin tx volume is gambling, if SatoshiDice was using off-chain it'd be much harder to prove that even if you could get a list of addresses they were using.

[jdillon]

I finally got a chance to see your new video. It's solid professional work, you have done a great job. You'll soon get another 2.5BTC from me by the same method I used before. (https://bitcointalk.org/index.php?topic=189792.msg1968200#msg1968200)

Nice to see that big 10BTC donation you got, and from an address with 125BTC! It really says something how many of the donations you have been getting all come from addresses with large balances of Bitcoins, about 250BTC and counting right now. It just goes to show how the people most heavily invested in Bitcoins are the ones with the most to lose from centralization and regulation.

Keep up the fight.

[tvbcof]

Thx for the reminder JD.  I sent a few more.

I was going to donate around .3 to someone, but I cannot for the life of me remember who.

I made a slight contribution to Jon Matonis while I was at it.  Not enough to warrant a dick sucking though...at least not enough to get a solid.

Ah...right.  Armory!  I don't actually use it, but they gave me a shirt at the 2013 conference.

[tvbcof]

Like I've said before, a healthy and active solution of off-chain activity would provide the necessary user numbers to give the protection you (supposedly) seek.  Any regulatory attacks are going to be a broad enough brushes to impact all such economic activity.  You seemed to have no answer for that.

What "healthy and active solution of off-chain activity"? All off-chain activity occurs on centralized third party payment processors that you have to trust your coins with. ...

Wrong.

It leaves a solution where my 'savings' are widely distributed all over the world on a system which is trustable because it is tight enough to be operated by a myriad of small players and enthusiasts.  (That would be Bitcoin.)

I would distribute out my spending money to various off-chain solutions which seemed the most promising to me.  Some might be attacked successfully and buried, and I might mis-judge some and they might rip me off.  Even if all of them failed, I would still have the bulk of my value in BTC.  I would probably only make several Bitcoin transactions per year.

This contrasts sharply with the vision of Bitcoin as a monolithic one-world currency solution which only large and well connected entities can profitably operate.  In that case a failure means the loss of all of my value*.

(*) Actually, my value is still in the block-chain and that, huge though it may be, would very possibly form the basis for value follow-on solutions, but the interim period would be protracted, stressful, and fraught with questions.  I'd rather not have to see it.

[d'aniel]

This contrasts sharply with the vision of Bitcoin as a monolithic one-world currency solution which only large and well connected entities can profitably operate.

I keep hearing this statement, but never any numbers with it to back it up.  Would you mind sharing how you estimated this operating cost?

[jdillon]

This contrasts sharply with the vision of Bitcoin as a monolithic one-world currency solution which only large and well connected entities can profitably operate.

I keep hearing this statement, but never any numbers with it to back it up.  Would you mind sharing how you estimated this operating cost?

Focusing on the operating cost is wrong. The issue is the availability of anonymous bandwidth and that does not scale with technology. Unless you can mine anonymously the government has control over you, and therefor Bitcoin, and Bitcoin is not really decentralized.

But some people are happy with Bitcoin being ultimately under government control. That is why this is a political issue rather than a technical one, as Peter Todd keeps on pointing out.

[tvbcof]

This contrasts sharply with the vision of Bitcoin as a monolithic one-world currency solution which only large and well connected entities can profitably operate.

I keep hearing this statement, but never any numbers with it to back it up.  Would you mind sharing how you estimated this operating cost?

I just did a quick calc and if 1/2 the people in the world did 1 transaction per day with a single solution, I get around 50,000 operations per second.

Someone at the conference (I heard, but did not see) mentioned that the number of cell phones vs. the number of bank accounts is about 5/1.  So, if Bitcoin became popular, and the movement toward more capable phones proceeds, it would likely result in a pretty large deluge of traffic.

I became concerned about scalability when reading the document on the main Bitcoin.org wiki.  Later (recently) I heard Dan Kaminsky refer to it as "the funniest document in the history of software engineering."  Which I consider one of the funniest comments in the history of software engineering.  And one of the saddest.

Now, I don't doubt that large corporations (who I can now name but will refrain from doing so anyway) can process that kind of data and much more.  And the intelligence information they could obtain in doing so would vastly exceed the cost of operating the system.  But it's not the Bitcoin which attracted me several years ago.  It is, in fact, the exact opposite.

[d'aniel]

Focusing on the operating cost is wrong. The issue is the availability of anonymous bandwidth and that does not scale with technology. Unless you can mine anonymously the government has control over you, and therefor Bitcoin, and Bitcoin is not really decentralized.

