Why are transactions much faster now?

[exstasie]

(IIRC) the hashing power increased by ~20% in the last few weeks, lower transaction volume and the mempool is empty again[1].

P.S: That's a good time to move your funds to a Segwit address. Do it!

[1] https://dedi.jochen-hoenicke.de/queue/#24h

Yes, this can't be stressed enough. It's a very good time to merge outputs in preparation for when fees rise again....and they will! The recent congestion definitely taught me to take advantage of these periods of lower network activity. I'm in the process of migrating to Segwit wallets as well.... a little extra fee savings never hurt anybody.

If anyone's thinking "but low fees are confirming fast, so that's wrong", then try paying less than 10 satoshis per byte. It'll still take many hours, or even longer than that, for your transaction to confirm.

There's still more than 100 MB of less than 10 satoshi per byte transactions queued in large mempools: https://jochen-hoenicke.de/queue (check the bottom graph)

Yep, I sent out lots of low-fee transactions over the past week. All my 12-15 satoshi/byte transactions have confirmed by now, but all the 5-10 satoshi/byte transactions are still unconfirmed after several days. Oh well, no hurry here!

[1715387676]

1715387676

[1715387676]

1715387676

[ruletheworld]

There's still more than 100 MB of less than 10 satoshi per byte transactions queued in large mempools: https://jochen-hoenicke.de/queue (check the bottom graph)

I get paid in £-Notes and we measure data in megabytes not bytes these days so dress it up what ever way
you like but the fees are high

See https://bitinfocharts.com/comparison/bitcoin-transactionfees.html
Says the fees are $12.56 for just storing 256 bytes of data and it's not my fault that 20,000 fulls nodes
repeat the same work because I did not ask for them all to be here and 1000 would do the trick

You're wrong on many different levels, so let's analyze what you're saying.

• First, according to https://bitinfocharts.com/comparison/bitcoin-transactionfees.html, the average fee is already around 12.5 sat/byte, which is significantly lower than in the past. However, that's not the whole story. Take a look at the mempool stats here: https://jochen-hoenicke.de/queue. This will give you a better idea of when your transactions can confirm. If they aren't urgent, you can put them as lower fee transactions hoping the miners will include them soon but not the next block.
• Your reading of the fees as "$12.56 for just storing 256 bytes" is completely wrong. See the description above to learn what it means.
• I think by 'work' you mean 'proof of work'. This isn't done by the nodes. The nodes simply check for transaction validity, and the honest nodes will only forward correct transactions on to the network, and reject the incorrect ones. This also includes accepting/rejecting blocks mined by miners. The miners are the ones that extend the blockchain, not the nodes.
• Changing the node count from 20,000 to 1,000 as you say will not have any effect on the fees. This suggestion also tells me you have no idea how Bitcoin works or what makes it valuable.

You need to do a little more research into how Bitcoin works.

[jinksters09]

Lucky for you guys because I didn't experience this yesterday when I sent some of bitcoin into my other wallet to cash them out but until now the transaction has not yet been confirmed, but I set at a low priority with a low transaction fee amount it sucks maybe because of the wallet I have used.

[ruletheworld]

Lucky for you guys because I didn't experience this yesterday when I sent some of bitcoin into my other wallet to cash them out but until now the transaction has not yet been confirmed, but I set at a low priority with a low transaction fee amount it sucks maybe because of the wallet I have used.

What's the transaction ID? What wallet did you use to send the transaction? If you post these details, perhaps the community can help.

[SamsungBitcoin]

Lucky for you guys because I didn't experience this yesterday when I sent some of bitcoin into my other wallet to cash them out but until now the transaction has not yet been confirmed, but I set at a low priority with a low transaction fee amount it sucks maybe because of the wallet I have used.

I guest you are using mycelium mobile wallet, because that is the same scenario in the wallet of my wife she send her fund to our local bitcoin wallet for cashing out because she have things want to buy but the confirmation was too slow i think 2 or 3 days to confirm that is sucks it was first time to experience.

correct: edit your post and put the tx id if someone see the unconfirmed transaction they can easily helps you about that, much better if that services is for free.

