Call for reconfiguration of nodes to relay transactions with fee-rates <1 sat/vb

[Greg Tonoski]

The minrelaytxfee default value of 1000/kvB (1 sat/vbyte) has been set in Bitcoin nodes since 2013 or earlier. In 2013, 1000 sats were priced at $0.01, today it is about 100 times more. Wouldn't you like to consider adjusting your node configuration and lower the hurdle, perhaps? I would suggest setting

Code:

minrelaytxfee=0.00000001

(which translates to 0.001 sat/vbyte) as a rule of thumb. The value could be set in a bitcoin.conf file. There is negligible number of such reconfigured nodes relaying transactions with fee-rates below 1 sat/vbyte currently.

There is the intention to unblock transactions with fee-rate below 1 sat/vbyte in Bitcoin. Probably consolidation and multisig types of transactions would benefit the most.

[SilverCryptoBullet]

1 satoshi/vbyte is like a defaul minimum fee rate by Bitcoin nodes that run by Bitcoin miners. They can config their nodes with lower fee rates but think about their income from mining. It comes from two sources: Bitcoin block subsidy that is fixed, and Bitcoin transaction fees that is flexible and depended on fee rates and transaction sizes of transactions they confirm.

If I am a Bitcoin miner or a Bitcoin mining pool, I have no interest to reconfig the minimum fee rate in my Bitcoin mempool to lower than 1 satoshi/vbyte.

[Greg Tonoski]

If I am a Bitcoin miner or a Bitcoin mining pool, I have no interest to reconfig the minimum fee rate in my Bitcoin mempool to lower than 1 satoshi/vbyte.

I disagree. Miners and mining pools may either accept low fee-rates or lose opportunity to earn them otherwise.

[ABCbits]

0.001 sat/vB sounds too low for me, where i would suggest 0.1 sat/vB instead. But if you want to see this happen, i believe you have better chance by convincing developer of Bitcoin Core and other full node software to change default minrelayvalue on their software instead.

[hosemary]

0.001 sat/vbyte would be too low.
Even if the minimum relay fee is 0.001 sat/vbyte, we will usually have to use higher fee rates.

Assume that you want to make transaction with 1 segwit input and 1 single output.
If nodes decrease the minimum relay fee rate to 0.001 sat/vbyte, the minimum fee for your transaction would be around 0.11 sat.
Since the smallest unit we have is 1 sat, you would have to pay the minimum fee of 1 sat for your transaction and that would make the fee rate around 0.01 sat/vbyte.

[d5000]

I would probably support this move (not totally sure still, have to think more), although in my opinion ABCbits' suggestion of 0.1 sat/vbyte is better. This is not something set in stone but simply a default configuration so it can be played around a bit with it.

I am an advocate of a free fee market and miners should be able to live from transaction fees in 2140 (and already several decades before, tx fees should become their main income). And I'm also pro incentives to move off chain (to LN, Ark or sidechains/rollups).

However, in the current configuration there is an artificial barrier for the market value to form, and there should also be incentives to use the network in low-congestion phases (like for consolidations). If the low fee phase with 1 sat/vByte "as new normal" continues, users will set lower fees "just to try", and eventually single nodes and mining pools will probably move to a lower value to maximize their income if enough users set a lower fee.

This would look like "free market is working", but it also would it make it more difficult for users without that much technical knowledge to know what is the lowest possible fee. Okay, with 1 sat/vByte they will always be at the safe side, but I think with a new default value of 0.1 the whole process would look cleaner and more transparent.

[Bitcoin Smith]

I would probably support this move (not totally sure still, have to think more), although in my opinion ABCbits' suggestion of 0.1 sat/vbyte is better. This is not something set in stone but simply a default configuration so it can be played around a bit with it.

If we have 0.1sat/vb then after another 10 years we may feel that it is huge value according to the price situation at that time so setting low as possible can be a good thing so we don't need to adjust every time. And since this is a free market everyone can opt to pay the high fee which can be 0.1 or even 1 sat/vb but I feel having the opportunity to pay as low as possible is kind of right to everyone who owns bitcoin and that paves the way for bitcoin to be used for actual payments since the most criticism that bitcoin receives is about expensive and slow transfers.

