Removing OP_return limits seems like a huge mistake

[Wind_FURY]

This BIP-110 supporter actually said what we have been saying about dick pics/fart sounds and the fee market. It's laughable how they debate against the Core Developers without understanding themselves.



He's admitting that we're right.

That's fine, but jpegs are going to get priced out anyway. They're not a threat to the network
https://x.com/mattkratter/status/1732812053122064795

This however might not the case forever. We cannot just hope ordinal craziness will simply go away.

  ¯\_(ツ)_/¯

Spam prevention is the one of the main functions of the fee market.

Sure, but this is quite a simplified view on it and is contrary to what Bitcoin is supposed to be. We can't expect it to be able to act as a digital peer to peer cash and simultaneously only rely on fees to price out undesirable behaviors.

I have already posted/debated about my viewpoint on that many times in BitcoinTalk. I am not going to keep repeating myself like a broken record. BUT if your viewpoint is that Bitcoin is, "Peer To Peer Cash" because "It's in the white-paper", then OK. You do your own sort of thinking.

Plus if I'm going to nit-pick because I believe that the "because white-paper" debate is a nit-pick, then I'll tell you, technically Bitcoin isn't a peer to peer network. It's a broadcast network.

[gmaxwell]

Sure, but this is quite a simplified view on it and is contrary to what Bitcoin is supposed to be. We can't expect it to be able to act as a digital peer to peer cash and simultaneously only rely on fees to price out undesirable behaviors.

Some people have some internal definition of what "digital cash" is and can't tell the difference between their own meaning of a term and reality.  Imagine someone howling that bitcoin has to be made of paper because the white paper uses the word cash and cash is paper-- except that complaint is too obvious, and those people didn't make it that far because they first got hung up on the white paper itself not being paper.

Bitcoin is what it is, the analogies used in the white paper are to help people generally understand what it's talking about.  They're not an imposition of a definition and are absolutely not a imposition of a very specific definition that exists only in your head.  Bitcoin has always had fees as part of the design, and fees have always been the proposed mechanism to fund security absent inflation.  Cash by any conventional usage doesn't have fees, one of its most important differences compared to other forms of payment.

Oh no!  Satoshi says bitcoin is cash, but cash doesn't have fees, so Bitcoin can't be cash, so if Satoshi is right then satoshi is wrong which makes him right which makes him wrong ... Can god make a rock so heavy that even god can't lift it??!?!    What happens when an unstoppable force meets an unmovable object?



Dog Pig Dog Pig Dog Pig loaf of bread S̷̖̒͊̔̎͐̈́̆͌͌̏͊͝͝Y̴̢̜̪̪͍̫͇͈̦̭̣̘̣̤͍̖̆̉̑̓͒̈́͌͐̿̈́̚͜͠S̷̢̛͈̰̙̳̟͓̟̲̰̟̥̎̒̈́́̀͐̌̊̍͗̿̾͆͝ͅT̷̫͛̇̏̀̏͘É̸̡̡̺̦̫̘̯͍̆͗̀͒̊͆́̉͌͜M̵̡̢̝̲̯͍̖͕̰̳̮̖͍̬͐͑̐̂͑͗̈́̅͐͂͑͘͝ ̷̧̢͓͎̤̘̻̩͕̞͉̰̼̻̇̇͊̈́͝E̷̳̜̰̠͎͔͂̽̑͐́͐͑̽̍̾̕͜Ŕ̶̡͕̝͔̠̺͙̻̲̖̈́͆̒̊̈̽̈́R̸̤̋͒̑̽̓̈́͑̚Ǫ̷̮̝͔͇̻̬̥͇͍͓͂̀Ŗ̷̨̧̹̹̠̪̩̫̰͓͙̹̀́ͅ.



Your head will only stop exploding when you realize overreading meaning into other peoples writing that was not ever intended and would never have made sense.  Bitcoin is peer to peer electronic cash.  Does that mean it needs to behave exactly the same as some other cash system (which is not peer to peer, or not electronic, or just not Bitcoin) or even behave particularly similarly? Not at all.  Terminology games are for idiots.  Bitcoin just needs to be Bitcoin, it needs to be useful-- and it is.  On day one anyone could have said "hey wait a minute, you're saying this has fees as part of the design, how can it be cash?"  But no one was saying that then and so it makes even less sense to raise it now. The feedback of online discussion must have created idiocy today more profound than existed in 2008.  Satoshi had to refer to existing concepts to teach people about Bitcoin was but in doing so he wasn't making any deep, subtle, or overly prescriptive remarks-- his audience was people who had no idea of what Bitcoin was or that it was even possible.  They just had to get the general idea.  Satoshi did not write that Bitcoin is a "functionally identical online replica of the argentenian peso", he simply said it's a peer to peer electronic cash.  He needed to say that or otherwise some readers might think he was writing about tacos or something.

