[POLL FINISHED] The block size limit controversy: a proper poll (30 days)

[Akka]

Self adjusting algorithm like difficult. We don't know what the future brings (hardware and demand wise) changing it again might be even harder, so a solution that possibly fixed this once and for all.

Difficulty adjustment is to make hashing competitive. If you allow for bandwidth adjustment you make that competitive as well. All adjustment methods are exploitable. Do you want people to compete on bandwidth?

Well not necessarily. Pools will always have to run full nodes. The decision which way to support could be made with on which pool you mine.

Also a calibration could be applied to the max Blocksize.

Like:

All 2048 Blocks, new max. Blocksize.

Target: Average Blocksize of the largest 1024 Blocks = 80% of max Blocksize. Max Blocksize increases or decreases by max 20%.

Therefore over 50% of all miners would need to produce constant Blocks at least at 80% the max Blocksize for it to increase. Ensuring Transaction space is always scarce and not enough for all transactions during peak times.

This is only a quick Idea. Please don't pick on the details.

[1713578663]

1713578663

[Timo Y]

I support misterbigg's idea of miner-voted limit changes [...]

Why miner-voted as opposed to stakeholder-voted?

Because "stakeholder-voted" is not technically or politically feasible. After all, what is included in a block is up to miners. Also, I think in the long term what's good for miners is good for the network, and good for anyone with Bitcoins.

When the hard limit starts to constrain Bitcoin's usefulness, this will tend to cause a decrease in Bitcoin's value, and thus, a decrease in miners' real profit. So, eventually, miners are going to want to increase the limit.

Technically, it's feasible. This is how it could be done: Create a non-standard "vote transaction" that sends 0 bitcoins and stores a vote in the script.  Every X blocks, the protocol mandates that the block size changes according to all vote transactions since the last change, ignoring duplicates.  Votes are weighed according to how many bitcoins are stored in the vote transactions' inputs.

If the majority of users, including large bitcoin businesses, use a hard-forked client with above rules, the miners would be forced to accept them.  Yes, what is included in a block is up to miners, but whether that block gets rejected by the users' clients is NOT up to them.  So politically, it's feasible too.

[bullioner]

If the block size can adapt variably and in a self-regulating way, similar to how difficulty works, then it will just be part of how the business of bitcoin is done. It will also have a certain elegance and symmetry, which is very appealing.

Strongly agree.  This means stating the invariant you want to keep the equilibrium around.  My favourite invariant is "long term, the amount paid by transactors to those securing the transaction log should work out out at 3% of the total monetary base per annum".  This boils down to: a recent mean of over 12 BTC tx fee per block raises the max size a little; a recent mean of under 12 BTC tx fee per block lowers the max size a little (with 1 MB remaining a floor).  This equates to a mean of 0.0029 BTC per transaction to get off the 1 MB floor.  It seems pretty scalable while still being self-regulating.

[deepceleron]

Where is the option: /24 ban the two mofos (or one and his socky buddies) that keep posting these stupid useless threads about a non issue.

[Piper67]

Where is the option: /24 ban the two mofos (or one and his socky buddies) that keep posting these stupid useless threads about a non issue.

It isn't a non-issue. It's an issue that will become somewhat important in about 6 months, a bit more important in about a year, and truly critical in about two years. But, as Joel Katz will often say, you don't turn the steering wheel when you're 5 miles away from a bend in the road. You just know the bend is there, and you make a plan for when you'll turn the wheel.

[nevafuse]

Can "No limit" be added?  Thanks.

[ShadowOfHarbringer]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

[Draino]

In time, the block size may need to change. But there's absolutely no rush.

Let the block size stay at the current size until the limit is hit; then let us run with that for a while. Let us see how easy or hard people find it to work within the limit; let us see what effect it has on transaction fees and confirmation times.

Eventually, a strong consensus might form around an appropriate change. Then, only then, is it worth working out how to implement the change. Equally, we may find that some other solution arises which negates the need for any hard change.

i agree, perhaps a clever solution will make itself known in the meantime

a hack that defies the elegance of Bitcoin would not help my confidence in the developers, to be honest.  i'm cynical by nature so maybe this is out of line, but i wonder if being funded by the btc foundation creates some weird new incentive.  "guys guys, we're worth all these bitcoins, we're doing stuff and stuff"

call me crazy, but an interface to be proud of would be the top priority, not causing drama about protocol changes for problems that might be imaginary

of course, if this idea wasn't even floated by the developers to begin with, i recant.