[d5000]

If we have 0.1sat/vb then after another 10 years we may feel that it is huge value according to the price situation at that time so setting low as possible can be a good thing so we don't need to adjust every time.

I disagree. Fees of 0.001 sat/vb are not realistic in the short to medium term, so there is no urgency at all to set that default value so low. And as I wrote, this is a configuration option that can be changed at any time by the nodes, not a softfork or so everybody must agree on, so there is no need to set it taking into account the next 10 years.

If the fee level is allowed to fall too low during some times, this would also probably again incentive behavior which will lead to more congestion in FOMO times. Runes or Ordinals could then practically be minted without any cost (and thus without a realistic "profit expectation"), as it occurred during the Litecoin and Dogecoin Ordinals waves (see this graph) which were quite "useless" as they didn't really generate much value for their networks.

The higher incentives for data transactions is also the reason why I wrote that I have to think more about that proposal: perhaps even 0.1 sat/vbyte would already have these unintended consequences.

[Bitcoin Smith]

^ there's no need for fork to implement the changes but I doubt it's an easy thing either convincing the nodes to change their default settings whenever we want to so I just thought why not set to the lowest possible value and let the users give an option to choose the lowest or priority fee according to their needs. Anyway 0.1sat/vb is more than welcome as well as of current situation.

Orginal spammers doesn't really care about the fee because the 0.1 or 1sat doesn't stop them from minting their useless tokens but I agree that lower the fees the more chance of spamming the blockchain.

[Greg Tonoski]

If the fee level is allowed to fall too low during some times, this would also probably again incentive behavior which will lead to more congestion in FOMO times. Runes or Ordinals could then practically be minted without any cost (and thus without a realistic "profit expectation"), as it occurred during the Litecoin and Dogecoin Ordinals waves (see this graph) which were quite "useless" as they didn't really generate much value for their networks.

The higher incentives for data transactions is also the reason why I wrote that I have to think more about that proposal: perhaps even 0.1 sat/vbyte would already have these unintended consequences.

I agree that spam should be eliminated from Bitcoin. I think that the minrelaytxfee is not an effective tool for that. In fact, genuine monetary transactions are the most sensitive to fee levels while spam is not - as evident from blocks full of spam and the highest fees in the last 2 years.

For spam protection, there is the

Code:

datacarrier=0

which is much more effective. Also, there is the (OP_FALSE OP_IF) filter bugfix implemented in Bitcoin Knots.

[d5000]

Orginal spammers doesn't really care about the fee because the 0.1 or 1sat doesn't stop them from minting their useless tokens but I agree that lower the fees the more chance of spamming the blockchain.

With 1 sat/vByte, they need to pay at least 50 cents for an Ordinals mint (as it takes basically 2 transactions) and about 15-20 cents for Runes. You can do that once compĺetely without profit expectation but not 1000x. For 0.1 the same basically would apply, it would still not be "free" in the sense you can create hundreds of mints for a few cents, because every mint would take a few cents. 0.001 is 100 times lower, this is basically the fee level on Litecoin.

So I think it makes a difference.

For spam protection, there is the datacarrier=0 which is much more effective. Also, there is the (OP_FALSE OP_IF) filter bugfix implemented in Bitcoin Knots.

Unfortunately this only targets some types of spam. There were several protocols published to evade such restrictions, simply encoding the token data in "normal-looking" public keys (basically what the first wave of coloured coins did in 2014 approximately). The probably most well known example is Stampchain. Fortunately they didn't really get a following but that can change in the future.

And it seems the datacarrier option isn't used enough to deter Ordinals who have grown again a bit in the past 2 months.

[stwenhao]

Wouldn't you like to consider adjusting your node configuration and lower the hurdle, perhaps?

How many low fee transactions can you see today? There were much more of them in the past: https://jochen-hoenicke.de/queue/#BTC,1y,weight,0

See the grey rectangle at the bottom? It was practically reduced into zero in July 2024.

