5 ms·

There are 100 MB of transactions which want to get in the next block: https://blockchain.info/charts/mempool-size?timespan=all https://blockchain.info/charts/me

by 21 9y ago

There are 100 MB of transactions which want to get in the next block: https://blockchain.info/charts/mempool-size?timespan=all https://blockchain.info/charts/mempool-size?timespan=all But the block is only 1 MB. So 100 times more transactions want in than there is space, thus you need to pay $10 to secure a place for your transaction. Fewer transactions or increased block size would decrease the transaction fee.

Empact 9y ago
More than 100 megs according to this, which also shows the fee strata: https://dedi.jochen-hoenicke.de/queue/more/#24h https://dedi.jochen-hoenicke.de/queue/more/#24h The good news is that at current volumes, the miners are making progress clearing the lower fee transactions. In the next day or two we should start clearing transactions at < 100 satoshis/B. That's one way that average / median transaction fees are misleading in Bitcoin: people paying much lower fees are making transactions, but traders and those moving large sums seem to be willing to over-pay to move their money. In some ways the minimum transaction fee processed is more instructive than the median, because it shows how much did one have to pay to get processed, vs how much dod people in general overpay to ensure their payment was processed.
- icelancer 9y ago
  Furthermore, tons of "exchanges" in Korea were spamming the mempool with 0-fee transactions, as were some other notorious figures, driving these relatively false mempool statistics way up in order drive up the value of other coins.
  - oh_sigh 9y ago
    What benefit do they get from spamming the exchanges? Is there ever an reason to mine 0-fee transactions?
    - icelancer 9y ago
      To get threads like this created about how the mempool is packed and it costs $20 to move BTC from peer-to-peer and to drive media in that direction. For what second-order reason would someone do this? Well, I can think of a few....
      - xocyabencl 9y ago
        How can a bunch of no fee transaction attempts cause a $20 market price? It seems like the block would need to be full of transactions with high fees.
        rocqua 9y ago
        By increasing the size of the backlog, you make it seem like high transaction fees will stick around longer. It also exaggerates the immediate issue of bitcoins low transaction rate.
      - qsucvatz 9y ago
        Because Bitcoin used to work, and now it doesn't. Bitcoin Cash scales appropriately.
    - xorcist 9y ago
      Some miners are likely paid outside the blockchain to pick up certain transactions. It is clear that at least one miner do it systematically but I don't think they have ever divulged who their client is and why they do it.
    - ncallaway 9y ago
      I don't know the answer to the first question. The second question (Is there ever an reason to mine 0-fee transactions?) is "yes, for a while". Until approximately 2040 new Bitcoin are introduced into the system as a reward to the miner who discovers a block. So, even if a miner mines a block with only 0-fee transactions, they are still rewarded with a small amount of Bitcoin "from nowhere". Bitcoin will be introduced in this manner until there are 21 million BTC in existence, at which point they will stop being introduced "from nowhere". This is expected to take until about 2040 (though I don't know how that projection is made). After that point there won't be much of an incentive to mine 0-fee transactions.
  - qsucvatz 9y ago
    0 fee txns don't cause high fees. They are dropped from the mempool as needed.
    - icelancer 9y ago
      Correct, I am aware. It generates pretty graphs, though.
      - meltsice 9y ago
        https://jochen-hoenicke.de/queue/#30d https://jochen-hoenicke.de/queue/#30d There are 100,000 transactions with fees greater than $1. Where are these 0-fees? (15 sat/Byte, ~443 B/tx)
- qsucvatz 9y ago
  Median fee is a good metric, it avoids the over-paying txns and the miner-accelerated txns.
aviv 9y ago
And people expect Amazon to announce Bitcoin acceptance any day now... lol.
xorcist 9y ago
> But the block is only 1 MB. Do note that this was the maximum allowed size. Many blocks are under that, or even empty. It's completely up to the miner that mines the block. This is probably not something that can easily be changed. The 1 MB maximum size was also true only up to August. New style transactions are allowed to go above that, and inputs and outputs are counted differently towards the limit in order to incentivize transaction defragmentation. It's not relevant to speak of a fixed size cap anymore. If everyone changed transaction type overnight we'd see somewhere around 2 MB blocks, but it is expected that the rules would lead to a change in what transactions are made which would make blocks larger. Especially if things like Lightning transactions are more popular in the future. > thus you need to pay $10 to secure a place for your transaction. That's probably something like the average fee for the past few blocks, but the minimum fee you need to pay is much lower. The exact number is not very relevant however since Bitcoin fluctuates in value, and so does the transaction volume. Note that fees are denominated in Bitcoin. Fees have actually been pretty stable over the long term, but the skyrocketing price of Bitcoin has lead to it not being practical for low-value USD transactions. This was always true, but the minimum value for which it is viable increases as Bitcoin is more valuable.
- qsucvatz 9y ago
  Fees are skyrocketing in Satoshi: what used to be 1 sat/Byte is now 1000 sat/Byte.
  - xorcist 9y ago
    Not really, unless you count the period that blocks weren't full when you could even send transactions without a fee. Fees today are about 100 sat/b, a multiple of that in times of congestion and lower in the weekends. You never have to pay 1000 sat/b, but you are of course free to subsidize miners however you want. This is roughly the same the fees were in 2013, only that those satoshis of yours are worth so much more USD now. That isn't the whole truth, and those spikes in fees come more often now and spike higher, but the minimum fee to get into a block averaged over a year isn't that much worse now than in 2013. That probably isn't very helpful if you need to transact in times of congestion, or make low value transactions, but it's important to know that netiher 1 nor 1000 sat/b has been reality for some time now, and that you should never talk about fees in USD because that doesn't describe how it works. Another common mistake is to include the value of the transaction when describing fees which misleads people into thinking fees are a percentage of the value transacted which is not the case.

