[hudon]

A fork like Bitcoin Cash will be a boon for the blockchain community.

There is a pretty big subset of the community that believe Bitcoin is a panacea that will replace currencies, payment networks and so on, regardless of how technically inferior the blockchain is to other more mature distributed databases and networks. They believe one day, the blockchain will be just as efficient or even more so. This subset is vocal about the urgency of increasing Bitcoin's block size limit so that we can increase transaction throughput as much as possible and as soon as possible.

On the other hand, most developers who have worked on Bitcoin proper (either protocol development or Core node development) believe that Bitcoin is more about financial sovereignty and censorship resistance, not as an in-place replacement of PayPal or VISA. This group wants to find as many ways to scale the blockchain without increasing the block size limit because increasing the size of blocks puts at risk users' ability to validate the chain. This is because larger blocks means more resources required to transmit, validate and store blocks and if you cannot validate blocks, then you are trusting transaction validators (miners) just like you trust PayPal. Risking the ability to validate the chain is risking the financial sovereignty or censorship resistance they value so much.

Regardless of how well some claim SegWit2x is doing, truth is once SegWit is activated, we still have to face the 2x hard fork which many people in the second camp will simply refuse to support.

Having said all this, Bitcoin Cash represents an earnest understanding that there are two camps in Bitcoin and because of the differing economic visions, they will have different technical visions, so why not have two chains and evolve them independently? Rather than playing tug-of-war and both parties being dissatisfied?

[jgord]

Its a commonly voiced, but false dichotomy - increasing blocksize 4x or even 8x will have negligible impact on the security or decentrality of Bitcoin.

It will just allow more transactions through, which will allow fees to fall to levels where normal people can use it for normal transactions.

The use of bitcoin as a currency is being throttled by the artificially small blocksize.

We can have more than 3 transactions per second, while still maintaining the fundamental properties that are Bitcoin.

Hopefully SegWit2X will be the sane compromise everyone hopes for - ie. that we see the bottleneck alleviated with segwit and 1MB blocksize contributing to a near 4x improvement in transaction throughput, and lower usage fees.

[Frogolocalypse]

> increasing blocksize 4x or even 8x will have negligible impact on the security or decentrality of Bitcoin.

Incorrect. It will make running a full node for almost all consumers in Australia impossible at any price.

> hopefully SegWit2X will be the sane compromise

Segwit is the compromise and we are finally going to get it. If you want to install the 2x hard-fork to china-coin, you are welcome to. Everyone else will just continue using bitcoin. Now with segwit.

[CyberDildonics]

You've made a lot of assertions like this but you aren't backing them up with any numbers.

8MB every 10 minutes is 13KB/s. Do Australians not look at youtube either, because when I was there well over a decade ago it worked pretty well. Even a full node syncing multiple people -while the blocks are full at 8MB- is no problem. Also there are many orders of magnitude left before running a full node exceeds the capacity of a $15/month VPS and many orders of magnitude before a person at home could no longer sync with a full node off a basic internet connection. That is also ignoring the fact that most people don't need or want to sync with the full chain, but to be clear, hundreds of millions can and would still be able to.

I see arguments like this from time to time, usually from the same small group of people, but there is nothing but crickets when I show them basic numbers and ask for how they came to this conclusion.

[Frogolocalypse]

Here's the source of the justification of my opinion :

https://iancoleman.github.io/blocksize/#_

Where's yours?

The issue is the upload bandwidth requirement, which demonstrates that your understanding of the constraint, with your comparison to YouTube, is completely incorrect. And centralizing of all of these nodes onto service provider infrastructure is the worst possible example of decentralization, which is what we're talking about in the first place.

What I've found in the past is the people who hold your views are only cursorily aware of the constraints of decentralization, and use unrelated justifications in order to hand wave away problems. Thankfully, the people who actually secure bitcoin, through their use of nodes, spend more time to understand how bitcoin works, and have repeatedly rejected 'solutions' that will reduce the security of the network.

