Will small UTXOs become a problem in a mempool-collapsed future?
1898 sats \ 36 comments \ @pillar 30 Nov 2022 bitcoin
I would like to spark some discussion with the bright minds around here.
Today, sats are cheap and the mempool generally tends to be almost empty. Many people don't really even care about lightning because they only make infrequent on-chain transactions and usually handle amounts that are at least in the hundreds, so paying 2 cents here and 10 cents there is not an issue for them.
But if adoption goes parabolic, two changes can be expected:
  • Sats will become way more precious.
  • Blockspace will experience a dramatic surge in demand along with the increase in the user base.
In such an environment, making an on-chain transaction would have a dramatic cost in comparison to today, and the days when we happily sent sats on-chain to each other in a casual way could be over. On-chain would become the natural environment for large settlements between whale-ish parties, where the high but fixed cost of miner fees wouldn't be a problem. I like Lyn Alden's comparisons to Fedwire, the US banking settlement layer that has an average transaction size of $5,000,000 (see this and this long-form masterpiece by Lyn).
We don't know how long it will take to get there or how expensive on-chain will eventually become. But it makes me wonder: are small UTXOs a long-term problem because spending them will come with a massive overhead in terms of miner fees?
Imagine I have a UTXO with 100,000 sats. Today, I wouldn't pay much attention to it. Now let's go to this hypothetical future. The price of BTC is $1,000,000, so my 100K sats are worth a thousand bucks. But the mempool is always packed, so paying tens of thousands to get a transaction done in days is normal. Transferring these sats into another address means I need to spend a double-digit percentage in miner fees. Ouch.
Do you think we are heading into this reality? If so, how do you think UTXO management will change as time passes? Do you have a plan for your own stack to get ready for this situation?
Looking forward to your ideas.
write
preview
related posts
101 sats \ 13 replies \ @justsmile 30 Nov 2022
Zero Fee Routing on Lightning Network is the way.
It may take some time (for others to realize) It may be hard to find But eL eN route for free.
CLN:
--fee-base 0
--fee-per-satoshi 0
reply
348 sats \ 12 replies \ @pillar OP 30 Nov 2022
As much as Zero Fee Routing is nice, I don't know think it's sustainable long-term, and it doesn't really address the issue I'm talking about.
reply
1 sat \ 11 replies \ @selfish_gene 1 Dec 2022
проблема большой комиссии закончиться на 21 млн монет в мем пул...
reply
0 sats \ 10 replies \ @justsmile 1 Dec 2022
Exactly. It is sustainable. The units you route are worth more and more.
Or what problem in sustainability you see, @pillar?
reply
10 sats \ 9 replies \ @pillar OP 1 Dec 2022
I mean, if there is no incentive to become a routing node, there won't be enough capacity in the network. Thus, the need for fees. Which will probably be dramatically low since it will be an extremely competitive market and the cost of routing is very low, but definitely not zero.
reply
0 sats \ 8 replies \ @justsmile 1 Dec 2022
I see. And what's the incentive to run a Bitcoin full node (incl. pruned ones I mean)?
For example running a CLN node requires one to run a Bitcoin Core node talking to the network and knowing what is the bestblock. The incentive to not track any other but only the bestblock chain is that when you want to stack more (or do business with it) you need to be on the chain which is alive, which has the most mining power. There your stack just holds, but in time it becomes more and more valuable. What's the difference when LN node runs alongside Bitcoin Core?
Another example. What's the incentive to run a Tor node? When anyone runs a Tor node, they are routing onions for free.
reply
13 sats \ 1 reply \ @pillar OP 1 Dec 2022
The incentive to run a Bitcoin full node is enjoying complete privacy plus access to the Bitcoin network without censorship nor permission.
A Lightning node might be run simply because the node runner wants to interact with the Lightning Network. In that case, fees are irrelevant because it's not a Routing node. A Routing node needs significant amounts of capital and a way more active management than a node that simply exists because the operator wants to send and receive sats in the Lightning Network. Again, I doubt people will commit millions in $ value and all the human effort it takes to maintain a Routing node to altruistically route for the network.
As for the Tor network, the only incentive is having an interest in the network being alive. That could explain why there are so few exit nodes.
Do you run a Lightning Routing node?
reply
10 sats \ 0 replies \ @justsmile 1 Dec 2022
Thanks. I see your point better now and it fully makes sense. Reminds me of my historical chats with @ZeroFeeRouting.
