Why was there no hard cap built in for tx fees?

[cloverme]

Why didn't satoshi or core developers set a max cap for transaction fees based on total sent? Seems like something that would keep miners from being able to exclude low-fee transactions at the same time keep the runaway inflation in check.

[1714470149]

1714470149

[1714470149]

1714470149

[ranochigo]

It shouldn't be. The fee market works in a way to deter people from spamming up the blockchain with low fee transaction. If fees were to be regulated, there would be a fixed cap and everyone would be paying the same fee. By doing so, more transactions would take even longer to be included in a block.

Miners are still able to include only high fee transactions if they want. If you were to set a fee limit, miners wouldn't be having sustainable profits.

[Kakmakr]

Miners fees are crucial for the survival of the network, once the Block rewards stop. We should not mess with that. The miners have found a way to exploit people with their strategy to jump between forks and leaving the MemPool to grow and in turn pushing up the fees.

The solution for that, would be to prevent that from happening. ^hmmmmm^

[nullius]

Why didn't satoshi or core developers set a max cap for transaction fees based on total sent? Seems like something that would keep miners from being able to exclude low-fee transactions at the same time keep the runaway inflation in check.

It would be wise to know the purpose of fees before attempting such proposals.  Fees have nothing whatsoever to do with the value sent.  Rather, they bid on a share of a limited global resource:  Block capacity.  It is easily possible for a huge transaction to transmit no value at all, via OP_RETURN; whereas a transaction which sends one huge input to one output could transmit a king’s fortune while using only a few hundred bytes of network resources.

Moreover, added to block reward and multiplied by the market value of Bitcoin, fees increase the total value of mining; and the value of mining is critical to Bitcoin security:

Technically if every single transaction was sent with 1 sat/byte fees then that will become the normal fees.

Technically, if every single transaction was sent with a 1 sat/byte fee, then the large mining farms would shut down, difficulty would plummet—and Bitcoin would lose its Byzantine fault-tolerant security against double-spends and rewriting of the blockchain history.

FEES HELP KEEP THE NETWORK SAFE.  The higher the value of Bitcoin, the more safety it needs.  The higher the potential profit or miners, the more heavily miners compete with each other—thus raising difficulty—thus making it more difficult, and astronomically more expensive for any entity to perform a “51% attack” and the like.

Bitcoin did not need high mining difficulty, when it was a toy and every bitcoin was worth 30¢.  Then, yes, it was fine for difficulty to be low so that anybody could mine on a desktop PC’s CPU.  Now that the network protects the exchange-value equivalent of hundreds of billions of dollars, it needs commensurately high hashrate.

Most people do not understand how the fee market interacts with the security of Bitcoin.

It shouldn't be. The fee market works in a way to deter people from spamming up the blockchain with low fee transaction. If fees were to be regulated, there would be a fixed cap and everyone would be paying the same fee. By doing so, more transactions would take even longer to be included in a block.

Actually, the whole network would implode and grind to a halt.  People think it’s bad now?  Imagine if a max fee were set by a ratio to the value sent, per OP’s suggestion.  Somebody out there would generate a bunch of OP_RETURN transactions containing cute cat photos (or much worse—I mean of course, photos of cryptokitties).  0×cap_ratio=0, for any value of cap_ratio; thus, those transactions would be free.  Ultimate results are left as an exercise to the reader.

The miners have found a way to exploit people with their strategy to jump between forks and leaving the MemPool to grow and in turn pushing up the fees.

The solution for that, would be to prevent that from happening. ^hmmmmm^

Time to go nuclear with MR POWA?  Nah.  I wouldn’t want to hurt the decent miners who pool up with p2pool, slush, et al.

In my opinion, the solution to huge, centralized malicious miners is to commoditize SHA-256 ASICs.  I’ve said before that a reasonably powerful ASIC should be inexpensive and readily available as a mass-market GPU.  Now I will go one step further:  The damn things should be built into novelty USB coffee warmers.  Do you really want to sock it to Jihan?

