Topic: a dumbed down version of the 1MB point of view (Read 2395 times)

Though such a notification system for possible cheater nodes could again be misused by someone holding a big part of nodes and wanting to weed out competition for rewards.

And i still don't see why the fake nodes should be a problem. They can't really do anything.

You worked on neural networks? I had some experiments coded some years back. I believed that you really can get surprising results but had to find out that all they can do is doing what you say them exactly to do. Only difference to a normal sourcecode is that they can find out similar things. That's all.

Only about the protection. Some years ago i was using Emule, which was a p2p adaption from edonkey. Used for filesharing. Since copyright groups did lobbying and the ISP's claimed filesharing hurts their business model because of the big data amounts transferred, the ISP's started to block that traffic.

Emule then tried to go against with constantly changing ports, encryption of the traffic so that deep packet inspection was not possible anymore and other things. The effect was that the ISP's started to throttle traffic that looked like p2p-traffic.

What i wonder is if bitcoin could easily overcome actions against the network. If there are enough interested people behind at all. I mean i know many will give up when bitcoin gets forbidden.

Anyway... the nodes taking part in the network probably would be relatively easy be identified. I mean they only need to do it like the copyright protecters who ran modified emules to find the other users in the network, see what they upload and sue them for that. It was not illegal to run emule but if it would be illegal than whole countries nodes could be switched off that way.

So i really don't know if bitcoin could survive that way since it needs a healthy network of nodes.

Of course that's a theoretical problem only that won't happen most probably.

Back to a few points that I didn't address above...


I would envision weightings when deciding how to do re-orgs.  Thus, a miner would have incentive to produce as 'valuable' a block as possible where on factor of value is how much 'diversity' it has.

As I mentioned in my last post, if the coinbase and transaction fees went onto a sidechain for collection by all block formation participants, the miner really would not have the opportunity to keep the entire reward.  Probably 'old style' blocks would be allowed under certain conditions (e.g., no blocks for an hour.)  This would provide a backup should there be a reward system failure.


On the topic of mock-distribution (e.g., 1000 VM's in Amazon's datacenter) it drives me nuts...

Kaminsky's 'nooter' program measures timings to attempt to analyze whether a packet is from a node behind a proxy or is being deceitful in certain ways.  I mentioned it as one potential way to try to discover if a well funded attacker had tried to amass an army of verifying nodes, but there are certainly others and I would anticipate using multiple of them.  When I said 'compare notes', I mean that if (a) analyzed it's connection with (b) and (c) analyzed it's connection with (b)  then  (a) and (c) compared notes on (b), they might be able to pick out issues.  If they found problems and warned the network of a possible cheater, it could help keep such problems weeded down and discourage people from trying to cheat (in this way) since their nodes (many) would probably end up blacklisted.

If many nodes have the same 'fingerprint' (e.g., they are all VM's in the same Amazon datacenter) they all lose value as desirable 'neurons'.  A relatively valuable node would be one which is sitting in someone's basement (which may be partially verified as the IP being part of a netblock used by and ISP for customer service for example.)  A super desirable 'neuron' would be one accessed over a radio link and situated in Outer Mongolia.  It's beyond the scope of this note to further describe some of the ways 'synapses' form and are valued, but suffice it to say, I look to the biological world as a model...in my musings about this stuff.

Switching to a discussion of mining, that never bothered me much.  Corp/gov could mine for all I care as long as they are not the only one's doing it.  A hash is a hash.  The ideas about weighting 'valuable' blocks based on the verification path metrics might provide some framework to deal with mining consolidation and abuse problems if/when they occur.

^^^^ This X 1 billion. If you ever forget it, read it again. In case of doubt, the same applies. Cut, file it, print it in the BTC bible, and preach it to the masses for it is gold.

For brevity, allow me to combine a few responses.  (BTW, thanks for giving Adam at least a fair-ish shake...it made me less down on Epicenter after what I considered to be a total knob-job with Hearn and inclined to discuss things with out.)

Mining algo:  It would be a hard fork in more-or-less the same class as changing the 1MB blocksize.

Me core dev?:  Nope.  Just a hodler.  I've not written a line of code 'in anger' (for money) for years.  I don't have any correspondence of significance with any core devs and have met a few only enough to mostly just exchange a few pleasantries.  Not associated with anyone or any organization related to Bitcoin.

Goal:  I see a broadly distribute and autonomous support infrastructure as the main pillar upon which Bitcoin can provide value.  I see this as being distinctly in the transfer of transactions around the network.  POW is important, but secondary to me.  I mention this because it helps explain why I am fixated on verifying transfer nodes and the matrix that they form.

