Topic: Please do not change MAX_BLOCK_SIZE - page 15. (Read 13023 times)

Can you use/fork altcoins (maybe a temporary blockchain?) to handle microtransactions with minimal impact on the blockchain?  Is this completely insane?

It sounds like any solution for microtransactions AND not bloating block size requires some degree of centralization.  Like, microtransactions to company A require that all people who want to microtransct with A use a pooled, shared BTC account.  then it batches together transactions for many people, keeping the data of who gets what off the blockchain.

but it would be one-way, only many people to one company.  one company to many people (ASIC miner shares for example(?)) seems to be a pain in the ass

The risk of a 51% attack on a lower hashrate network is mitigated by the fact that the transactions on said network are significantly less valueable.


i.e. the Y $ invested to pull off a 51% attack on the main network produces X $ expected value from a double-spending attack, X/Y is the same for the alternate chain.  Of course calculated in BTC for these purposes

it just seems like altcoins (especially with faster blocks, and for transactions where the value is low that you don't really give a damn about the [hopefully] small chance for a 51%/double spend attack ) would be a good solution for micro-transactions.

I think it's more likely that the community would let a fork go, rather than trying to beat governments on hashing power, resources, or (lack of) anonymity in order to run a node, or a miner...

People can adopt easily and they can easily switch to another branch, if they find an actual need to do so.
In fact, knowing life, I wish they didn't, but they probably will.
But I don't think it will be the end of the world.
If we managed to gather so much power to build this net, we can as well manage to move this money into the right branch, after a potential fork.
When the chain splits, the hashing power must also split - so the risk of a 50+% attack gets actually lower for you, because you get to choose to go with a branch that you personally consider less vulnerable.

The government level attacker would have to match that spent by the new miners, which is roughly proportional to the expected value of their block rewards, which is roughly proportional to the demand there is for the censorship-preserving fork.  Are you doubting this demand?  The effort can be trusted to be as real as the ASICs produced by chip makers, which would be being produced well in advance of the fork, and purely as a for-profit enterprise based on expected demand for the fork.

If demand for the censorship resistant fork is high, but the government level attacker is still willing to outspend the miners, well then the problem is completely orthogonal to the issue of block sizes.

And just to be clear, all of this matters only after 1) all governments across the world have successfully colluded to eradicate all non-hidden, censorship-resisting mining pools everywhere, and 2) mining pools aren't able to find any technical means to hide the location(s) of their machines, despite the funding available from parties (good and bad) with financial interest in resisting censorship.

Not to mention there are other, non-technical, ways to deanonymize miners: "Just register with us, the government, to collect your mining subsidy."

I hypothesis that the competitive parties in mining will be 'the masses' who, like you say, don't use enough power on an individual level to consider, and large entities with a big enough footprint to harvest high value business intelligence.  The latter can subsidize their power costs.

New as Bitcoin is, we've already seen that power costs can influence mining effort, so it is not a wild theory at this point.  I do suspect that small mining operations will vanish unless they can monetize the data they can obtain and sell it up-stream to entities who can use it.  Servicing SPV clients or offering stratum server services may be one way to do this.

I strongly agree that 'ASICMiner style USB' devices in the hands of ordinary market participants are a powerful and positive thing.  I do have some qualms about buying sha256 ASIC gear myself or inducing others to do so however.  It may not happen this year like Kaminsky tried to claim (or ever), but it could be that a different and possibly varying proof of work algorithm is in the interest of a 'free' Bitcoin.

Relatedly, I've been musing about whether it might be possible to favor work done by demonstrably 'network edge' participants.  These players have the 'strength of numbers' which make the concept of voting based on choosing software work.  I hope that smarter persons than myself have been pondering that as well.

Heck no. How on earth do you think you are going to manage to come up with enough funds to outspend potential government level attackers with a brand new ASIC and hashing algorithm, and on top of that, have people trust you that the effort is real?

In the future, we may need a lot of ASICMiner style USB miners which can run all day at high hashrate, yet doesn't cost you anything noticeable in electricity, yes it will not give me a steady income of mining revenue, but should be a worthwhile investment to secure the network if the power cost can be driven to low enough that you can just plug it into your laptop and stop caring.

I can only speak for myself, and I don't think that it should be a big factor in this decision, but...

I will definitely be liquidating a fair fraction of my BTC holdings if I don't feel that the solution is on a trajectory to withstand very significant growth and very significant attacks.  It is in fact something I've been actively planning but a serious discussion about block size and off-chain frameworks has given me new hope.

Now, it is fair to say that I'm an unusual thinker in my system analysis.  I was buying with both hands in late 2011 based on analysis and intuition, and with a few exceptions it seemed like many/most of the people on bitcointalk.org were kind of on death watch.

Excellent.  I think it's important for the small blocks crowd to enumerate these hypothetical difficulties then, so they can be addressed.  Were you convinced by my response to the one you already brought up?

Exactly. It is wishful thinking that Bitcoin can be parked in a state of suspended animation with "just the right amount" of market-share until the blockchain becomes a replacement for Fedwire, and off-chain solutions of the "just the right type" for all non-bank users mushroom around it.

Market forces mean it will either soar and succeed or crash and burn. There is no safe middle-ground.

d'aniel: Of course.