Any open-source hardware experts want to cook up a design and crowdfund an initial production run?  Yes, I know it’s not so easy.  But there could be much money in this, too.  Somebody with an appropriate professional skillset should work on it.

[LoyceV]

Miners fees are crucial for the survival of the network, once the Block rewards stop. We should not mess with that.

The block rewards are sufficient for decades to come, but despite that, last week fees got as high as the block reward already.
Of course fees could replace the block reward in the far future, when many more people make many more transactions. Current high fees only assure miners don't want larger blocks. I don't buy the reasoning that small blocks are needed for nodes, disk space is cheap.

A limit on fees won't solve the problem, it should have been the other way around: if x consecutive blocks are full, and/or if x consecutive blocks have fees higher than y, blocks should get bigger. Core could have implemented this years ago, Satoshi planned for larger blocks in the future already in 2010. Instead, Bitcoin got crippled for financial gains of a few.

Eventually lower fees and zero block reward could even be a huge environmental benefit! Currently, miners get about $40 million per day, which leads to a large difficulty, huge hashing capacity, and a few GW energy consumption. If miners get less money, the hashrate doesn't have to be this high, and they'll consume less energy.
This $40 million per day adds up to $14 billion per year now. That's money leaving the Bitcoin ecosystem, it's a drain that requires continuous inflow of new capital.

Many people don't realize more miners don't contribute to making more transactions. If Satoshi would have been the only one mining all these years, his 9 year old computer could easily process all current transactions on his own.

[cloverme]

I don't disagree with any these points... it's a an interesting discussion. The network was designed for fair use and under that, yes, transaction fees and allowing miners to choose transactions based on that makes sense. However, this assumes that the people using the network have the same goal to use the network fairly. If Alice and Bob have significant hash power to mine and also collude to undermine system by filling blocks with high transaction fees, Chuck as a fair-use player may pick a high transaction fee not realizing Alice and Bob are not playing fair.  The things that keep bitcoin safe like higher difficulty also keeps out miners who want to clear out the reasonable tx fee mempool because they don't have enough hashpower. I do believe that we have bad actors today working to try to damage the network and at the same time, promote other forks as better alternatives. I get it is a wait-and-see approach at this point to get the bitcoin industrial sector to adopt and use segwit as well as wait for lightning. Hopefully we're not "stuck" here for long, watching this attack is gut wrenching.

[nullius]

A limit on fees won't solve the problem, it should have been the other way around: if x consecutive blocks are full, and/or if x consecutive blocks have fees higher than y, blocks should get bigger. Core could have implemented this years ago, Satoshi planned for larger blocks in the future already in 2010. Instead, Bitcoin got crippled for financial gains of a few.

Three questions (not to beat a horse long dead):

0. How do you prevent malicious parties, especially malicious miners from gaming your system with spam attacks to raise the blocksize?  N.b. that a malicious miner with huge hashrate (such as Jihan) has relatively low cost for spam attacks, since any fees for the spam go back to him for blocks he himself mines.

1. What does this even solve?  A blocksize increase increases linearly.  But Bitcoin needs capacity increase by orders of magnitude—I think at least 10000x, to provide adequate room to grow in the next decade or so.  An 8MB or even a 32MB blocksize would promptly fill today, with no room for tomorrow.  If BCH had Bitcoin’s popularity and demand, then BCH would have high fees, too (and if I had cybernetic wings welded into my back...).  The problem as you state it is really one of scaling.  Linear blocksize increase is a non-solution to scaling.

2. How does your idea solve problems caused by increased blocksize, such as UTXO set growth, increased orphan rate, etc., etc.?  Blocks larger than 4MB would not be safe for the network.  With Segwit’s 4MB block weight limit, we already have the potential to approach that limit.  N.b. for those in the peanut gallery, increased disk use for a growing blockchain is the smallest problem with increased blocksize.

