Why the Bitcoin rules can't change (reading time ~5min)

[ElectricMucus]

Even if you don't run a full node, you can still exert you sovereignty to some extent by refusing to accept transactions that don't agree with your rules.  

If you are a major bitcoin business such as Mtgox, this alone will exert considerable pressure on the rest of the community to run full nodes that agree with your rules.

yay decentralization! 

[greyhawk]

The client in it's current state is not a viable option for most people to run as a node in the not too far future. Both bandwidth and space issues will only intensify with further acceptance of Bitcoin. Thus either the client has to change with all that entails (such as developer decided rules changes) or more and more people will have to stop running a full node, when it's no longer viable for them.

The sky is falling! Hard drive capacities are only 4TB and falling! Total internet bandwidth can only go down from here!  

Congratulations on being part of the 1%.

[hazek]

The client in it's current state is not a viable option for most people to run as a node in the not too far future. Both bandwidth and space issues will only intensify with further acceptance of Bitcoin. Thus either the client has to change with all that entails (such as developer decided rules changes) or more and more people will have to stop running a full node, when it's no longer viable for them.

The sky is falling! Hard drive capacities are only 4TB and falling! Total internet bandwidth can only go down from here!  

Congratulations on being part of the 1%.

1% out of 7billion is enough to maintain the rules of Bitcoin. A few super nodes isn't.

[ArticMine]

I'm confused about what you are arguing. You want everyone to be able to always run a full node, correct?

Wrong. I want me to be able to validate the rules I consented to are followed and right now that means running a full node, I don't care about anybody else or how that is achieved.

Which includes a 1Mb block size limit?

Given the arguments on both sides that I have heard, yes.

But must one validate it on a mobile feature phone or on a computer with a 28.8kbs modem connection to the Internet? On can one use a thin client on those devices to connect securely to a server under ones control in another location to perform the validation? Raising the 1Mb block size limit is an absolute must in order for Bitcoin to be useful for any significant proportion of the world's population.

[World]

I would like to see a special box (bitcoin hardware full node) separate from the computer and work just as node storage and distribution blockchain with large HD connected to the network much like Samsung Chromebox but much cheaper with a very small power consumption and I can turned on 24 hours a day.

[hazek]

I'm confused about what you are arguing. You want everyone to be able to always run a full node, correct?

Wrong. I want me to be able to validate the rules I consented to are followed and right now that means running a full node, I don't care about anybody else or how that is achieved.

Which includes a 1Mb block size limit?

Given the arguments on both sides that I have heard, yes.

But must one validate it on a mobile feature phone or on a computer with a 28kbs modem connection to the Internet? On can one use a thin client on those devices to connect securely to a server under ones control in another location to perform the validation? Raising the 1Mb block size limit is an absolute must in order for Bitcoin to be useful for any significant proportion of the world's population.

That's your opinion and you are free to consent to any rules you wish but I wont.

I would like to see a special box (bitcoin hardware full node) separate from the computer and work just as node storage and distribution blockchain with large HD connected to the network much like Samsung Chromebox but much cheaper with a very small power consumption and I can turned on 24 hours a day.

Interesting idea.

[markm]

The client in it's current state is not a viable option for most people to run as a node in the not too far future. Both bandwidth and space issues will only intensify with further acceptance of Bitcoin. Thus either the client has to change with all that entails (such as developer decided rules changes) or more and more people will have to stop running a full node, when it's no longer viable for them.

That seems precisely reversed / backward.

It is only if YOU allow a change to the rules YOUR full node accepts which change permits bandwidth and space attacks which the current rules absolutely prevent that further acceptance of Bitcoin could lead to bandwidth and space issues beyond the scale your normal everyday class of equipment running your existing full node can handle.

That is precisely because the current rules absolutely, totally, utterly do NOT permit ANY block to be larger than one megabyte and even with massive massive uptake of ASICs spewing out blocks faster than one per ten minutes that can only result in so many gigabytes per year that your full node needs to be ready to handle, while whatever class of machine and connection you are already successfully using to run a full node continues to increase in power.

For example if your machine has the normal mass-retail-system disk drive, check how large a normal mass retail system disk drive is today compared to how large a normal mass retail system was one year ago, two years ago, and so on. Basically the price of a "normal consumer grade machine" tends to stay about the same while what it is capable of is more and more each time you go shopping for a normal consumer grade machine.