Bitcoin's nature creates some limitations-- e.g. such as floods of high fee transactions can delay your transactions if you're not willing to outbid them.  I think it's still useful in spite of those limitations.  Somehow before Bitcoin we managed to survive since time immemorial with person to person money that couldn't even be sent online at all.  Compared to that I think that irreversibility when sending online having potential delays isn't a particularly serious limit.  In particular because the mechanisms that people have used to be assured that payment would be completed can still also be used when Bitcoin itself isn't settling quickly.

[stwenhao]

On day one anyone could have said "hey wait a minute, you're saying this has fees as part of the design, how can it be cash?"

On day one, many transactions were free, and it was standard to pay zero satoshi, to send some coins. And you can probably remember times, when 0.01 BTC fees appeared, and how some people reacted to that, how some tried to avoid it, build an alternative client without it, and so on.

I don't think the threshold should ever be 0.  We should always allow at least some free transactions.

I wonder, if it will ever be implemented in that way, as the current default implementation provides no room for free transactions (even if they can fit into a block). Because now it seems there will always be at least one satoshi per transaction (when fees will decrease from 1 sat/vB or 0.1 sat/vB into 0.001 sat/vB or similar). And even if technically there will be some transactions with zero fees, then still: they will be just a part of a bigger package, where some transaction with zero fee will be pushed to the block, by some CPFP transaction with non-zero fees, and they will form a package.

[Wind_FURY]

On day one anyone could have said "hey wait a minute, you're saying this has fees as part of the design, how can it be cash?"

On day one, many transactions were free, and it was standard to pay zero satoshi, to send some coins.

And you can probably remember times, when 0.01 BTC fees appeared, and how some people reacted to that, how some tried to avoid it, build an alternative client without it, and so on.

I don't think the threshold should ever be 0.  We should always allow at least some free transactions.

I wonder, if it will ever be implemented in that way, as the current default implementation provides no room for free transactions (even if they can fit into a block). Because now it seems there will always be at least one satoshi per transaction (when fees will decrease from 1 sat/vB or 0.1 sat/vB into 0.001 sat/vB or similar). And even if technically there will be some transactions with zero fees, then still: they will be just a part of a bigger package, where some transaction with zero fee will be pushed to the block, by some CPFP transaction with non-zero fees, and they will form a package.

OK, then in that "Zero-Fee Decentralized Bitcoin Network", how do you effectively create a spam-protection mechanism? Filter and hope?

In my personal opinion, if Satoshi came back he'll probably change some of his/her viewpoints about Bitcoin, especially after learning how the network is working in practice.

This is the first time the world has ever had something like Bitcoin. Satoshi probably didn't actually have all of the answers. The smartest developers in the Bitcoin community are probably still learning as we all are also still learning.

[stwenhao]

how do you effectively create a spam-protection mechanism? Filter and hope?

Proof of Work is a quite good filter. Today, you can pay only with fees. But it is possible to allow lower fees, if you provide more Proof of Work instead. And I guess there is some equilibrium somewhere, where users could accept paying N satoshis fee, with M difficulty. And then, you can decide, if you want to provide more coins, or more hashrate, to get your transaction confirmed.

Also, as long as OP_CHECKSIG, or its equivalent, is used almost everywhere, enforcing it is quite easy: the smaller your DER signature is, the more Proof of Work is needed to make it. Which also makes the whole transaction smaller, and can save you some satoshis on fees.

So, it is even possible now, to some extent. The main problem is that doesn't contribute to the whole network's chainwork. But well, if someone is against Merged Mining, then it can be used as it is: if you require a signature smaller than N bytes, then everyone, who will want to double-spend it, will need to re-mine it, by solving the same Proof of Work challenge from scratch.

Also, we can look at some numbers, to estimate, how many hashes are needed, if there is a lot of competition:

Code:

hash=00000000000000000000eaa0b9031bf81c368df29bb62a27794f7d7a38308f36
difficulty=1701f303
target=00000000000000000001f3030000000000000000000000000000000000000000
satoshis=314736509=0x12c27f7d
target*satoshis=000000000000249156c825770000000000000000000000000000000000000000

And then, some users will never meet 0x1701f303 difficulty. But more of them could meet 0x1a249156 difficulty instead, which could be a price for lowering your fee by one satoshi, if there will be many people, trying to do the same thing.

Which means, that if fees will be high, then it may be worth grinding your signatures, and making some satoshis out of it, by sending your transaction with the same fee rate, which could take less bytes.