Litecoin uses scrypt in a way that can be implemented on FPGAs and ASICs in a cost-effective manner. It's harder to do than the Bitcoin PoW, but that's actually a decentralization disaster because if one player ever does make a ASIC for litecoin mining it'll be much easier for them to get 100% of the market. Bitcoin's SHA256 PoW is really easy to implement on an ASIC, so we've already seen 4 different ASICS implementing it.

BTW one of the mysterious things about people wanting to optimize Bitcoin for ease of running a mining node, at the risk of losing in the marketplace to another currency optimized for cheap payments, is that some other coins, notably LiteCoin, are actually already much better optimized for what _they_ are worried about than Bitcoin is.

The big issue here is the proof-of-work, which is far more important to the economics of mining than the block size. If you want a radically distributed network rather than one dominated by a bunch of big pools or companies, you need the economics of mining to be a good fit for a cottage industry, rather than a capital-intensive commodity business with large economies of scale.

This isn't an easy thing to design, but one of the main features would be that you'd be able to mine competitively on hardware that you already own, using spare CPU cycles, rather than needing to buy a lot of fast-depreciating specialized hardware, which you then need to install and operate highly efficiently to get a good return on your capital. This is something LiteCoin's proof-of-work is actually designed to do, and Bitcoin's isn't. It's not clear whether their current approach will work in the long-term, but in Bitcoin's case, as someone said upthread, the ship has long since sailed.

So announce the PoW algorithm switch/reimplementation of the block size limit well in advance, giving ASIC manufacturers enough time to get their chips widely distributed.  New alt-coins haven't been able to survive 51% attacks because they've been worthless, and therefore don't command a significant enough mining force.  It can be assumed that this Bitcoin fork would not be worthless. and would thus command a significant mining force.

I'm curious though, suppose all of the reasonable hypothetical difficulties with reimplementing a block size limit could be addressed, would you then switch sides in this argument?

I don't.  LTC (or some alternative) also would serve as a store of value.  So if given the choice between:

a) store of value with high transaction costs, low liquidity, and difficulty directly spending
vs
b) store of value with low transaction costs, high liquidity, and ease directly spending.

which is likely to be more popular?  Given the friction between crypto-currencies is very low (compared to crypto to fiat) I would expect a massive decline in the value of a relative to b.

TL/DR Bitcoin's value as a store of wealth is linked to its utility as a wealth transfer mechanism.  If the later if limited so if the former and the price will reflect it.

That's a very - ahem - optimistic take.

It would be interesting to hear what those various people you mention think about it, but for example Falkvinge's piece on the target value of Bitcoin is specifically premised on being used as a payment system:
http://falkvinge.net/2013/03/06/the-target-value-for-bitcoin-is-not-some-50-or-100-it-is-100000-to-1000000/

There would be an exchange rate between BTC and LTC. Because of BTC's properties combined with a commitment to a limited blocksize I would guess that it would maintain its purchasing power much like its analog equivalent, gold.

Let's assume the transaction fees go up to $20 and Wordpress, Mega, Silkroad and all BitPay's clients sod off to LiteCoin or some other coin without this restriction, as they obviously would, so you were no longer able to actually buy anything with your Bitcoins.

Do you think the value of your Bitcoins would go up or down?

I'm thinking it would go down, and more than marginally.

What the world desperately needs is a better store of value. We need a way of combating the inflationary policies of central banks worldwide.
Although gold has historically served this role, with the advent of the paper market, gold's true value has been obscured and manipulated by bankers. Bitcoin provides us with a new tamper-resistant way of saving our purchasing power over time.
Any increase in the blocksize will marginally decentralize and weaken this store of value property.
Personally, I'm all for storing my retirement savings "on the chain" and doing small, low-risk, day to day purchases "off the chain".

I'd love to hear what guys like Tuur D., Andreas A., Voorhees, and Surda think about the blocksize limit. I believe that ultimately this is more of an economic issue than a technical one.

This discussion is obviously going to go on and on, so a request that would make it a bit clearer and more meaningful:

When we talk about who will and won't be able to mine under what circumstances, can we be clear on whether we're talking about being able to do it at all or whether we're talking about being able to do it _competitively_, and thus at a profit rather than a loss? They're both potentially important in different contexts, but the latter is what drives who controls 51%+ of the network.

we run a few tor exit nodes at our colo and either (A) our colo provider or (B) the USG have substantially down-regulated our tor exit node bandwidth in the past couple month. we are capable of pushing up to ~1 Gbps of tor traffic but can only push ~20 Mbps at the moment.

i would expect to see a lot more manipulation of tor exit nodes in the future.

Speaking of fantastical, there's nothing I've seen more fantastical than your claim that turn-key engines will be created that will allow anyone to create a secure BTC-accepting site at low cost:


And retep's claim that if block sizes are large, governments would identify all miners in the world and force them to use their hashing power to attack a parallel block size limited blockchain:


The fact remains that with existing technology, my site, btctip.com, and others like it which have low volume would not be possible with $20 transaction fees. Bitcoin's low transaction fees are what make most of what Bitcoin is used for possible.

Claiming that as of yet non-existent technologies will be created that will solve all of these problems created by $20 transaction fees is fantastical.