But you can now prove me wrong! Instead of using bitcoin, use bitcoin cash. If what you say is true, it will be valued significantly higher than bitcoin. If it stops you trying to break bitcoin, even better.

[CyberDildonics]

> with your comparison to YouTube, is completely incorrect

Not for syncing. For running a full node, a VPS costing $12 - $15 USD per month has orders of magnitude more bandwidth. I already have you the math of full 8MB blocks, there is no way you can say that these aren't trivial bandwidth numbers.

> What I've found in the past is the people who hold your views are only cursorily aware of the constraints of decentralization, and use unrelated justifications in order to hand wave away problems.

That's big talk when:

1. You are talking about bandwidth and the numbers are simple to calculate and come out to trivial amounts.

2. Bitcoin is literally working right now with 0 hiccups except for full blocks and high fees. The people working on it right now didn't do that, they kicked out the people who did.

3. You say hand waving while not backing up a single thing you've said or confronting what I've said. 8MB block when 100% full takes 13KB/s to sync. Is that wrong? Twitch streams are 250KB/s. Most people don't even want to or need to sync.

> But you can now prove me wrong! Instead of using bitcoin, use bitcoin cash.

This is about how ridiculous the 1MB limitation is, you can try to divert from that, but lets see you confront it.

[Frogolocalypse]

> For running a full node, a VPS costing $12 - $15 USD

So your solution to centralization pressures is to use a centralized service for decentralization?

> You are talking about bandwidth and the numbers are simple to calculate and come out to trivial amounts.

I provided more evidence on the bandwidth requirements than your incomplete understanding of how bitcoin works multiplied by your flawed opinion.

> 2. Bitcoin is literally working right now with 0 hiccups except for full blocks and high fees.

No-one is forcing you to use it. There are a multitude of other options if you feel that bitcoin isn't meeting your use-case.

> You say hand waving while not backing up a single thing you've said

I provided evidence that backed up my opinion. You should try it.

> This is about how ridiculous the 1MB limitation is, you can try to divert from that, but lets see you confront it.

Nah. You can use another service if you feel that the 10's of thousands of nodes somehow don't have a better understanding of the constraints of running a node than you. No-one is holding a gun to your head. Ever heard of Bitcoin Cash? Perhaps that might better suit your needs.

[h1d]

The reason is not because of the vision. Also if anyone thinks about it as a replacement of paypal/visa, they know scaling via increased block size is not going to cut it. 7tx/s doen't do much good if scaled linearly.

It was a flaw in the Bitcoin specification that there was a conflict of interest between miners and users that when more blocks are filled, the more fee the miners get, thus miners have no incentive to make the blocks less congested.

Not sure how doubling the block size suddenly disallows users from validating the blocks, besides, as the entire data is already around 160GB or so, none of the mobile wallets contain the full data and are delegating much of the credibility of the chain to others' full nodes.

Bitcoin Cash is mostly built around the Chinese mining community who have been rejecting to activate segwit for a while since their asicboost that helps their mining speed by 20-30% was about to become unusable, though they claim it's not used.

And with 8MB for Bitcoin Cash it would make the chain data so large after several years, if block spaces get filled you will be able to barely host a node on a desktop with TBs of storage not to mention hosting on clouds would be pretty expensive leading to centralization.

[taysic]

"7tx/s doen't do much good if scaled linearly"

Why not? Bitcoin Cash may be more vocal in the Chinese miner community but there are plenty of Americans who believe in it too. The storage issue both chains will have to deal with if they are used in any big capacity

[PeterisP]

Because solutions that are practical for getting 2x or 10x the capacity aren't sufficient, since getting used in big capacity requires a vision that can give a thousandfold increase, and increasing the block size can't ever provide that in a practical manner.

If you want to scale a huge cliff, going to fetch a 20 feet ladder will not help you.

[jgord]

I think that its quite possible that we can see scaling of 1000x on the linear blockchain, without fundamentally changing the economic/security properties.

Increasing blocksize is just the first thing to do of many engineering optimizations. If you want to climb Everest, you first need to reach base camp.