Incentives/overhead:  I envision the coinbase being issued on a multi-sig basis where each participant in the block formation can collect their share.  It dawned on me last night that if the reward messiness were handled on a dedicated sidechain almost all of the overhead could be borne only by those who were actively participating (for a monetary reward) and Bitcoin could remain similar if not identical to the way it is now.

Value:  The transfers of wealth within the mainstream and shadow banking systems worldwide is gianormous.  If Bitcoin could tap into that, the support reward could be massive and sustain a very robust system.  I personally don't even use native Bitcoin in an exchange capacity simply on principle.  I personally don't think it is appropriate.  I'm looking forward to something like sidechains so I can start using Bitcoin through that proxy.  That makes sense to me as the only rational way to do scaling.  I have no guilt about 'forcing' other coffee drinkers to use a proxy by pricing them out of the native Bitcoin market because it is something I would gladly do myself.  And because I see almost nothing but advantages in doing so.

Cracks:  Since I got started, the need to shift to steganographic methods at some point has always been in my mind.  For those who do not know, it would be something like taking 1% of a 'cat vid' and hiding data within it.  These methods could 'slip though the cracks' of even an intensely controlled network of the likes most of us have not yet seen.  We do not necessarily have to give up all hope even if Bitcoin is attacked by everything corp/gov can throw at us.  Adam says it best:  "At the margins, steganography wins."

Bitcoin being a non-realtime system means that latency can be used as a defense mechanism.  Latency is the enemy of monitoring and stream analysis.  Additionally, a transaction is a tiny bit of data (relatively speaking) and a 'one-shot' thing.  These inherent strengths of the Bitcoin protocol could make it a very thorny problem for censors even without resorting to steganography.  One of the reasons I would like to see constantly changing data paths (my 'j->h->f->miner' illustration) is that it would further complicate analysis and censorship methods.

Yes, censorship can and would happen on sidechains, but they are not a single-point-of-failure and thus do not need the same level of security.  They are the classic whack-a-mole.  If one is killed off, several better ones will pop up to replace it.  Only the core backing store (hopefully Bitcoin) needs complete protection.  The Blockstream guys have gone one better than I imagined and seem to have plans such that even if a sidechain froze and dies, the money would come back to the owner eventually.  Awesome!

Um... yes, in the new altcoin. The miners would still use their old wallet and mine on that. The new algo only would become "the" new bitcoin when all users switch to it. And will they? Might be, when you promise that they can earn from mining without having an expensive miner.

I wonder how the value of the original bitcoin would develop. Would the electricy cost put into it make it a stable price or would it be like riding a dead horse?


I can't connect your username but are you a core developer?

I guess the question is which fork would win. Some people say the worth of bitcoin is given from mining. That is nonsense. You can put all the workpower you have into shuffling dirt from one side of the street to the other and back. Still, it would not create value. Value might be in securing the network, so the trust in bitcoin creates value.

So trust is one factor.

Next factor might be lazy people. It's very sure that a huge part of bitcoin users would not switch out of being lazy and or slow. Still it might not be a big problem over time.

The miners, with a new algo, would not be a problem. They would not be the ones anymore who could decide which forks win with their hashpower.

Believe... people would need to believe that one of the forks is the one having the value.

Bribery... given users a reward might mean they get convinced. Though it might be that they use it and still hold the old bitcoin.

I think it would be a hard step to reach. But i'm not yet convinced that the new system would be as safe and would have no disadvantages. I think at the very least the new coin is not allowed to be inferior to the current bitcoin. Speed, fees and so on would have to be the same or better. Otherwise the fork will surely not win.


I would not see it as black. Amount of transactions is rising. And adoption is slowly progressing. Surely... all the scams held bitcoin back for a big part. And yes, i imagine bitcoin like a future currency that is widely used. The missing amount of bitcoin users and the missing amount of accepting businesses is what makes bitcoin inferior at the moment. But that slowly and steadily will change. I'm very sure of it.


That would be an incentive that the miner would have to give? Then what happens when the miner keeps the rewards? Or is that cryptographically solved somehow?

And... that is a reward only. It might make it rewarding to run a node but it would be a problem to still run amazon cloud nodes. Next thing is that the reward can't be very high to make it worth to run a node. Plus, if it would be rewarding then it would get saturated to the point where it is not rewarding anymore. It would be like miners who run 100 miners or more. All owned by the same person. It might even happen that someone owns a huge part of it and holds back certain transactions. It would not be a free service anymore so misuse might happen more easy.


So you say you can preven someone from running 10k nodes?

Besides that... is the amount of non usefull nodes a problem at all?