[digaran]

There's still more than 100 MB of less than 10 satoshi per byte transactions queued in large mempools: https://jochen-hoenicke.de/queue (check the bottom graph)

I get paid in £-Notes and we measure data in megabytes not bytes these days so dress it up what ever way
you like but the fees are high

See https://bitinfocharts.com/comparison/bitcoin-transactionfees.html
Says the fees are $12.56 for just storing 256 bytes of data and it's not my fault that 20,000 fulls nodes
repeat the same work because I did not ask for them all to be here and 1000 would do the trick

You're wrong on many different levels, so let's analyze what you're saying.

• First, according to https://bitinfocharts.com/comparison/bitcoin-transactionfees.html, the average fee is already around 12.5 sat/byte, which is significantly lower than in the past. However, that's not the whole story. Take a look at the mempool stats here: https://jochen-hoenicke.de/queue. This will give you a better idea of when your transactions can confirm. If they aren't urgent, you can put them as lower fee transactions hoping the miners will include them soon but not the next block.
• Your reading of the fees as "$12.56 for just storing 256 bytes" is completely wrong. See the description above to learn what it means.
• I think by 'work' you mean 'proof of work'. This isn't done by the nodes. The nodes simply check for transaction validity, and the honest nodes will only forward correct transactions on to the network, and reject the incorrect ones. This also includes accepting/rejecting blocks mined by miners. The miners are the ones that extend the blockchain, not the nodes.
• Changing the node count from 20,000 to 1,000 as you say will not have any effect on the fees. This suggestion also tells me you have no idea how Bitcoin works or what makes it valuable.

You need to do a little more research into how Bitcoin works.

If you had researched Bitcoin you could know what he is trying to say, is that when you add 1 MB data to the blockchain, you are adding 10,000 MB if you have 10,000 full nodes. controlling the majority is controlling the longest chain.

[ranochigo]

See https://bitinfocharts.com/comparison/bitcoin-transactionfees.html
Says the fees are $12.56 for just storing 256 bytes of data and it's not my fault that 20,000 fulls nodes
repeat the same work because I did not ask for them all to be here and 1000 would do the trick

If you interpret it that way, yeah it would be true. The problem here is that if you're not running a Bitcoin node, Bitcoin won't be trustless. You're basically relying on someone else to help you verify the transaction and blocks. You would have to run a node too and it wouldn't be fair for people to spam as much transactions as possible to occupy all your storage space.

Unless you somehow come up with a method to ensure decentralisation and redundancy with 1000 nodes, then its impossible to blame it on others running a node.

Tbh, I don't know how to interpet your question. The fees does not go to the nodes at all. Your interpretation could be somewhat logical but isn't correct.

[pebwindkraft]

See https://bitinfocharts.com/comparison/bitcoin-transactionfees.html
Says the fees are $12.56 for just storing 256 bytes of data and it's not my fault that 20,000 fulls nodes
repeat the same work because I did not ask for them all to be here and 1000 would do the trick

If you interpret it that way, yeah it would be true. The problem here is that if you're not running a Bitcoin node, Bitcoin won't be trustless. You're basically relying on someone else to help you verify the transaction and blocks. You would have to run a node too and it wouldn't be fair for people to spam as much transactions as possible to occupy all your storage space.

Unless you somehow come up with a method to ensure decentralisation and redundancy with 1000 nodes, then its impossible to blame it on others running a node.

when you follow the entries of Anti-Cen in the forum, you naturally come to the conclusion, that he is trying to throw his fud on bitcoin/Lightning. He references webpages with 12$ to store 220bytes of data. And also wants to measure Satoshi/Byte in Megabytes. Well, he is free to do so (IMHO a good laugh). He is trying to through his torsioned view at the community. He can do so, it's a free world (somehow)! From the posts one can derive a limited level of comprehension of the underlying logic (well, as every low level people). In his view Lightning shall be used for large amount tx, and the number of hosts shall be reduced to 1000. Which in his logic proofs, that the system cannot work .
I haven't seen a single constructive proposal. He is probably paid by government to disturb fruitful discussions. Those who got it, understand the fud. @ruletheworld got it... the others discuss at the same level. Kind of hopeless. 