[Satofan44]

Sure, but this is quite a simplified view on it and is contrary to what Bitcoin is supposed to be. We can't expect it to be able to act as a digital peer to peer cash and simultaneously only rely on fees to price out undesirable behaviors.

Some people have some internal definition of what "digital cash" is and can't tell the difference between their own meaning of a term and reality.  Imagine someone howling that bitcoin has to be made of paper because the white paper uses the word cash and cash is paper-- except that complaint is too obvious, and those people didn't make it that far because they first got hung up on the white paper itself not being paper.

I don't have a strict definition of what that is, I was using terminology that is familiar to most readers. You can call it many other things that it it usually called or bring up many others functions that it is supposed to fulfill, and a similar point can be made since it impacts a lot of these things.

Bitcoin's nature creates some limitations-- e.g. such as floods of high fee transactions can delay your transactions if you're not willing to outbid them.  I think it's still useful in spite of those limitations.  Somehow before Bitcoin we managed to survive since time immemorial with person to person money that couldn't even be sent online at all.  Compared to that I think that irreversibility when sending online having potential delays isn't a particularly serious limit.  In particular because the mechanisms that people have used to be assured that payment would be completed can still also be used when Bitcoin itself isn't settling quickly.

The conclusion is, nothing at all can be done or should be done to improve this aspect (I am not arguing for any particular idea)? That's not a very forward way of thinking on topic of usability, is it? It is a bit strange to see this response, perhaps we don't understand each other's intent clearly -- when LN itself is an improvement on the issue that I am talking about (and many improvements to LN itself are as well) -- it just does not work if you are not already in LN and the chain is congested. So why would we stop there or not want more? In essence we have made huge progress for certain scenarios, but for others we are stuck in the same place where we were at the time of spam attack a decade ago or so.

[gmaxwell]

If you have a solution-- by all means suggest it.  I think it's fundamental:  the resource needs to be limit in supply to control operating costs and to drive income for security.  Free market action turns that into a market.  If someone is outbidding what you're willing to pay that is a problem for you but it's also a success for the system.

Essentially this is just repeating the age old arguments against free markets-- they don't necessarily optimize for human flourishing to the extent a ideal spherical dictator in simple harmonic motion might be conjectured to do.   But we find that e.g. communist regimes tend to create appalling living conditions when they're not busy commit massing murder on unimaginable scale ... while dirty amoral capitalism has reliably produced the best living conditions ever known to humanity in spite of the fact that it regularly screws up some stuff.

Part of the issue is that there already are other solutions-- people just want to ignore them perhaps because stuff that works is boring.  You mention lightning but it's also perfectly possible for altruistic miners to skip over higher fee paying transactions and prioritize stuff they think is good for the welfare of the system.   And that works fine without coercive consensus changes or mob action that deprives people of individual choice.

And of course this also gets ignored because all this gnashing of teeth is over a problem which is currently inactive.   Knots supporters are a bored immune system that has turned on the host because Bitcoin's success has for the moment robbed them of any real enemies they're capable and willing to fight.

[Wind_FURY]

how do you effectively create a spam-protection mechanism? Filter and hope?

Proof of Work is a quite good filter. Today, you can pay only with fees. But it is possible to allow lower fees, if you provide more Proof of Work instead. And I guess there is some equilibrium somewhere, where users could accept paying N satoshis fee, with M difficulty. And then, you can decide, if you want to provide more coins, or more hashrate, to get your transaction confirmed.

OK, but the cost of spam-protection must come from somewhere. Why don't we choose the simplest path and let users pay fees in Bitcoin instead? I believe the miners would appreciate it greatly, and for them to keep giving security for the network.

Although for your idea, you probably should start your own topic for it. It could be something good to learn from the discussions that could come from it.

  👍

[Satofan44]

If you have a solution-- by all means suggest it.  I think it's fundamental:  the resource needs to be limit in supply to control operating costs and to drive income for security.  Free market action turns that into a market.  If someone is outbidding what you're willing to pay that is a problem for you but it's also a success for the system.

I do not, that is why I am asking. I also don't see this system as the end solution to it all. 4 MB blocks in 2030? 2050? 2100? Same fee market issues 100 years from now? Problem for someone else from the future? Sure we can take that view, but I'd like to discuss these things. Do w tell new users that they are responsible for not being able to use Bitcoin at such at time and they should have known better? They should have, but they didn't and they won't. We are talking about probably 90-99% of the humans on the planet.