Even if LN works great and scales beautifully, we will still need that 100x increase in the blockchain that it settles on. LN itself will create transaction demand - people aren't paying $2.50 in fees now to buy a $3 coffee with bitcoin.

[Frogolocalypse]

> I think that its quite possible that we can see scaling of 1000x on the linear blockchain

In order for your fantasy to become a reality the tens of thousands of current users of the core ref nodes would have to uninstall their node clients and permanently remove their ability to ever validate their transactions on the blockchain. That is never going to happen.

[CyberDildonics]

8MB blocks means 13KB/s to sync with the chain -when blocks are full-. The numbers for this argument don't work out at all.

[DougN7]

Thank you for your explanation. This is the best description of WHY there is such a battle going on that lay people like me can grok :)

[vgb2k11]

But its still the wrong explanation. Think about it:

(i) Miners, not average users, hold the voting power.

(ii) When the 1 MB block-size is congested (like it is now), miners make significantly more money on transaction fee's.

(iii) Increasing the block-size increases supply and reduces demand for priority processing in the block, hence, reduces transaction fee's.

Which explanation truly matches your view of reality here. People (miners) are motivated by "A shared vision" or a dollar in the bank account? Look at it closer (quote from parent comment):

>

> "larger blocks means more resources required to transmit, validate and store blocks and if you cannot validate blocks, then you are trusting transaction validators (miners)"

>

Can you see the problem with the explanation now? Its written from the view of an end user (of bitcoin), not a miner. But its miners who vote on the fork, hence, the above quoted text is mostly irrelevant to understanding the "WHY" of the fork battle.

[jgord]

Its a transient equilibrium during growth, not a zero-sum game. We could see lower fees and miners earning more :

Hypothetically, if blocksize went up tomorrow by 8x, then potentially fees might drop by 4x and volume of transactions go up by 8x .. then miners will be making 2x in fees per block, for maybe 2% extra cost.

That has other flow on effects - if fees are lower bitcoin can accommodate people using it as a day to day currency, then its utility and user base goes up, then the valuation in USD goes up, the value of the bitcoins miners earn/save goes up, etc.

I think there are miners who understand that there is a sweet spot of fees being low enough to facilitate higher daily usage and growth. Also, the majority of their current income is not in fees, but rather in 'coinbase', the reward for mining the block [ which is how new bitcoin money supply is injected into the system ]. They are highly vested in the valuation of USD/BTC, so they will do well if the user base of bitcoin grows.

[Frogolocalypse]

None of the nodes users, which define and police consensus in bitcoin, is interested in reducing the security of the network to allow miners more control. This is the fourth failed coup attempt of the bitcoin network, and the only thing that it has demonstrated, is the resilience of the network against large corporate centralization attempts. Segwit might very well be the last architectural change of the protocol.

[hn_throwaway_99]

But miners aren't stupid, and they know that if they are perceived as being a huge bottleneck and impediment to the growth and viability of Bitcoin, everyone else will just follow a different chain.

Currencies only have value if people are willing to trade you actual goods/services for it, and if the larger community decides the value lies with another fork, miners will lose all their power.

[kalleboo]

> (ii) When the 1 MB block-size is congested (like it is now), miners make significantly more money on transaction fee's.

One thing I've not understood is, if the miners want more money from fees, why don't they just say "we're not accepting into blocks any transactions with fees lower than X"?

If just one large miner with 10% of the hashrate does this, that instantly puts any transaction with a lower fee than that at a 10% chance of being delayed, and if the fee is still reasonable people will pay it just in case.

And it's not like there's a market of miners. A user can't refuse to deal with a miner or choose who gets to confirm their transaction.

[manquer]

This is how priority transactions work.Increasing block size would allow more transactions per second. Less people would then pay for the premium of fast transactions as the normal transaction is fast enough

[ithought]

Can you elaborate further?