I think it would not be possible to slip through regulations. They would make sure to rule over them all. If that really happens then yes, it would be a big hit. But bitcoin would be used anyway, at least the way it is working now. I don't think that a small incentive would be needed or change much.

Even with big regulations i think the advantages of bitcoin would attract users. A reward system could never be rewarding enough to make a difference. If it could then bitcoin would be as expensive that practically no one would want to use it.

How do you get that idea? Centralization will happen anyway. If or without high fees. With higher fees miners to buy or rent will be more expensive. There is no reason to believe that a different fee would change something regarind centralizaton.

I believe i have read that even satoshie foresaw the centralisation of mining. I wonder if he viewed that as a problem or not.

If there were an update to the POW algo, existing miners would suddenly and abrutly have not jack shit to say about the course of Bitcoin.

Things which are rewarded economically are HIGHLY successful.  More than I and more than probably Satoshi would have guessed.  Unfortunately the only thing which was economically rewarded (in the core system at initial release) was sha256 hashing.  ASIC came at least a year before my own best case estimates.  Now we have absurdities like you pointed out where every transaction costs > 1 daily household energy cost and no clear way out of the maze.

[You might say that the ratio can be improved by somehow getting Bitcoin to be used for every coffee purchase to which I would point out that transactions have been basically free (to end users) for half a decade and Bitcoin has not gotten traction in this role.  I would further say that it is because Bitcoin sucks in this role, and making it get somewhere to par with dedicated exchange currencies would be a bigger re-design than some of the adjustments I suggest.]

Let me explain my 'matrix' and 'path' valuations.  Say there are 10 verifying transfer nodes (a-j).  Say that prior to being mined, a transaction needs to be verified three times before being mined.  Say you want to monetarily reward good distribution.  What you do is statistically prefer the least noted verification paths:

Here's an example of three different paths:

  a->b->c-> mine
  f->a->j-> mine
  d->e->j-> mine

If a particular verification path (e.g., f->a->j-> mine) is continually repeated for many transactions, that path loses weighting.  A miner will likely see the same transaction come in through various paths and can choose the one to insert into the block they are mining.

In my vision of things, a fair fraction of the block reward is distributed to each node who participated in transferring the transactions to the eventual miners thereby producing an economic incentive both to transfer transactions, and to be 'weird' (e.g., not one of 10,000 nodes operating within Amazon's cloud which offers no real value to the network.)

The more complex the system, the more easy it is to game.  The real challenge is to try to make sure that the nodes are legitimately operated by independent players ('Joe Enduser' as much as possible) and are not cheating.  We also want to do this while preserving anonymity as much as possible.  A cheap USB device which only becomes unique in the enduser's hands can achieve this.  If the nodes in the transfer network matrix are randomly making connections to one another and are cross-checking one another (see Kaminski's 'nooter') and comparing notes, liars and cheats can be identified and ejected.

This sounds complicated, but functionally it is actually not a very big shift from what we have in today's Bitcoin.  Most users would not even notice a difference.

Most semi-technical people can appreciated the overhead that keeping such a distribution network matrix healthy would add.  If it becomes necessary to operate Bitcoin under steganographic methods, the overhead of that would dwarf the functional overhead which is why the current 1MB block size is not something I want to see vanish.  Should the internet tighten up with global network filtering and regulations on distributed crypto-currencies, the choices would be basically one of three:

 - Forget about distributed crypto-currencies like Bitcoin as a footnote in history.
 - Comply with any regulations mandated to use them.
 - Adapt crypto-currencies to slip through the cracks and provide economic incentives strong enough to incent people to try.

From what I've read and i've read quite a bit, the limit is not an artificial one because some guys have an agenda to push Bitcoin into a fee market model, the thing is, the problem surrounding Bitcoin now leaves us no other choice but to do this unless we want to jeopardize the decentralization of nodes + other security problems.

You know there exists other actors in an economy than the "mainstream consumer"?

I think it is artifical since it only was implemented to prevent spam. Though the spam never happened. Only now when it is possible to pay for the spam it comes out. So the spam did not be prevented by the limit though in fact it is made possible now. You can fill the block rather cheap.

The 21m is artificial too, though the currency can work with it without problems. That is not the case with the blocksize limit. You can't triple the userbase including trippeling the transactions and await that bitcoin still won't work. It will not. Inevitably around half of the transactions can't be confirmed. And the race for fees are not able to change anything with it. The fees can't make the transactions less. And a bitcoin where you don't know if your sent money is really sent is so unreliable that only a few would use it.

10m block is artificially too, it's a compromise that had to be made in order to make bitcoin work.