Even if every block is the absolute maximum size, and uptake of ASICs keeps accelerating so that the maximum rate of increase of difficulty cannot keep blocks from being spewed out at an average rate that never manages to reach the full ten minutes the software tries to target, there is still every reason to suppose that whatever class of machine you are running a full node on, the machines of that class, of whatever future time we project forwards in time to, will be well able to handle it all.

The danger is that someone will lure you into agreeing to download and use a node that permits larger blocks, or that targets a faster time between blocks, or both, so that the fastest speed largest possible blocks scale outpaces the rate at which the class of equipment you are accustomed to using to run your full node grows naturally in the course of your normal day to day obsoleting of your computer hardware.

As it is, if you do not get sucked into changing the crucial limits in the code, you can probably expect to handle your node with the hardware that is handling it now for another few years, by which time normal consumer machines will be so much more powerful than what you are currently using that simply putting in whatever connection and disk drive is then the norm should do you for another few, or maybe even several, years.

It is all very safe currently. It is only if the lure of putting more and more eggs in one basket, as in people onto one blockchain, lures you into hacking at those hard limits built into the current software that you are in danger.

If instead of butchering the existing system you deploy more and more blockchains (which your node's ability to serve as primary chain for the merged-mining of is already programmed in), as many blockchains can be deployed as adoption of the blockchain based currency concept might require in order that all people who need or desire to perform transactions on a blockchain get to do so.

Very likely the more they pay for their transactions the more ancient and respected and secured - and full to the brim every block - the specific chain their comfort level of fees affords them the use of will be. The elite might prefer to pay massive fees for the cachet of utilising the primary chain, the original bitcoin; others might feel the fee level of that chain is too high thus prefer to use the first secondary chain, and so on right down to those who use the latest and leastest chain that demand has caused by any given date to have been added to the panoply of merged mined chains (that the biggest, richest miners are able to mine every one of all at once despite the large number of them...)

Basically, putting the chain you use now out of your reach can pretty much only be done by luring you into hacking at the hard limits that are there to ensure it remains forever within your reach; if at some level of mass demand for blockchain based currency some folk feel it is not fair that there is no room for them on that chain, so what, plenty more chains can be created and all of them secured together using the merged mining ability that is already in the code.

Be very very very cautious of any attempt to make you vulnerable to untenable-for-you bandwidth and space problems that people unwilling to pay the price of the adoption rate pressure they try to apply to you might try to convince you to let yourself in for by changing the limits that currently protect you from any quantity of "adoption pressure" they might try to pressure you with. There is no need to let the masses force you out of "your own" chain, the chain you are already running a full node of. If they need more transactions per day, let them deploy more merged-mined chains...

-MarkM-

[justusranvier]

The danger is that someone will lure you into agreeing to download and use a node that permits larger blocks, or that targets a faser time between blocks, or both, so that the fastest speed largest possible blocks scale outpaces the rate at which the class of equipment you are accustomed to using to run your full node grows naturally in the course of the normal day to day obsoleting of computer hardware.

The risk of blockchain size outpacing hard drive capacity is only a danger if you assume that all development currently underway to address this problem ceases: https://bitcointalk.org/index.php?topic=88208.0

[markm]

The danger is that someone will lure you into agreeing to download and use a node that permits larger blocks, or that targets a faser time between blocks, or both, so that the fastest speed largest possible blocks scale outpaces the rate at which the class of equipment you are accustomed to using to run your full node grows naturally in the course of the normal day to day obsoleting of computer hardware.

The risk of blockchain size outpacing hard drive capacity is only a danger if you assume that all development currently underway to address this problem ceases: https://bitcointalk.org/index.php?topic=88208.0

Or if the hard limits increase faster than those efforts bear fruit.

-MarkM-

[ArticMine]

I'm confused about what you are arguing. You want everyone to be able to always run a full node, correct?

Wrong. I want me to be able to validate the rules I consented to are followed and right now that means running a full node, I don't care about anybody else or how that is achieved.

Which includes a 1Mb block size limit?

Given the arguments on both sides that I have heard, yes.

But must one validate it on a mobile feature phone or on a computer with a 28kbs modem connection to the Internet? On can one use a thin client on those devices to connect securely to a server under ones control in another location to perform the validation? Raising the 1Mb block size limit is an absolute must in order for Bitcoin to be useful for any significant proportion of the world's population.

That's your opinion and you are free to consent to any rules you wish but I wont.

I would like to see a special box (bitcoin hardware full node) separate from the computer and work just as node storage and distribution blockchain with large HD connected to the network much like Samsung Chromebox but much cheaper with a very small power consumption and I can turned on 24 hours a day.

Interesting idea.