When Mike Hearn quit bitcoin (Jan 2016), he wrote the Chinese miners were worried about bitcoin getting too popular because of their limited access to the Internet. And said they were actively trying to supress its popularity. But obviously that isn't true now?

https://blog.plan99.net/the-resolution-of-the-bitcoin-experi...

[vgb2k11]

Check out replies by zkSNARK and stale2002 (just down-page at time of writing). Screen cap here for reference: http://imgur.com/a/xBJUW

[empath75]

Once there are multiple viable forks of Bitcoin proper, the idea that bitcoins are rare assets worth investing in will collapse.

[bdcravens]

Aren't most alts a fork of a fork? How many of them have truly original codebases? (obviously operative word in your comment was "viable", but at a minimum, Litecoin is one viable fork)

[placeybordeaux]

Those aren't forks, they are new chains.

[curyous]

FYI, a network with 1000x more on-chain transactions, that everyone uses, is much more difficult to censor, than one that requires bank-like organisation with IOUs to the main chain.

[hudon]

Two problems with your argument:

1- You imply that only one camp wants to scale. Actually, both camps want to scale. The only difference is that the "big blocks" camp do not think a user's ability to validate the blockchain is worth keeping.

2- The size of the network is irrelevant to the government's ability to censor transactions. The only thing that stops them from having power over the network is decentralization. If you as a user have no say whatsoever in what goes into blocks, you are forfeiting your power and handing it over to the miners. Then, the miners become transaction validators that can be compelled to censor certain kinds of transactions by their government. As a user, you will have no means to stop this because you can't even validate the chain they are producing (because it costs you too much to transmit, store and validate it in computing resources).

I'm not saying the argument for the "not-big-blocks" camp is to never increase the block size, but it is to try everything they can to keep the power in the user's hands while growing the network. The minute you give up on decentralization for the purpose of scale, Bitcoin loses its value proposition as it looks less like digital cash and more like paypal with central transaction validators and a central database.

[zkSNARK]

You are either lying or misinformed. The "big block camp" absolutely does think a user's ability to validate the blockchain is worth keeping. This is why proposals that scale blocks according to Moore's law have been created. This is also why proposals exist which allow flexible block sizes based on need.

You are also ignoring the fact that for most of bitcoin's existence, block sizes were soft limited by miners as appropriate. With soft limited or flexible block sizes, just because big blocks are used sometimes when transaction volume is high does not mean that we will see gigabyte block sizes every 10 minutes for all of time.

The root of the debate is that small blockers have conveniently chosen to ignore the reality that markets and market participants can regulate themselves.

In the small blocker's mind, keeping the power in the user's hands means making bitcoin so expensive that the only people who can use it directly are banks and governments that offer sidechain connections to other networks. They blather on about decentralization while the reality is they are the ones creating centralization.

It is a questionably convenient choice for small blockers because it forces centralization to things that aren't actually bitcoin which is the most logical desire of governments or other bodies who wish to subvert bitcoin.

[stale2002]

"You imply that only one camp wants to scale."

No. Go read what official Core developers like Luke-jr and Blue Hair Matt are saying.

They think that the blocksize might ALREADY be too large, and actually want to DECREASE the blocksize.

They also explicitly do not want to scale, because that would mean that transaction prices go down. They literally want block fees to be high so as to create a "fee market".

[Frogolocalypse]

> They think that the blocksize might ALREADY be too large

That's because the block-size is only a small part of scaling.

[olegkikin]

> the miners become transaction validators that can be compelled to censor certain kinds of transactions by their government

That's only true if

1) >50% of the miners are in said county (currently China)

2) >50% of the miners agree to this censorship

3) They actually know the meaning of transactions in order to filter them. Most transactions are anonymous, even the big ones. We know identities of a tiny percentage of the addresses, and most of them are not hiding. For instance, we can't figure out where Mt Gox money went - I've seen some attempts of tracing them, but it went nowhere.

[vertex-four]

I am unaware of any current proposal which requires putting in place a bank-like organisation. If you're talking about Lightning Network: routing things across a network without requiring a central controller is a very well-studied field.