[dree12]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this. A possible system is one where the client limits the download speed for any particular block, but not the actual block size. If there is a tie between two blocks, the smaller one wins. Miners will not intentionally create small blocks because there is no income when the subsidy falls, and they will not intentionally create large blocks because they waste their own bandwidth and the block will never get in.

So far, nobody has refuted this system.

[mp420]

When the hard limit starts to constrain Bitcoin's usefulness

The hard limit on block size is what gives Bitcoin its most important attribute: The blockchain with the largest amount of hashing power. Increasing the block size reduces total hashing power (because it drives fees down).

Well, I think there are other attributes that give value to Bitcoin. Such as scarcity and fungibility (which are not affected by block size) and liquidity (which is).

High fees make Bitcoin less liquid, and thus, less valuable. And, as a consequence, smaller block size limit does not necessarily mean more real income to miners.

this will tend to cause a decrease in Bitcoin's value, and thus, a decrease in miners' real profit. So, eventually, miners are going to want to increase the limit.

Wrong on all counts. Its in the miners' best interests to NOT raise the limit since the limit is what drives fees. The only miners who benefit from an increase in the block size are the ones with the most bandwidth, since a limit increase will kill off the smallest competitors.

This is assuming mining is always primarily bandwith-constrained (as opposed to energy price constrained).

I think mining ops will in any case tend to either grow bigger or die. Economies of scale.

[ShadowOfHarbringer]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this.

1. After the last Bitcoin will be minted, there will be no incentive for miners to mine, except for transaction fees. When block can be of any size, then any number of transactions can make it into every block, therefore theoretically no transaction fees need to be paid.

2. How do you protect the network from dust spam & DDoS without block size limit ?

[jl2012]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this.

1. After the last Bitcoin will be minted, there will be no incentive for miners to mine, except for transaction fees. When block can be of any size, then any number of transactions can make it into every block, therefore theoretically no transaction fees need to be paid.

2. How do you protect the network from dust spam & DDoS without block size limit ?

Your theory is wrong. Miners have no obligation to include free transactions

[ShadowOfHarbringer]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this.

1. After the last Bitcoin will be minted, there will be no incentive for miners to mine, except for transaction fees. When block can be of any size, then any number of transactions can make it into every block, therefore theoretically no transaction fees need to be paid.

2. How do you protect the network from dust spam & DDoS without block size limit ?

Your theory is wrong. Miners have no obligation to include free transactions

Correct, however if any number of transactions can and should theoretically make it into the block, won't the miners conspire and create their own limits anyway in order to save their profits ?

So won't theoretical "no limit" result in practical "miners decide what the limit is" ?

[Piper67]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this.

1. After the last Bitcoin will be minted, there will be no incentive for miners to mine, except for transaction fees. When block can be of any size, then any number of transactions can make it into every block, therefore theoretically no transaction fees need to be paid.

2. How do you protect the network from dust spam & DDoS without block size limit ?

Your theory is wrong. Miners have no obligation to include free transactions

Correct, however if any number of transactions can and should theoretically make it into the block, won't the miners conspire and create their own limits anyway in order to save their profits ?

So won't theoretical "no limit" result in practical "miners decide what the limit is" ?

There is also no indication that the block size limit will be left to the miners. The prevailing thought is that it will be regulated by an algorithm, much the same as difficulty. Miners would also love to have a lower difficulty, but that's not their decision. It will be something similar with the block size. Large enough to accommodate all transactions as required, small enough to still prioritize those with fees so the miners are happy.

Incidentally, any miner who isn't happy is welcome to switch off their rigs. Then the difficulty will drop and whatever fees there are available (and there will always be some) will go to the remaining miners.

[SimonL]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this. A possible system is one where the client limits the download speed for any particular block, but not the actual block size. If there is a tie between two blocks, the smaller one wins. Miners will not intentionally create small blocks because there is no income when the subsidy falls, and they will not intentionally create large blocks because they waste their own bandwidth and the block will never get in.

So far, nobody has refuted this system.

I can offer a possible avenue of attack on the system if no limit is imposed on blocks. If no max block size is imposed it is possible to force on the network a massive block that takes longer to propagate than the 10 minute window between block broadcasts. If this occurs enough times the difficulty calculated based on how many blocks were broadcast in the last difficulty period will result in a lower difficulty. Thus massive blocks that the network can't handle will attack the metric used to determine difficulty. This will result in a shorter window so that miners are incentivised to broadcast massive blocks. This would inevitable massively bloat the blockchain in a short space of time. This could also push out smaller miners and make the network more centralised.

