[Apr 2024] Fees are HIGH, wait for opportunity to Consolidate your small inputs

[jackg]

The transaction eventually confirmed in about 11 hours. So still within range of the day. I sent it in the morning of that day, and by the time I finished my dinner it already had 2 confirmations. 1 sat fee per byte. (0.00000113 BTC to be exact.)

I sent a transaction yesterday and it took less than an hour to confirm at 1 sat per byte (exactly).

[1714021010]

1714021010

[1714021010]

1714021010

[1714021010]

1714021010

[jackg]

For a fast confirmation, at the moment all you need is about 4 sats per byte for the fastest confirmation (10-20 minutes normally).

Consolidate now while you have the chance!

[LFC_Bitcoin]

1 sat/byte now the recommended fee

Hurry whilst it’s so low.

https://btc.com/stats/unconfirmed-tx

[Stedsm]

While the transaction I'm going to put here has paid around 23 SATs/byte, I still think I should bump this thread with the very best example of how crypto has changed the movement of money totally, and yeah, a lot of money actually.

https://www.blockchain.com/btc/tx/0f142e64e995a9dc2281d18061fefc7bfb813358840ff206846550be8ead210d

Around 125k BTC worth $1.1 billion were moved by Bitfinex from their wallet for just a measly 5578 Satoshis (just 48 cents at the time it was sent). Such big transactions being sent with least fee and still getting confirmed almost immediately is a win-win for this technology where we used to pay 100s of dollars to send money abroad. Hats off to crypto. Anyways, this event is a perfect bumping example for me.

[LoyceV]

Around 125k BTC worth $1.1 billion were moved by Bitfinex from their wallet for just a measly 5578 Satoshis (just 48 cents at the time it was sent).

I've seen similar cases before, and now too I don't think they moved 125 kBTC. I think they moved "only" 1.5kBTC, the rest was just "change". If this would be fiat, you wouldn't see the change that's left in the account, but because of the way Bitcoin works, you get to see it.
Think of it as if someone Scrooge McDuck moves a huge pile of money from one corner of his money bin to another corner, and also uses a few buckets of coins to make a payment.

[2double0]

Bumping the good work but want to add a caution that some fee value spikes are happening these days where I have seen some users paying 100 satoshis per byte to get their transactions win in the race of getting involved in a block. Does the low fee 'like 1-5 satoshis per byte' remain during some specific timings?

[LoyceV]

Does the low fee 'like 1-5 satoshis per byte' remain during some specific timings?

Your best odds for a low fee are on Sundays, but the point of this topic is to consolidate your inputs when you're not in a hurry and when you don't need your funds for a while. So if today's transaction confirms next Sunday, that shouldn't matter.
That being said: My Bitcoin Core estimates a minimum fee transaction to confirm in 4 to 8 hours.

[2double0]

Does the low fee 'like 1-5 satoshis per byte' remain during some specific timings?

Your best odds for a low fee are on Sundays, but the point of this topic is to consolidate your inputs when you're not in a hurry and when you don't need your funds for a while. So if today's transaction confirms next Sunday, that shouldn't matter.

I get that.
So basically, it is not a high or low fees thing and just an angle/ a way of looking at it as when you send a low fee transaction and unfortunately network clogs down, you will wait while higher fee payers will enjoy their transactions being confirmed earlier than yours and if the fee drops immediately after some transactions and you just wait till that time for that fee drop to come, then I think there is no need to send a transaction when fees are high.

That being said: My Bitcoin Core estimates a minimum fee transaction to confirm in 4 to 8 hours.

And what happens if the prediction fails due to a mempool clog?
I will not bother to see it for 24 hours too but will you take it for 2 days after trusting the wallet's fee prediction? We have seen that happening in the past days already.

[LoyceV]

And what happens if the prediction fails due to a mempool clog?

Then it can take weeks, you can drop the transaction, use CPFP, or use RBF.

This will probably happen at some point again during the next bull run, and that's all the more reason to consolidate your small inputs before that happens. Which reminds me: I haven't done that in a while.

[Artemis3]

And what happens if the prediction fails due to a mempool clog?

Then it can take weeks, you can drop the transaction, use CPFP, or use RBF.

This will probably happen at some point again during the next bull run, and that's all the more reason to consolidate your small inputs before that happens. Which reminds me: I haven't done that in a while.

I always use 1 sat/B and this is never a problem. Sure, have RBF around just in case.

Why do you even let wallets decide? Always use 1 sat/B. Does it make a difference if it takes 6 hours, 11 hours or 20 hours?

As far as i know, the only tx that are "dropped" are those using 0 fee. 1 is safe, no matter how long it takes. Yes i know some nodes may drop them, but others will retain it and they won't ever leave the network until processed.

I love how Electrum shows the "size" of (mempool?) transactions to go before an unconfirmed gets confirmed, the number is quite useful to give you a rough idea how the traffic currently is.

Actually, the fact that the prediction can fail (a sudden traffic spike right after you broadcast) is even more reason to NOT let wallets decide the fee.

BTW: The only time i had a transaction take "weeks", was on Jan 2018, i have never seen that again. In the last two years, my 1 sat/B tx would took a day or less, sometimes even 1 hour.