Yes, I run a lightning node. On mainnet since 2020. And I run with zero fees. But as you correctly point out, I do not have much stake in it and even in times when I had ~10 channels, it was not very busy. CLN, always latest master branch. I do it mostly for CI-reasons to see it compiles and is in working condition (on musl-libc based system).
reply on another page
0 sats \ 5 replies \ @hades 1 Dec 2022
Running core and tor don’t involve a scarce asset. Spread the existing bitcoin across the entire planet and then have everyone open a lightning channel. If it was today, that’s 225,000 sats per person. Miners will be running mainly on fees in the future. I disagree with @pillar I think fees go up and LN moves to centralization. The average family will not have this many sats and with a planet running on bitcoin they certainly couldn’t afford opening up a channel each. We’d likely share channels or some larger entity with resources would let you pay to access.
reply
11 sats \ 4 replies \ @pillar OP 1 Dec 2022
The key is that the LN is an open market. The same way you don't need permission on-chain, you don't in the LN.
If LN gets centralized and the big nodes get greedy, their too-fat-margins become an incentive for other players to enter the network and set slightly lower fees. Competition is just brutal, since every node that wants to send a transaction is the most disloyal customer you can have: the minute the Routing node next to you has a fee 1ppm lower, routing will abandon your channel and move to his.
If you study perfect competition markets, you will realize that the routing services market is very close to a perfect one. Thus, we can predict that the average routing fees in the network will tend to match the cost of running a routing node (that's the human and equipment capital plus the opportunity cost of locking the sats in the node).
How high do you think fees will go? I have a hard time picturing anything above 0.01% long-term (10-15 years).
view replies
51 sats \ 2 replies \ @hades 1 Dec 2022
You bring up an excellent point. Let's take it to an extreme and then attempt to disprove it. Consider a future of hyperbitcoinization where Bitcoin price in dollars as we know them today is astronomical. Only large corporations or the elite could afford an on-chain transaction. You would not be able to open a lightning channel due to lack of mempool space and funds. Channels would all be owned by banking types (which are the grandchildren of hodlers today). Governments begin to legislate regulations regarding mempool priority. Governments may force miners to mine a certain percentage of underfunded txs in an effort to quell the satless. And the Governments could enforce it because they will be backed by the masses without Bitcoin demanding relief from mining farms and channel owners. Because no one fought for privacy improvements in the 2020s, employers pay wages in Bitcoin but to do so you must register your zpub with them so they can pay a new address each time and track their payouts for taxes. I suspect there is a point of expense where users would create fiat IOUs again or resort to trading keys instead of transactions. Maybe your small denomination addresses would be the best result because you wouldn't need as much change?
reply
2 sats \ 1 reply \ @Wil 1 Dec 2022
Everything tends towards centralization.
reply
0 sats \ 0 replies \ @justsmile 1 Dec 2022
Eh, was stacker.news stolen away by a group of trolls trying to prove their point by copy&pasting BS and posting it from different ("unique") accounts?
reply
8 sats \ 0 replies \ @snakedye 1 Dec 2022
Part of why the mempool is so "empty" is because of lightning. By the time we get to that point people making these 100k sats transactions will be on lightning and if they aren't they better get on it because they're losing money.
reply
5 sats \ 2 replies \ @01xCc 30 Nov 2022
заморозить все биткоины в транзакциях? Вы считали сколько нужно сделать транзакция когда предложение ограниченно? это максимум существует по причине ограниченного предложения монет. Да , в момент можно и платить 1 000 000 сат за транзакцию в 14 дней. но все равно это приведет что 90% будут ожидать очереди. и кто станет еще больше платить?
Давай развивать это вопрос в теории.
практика покажет что спам самый дорого!
но самое интересное это то, что . будет проще создать канал для обмена с другим соседом по торговле. все просто...
Биткоин = инструкция
reply
21 sats \ 1 reply \ @justsmile 1 Dec 2022
It's kind of funny you keep writing the way you do. Everyone else accepts the basic language of stacker.news is English. It does not help the discussion even though your points make sense, but many can not decypher them.
So no matter where I come from, to an English article I do my best to answer in English. Please try to do it yourself as well.
reply
0 sats \ 0 replies \ @01xCc 1 Dec 2022
не возможно... мой язык переносит более глубокий смысл слов в короткой фразе. я пытался перевести на английский, но обратный перевод на итальянский получается очень сломанным. поэтому приходиться писать как я это понимаю . переводчик не совсем идеален. поэтому думаю писать просто для себя. если кто то захочет узнать что то по подробнее придется обьяснять через переводчик.