[TheQuin]

when you follow the entries of Anti-Cen in the forum, you naturally come to the conclusion, that he is trying to throw his fud on bitcoin/Lightning.

It is worth checking out this post https://bitcointalk.org/index.php?topic=2821298.msg28929402#msg28929402 and then reading DannyHamilton's dissection of it two posts below before taking anything Anti-Cen says at face value.

Anyway, in an update to my earlier post above it looks like the difficulty retarget hasn't stopped the mempool clearing out. Currently the 8 and 9 satoshi/byte tx are going through.

[AGD]

There's still more than 100 MB of less than 10 satoshi per byte transactions queued in large mempools: https://jochen-hoenicke.de/queue (check the bottom graph)

I get paid in £-Notes and we measure data in megabytes not bytes these days so dress it up what ever way
you like but the fees are high

See https://bitinfocharts.com/comparison/bitcoin-transactionfees.html
Says the fees are $12.56 for just storing 256 bytes of data and it's not my fault that 20,000 fulls nodes
repeat the same work because I did not ask for them all to be here and 1000 would do the trick

Sorry, but this makes me laugh. You obv. can't accept the fact that Bitcoin fees are actually pretty low these days.

[DexterRock]

Today I set a transaction with the lowest fee and I was warned with a notice that with such a fee, the transaction could take days to be confirmed. I confirmed it because I wasn't in a hurry at all. That was about a couple of hours ago and it already has 8 confirmations.

I've checked mempool's unconfirmed transactions and there are about 72.000 whereas that figure hasn't been below 200.000 for quite a while.

Any specific reason for that? Has the price anything to do with it?

When BTC hit the news in December and hit heights of 19000USD many people jumped in trying to get a piece of the pie, created accounts on exchanges etc etc and new users = new wallets = more transactions.
BTC transaction fees went sky high as a result, high fees caused people to reconsider making small transactions , less small transactions means lower fees. BTC price dropped, interest dropped off, number of small transactions decreased = back to normal.

[reliable]

Lucky for you guys because I didn't experience this yesterday when I sent some of bitcoin into my other wallet to cash them out but until now the transaction has not yet been confirmed, but I set at a low priority with a low transaction fee amount it sucks maybe because of the wallet I have used.

The unconfirmed transaction has being going down which is a good sign and due to this even the network fees what block chain use to charge before and now has a drastic change. May be from next week if it continues this way even with the low fees our transaction should get confirmed in a day or 2 max which will help us to save huge fees while transferring .

[Carlton Banks]

the network fees what block chain use to charge before and now has a drastic change.

The network isn't setting the price of fees, it's users are.

This is either the result of people using badly written fee estimator websites, or bad wallet software estimates, or even just users blindly picking a fee without knowing or caring how low fees are (people are still choosing 200 satoshis/byte and even 1000 satoshis/byte in the last week, when they could've paid from 10-50 various points throughout the week and gotten the same speed of confirmation)

https://jochen-hoenicke.de/queue is the best website to use, the other popular websites are misleading or poorly laid out.

[Samarkand]

the network fees what block chain use to charge before and now has a drastic change.

The network isn't setting the price of fees, it's users are.

This is either the result of people using badly written fee estimator websites, or bad wallet software estimates, or even just users blindly picking a fee without knowing or caring how low fees are (people are still choosing 200 satoshis/byte and even 1000 satoshis/byte in the last week, when they could've paid from 10-50 various points throughout the week and gotten the same speed of confirmation)

https://jochen-hoenicke.de/queue is the best website to use, the other popular websites are misleading or poorly laid out.