Essentially this is just repeating the age old arguments against free markets-- they don't necessarily optimize for human flourishing to the extent a ideal spherical dictator in simple harmonic motion might be conjectured to do.   But we find that e.g. communist regimes tend to create appalling living conditions when they're not busy commit massing murder on unimaginable scale ... while dirty amoral capitalism has reliably produced the best living conditions ever known to humanity in spite of the fact that it regularly screws up some stuff.

I am not arguing against it, I am just interested in ways with which people who -- for whatever reason -- did not prepare themselves for times like that are still able to use Bitcoin in many ways. Don't tell me that is a bad idea.  

Part of the issue is that there already are other solutions-- people just want to ignore them perhaps because stuff that works is boring.  You mention lightning but it's also perfectly possible for altruistic miners to skip over higher fee paying transactions and prioritize stuff they think is good for the welfare of the system.   And that works fine without coercive consensus changes or mob action that deprives people of individual choice.

Someone could write a thread summarizing those issues. I don't think more people know many solutions or any solutions at all for this to be honest. The most often cited advice, is simply wait it out or use LN next time which shows an absence of such knowledge.

And of course this also gets ignored because all this gnashing of teeth is over a problem which is currently inactive.   Knots supporters are a bored immune system that has turned on the host because Bitcoin's success has for the moment robbed them of any real enemies they're capable and willing to fight.

Sure, it is not a problem right now but it might be again one day. I see no reason why not to use peaceful time to discuss ideas and solutions, rather than generate rushed and stupid ideas when times are or seem tough -- like the Knots crowd did. They essentially brought forth a "solution" that does not solve anything at all and makes many things worse. For usability things, but for other things too, there is no reason to approach them this way.

[Wind_FURY]

I do not, that is why I am asking. I also don't see this system as the end solution to it all. 4 MB blocks in 2030? 2050? 2100? Same fee market issues 100 years from now? Problem for someone else from the future? Sure we can take that view, but I'd like to discuss these things. Do w tell new users that they are responsible for not being able to use Bitcoin at such at time and they should have known better? They should have, but they didn't and they won't. We are talking about probably 90-99% of the humans on the planet.

?

But CURRENTLY, what do YOU believe is the right block size for Bitcoin? Please consider current usage + what the Core Developers are trying to achieve, in that it's not to sacrifice decentralization/security for as much as possible.

Because that has always been a problem in "scaling" Bitcoin. How to increase throughput without sacrificing security/decentralization.

[stwenhao]

But CURRENTLY, what do YOU believe is the right block size for Bitcoin?

It is good as it is. And the more time will pass, the harder it would be to change it in the future. Also, users pushing non-monetary transactions on-chain, will prevent any block size increases. Because then, if you will pick for example 4 GB, instead of 4 MB, then guess what: you will still have the same amount of payments, but you will also have videos on-chain, taking the rest of that space.

Also, currently people don't test bigger block sizes, but smaller. For example, signet went from 4 MB to 1 MB witness in blocks.

How to increase throughput without sacrificing security/decentralization.

By having more users per UTXO. Today, you have 4 MB blocks, and one user can have one or more UTXOs. In the future, if you will have the same amount of data, and more users, then you will have to handle more users per single UTXO, to make things tick. In the Lightning Network, you have two users per UTXO. In sidechains, you can have more, and even the whole network, with thousands of users, could run under a single UTXO. So, they will be necessary in the future, or else users will do, what they always do: if no decentralized solution is available, then they pick centralized ones. Not because they are better, but because they are implemented.

[Amaro958]

I don't understand why these developers are risking it by playing with people's money. Why overcomplicate something that just should be as lean as possible and as simple as possible? You are supposed to keep the money safe, Bitcoin is not some toy app you can add "cool things" to it, this is like updating a nuclear reactor on the fly. You are playing with people's money, and you don't to gamble with people's savings, so anything added into the code that adds extra layers of complexity, anything that isn't to guarantee everything remains stable and money to be able to move from A to B or to stay frozen for as long as the owner wants, does not do anything good for BTC. I wouldn't want to be the ideas guy that screws up Bitcoin eventually. All these innovations and ideas should be happening on top of Bitcoin, not on the Core. That should be the point of Bitcoin Core, remain the Core, stick to the basics, and do not add any complexity in the name of innovation. Bitcoin's value is because it's reasonable to expect that your money will not disappear in many decades, but at this rate if they keep adding a bunch of crap on top of the monetary use case, I don't see how you could trust this long term. I have not seen any convincing arguments as to why this is not the case at all. They always sound like developers having an ego trip about how their great idea must be merged into Bitcoin. Well good luck to them.

To what additional complexity in the Bitcoin core code are your reffering?
Removing the limit technically reduces the complexity