Though there is no "proof", I've yet to hear anyone say that this kind of attack is not possible. An algorithmic and gradual raising of the max block size based around what the network can handle is far more sane and reasonable. It should not be based on miner incentives, economic principles or any such justification. It should only be raised if the network of nodes as a majority, are able to comfortably process larger blocks.

[dree12]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this.

1. After the last Bitcoin will be minted, there will be no incentive for miners to mine, except for transaction fees. When block can be of any size, then any number of transactions can make it into every block, therefore theoretically no transaction fees need to be paid.

2. How do you protect the network from dust spam & DDoS without block size limit ?

Neither of these is a problem if large blocks are orphaned by small ones. It makes mining more of a skill, as they must make the largest block that is not too large such that other miners will receive it fast enough.

[ShadowOfHarbringer]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this.

1. After the last Bitcoin will be minted, there will be no incentive for miners to mine, except for transaction fees. When block can be of any size, then any number of transactions can make it into every block, therefore theoretically no transaction fees need to be paid.

2. How do you protect the network from dust spam & DDoS without block size limit ?

Neither of these is a problem if large blocks are orphaned by small ones. It makes mining more of a skill, as they must make the largest block that is not too large such that other miners will receive it fast enough.

You are giving total control to the miners.
And sooner or later they will conspire to decide the block size between themselves.

[nevafuse]

Can "No limit" be added?  Thanks.

No it cannot.

Without the limit, Bitcoin network can't function.

There has been no proof of this.

1. After the last Bitcoin will be minted, there will be no incentive for miners to mine, except for transaction fees. When block can be of any size, then any number of transactions can make it into every block, therefore theoretically no transaction fees need to be paid.

2. How do you protect the network from dust spam & DDoS without block size limit ?

Neither of these is a problem if large blocks are orphaned by small ones. It makes mining more of a skill, as they must make the largest block that is not too large such that other miners will receive it fast enough.

I couldn't agree more with dree12.  Placing a limit is unnecessary.  Miners have no incentives to push unnecessary large blocks because they run the risk of them not propagating fast enough & being orphaned.  Even if the orphan risk is low enough to include every transaction & the block reward is gone, the cost of mining is still an incentive to only include transactions with fees.  Of course there will be COMPETING miners out there that include transactions without fees, but they probably won't have the hash rate of the ones that do require fees.  Therefore you will have choice whether to include a fee to get it confirmed faster or not which will keep the transaction fee from being monopolized.  The same could be said for the divisibility limit IMO.

[niko]

I voted "I don't know," but changed my mind after reading - can you edit the poll so we can change our vote (not sure if forum software lets you do that)?

[Timo Y]

Please stop asking for a hard fork, people doing this are destroing Bitcoin image and credibility. No one of you are entitled to make a poll in the name of the comunity, a few people on these forums don't represent the whole comunity.

¿Do you realice how many people have invested their live savings here? We are speaking at the moment of 318,709,754$, you are all playing with fire and people lives.

Do you expect, all of us selling our Bitcoins and switch to Bitcoin 2 at no loss? People will lose money, a lot o money if a panic sell occurs. Bitcoin will lose credibility for YEARS and investors will flee for ever.

If you want a hard fork, no one is stoping you taking the source code and doing an alt currency but don't name it Bitcoin.

What gives Bitcoin it's value are the rules. Protocol rules must be set in stone, unalterable for ever.

There are good alternatives around there if anyone whant instant transactions like VISA/Mastercard and Ripple.

Sorry to be rought, no offence, but these are my feelings.

Regards.

We should be very careful about the hard fork, but we shouldn't be paralyzed by fear either.

One could equally argue that NOT upgrading is playing with peoples' lives.  Nobody knows what's going to happen.

I am still undecided on this issue, but I doubt that a conservative hard fork will trigger panic selling. Eg. increasing the limit from 1 MiB to 10 MiB, as sugessted by Peter Wuille, will bring benefits of a higher transaction rate while still allowing normal users to run a full node.  Investors understand this.  

The switch doesn't need to be catastrophic. It can be announced 6 months in advance and it can be done in a way that it only happens once 99% of users have upgraded to the new client already.

As for people putting their life savings into bitcoin, well, they have been repeatedly warned against this by Gavin and other community members, and you can't really blame the developers if people invest money that they can't afford to lose.