It might be worthy for that case but it would definitely not be the alternative to the banking system. It would be a very limited thing. And there is no real reason to limit it at the moment. You don't imagine bitcoin becoming a world wide used currency? Maybe in 50 years being the mostly used one? I think the advantages of bitcoin can bring this. So taking away advantages of bitcoin, or making it unreliable, will let it stay in a niche. At least i don't want that.


Sorry, but a currency that is not used will die. Why should bitcoin remain being a geek currency? Yes, scams happen but every new thing born had it's birth aches. Even when you don't want that... this will not stop. Since always some will leave and some will come. Only education and learning will help there. No reason to be in fear and restrict everything. Being in fear is the first step for dying. In fact it is the fate of conservatism. There will always be change. Don't fear it or you will be hurt. And since when did conservative thoughts stop the evolution really? North Korea? Will go down at one point. Maybe when they don't have enough money anymore like all socialist countries.

Hah, yeah, the forum can eat a lot of time.

I don't understand fully what your vision is but i think there is zero chance that bitcoin will be the currency that implements that. Miners are the one to decide and they will never do that.

I'm not sure if the network matrix and black listing would be a good think for decentralization though.

POS never made any sense to me and does not now.  I was implying that it would be relatively straightforward to change the POW algo from sha256 to other things, and that could (relatively) easily happen if miners acted in a manner which was to dickish.

I would like to see rewards base on 'trasfer node path'.  That is, a unique and valuable 'path' of verification would be weighted in reward, acceptance, or both.  Some CPU-friendly per-verification-node pre-mining and/or optimization might be worthwhile both as a means of enhancing efficiency and verifying the distributed nature of the 'transfer path'.

TPM-like technology got Todd, Hearn, and myself sporting tech boners some years ago.  The reason it had that impact on me is that it might be a way to enhance the credibility of a unique verifying transfer node.  That is to say, it would be cheap and easy enough to obtain one (envision a thumb-drive miner form factor) but difficult enough to discourage well funded attackers from trying to flood the network in sybil mode.  Todd's 'scratch off' power-on key generation initialization makes the idea really workable.  I would envision 'black listing' of mis-behaving devices as an aspect of the transfer network.  So, say, a node were not verifying transactions (a current problem) the device would be obsolete and the operator would lose their $3.00 investment and ranking in the network matrix.

As I said, the technology needed to get a sustainable system which fosters good geo-political distribution is not trivial.  But it very well could be worth doing and it might be the case that a solution which gets it right may be the one which eventually bubbles to the top of the heap.

I'll maybe get back to some of your other points later (since you don't seem to be a total shill), but I better get something useful done today.

Exactly my thoughts. The maximum blocksize limit is an artificial limit from the start. You can't say it is organic when fees develop in such a cage.

And yes, miners will be the new banks. Unfortunately at that point satoshi might realize that not much is left of his idea.

So do you prefer bitcoin switching to POS-Mining or something similar? Surely that would shut up all the persons claiming bitcoin is bad for the environment. But it surely will never happen since miners decide which chain wins and you can bet that most bitcoiners will switch to use that chain. I'm not sure what must happen that a pos-chain gets the new real bitcoin chain.

And yes, the risk exists that miners will become the new banks. It's already very bad that mining is so centralized but they have all the might. They can raise fees and do whatever they want. I doubt that something like visa will happen now. The miners are still idealists for the most part and they have to fear losing users. Maybe some years in the future...

Unfortunately i think miners might really tend to prefer high fees. Since that is how they earn. The only thing we can do now is hope that they are sane enough to not do this. They would hurt themselfes. But since businesses nowadays often enough life after the ruling "Fast money now, press as much out of it as possible and let it die. Move to the next." then i don't have the best feeling about that too.

I really am not sure about blockstream. How many trust will i give some group that looks like they prevent bitcoin from solving it's problems to make their own project more successful? I don't know if it is the case, but it looks so. I'm really unsure that i will support such persons or network. And they don't do it for free. And how decentralized will it be? Will the investors have more power about how that network works or will it be less? Anyway... seeing this drama iam VERY cautious with that.

If you really transfer mostly such high amounts then you are a very very rare person in the community. If you think it would be enough to set on these high amounts then surely this will not work. Most people could not use bitcoin anymore and a bitcoin that can't be used for every usecase is very very very limited in it's usefullness. Whole countries could be blocked from using bitcoin. Surely that can't be a solution. And it surely is not "replacing fiat money" in the way satoshi wanted to build with bitcoin.