by 21 9y ago

Empact 9y ago

More than 100 megs according to this, which also shows the fee strata: https://dedi.jochen-hoenicke.de/queue/more/#24h https://dedi.jochen-hoenicke.de/queue/more/#24h The good news is that at current volumes, the miners are making progress clearing the lower fee transactions. In the next day or two we should start clearing transactions at < 100 satoshis/B. That's one way that average / median transaction fees are misleading in Bitcoin: people paying much lower fees are making transactions, but traders and those moving large sums seem to be willing to over-pay to move their money. In some ways the minimum transaction fee processed is more instructive than the median, because it shows how much did one have to pay to get processed, vs how much dod people in general overpay to ensure their payment was processed.

icelancer 9y ago
Furthermore, tons of "exchanges" in Korea were spamming the mempool with 0-fee transactions, as were some other notorious figures, driving these relatively false mempool statistics way up in order drive up the value of other coins.
- oh_sigh 9y ago
  What benefit do they get from spamming the exchanges? Is there ever an reason to mine 0-fee transactions?
  - icelancer 9y ago
    To get threads like this created about how the mempool is packed and it costs $20 to move BTC from peer-to-peer and to drive media in that direction. For what second-order reason would someone do this? Well, I can think of a few....
    - xocyabencl 9y ago
      How can a bunch of no fee transaction attempts cause a $20 market price? It seems like the block would need to be full of transactions with high fees.
      - rocqua 9y ago
        By increasing the size of the backlog, you make it seem like high transaction fees will stick around longer. It also exaggerates the immediate issue of bitcoins low transaction rate.
    - qsucvatz 9y ago
      Because Bitcoin used to work, and now it doesn't. Bitcoin Cash scales appropriately.
  - xorcist 9y ago
    Some miners are likely paid outside the blockchain to pick up certain transactions. It is clear that at least one miner do it systematically but I don't think they have ever divulged who their client is and why they do it.
  - ncallaway 9y ago
    I don't know the answer to the first question. The second question (Is there ever an reason to mine 0-fee transactions?) is "yes, for a while". Until approximately 2040 new Bitcoin are introduced into the system as a reward to the miner who discovers a block. So, even if a miner mines a block with only 0-fee transactions, they are still rewarded with a small amount of Bitcoin "from nowhere". Bitcoin will be introduced in this manner until there are 21 million BTC in existence, at which point they will stop being introduced "from nowhere". This is expected to take until about 2040 (though I don't know how that projection is made). After that point there won't be much of an incentive to mine 0-fee transactions.
- qsucvatz 9y ago
  0 fee txns don't cause high fees. They are dropped from the mempool as needed.
  - icelancer 9y ago
    Correct, I am aware. It generates pretty graphs, though.
    - meltsice 9y ago
      https://jochen-hoenicke.de/queue/#30d https://jochen-hoenicke.de/queue/#30d There are 100,000 transactions with fees greater than $1. Where are these 0-fees? (15 sat/Byte, ~443 B/tx)
qsucvatz 9y ago
Median fee is a good metric, it avoids the over-paying txns and the miner-accelerated txns.

aviv 9y ago

And people expect Amazon to announce Bitcoin acceptance any day now... lol.

xorcist 9y ago

> But the block is only 1 MB. Do note that this was the maximum allowed size. Many blocks are under that, or even empty. It's completely up to the miner that mines the block. This is probably not something that can easily be changed. The 1 MB maximum size was also true only up to August. New style transactions are allowed to go above that, and inputs and outputs are counted differently towards the limit in order to incentivize transaction defragmentation. It's not relevant to speak of a fixed size cap anymore. If everyone changed transaction type overnight we'd see somewhere around 2 MB blocks, but it is expected that the rules would lead to a change in what transactions are made which would make blocks larger. Especially if things like Lightning transactions are more popular in the future. > thus you need to pay $10 to secure a place for your transaction. That's probably something like the average fee for the past few blocks, but the minimum fee you need to pay is much lower. The exact number is not very relevant however since Bitcoin fluctuates in value, and so does the transaction volume. Note that fees are denominated in Bitcoin. Fees have actually been pretty stable over the long term, but the skyrocketing price of Bitcoin has lead to it not being practical for low-value USD transactions. This was always true, but the minimum value for which it is viable increases as Bitcoin is more valuable.

qsucvatz 9y ago
Fees are skyrocketing in Satoshi: what used to be 1 sat/Byte is now 1000 sat/Byte.
- xorcist 9y ago
  Not really, unless you count the period that blocks weren't full when you could even send transactions without a fee. Fees today are about 100 sat/b, a multiple of that in times of congestion and lower in the weekends. You never have to pay 1000 sat/b, but you are of course free to subsidize miners however you want. This is roughly the same the fees were in 2013, only that those satoshis of yours are worth so much more USD now. That isn't the whole truth, and those spikes in fees come more often now and spike higher, but the minimum fee to get into a block averaged over a year isn't that much worse now than in 2013. That probably isn't very helpful if you need to transact in times of congestion, or make low value transactions, but it's important to know that netiher 1 nor 1000 sat/b has been reality for some time now, and that you should never talk about fees in USD because that doesn't describe how it works. Another common mistake is to include the value of the transaction when describing fees which misleads people into thinking fees are a percentage of the value transacted which is not the case.