[mhluongo]

Lightning's network topology is still up in the air. Saying "routing is well studied" is well and good, but from what I've read doing it with payment channels without devolving to hub and spoke isn't trivial.

[vertex-four]

I would argue that that is mostly because cryptocurrency developers are mostly incapable of reading CS papers, but that's neither here nor there.

[mhluongo]

I'm sure your input would be welcome :)

There have been a couple posts recently debating different network configurations [0]. It seems like the difference between Lightning's routing and other networks is that channels need re-balancing, but I don't know the literature well enough to know if there's an existing, well-studied analog.

[0] https://hackernoon.com/simulating-a-decentralized-lightning-...

[Frogolocalypse]

If it leads to a small group of people choosing an alt based upon the bitcoin protocol such that they stop trying to kill decentralization in bitcoin, good luck to them.

[stale2002]

Kind of.

You are correct that segwit2X is a hard fork, and therefore the other side could potentially stop it from happening.

But the thing is that it is already established that it is possible to do a blocksize increase via a soft fork. Segwit itself is literally a blocksize soft fork.

So if the hard fork fails, the big blockers could instead just blocksize soft fork instead of hard fork.

The war won't be over if segwit2X fails. That is the beginning, not the end, of the real Bitcoin civil war that will happen.

[Frogolocalypse]

> therefore the other side could potentially stop it from happening

No one can stop you from forking bitcoin. It's open source. Anyone can do it. What you can't do is force all of the existing users to stop using bitcoin, and instead use your china-coin fork.

I wish you woukd fork. But you won't. Because as soon as you do, you'll find that none of the existing users of bitcoin will follow your fork. So a week or so later, you'll blame core.

[Jabanga]

>There is a pretty big subset of the community that believe Bitcoin is a panacea that will replace currencies, payment networks and so on, regardless of how technically inferior the blockchain is to other more mature distributed databases and networks

But they don't believe it's technically inferior to more mature distributed databases and networks..

You're transposing your own opinion onto theirs.

[hudon]

It's not opinion, it's fact. VISA can easily process over 40,000 TPS today. A single MySQL instance can process 10,000 TPS with very little electricity consumption compared to the Bitcoin or Ethereum networks, which currently use a ridiculous amount of electricity to process somewhere between 2 TPS to 20 TPS.

None of these numbers are hard to find and they are widely accepted as fact, not opinion.

> They believe one day, the blockchain will be just as efficient or even more so

^This is indeed their opinion and not a fact. Do you disagree?

[CyberDildonics]

This is a fundamental misunderstanding that assumes a linear relationship between transaction capacity and electricity usage even though they are completely decoupled.

Here is a comment I made earlier addressing the comparison to Visa with actual resources it takes:

-They don't require everyone to hold the whole blockchain, you can have a thin wallet where the blockchain is held elsewhere.

-An ecosystem of multiple different currencies does create sharding.

-Merkle trees are a hierarchy of hashes so that someone can hold only the parts of the chain they want to look at and know that they are on the right chain by syncing large parts of it with their hash as a whole.

-At Visa level transaction rates of 300 Million transactions per day and the minimum size of about 4 transactions per KB in bitcoin, that means that at current 8TB hard drive prices it costs about $900 USD per year to store all the transactions. This would also require a steady connection of at least 870 KB/s to sync with the chain.

In short, there are plenty of solutions with even the extreme examples of involvement being very obtainable by the average person in a developed country.

[Jabanga]

>which currently use a ridiculous amount of electricity to process somewhere between 2 TPS to 20 TPS.

The electricity consumption is a result of the mining subsidy, and artificially high fees from block space constraints, not fundamental costs for transaction processing.

If you break down the necessary costs (storage, bandwidth, CPU cycles for verification), it's actually very low, at about $0.001 (a tenth of a cent) a transaction:

https://bitcointalk.org/index.php?topic=3332.0

The mining subsidy is a fixed cost that will decrease in relation to the total number of transactions as transaction volumes increase.