The current 1Mb block size limit places a limit on the Bitcoin network of 7 transactions per second or approximately 604,800 transactions a day. https://en.bitcoin.it/wiki/Scalability. If say 1% of the world's population say 90,000,000 people were to use Bitcoin they would have to wait on average 148 days for their transactions to clear. And this is for what purpose so that once can validate the blockchain over dialup? It is time to do some simple math here.

I like the dedicated box idea but I would prefer a client server model where one would connect to the box remotely using a dailup modem while keeping the sensitive encryption keys on the local device.

[hazek]

I'm confused about what you are arguing. You want everyone to be able to always run a full node, correct?

Wrong. I want me to be able to validate the rules I consented to are followed and right now that means running a full node, I don't care about anybody else or how that is achieved.

Which includes a 1Mb block size limit?

Given the arguments on both sides that I have heard, yes.

But must one validate it on a mobile feature phone or on a computer with a 28kbs modem connection to the Internet? On can one use a thin client on those devices to connect securely to a server under ones control in another location to perform the validation? Raising the 1Mb block size limit is an absolute must in order for Bitcoin to be useful for any significant proportion of the world's population.

That's your opinion and you are free to consent to any rules you wish but I wont.

I would like to see a special box (bitcoin hardware full node) separate from the computer and work just as node storage and distribution blockchain with large HD connected to the network much like Samsung Chromebox but much cheaper with a very small power consumption and I can turned on 24 hours a day.

Interesting idea.

The current 1Mb block size limit places a limit on the Bitcoin network of 7 transactions per second or approximately 604,800 transactions a day. https://en.bitcoin.it/wiki/Scalability. If say 1% of the world's population say 90,000,000 people were to use Bitcoin they would have to wait on average 148 days for their transactions to clear. And this is for what purpose so that once can validate the blockchain over dialup? It is time to do some simple math here.

I like the dedicated box idea but I would prefer a client server model where one would connect to the box remotely using a dailup modem while keeping the sensitive encryption keys on the local device.

As I've said, I don't care how it's done, all I care is that I am able so be a sovereign Bitcoin user, meaning that rules can't change without my giving my explicit consent via downloading a new client. How this is solved technically is beyond my concerns.

[Walter Rothbard]

If instead of butchering the existing system you deploy more and more blockchains (which your node's ability to serve as primary chain for the merged-mining of is already programmed in), as many blockchains can be deployed as adoption of the blockchain based currency concept might require in order that all people who need or desire to perform transactions on a blockchain get to do so.

Very likely the more they pay for their transactions the more ancient and respected and secured - and full to the brim every block - the specific chain their comfort level of fees affords them the use of will be. The elite might prefer to pay massive fees for the cachet of utilising the primary chain, the original bitcoin; others might feel the fee level of that chain is too high thus prefer to use the first secondary chain, and so on right down to those who use the latest and leastest chain that demand has caused by any given date to have been added to the panoply of merged mined chains (that the biggest, richest miners are able to mine every one of all at once despite the large number of them...)

And so finally it turns out that Bitcoin's gold might need a silver after all, and a nickle, and a copper, ...

[Spekulatius]

@ OP: +tip 0.01 BTC verify

[nevafuse]

But what if all miners suddenly switch to new rules? Can they do this? Is there something that stops them? Answer: YES, EVERY NON MINING FULL NODE

Hashing power is only a piece of the large puzzle that makes up a successful fork.  Ultimately, it is about the exchange rate.  Every fork will have its own exchange rate.  Whichever has the higher value, people will migrate to.

Here's an interesting scenario.  Create a fork with no more block reward.  Given we are at about the half way mark to 21M, it should be worth double the value of the current bitcoin version.  Key word should be.  It would definitely get people's attention.

[wtfvanity]

I'm confused about what you are arguing. You want everyone to be able to always run a full node, correct?

Wrong. I want me to be able to validate the rules I consented to are followed and right now that means running a full node, I don't care about anybody else or how that is achieved.

Which includes a 1Mb block size limit?

Given the arguments on both sides that I have heard, yes.

Now you are listening to arguments? Previously it was merely you signed up and want to keep it that way.

It sounds more like you don't want Bitcoin to get big? You're still very confusing.

People in a low bandwidth community, could easily pool together, and rent a dedicated server that they control and run a full node. Then the members of the community with low bandwidth could easily run a lite client especially now that we have the bloom filters working in the Satoshi client. They could have control of their FULL node, how the rules are applied, and could deal with block sizes larger than a meg.