Also, currently, as long as blocks are full, miners are guaranteed to earn at least 0.01 BTC in fees per block. If you lower this value, then this minimum value will be lowered as well. Which is not a big deal now, but it will be after 12th halving, when the basic block reward will reach 0.01220703 BTC.

If you want cheaper transactions, then focus on batching instead. In the linked example, you can batch 560+561=1121 weight units into 685 weight units, and the end result will be the same (so if you keep the same fees, the feerate will increase).

Miners and mining pools may either accept low fee-rates or lose opportunity to earn them otherwise.

And their decision is to lose that opportunity, because otherwise, they would change the default settings long time ago. But they don't want to do that. And if you observe, how fees can suddenly increase, when the next block is not mined in 60 minutes, then you can see, that delaying transaction confirmations is in general profitable for mining pool operators, because then, users increase their fees with full-RBF.

Probably consolidation and multisig types of transactions would benefit the most.

Not really, Ordinals, Inscriptions, Runes, NFTs, and so on, would benefit the most. In case of regular payments, it is technically possible to strip in-the-middle transactions, and just use a batched version with a higher feerate. However, in case of data-pushing protocols, they don't want to strip anything, so decreasing fees will just make spamming easier.

By the way: it is good to keep the spam in testnet3, if it decreases the spam in mainnet. Currently, many spammers are sitting on different chains, and if you lower mainnet fees, then they will have an incentive to come back.

[Mia Chloe]

~snip

Yeah I get the angle you are coming from and where this your suggestion is driving towards. Now mining is a two way thing and the fact is you are making your suggestions towards favouring those broadcasting transactions forgetting the reverse effects on miners. Basically a miner setting his feet rate that low is basically increasing the load on his node. This is because his node will basically start scraping transactions that majority of other nodes dropped.

Now the default purge fee for most nodes is 1sat / byte and it's quite reasonable. Nobody would want to waste power Searching for transactions with lower fee rates when same power can be used to increase their chances of getting the block reward or hashing transactions with bigger fee rates.

[hosemary]

Now the default purge fee for most nodes is 1sat / byte and it's quite reasonable.

OP is talking about minimum relay fee and that's different from purging fee.

The minimum relay fee is set by the nodes and the default value is now 1 sat/vbyte. Any transaction paying less than 1 sat/vbyte is considered non-standard and would be rejected by any node with default minimum relay fee.
The purging fee rate isn't set by the nodes directly and that depends on their maximum mempool size and how congested the network is.

[hd49728]

You can check Bitcoin blockchain history and transactions with lowest transaction fees, but with more than only Coinbase transaction.
https://blockchair.com/bitcoin/transactions?s=output_total(desc)#f=block_id,time,input_count,fee_per_kwu,fee_per_kb,fee_per_kb_usd,output_total

As a Bitcoin user, I believe that everyone want to enjoy as cheapest transaction fee as possible so your proposal makes sense for Bitcoin users. With Bitcoin miners, it does not make sense because it reduces their mining income.

When you work, you want to receive as highest salary / income as possible so do Bitcoin miners.

[ABCbits]

I forget to mention it earlier, but there's at least 3 call to reduce default minrelayfee value in past.
https://github.com/bitcoin/bitcoin/issues/14012
https://gnusha.org/pi/bitcoindev/CAGHFe1BXdTkPZn4r_KTxYoz0sqcMsV830dm5JTTFURxDezBnDQ@mail.gmail.com/
https://www.erisian.com.au/meetbot/bitcoin-core-dev/2018/bitcoin-core-dev.2018-07-05-19.22.log.html#l-24

For spam protection, there is the

Code:

datacarrier=0

which is much more effective. Also, there is the (OP_FALSE OP_IF) filter bugfix implemented in Bitcoin Knots.

If both of them become default on many nodes, spammer will resort to use old technique (using 20-byte public key on address) which is worse on long term due to unprunable UTXO. IMO only making OP_FALSE OP_IF ... OP_ENDIF as non-standard is good compromise.

And it seems the datacarrier option isn't used enough to deter Ordinals who have grown again a bit in the past 2 months.

Ordinals use Taproot redeem script. So i think you actually meant Runes, which use OP_RETURN.