Another reason is that some online services/exchanges are poorly coded as well. E.g. some major
exchanges don´t even batch transactions and use the same fee irrespective of changes
in the mempool size.

It is unbelievable how some businesses that have a trading volume in the billions are getting
away with stuff like this, which perpetuates the fee problem unnecessarily.

If I were running a business like Coinbase I´d use a period of low fees to consolidate
inputs that are otherwise unspendable in an effort to increase the bottom line of my business
as well as the overall health of the Bitcoin network.

[hugeblack]

the network fees what block chain use to charge before and now has a drastic change.

The network isn't setting the price of fees, it's users are.

This is either the result of people using badly written fee estimator websites, or bad wallet software estimates, or even just users blindly picking a fee without knowing or caring how low fees are (people are still choosing 200 satoshis/byte and even 1000 satoshis/byte in the last week, when they could've paid from 10-50 various points throughout the week and gotten the same speed of confirmation)

https://jochen-hoenicke.de/queue is the best website to use, the other popular websites are misleading or poorly laid out.

can you explain why this site https://jochen-hoenicke.de/queue is better than https://bitcoinfees.earn.com/ "which 90% of users recommended it"? why 230 satoshis/byte isnot cheapest fee?
also, give us small tips how to us this site https://jochen-hoenicke.de/queue?(i checked it and only find Graphic forms).
General question:
-best way to calculate the fee for segwit transctions?

[hatshepsut93]

Today I set a transaction with the lowest fee and I was warned with a notice that with such a fee, the transaction could take days to be confirmed. I confirmed it because I wasn't in a hurry at all. That was about a couple of hours ago and it already has 8 confirmations.

I've checked mempool's unconfirmed transactions and there are about 72.000 whereas that figure hasn't been below 200.000 for quite a while.

Any specific reason for that? Has the price anything to do with it?

Today, the fees in satoshi/byte are as low as they were in November, but the demand for Bitcoin transactions is actually bigger, because it's main use-case - trading, has more volume now than in November, so this is a bit strange, although the price is also higher, so it can partially explain this. I suspect two things happened - the very expensive spam attack has finally stopped and exchanges have started focusing more on their fee estimation algorithms and batching (SegWit adoption is still low).

[ruletheworld]

Today I set a transaction with the lowest fee and I was warned with a notice that with such a fee, the transaction could take days to be confirmed. I confirmed it because I wasn't in a hurry at all. That was about a couple of hours ago and it already has 8 confirmations.

I've checked mempool's unconfirmed transactions and there are about 72.000 whereas that figure hasn't been below 200.000 for quite a while.

Any specific reason for that? Has the price anything to do with it?

Today, the fees in satoshi/byte are as low as they were in November, but the demand for Bitcoin transactions is actually bigger, because it's main use-case - trading, has more volume now than in November, so this is a bit strange, although the price is also higher, so it can partially explain this. I suspect two things happened - the very expensive spam attack has finally stopped and exchanges have started focusing more on their fee estimation algorithms and batching (SegWit adoption is still low).

Just seems like a normal decline in number of transactions on-chain as the price has fallen. Better fee estimation doesn't affect the number of transactions on-chain, although better batching definitely does. I do not know of any major exchange that went from non-batching to batching recently though. Coinbase - I am still looking at you.

SegWit adoption still seems low. However, exchanges need to remember that the best time to do this type of migration is when the markets are calm. The number of on-chain transactions are low, the fees are low. It is safer to do things now, and when the next run up in price and transaction volume happens, they can be prepared with better systems in place. Coinbase - I am still looking at you.

[weeman]

Is it likely that the drop in fees is because miner spammers are thinking twice about hurting the bitcoin name now that the entire crypto market is at a low point?

When BTC is at a high, it would be tempting to benefit from hurting its name. But now that everyone is hurting, it makes sense to ease off, rather than kill the golden goose.

[jnano]

can you explain why this site https://jochen-hoenicke.de/queue is better than https://bitcoinfees.earn.com/ "which 90% of users recommended it"?