Batchsystem? Well, that sounds like you think killing two advantages of bitcoin could be healthy somehow. Bitcoin only has a couple of advantages that it can use to draw adopters. It's nearly free, it's faster and you can transfer any amount. Bitcoin will die if you take one or more of that. Because who and why should use bitcoin? For illegal things? Surely bad for bitcoin. And legal things? Buying a car? Why should i pay with bitcoin, having to wait until night for my transaction to confirm and then i can take my bought car driving home? I would simply use fiat then. Bitcoin would have lost all advantage.

Really... at that stage i only can imagine bitcoin being usefull for illegal buys, transactions, money laundering, black money and such things. There simply would be no advantage using bitcoin for normal things anymore.

If by 'problems that have not happened yet' you mean issues where TPTB acting through core service providers enforce certain legal mandates on Bitcoin infrastructure providers, vendors, and users, I would agree with some of what you say.

As best I can tell, the reason given by people who claim that government interference with Bitcoin is impossible amounts to 'because: freedom'.  And secondarily, because TPTB have such respect for and fear of the masses that they would not dare...totally ignoring that the masses who have even heard of Bitcoin don't give a fuck about it by-n-large.  I find both of these suggestions to be ludicrous and expect that in the event of a monetary crisis in fiat-land both laughable assertions (and anything resembling unencumbered use of the global internet by the plebs) would vanish into a tiny puff of smoke.

I'm saying the issue is two-fold: incentivize the security of the network through fee pressure & limit the cost to participate in validation of the network (becoming a peer by running a full node). The block size limit addresses both.


Every one of these assets, given their physical nature, can be confiscated. It can also be said that in most cases their value is/can be manipulated.

Litecoin doesn't come close to offering the security of the Bitcoin network nor its liquidity or trust. Trust, in particular is paramount as this is from where value is derived. Obviously the value of Bitcoin is not guaranteed at this point but it certainly has a better shot than Litecoin. This goes for all other cryptocurrencies.


Banks already have their own payment networks. As for them bootstrapping their own crypto I do believe I said something previously about "monetary sovereignty"... not sure how this suggestion of yours fits in there..


Sorry I should have been more precise: by "security" I meant decentralization and it should be preserved by fitting the maximum blocksize to a rational & conservative observation and projection of "the cost of the option to create a full node". More here: http://www.truthcoin.info/blog/measuring-decentralization/

To think this whole debate began with, Its not a problem that there are problems because the problems only happen some of the time, of course the problem will become bigger problems in a few months but that's not a problem!

I think the bigger problem, is the problem that people are ignoring the problem until it becomes a bigger problem and then they increase the problem, by adding too much debate about the problem.

Increase the damn blocksize and end the damn debate.

So you are saying the limit is just the intermediate step, and the true basis for the fee pressure is the centralization risk?  That’s fair enough.  But that is a completely different take on the situation than saying ‘we need to limit TPS so a fee market can develop, so we can pay for security’.  I’m not saying you said that or didn’t say that, but the concept is repeated often enough in defense of keeping a block size limit, and it’s bunk.

Why do you think that only Bitcoin offers it?  What about Litecoin? Monetary sovereignty is important, but anyone who wants to already has wealth freedom / sovereignty.  Buy some gold or gems if you are worried about confiscation, or invest in land if you want protection from devaluation.

Bitcoin does have a big lead, but I think you are being too complacent in saying a new form of money is almost impossibly hard to build.  We have $3 Billion.  There are $100's Billions out there that banks could use to boot strap something, and most people in the world seem to like big shiny corporate logos.

Yes, most likely.

So what can the security of the network be left to?  Hoping that miners don’t collude?

A limit is a limit.  I’d call it an artificial limit if the purpose is to create a fee market.  If the limit is there by a technological necessity, than it would not be an arbirary limit, but one dictated by capabilities of the technology.  As for the examples of other limits you gave, the 21-quadrillion satoshis would only become relevant once the price of a single satoshi begins to approach commonly used minimum transaction values.  And the POW algo? The TPS limit dictates a maximum user base with direct access to the chain in a way that the POW algorithm has nothing to do with.
I understand the desire to have bitcoin be internet equivalent of a fallout shelter, and it can be that.  In fact, it kinda started out that way and was hardened over time, being able to run over TOR and such.  I think bitcoin has a better chance of survival going public, taking the Uber route, making itself economically indispensable such that governments can’t afford, plicately or economically, to ban it.  But that’s my amateur opinion.  As I suggest on the dev mailing list, it would be best to consult policy experts when trying to draw up the part of the threat model that includes governments.
I mostly agree with this.  BTW, the coffee argument is a straw man at this point.  Who is seriously suggesting that bitcoin in it’s current form, but with 1000x the TPS, would be a good system for people to buy coffee with?  No one, really.