I'm still confused on your argument. Why do you want Bitcoin not to change? You even started your post of confirming the fact that Bitcoin has already changed in the past. Shouldn't you just leave Bitcoin and make your own currency? One that can't change? If it has already changed in the past, why are you fighting so hard that it would never change again?

[paraipan]

If instead of butchering the existing system you deploy more and more blockchains (which your node's ability to serve as primary chain for the merged-mining of is already programmed in), as many blockchains can be deployed as adoption of the blockchain based currency concept might require in order that all people who need or desire to perform transactions on a blockchain get to do so.

Very likely the more they pay for their transactions the more ancient and respected and secured - and full to the brim every block - the specific chain their comfort level of fees affords them the use of will be. The elite might prefer to pay massive fees for the cachet of utilising the primary chain, the original bitcoin; others might feel the fee level of that chain is too high thus prefer to use the first secondary chain, and so on right down to those who use the latest and leastest chain that demand has caused by any given date to have been added to the panoply of merged mined chains (that the biggest, richest miners are able to mine every one of all at once despite the large number of them...)

And so finally it turns out that Bitcoin's gold might need a silver after all, and a nickle, and a copper, ...

Not at all, there could be an easier solution by reducing the new blocks propagation time.

https://bitcointalk.org/index.php?topic=145066.0

[justusranvier]

I'm still confused on your argument. Why do you want Bitcoin not to change? You even started your post of confirming the fact that Bitcoin has already changed in the past. Shouldn't you just leave Bitcoin and make your own currency? One that can't change? If it has already changed in the past, why are you fighting so hard that it would never change again?

Especially a change that has been anticipated and talked about since the very beginning:

https://bitcointalk.org/index.php?topic=1347

Raising the max block size was always part of the plan - it's what was always brought up whenever some bitcoin skeptic mention the low transaction rate. Now all of a sudden when the time is approaching to do it some people want to keep bitcoin limited to seven transactions per second.

[markm]

A lot of the appearances of the word bitcoin in this thread might be more usefully replaced by the term blockchain based currency, or possibly one might limit that to lbockchain based currencies that can be merged mined alongside bitcoin.

There is absolutely no reason to have the masses running in and out of fort knox every day, but there is every reason to have them able to fire up the primary chain to check that their gold is still there, that it still has the properties they selected it for, that no one has snuck in some subtle timebomb, a change of hard limits or something that will in the long run deny them access by increasing the technical limits that made that particular chain appealing to them in the first place.

Hopefully the primary chain coins will be so valuable anyway that putting your yearly retirement savings in there once a year will be plenty enough transactions on that chain.

Multiple chains seems to me far more scalable, with far more ability to stick to the constants assigned to the particular chain.

All this fuss has in fact led me to think that putting lower limits initially when first deploying a chain than one intends as the permanent and final limits of of that chain, is not a good idea.

At one time I had thought hmm maube to get people initially into a new chain one might get uptake faster by pretend, baby values, to bait and switch people into thinking aha here is a currency my particular level of hardware and connection capability can handle, that I can confidently commit to for life, might be reasonable so that spammers in the earliest days could not immediately push the thing to its max size of blocks and max speed of blocks and max transaction fees and so on to scare people way, making them think oh gosh sounds nice but there is no way I am the kind of person who tends all my life to have that much computer power and connectivity at my fingertips.

So now I think we should be careful whenever deploying a chain not to dress it up in baby values to sucker people into thinking it is something it is not.

If we want to make a blockchain based currency suitable for Google, Mastercard, Visa, Amazon, Paypal, and so on and so on to all use for all their daily transactions and all their customers' daily transactions while every adult and child on the planet also uses it for all their lunch purchases and pocket money and penny-ante wagers we can look at how large we would need block size to be to accomplish that, but I do not think we should pretend that that "establishment megacorp-to-megacorp network" is a grass roots person to person, small enough footprint any normal machine anywhere might be one of the backbone servers "the man" is looking to "take down" power to the people network such as bitcoin was sold to us as being.

-MarkM-

[wtfvanity]

From the mouth of satoshi himself!!

Applying this patch will make you incompatible with other Bitcoin clients.

+1 theymos.  Don't use this patch, it'll make you incompatible with the network, to your own detriment.

We can phase in a change later if we get closer to needing it.

It wasn't needed then, but we are getting closer. Hmmm..... long live satoshi.

[Walter Rothbard]

Here's an interesting scenario.  Create a fork with no more block reward.  Given we are at about the half way mark to 21M, it should be worth double the value of the current bitcoin version.  Key word should be.

Value depends greatly on adoption rate, though, so it would start with a pretty low value.