Because you can see the exact (more or less) fee distribution of the transactions currently waiting for confirmation, and you can see both the near and far historic trends. You can see if transactions in a certain fee range were picked up during the last hour, day, or week. You can figure out the overall trajectory to know whether the fees condition is flat, getting better, or getting worse. You can know if the current congestion is a temporary thing that just started and might disappear in a day, or if it's been going on for days or weeks.

If I were running a business like Coinbase I´d use a period of low fees to consolidate
inputs that are otherwise unspendable in an effort to increase the bottom line of my business
as well as the overall health of the Bitcoin network.

Coinbase are indeed amazing. Non-batched transactions, unneeded crazy high fees that just inflate the mempool's going rates because everyone else has to compete with them, and no option to select lower fees in case you don't need the quickest confirmation. Right now, when 10 sat/b will 99% get you in the next block, they charge approximately 200 sat/b. And why should they care? They pass on the fees to their captive users.

Do other web wallets and exchanges act the same? I think Coinbase might be the largest player, but if others are acting similarly I do wonder if these companies are the main reason for the very high fees whenever there's congestion.

Anyway, in an update to my earlier post above it looks like the difficulty retarget hasn't stopped the mempool clearing out. Currently the 8 and 9 satoshi/byte tx are going through.

The hash rate is still higher than the difficulty, which is good in such times, I guess.

[codewench]

give us small tips how to us this site https://jochen-hoenicke.de/queue?(i checked it and only find Graphic forms).

Look at the top sawtooth edge of the graphs. As transactions accumulate, the graph climbs. When a block is mined, that block's transactions are removed. This causes the graph to suddenly drop.* You'll notice that usually the sawteeth are uniform - climbing, drop back, climbing, and drop back. Sometimes the jaggy keeps climbing a while before dropping. This happens when a block is late. Other times there are two or more drops in quick succession. This means that blocks were mined (much) faster than 10 minutes apart.

* The software is obviously sampling at some interval detached from the block mining times. Some time after a block is mined, the charting software merely captures smaller numbers to graph. It then draws a diagonal line between the previous (higher) sample and the new (lower) sample. In reality, it should drop straight down.

Look at the bottom of a jaggy. Notice what color is found there. Look up that color in the legend. This tells you that transactions at that fee level were mined in that block. Look at the colors higher in the jaggy. Convert them to fee levels. That means some transactions paid that much to get in the block.

If you are looking to save on fees, look for the lowest bottom of a jaggy within the timescale of when you want the transaction to confirm. Translate that color to the fee. Beware, the graph is of past data. You don't know what's coming in the future. New transactions may flood onto the network. Your cheap fees may run away from you. You may have to wait weeks before fee levels come into the range of your transaction.

If your transaction must be in the next block, choose a color band at the top edge of the jaggies.

If you'd like your transaction in the next block, but are OK with waiting a bit, pick a color band that ends in middle of a sawtooth drop.

The top graph is the number of transaction. This plots the data without giving emphasis to any particular transactions.

The bottom graph is the size of waiting transactions. When (lower) color bands in this graph are bigger than the top graph, you can infer that the cheap transactions are large, involving many inputs and/or outputs. Exchanges or merchants consolidating transactions will be here. This graph is useful to estimate how many blocks it would take to mine everything: Imagine taking 1MB slices off of this graph. How many would it take? Keep in mind that new transactions are always arriving.

The middle graph is the amount of fees pending. This isn't as useful. Sometimes there is a sudden bloom of a higher color band that's totally invisible in the top graph and slightly visible in the bottom graph. This means that someone flooded the network with large moderate fee transactions. This could be an exchange that needed to consolidate UTXOs *NOW*. The reason for the bloom is that this is a nonlinear transformation of the bottom graph - the higher fee levels get magnified.

Sometimes you'll notice the bottom band reducing in size. It is possible that a miner mined their friend's cheap transactions. However, in almost all cases this instead means that the mempool was full and this node discarded the cheapest transactions. You'll usually notice the bottom band reduce as a higher band increases.