"Fees are low" is misleading, the fee is not arbitrarily decided by Bitcoin. You the user decides the fee. Wallets are prone to guess wrong (too late) or be victim of a sort of attack where a large traffic spike in turn makes wallet use higher fees which in turn makes the other wallets rise the fees which in turn...

Anyway, don't let the wallets decide. 1 sat/B is good for everything, unless you have an emergency. Then "2" would probably make it in half an hour...

I'm still waiting for when they let us use 0.1 sat/B...

[Dabs]

I think anything at or above the minimum relay fee will work. 0.1 sat/B isn't going to work since it won't get relayed at all. You can try to experiment and see how that would work by setting the fee manually.

[jackg]

I think anything at or above the minimum relay fee will work. 0.1 sat/B isn't going to work since it won't get relayed at all. You can try to experiment and see how that would work by setting the fee manually.

There was "speculation" surrounding a reduction in minrelayfee in a new release of bitcoin xcore down to 0.1 sat per byte but alas it hasn't happened yet and I'm not sure they're planning to afaict.

[Carlton Banks]

I think anything at or above the minimum relay fee will work. 0.1 sat/B isn't going to work since it won't get relayed at all. You can try to experiment and see how that would work by setting the fee manually.

There was "speculation" surrounding a reduction in minrelayfee in a new release of bitcoin xcore down to 0.1 sat per byte but alas it hasn't happened yet and I'm not sure they're planning to afaict.

the main developers discussed it, I think either on IRC or the bitcoin-dev mailing list. All agreed (or at least, noone present back then disagreed) that reducing the lower bound for minrelayfee was a good move, and 0.1 sats/byte was suggested as the new minimum. The code governing fee estimation needed changing to accommodate that extra order of magnitude in granularity, and although the work to do so was started, it was never finished (not sure why). There's a link to the pull request in the Bitcoin github repo... I think further back in this thread (in one of my posts)

[Dabs]

In that case, it's just a matter of time then when the relay fee gets reduced, and then, that would be our new "lowest" fee for transactions.

When that happens, there will be an eventual cascade effect and fees across the board will go lower, and while most people will need to wait 10 or 20 minutes (or a few hours), low fees would not be an issue as compared to the other "direct" bitcoin forks. They'll need to use something else to debate or argue or whatever.

[jackg]

the main developers discussed it, I think either on IRC or the bitcoin-dev mailing list. All agreed (or at least, noone present back then disagreed) that reducing the lower bound for minrelayfee was a good move, and 0.1 sats/byte was suggested as the new minimum. The code governing fee estimation needed changing to accommodate that extra order of magnitude in granularity, and although the work to do so was started, it was never finished (not sure why). There's a link to the pull request in the Bitcoin github repo... I think further back in this thread (in one of my posts)

It might have been a lot for them to try to edit, I'll have a look above as I've had to start with a bit of c/c++ so know a very limited amount of it... If they can't do it I probably can't but it might be worth looking at.

I remember the sat was the lowest unit Satoshi wrote and it was in a uint64 and care would have to be taken as I think that was near the max for the 21 million. 2^64 is around 8x10^14 so it's very close to the limit afaik (I might be out by a factor of 10).

[pooya87]

It might have been a lot for them to try to edit, I'll have a look above as I've had to start with a bit of c/c++ so know a very limited amount of it... If they can't do it I probably can't but it might be worth looking at.

I remember the sat was the lowest unit Satoshi wrote and it was in a uint64 and care would have to be taken as I think that was near the max for the 21 million. 2^64 is around 8x10^14 so it's very close to the limit afaik (I might be out by a factor of 10).

it would be a lot if they were trying to change the total supply (aka consensus rule) not a very optional local rule that is computing fee/byte. it is a simple matter of dividing the amount of fee (UInt64) by the transaction's raw byte size (Int32) and returning the result for comparison. there is no overflow in division.
21,000,000.00000000 = 0x000775f05a074000

[Dabs]

I did another transaction over the weekend, my not-so-good estimator said 2 sats was good for 1 hour, so I purposely made the transaction pay only 1 sat and I thought I'd wait a few hours. I think it confirmed under 1 hour anyway.

[jackg]

I did another transaction over the weekend, my not-so-good estimator said 2 sats was good for 1 hour, so I purposely made the transaction pay only 1 sat and I thought I'd wait a few hours. I think it confirmed under 1 hour anyway.

Was this Friday or Saturday? I think we saw volatility then so we must've shaken out the weak hands from the market already (fees normally zoom when it fluctuates)... I think I used electrum's prediction of within 25 blocks and saw a confirmation after five minutes (I needed to send a fast one) so it's definitely still better to use a flat fee when you can.

[DireWolfM14]

I did made a transaction this morning, Electrum's suggested minimal fee for a confirmation within 25 blocks was over 20 sats/bite.  Bitcoin was fluctuating up and down by a few hundred USD in short periods of time. 

Note to self; don't consolidate dust when the main-stream-media is in Chicken-Little mode over a SARS pandemic.

[LoyceV]

Note to self; don't consolidate dust when the main-stream-media is in Chicken-Little mode over a SARS pandemic.

Bitcoin Core now estimates a 1 sat/byte transaction to confirm within 24 hours. That's longer than I've seen in months. Panic on the markets caused it, but this is exactly the reason why you should consolidate small inputs when you don't quickly need them.