So we wouldn't have to fear that it'd require an expensive data center to run a node?

But some people are happy with Bitcoin being ultimately under government control. That is why this is a political issue rather than a technical one, as Peter Todd keeps on pointing out.

Do you think some of these folks might just disagree with the claim that larger blocks would result in Bitcoin being ultimately under government control?  Which ones do you believe really are happy with this outcome?

I just did a quick calc and if 1/2 the people in the world did 1 transaction per day with a single solution, I get around 50,000 operations per second.

Someone at the conference (I heard, but did not see) mentioned that the number of cell phones vs. the number of bank accounts is about 5/1.  So, if Bitcoin became popular, and the movement toward more capable phones proceeds, it would likely result in a pretty large deluge of traffic.

I became concerned about scalability when reading the document on the main Bitcoin.org wiki.  Later (recently) I heard Dan Kaminsky refer to it as "the funniest document in the history of software engineering."  Which I consider one of the funniest comments in the history of software engineering.  And one of the saddest.

Now, I don't doubt that large corporations (who I can now name but will refrain from doing so anyway) can process that kind of data and much more.  And the intelligence information they could obtain in doing so would vastly exceed the cost of operating the system.  But it's not the Bitcoin which attracted me several years ago.  It is, in fact, the exact opposite.

It seems like you got a bit distracted while explaining your estimate of the operating cost...

[amincd]

It leaves a solution where my 'savings' are widely distributed all over the world on a system which is trustable because it is tight enough to be operated by a myriad of small players and enthusiasts.  (That would be Bitcoin.)

Your savings would be locked in a p2p network that can't be withdrawn from without a $20 transaction fee. They would be safe from governments but not from competition that threatens to take over BTC's role as currency used for trade because of a myriad of reasons: lower transaction fees, status as 'coin of the realm' of a country in the case of fiat currencies, more merchants accepting the currency, etc.

I would distribute out my spending money to various off-chain solutions which seemed the most promising to me.  Some might be attacked successfully and buried, and I might mis-judge some and they might rip me off.  Even if all of them failed, I would still have the bulk of my value in BTC.  I would probably only make several Bitcoin transactions per year.

This is just my opinion, but I think the vast majority of people would prefer a fully automated p2p network that's made up of thousands of nodes operated by enterprises, than a fully automated p2p network with hundreds of thousands of nodes operated by hobbyists, but with the disadvantage that you need to trust third party payment processors for any kind of real world use of currency.

If we're going to use proprietary networks (and they have to be networks to avoid expensive on-chain transactions) to transfer bitcoin credit, instead of bitcoin itself around, we can just use banks, or some successor to e-gold or something.

A bitcoin economy where you are reliant on using bitcoin-credit controlled by third parties is very similar to modern banking, and would probably have almost all of the same shortcomings.

This contrasts sharply with the vision of Bitcoin as a monolithic one-world currency solution which only large and well connected entities can profitably operate.  In that case a failure means the loss of all of my value*.

Your vision requires large payment processors that are connected, in the classical banking sense of having trust-based credit relationships amongst each other, to form a proprietary network that allows you to actually use bitcoin in every day trade.

Bitcoin as a global scale network allowing millions of on-chain transactions per day, or as you put it, a 'monolithic one-world currency solution', was in fact the goal Satoshi Nakamoto set out for it:

http://www.mail-archive.com/cryptography@metzdowd.com/msg09964.html

At first, most users would run network nodes, but as the network grows beyond a certain point, it would be left more and more to specialists with server farms of specialized hardware.  A server farm would only need to have one node on the network and the rest of the LAN connects with that one node.

The bandwidth might not be as prohibitive as you think.  A typical transaction would be about 400 bytes (ECC is nicely compact).  Each transaction has to be broadcast twice, so lets say 1KB per transaction.  Visa processed 37 billion transactions in FY2008, or an average of 100 million transactions per day.  That many transactions would take 100GB of bandwidth, or the size of 12 DVD or 2 HD quality movies, or about $18 worth of bandwidth at current prices.

If the network were to get that big, it would take several years, and by then, sending 2 HD movies over the Internet would probably not seem like a big deal.  

Limiting Bitcoin to 7 tps in my opinion virtually guarantees it will never have significant impact, because I believe people will opt for government-regulated networks, or alternative blockchains, if Bitcoin, as a p2p network, loses its transaction fee advantage.

The 1 MB cap was put in place as a temporary measure, until a better way to control transaction spam was found. Trying to make the limit permanent is trying to change the vision of Bitcoin as originally conceived.