А в целом я просто оставляю вопросы для себя , в будущем могут быть найдены ответы. Спасибо что увидел смысл в сказанном!
reply
11 sats \ 2 replies \ @Rock 30 Nov 2022
I think you make a good point. I think this would lead me to believe the best strategy is a combination of regular consolidation of UTXO's before the fees become unmanageable as well as maintaining more of your smaller balances on layer 2, like lightning channels.
I do think that the cost of maintaining lots of small UTXO's will continue to increase, so better to learn about UTXO management sooner than later.
reply
210 sats \ 0 replies \ @asialeone 1 Dec 2022 freebie
the trade - off small vs big UTXO's might no be so simple with a focus on coin control and privacy.
A little consolidation of 0.001 UTXO's: In an hyperbitcoinization scenario 0.005 BTC might be worth in FIAT ~ 5000$ so for every spend ( on average between 70 - 200 $) we'll get back big changes. Right now we can 1)mix those changes, 2)swap them or 3)keep reusing them linking tx by tx.
leaving aside 3); either mixing them again or swapping will involve higher fees, IMHO rather than having 0.001 BTC ready to be spent in a higher fees scenario.
ATM the best option might be a wide range of UTXO's with a thought at the supposed hyperbitconization price of BTC - FIAT denominated
reply
0 sats \ 0 replies \ @justsmile 1 Dec 2022
IMHO the minimal fees will be 1sat per transaction, i.e. the small TXs will need to "overpay" to get to 1 final fee when the -minrelaytxfee is the minimum — 0.00000001.
But everyone will know that a TX with minimal fee will not live in the mempool for long and will never get mined. That's a real market. When the mempool is full all the time and minimal fees are minimal (not like the current default of 1000 sat perkvB).
But that's just my opinion after thinking about it and running nodes and explorers for years.
reply
0 sats \ 0 replies \ @01xCc 30 Nov 2022
быстрее создадут 1000 000 000 каналов в сети молния. ИМХО
reply
2 sats \ 0 replies \ @TheBTCManual 1 Dec 2022
Personally, I am torn on this one, and we don't know how the market will re-act, will we see more people moving on to liquid just for the fee incentive, will wee see more lightning channels on Liquid, could state chains become a thing and then use that state chain UTXO instead? Will we see channel factories reducing the need to hit the chain to rebalance and create channels, all this could offset some of the need to touch the chain
reply
0 sats \ 9 replies \ @justsmile 1 Dec 2022
So the block 765446 was only 27% full. At the same time there are thousands lower-than-usual transactions waiting to be mined.
Miners, please connect your Bitcoin node to be.anyone.eu.org and get some. It's better to mine a block with some extra fees than to mine pretending there is nothing except the default 0.00001 -minrelaytxfee.
https://ln.uk.ms/bitcoin.conf.html
reply
0 sats \ 8 replies \ @justsmile 1 Dec 2022
@k00b it's quite fun that the comment itself can not be easily referred to with an exact link. Or can it? Maybe I'm just missing something.
reply
1 sat \ 7 replies \ @k00b 1 Dec 2022
#100994
How hard did you try? Maybe I can make it easier?
reply
198 sats \ 0 replies \ @justsmile 1 Dec 2022
The clear way would be to make left empty space of the comment change color on mouse-over and clickable, which would lead to the same place where now "\ … replies " leads to.
reply
0 sats \ 0 replies \ @justsmile 1 Dec 2022
The unbelievable thing is that when I click "… replies" I get to the comment itself. The more obvious (but also horrible) would be Twitter-way where one can click somewhere on the text of the Tweet/comment and get to its own link. But when there are more active elements, it may be hard to do it.
reply
0 sats \ 0 replies \ @justsmile 1 Dec 2022
I mean it is quite easy. I will do it now and will remember (and I did before). But when I came back after some time, it's totally not resembling anything else.
reply
0 sats \ 3 replies \ @justsmile 1 Dec 2022
Let's say I want to get a link to this comment of yours. I see three dots. When I click them it allows me to Flag, which is not what I want.
On the right I see something, reminds me of GitHub where I can copy a link to this comment. But when I click it, the comment just disappears.
In my notifications panel I go to parent of your comment and get link to #101003. Never used it before and it's quite hard for me to see there are more links in the notifications than when looking at the real thing.
Really no idea how to get to link #100994.
reply
0 sats \ 2 replies \ @justsmile 1 Dec 2022
Aaaah, have it now. I have to click "# replies". Oops. Totally.
reply
1 sat \ 1 reply \ @k00b 1 Dec 2022
It’s also available if you click on the time
reply
190 sats \ 0 replies \ @justsmile 1 Dec 2022
Would never think of that either. I agree it's just about what one is already used to. So feel free to ignore my ideas. Just rambling and shilling sats ;-)
reply