Eventually lower fees and zero block reward could even be a huge environmental benefit! Currently, miners get about $40 million per day, which leads to a large difficulty, huge hashing capacity, and a few GW energy consumption. If miners get less money, the hashrate doesn't have to be this high, and they'll consume less energy.
This $40 million per day adds up to $14 billion per year now. That's money leaving the Bitcoin ecosystem, it's a drain that requires continuous inflow of new capital.

Many people don't realize more miners don't contribute to making more transactions. If Satoshi would have been the only one mining all these years, his 9 year old computer could easily process all current transactions on his own.

Most people don’t realize that competition to increase hashrate (thus difficulty) assures Bitcoin’s BFT security.  I noted this above.  If Satoshi would have been the only one mining all these years, what would now stop somebody from spinning up $100k in EC2 nodes, instantly grabbing 99.999999% of the global hashrate (forget a “51% attack”!), and arbitrarily rewriting the blockchain on whim?  To describe only one obvious and easy practical attack vector.

We don’t need for miners to consume less energy.  We need for them to consume more.  Part of the beauty of Bitcoin’s design is that the higher the total value of the network, and thus the higher its security needs, the competition and difficulty of mining automatically increases to adjust.  Satoshi’s home PC was an adequate miner when a bitcoin was worth 30¢.  Nowadays, anybody who could acquire ($$$) a majority of hashrate could steal a fortune—and worse, disrupt a market worth the current exchange eqivalent of $300+ billions; there would be all sorts of nefarious ways to profit from that.  We need for mining to be collectively expensive, prohibitively expensive from the perspective of anybody trying to acquire majority hashrate.

But that expense should never be centralized, which it is now; for then, that expense buys power....

I do believe that we have bad actors today working to try to damage the network and at the same time, promote other forks as better alternatives. [...] Hopefully we're not "stuck" here for long, watching this attack is gut wrenching.

Bingo.  And the long-term solution is to see serious competition in ASIC production.

How many people even realize that something like 70% of all hashpower is from ASICs produced by Bitcoin’s current most dangerous enemy?  Who currently accepts payment in their own pet scamcoin instead of Bitcoin, so you are coerced to help pump BCH if you want to buy from a supplier who in practice has something almost tantamount to a monopoly.  What an insulting dilemma for Bitcoin supporters.

I want to see stocking-stuffer ASIC toys which can be sold as a mass-market consumer item, and would each give (via a mining pool) about the same microincome as clicking ads on a “faucet” site.  Call that the ARM/Raspberry Pi level of ASIC.  One step up, I want to see ASICs as cheap and readily available as a commodity consumer CPU.  The equivalents of Celeron/i3/i5/i7 ASICS, for casual miners to moderate enthusiasts.  If ten million people each had an average of a measly 500 GH/s, the result would sum to 5 EH/s.  Current global hashrate is about 8.5 EH/s.  Jihan would cry.

[LoyceV]

0. How do you prevent malicious parties, especially malicious miners from gaming your system with spam attacks to raise the blocksize?  N.b. that a malicious miner with huge hashrate (such as Jihan) has relatively low cost for spam attacks, since any fees for the spam go back to him for blocks he himself mines.

I'm not sure how to stop that, but I doubt it's worse than paying $30 for a small transaction.

1. What does this even solve?  A blocksize increase increases linearly.  But Bitcoin needs capacity increase by orders of magnitude—I think at least 10000x, to provide adequate room to grow in the next decade or so.  An 8MB or even a 32MB blocksize would promptly fill today, with no room for tomorrow.  If BCH had Bitcoin’s popularity and demand, then BCH would have high fees, too (and if I had cybernetic wings welded into my back...).  The problem as you state it is really one of scaling.  Linear blocksize increase is a non-solution to scaling.

A blocksize increase gives breathing space, until a more permanent solution is in place. 1 MB blocks were sufficient for the first 6 years of Bitcoin, 8 MB blocks would give 8 times more space to grow.
I agree Bitcoin needs exponential scaling, but until that is available, larger blocks would at least make it possible to make transactions again.
Bitcoin.org still says "low processing fees". It's simply not true anymore.

2. How does your idea solve problems caused by increased blocksize, such as UTXO set growth, increased orphan rate, etc., etc.?  Blocks larger than 4MB would not be safe for the network.  With Segwit’s 4MB block weight limit, we already have the potential to approach that limit.  N.b. for those in the peanut gallery, increased disk use for a growing blockchain is the smallest problem with increased blocksize.

I wish I had all the answers It just hurts to see the Bitcoin I like being turned into something that I can barely use. I can no longer use it to order food, unless I pay 3 times more due to fees.
Something has to be done, but that "something" takes much too long.

Most people don’t realize that competition to increase hashrate (thus difficulty) assures Bitcoin’s BFT security.  I noted this above.  If Satoshi would have been the only one mining all these years, what would now stop somebody from spinning up $100k in EC2 nodes, instantly grabbing 99.999999% of the global hashrate (forget a “51% attack”!), and arbitrarily rewriting the blockchain on whim?  To describe only one obvious and easy practical attack vector.

I do realize the difficulty is needed to prevent attacks, but wouldn't Bitcoin be almost as secure with just 10% of the current hash rate?
Note that I don't mean turning off 90% of current hashing power and waiting for them to do a 90% attack, I mean the hypothetical situation in which 90% of current hashing power was never installed.

We don’t need for miners to consume less energy.  We need for them to consume more.  Part of the beauty of Bitcoin’s design is that the higher the total value of the network, and thus the higher its security needs, the competition and difficulty of mining automatically increases to adjust.  Satoshi’s home PC was an adequate miner when a bitcoin was worth 30¢.

I was speaking from an environmental point of view. The amount of power consumed is immense! If possible, try to get a tour in a power plant that produces "just" 1 GW, it's much more impressive than I can describe. You can actually feel the power.
If you're speaking from a network security point of view, I guess there is no upper limit of how much power you'd want to consume to secure the network.

Nowadays, anybody who could acquire ($$$) a majority of hashrate could steal a fortune—and worse, disrupt a market worth the current exchange eqivalent of $300+ billions

Thanks to the Bitcoin forks, there are now actually some parties which control enough hash power to disrupt Bitcoin when they switch to Bitcoin Cash. The Cash fork wouldn't have happened if Bitcoin would have had bigger blocks.

And the long-term solution is to see serious competition in ASIC production.

How many people even realize that something like 70% of all hashpower is from ASICs produced by Bitcoin’s current most dangerous enemy?

You mean the one with the remote shutdown backdoor?

I want to see stocking-stuffer ASIC toys which can be sold as a mass-market consumer item, and would each give (via a mining pool) about the same microincome as clicking ads on a “faucet” site.  Call that the ARM/Raspberry Pi level of ASIC.  One step up, I want to see ASICs as cheap and readily available as a commodity consumer CPU.  The equivalents of Celeron/i3/i5/i7 ASICS, for casual miners to moderate enthusiasts.  If ten million people each had an average of a measly 500 GH/s, the result would sum to 5 EH/s.  Current global hashrate is about 8.5 EH/s.  Jihan would cry.

Isn't that what Bitcoin Gold is trying to do (in it's own way)? I'm no expert on algorithms, but as far as I understand they switched to something that's more suited to mine on GPU again.
I'd love to see your idea implemented, and it's much closer to what Satoshi originally envisioned. If Bitcoin entirely relies on 10 million people each mining with a small device, each will mine 0.066BTC per year. In my country that's enough for approximately 350W of electricity (excluding the cost of the device itself).
But even if ASICS could be small and cheap for consumers, that also means they're cheap for large miners, who can then centralize them again.