[OUTDATED] The Lightning Network FAQ

[d5000]

Haha, don't give an inch to their narrative. I wouldn't give the fudsters any encouragement by saying "they share some properties with IOUs". Because you're not sending anything worthless - IOUs in Lightning.

If a discussion should have any value and not be only a game of propaganda and counter-propaganda, one should not try to hide disadvantages and tradeoffs. The main property that LN channel balances are sharing with IOUs is obviously that it is possible that one party runs away with the money of the other one (if the other isn't watching his/her channel and settling), because of a disagreement about "how much one owes to the other".

But that isn't of much relevance, because if you're watching your channels regularly, in the case this conflict happens, you can always call an extremely reliable arbitrator (the Bitcoin blockchain).

So what I'm saying basically is that "even if LN shares properties with IOUs, that is no point for FUDsters." So the complete argumentation of these people falls to pieces.

[Wind_FURY]

Haha, don't give an inch to their narrative. I wouldn't give the fudsters any encouragement by saying "they share some properties with IOUs". Because you're not sending anything worthless - IOUs in Lightning.

If a discussion should have any value and not be only a game of propaganda and counter-propaganda, one should not try to hide disadvantages and tradeoffs. The main property that LN channel balances are sharing with IOUs is obviously that it is possible that one party runs away with the money of the other one (if the other isn't watching his/her channel and settling), because of a disagreement about "how much one owes to the other".

Irrelevant. But that property also shares it with real Bitcoins sent to a special multi-sig wallet, that was confirmed on-chain, to fund the channel.

[d5000]

Irrelevant.

But that isn't of much relevance [...]

I see our opinions on this topic are not too far away one from another

But ...

But that property also shares it with real Bitcoins sent to a special multi-sig wallet, that was confirmed on-chain, to fund the channel.

... what is the difference between the scenario you describe and LN? The disagreement I wrote about isn't about the BTC  "parked" in the transaction, but about the channel state (the transactions that are never broadcasted). Dunno what you mean here.

[franky1]

windfury. can i just give you some advice
try not to get involved in LN discussions until you understand it and atleast used it.

for pages AND topics now you have shown lack of experience of using it and also lack of awareness of its function, and lack of research of how it works/how its coded.

i understand you want to promote it, but please just spend some time understanding what your promoting and stop trying to be a snake oil salesmen for something you dont fully know.

trying to be the utopian positive salesmen is unhelpful, especially an inexperienced salesman who's making it obvious you not only dont know, but dont wish to know.
so if you do want to positively promote LN IN THE DISTANT FUTURE atleast spend some time now actually learning, researching and experiencing it

its getting ridiculous that you spend pages saying 'there is no iou or pegs' but you have never proved your point nor shown any desire to prove it. you just want to rubber stamp your repetitive misunderstanding, which is becoming more suspect of doing so just to 'post count' some signature/footnote campaign income

[franky1]

Irrelevant.

anyway. the whole point of bitcoin is once confirmed the payment is done, irreversible, fully secure, no counterparty needed, no trust. its just done,

any stuff before a confirm is just blah. insecure untrusted not guaranteed, not complete, irrelevant

LN is IOU, reversible, insecure, requires others permission, no guarantee you get to keep what was suggested to be owed to you.

LN payments are not bitcoin payments from the point of view of:
actually owning the funds complete and in the clear(irreversible)
unit of account, LN=msats bitcoin=sats
bitcoins are push payments 24/7. LN are agreements when others are available
LN units of account are not even protocol locked tethered to bitcoin
LN units of account are not even forced to be collateral of confirmed bitcoin(turbo zero confirm)
LN units of account are not even recognised on the bitcoin network

LN is not a feature solely and exclusively for bitcoin. it is not a 'layer' it is a separate network

[LoyceV]

anyway. the whole point of bitcoin is once confirmed the payment is done, irreversible, fully secure, no counterparty needed, no trust. its just done,

None of that changes if people use the Lightning Network. I don't really get why you're so against it. Who cares if people use something they like?

unit of account, LN=msats bitcoin=sats

I've seen this argument before, and I don't know how LN handles the rounding on settlement, but honestly I couldn't care less about possibly losing 0.999 milli-satoshi.

I've played with it, and I like it! I know it's immature, experimental or even high risk to use, but considering the possibilities, I'm totally fine with risking a few bucks on it.

[Wind_FURY]

windfury. can i just give you some advice
try not to get involved in LN discussions until you understand it and atleast used it.

I will stop to get involved in Lightning discussions, if you stop with your misinformation rampage about the Lightning Network. Plus for the benefit of the people discussing in the topic, do you, or anyone else believe this person is correct in his opinion of the Lightning Network?

https://twitter.com/rogerkver/status/1135567225102905345

The #LightningNetwork requires trust for 99.99% of users because they are using custodial non-wallets similar to PayPal.

[LoyceV]

Plus for the benefit of the people discussing in the topic, do you, or anyone else believe this person is correct in his opinion of the Lightning Network?
https://twitter.com/rogerkver/status/1135567225102905345

The #LightningNetwork requires trust for 99.99% of users because they are using custodial non-wallets similar to PayPal.

Even if that would be true, I'd be totally fine with a trusted system for small Bitcoin payments. Exchanges are trusted with billions worth of crypto, I'd love to see them develop a standard to make small, low-fee transactions to other users.

[user512]

Hello,

may I ask you for some logical question about Lightning Network?

I use the channel explanation on this site:
https://bitcoinmagazine.com/articles/understanding-the-lightning-network-part-creating-the-network-1465326903/ (https://bitcoinmagazine.com/articles/understanding-the-lightning-network-part-creating-the-network-1465326903/)

Main question: But this seems that middlemen Bob has to broadcast the transaction which gives to him the money on chain if he provides the value otherwise the CLTV on the third output option of commitment transaction ensures that after a certain time Alice receives the bitcoin and cancels all this. That can't be understand well, because this means far more than only maximum two on-chain transactions per channel.

Q2: How is ensured the autenticity of inner communication in the LN? It seems like all data pushed through the channel must be signed to be verify by the counterparty.

Q3: How the nodes start multihub transfer? How can be Alice sure that the pre-image which she obtained is really from Carol (in LN example scheme)?

Q4: What is the main difference between CLTV and CLV timelocks?

Q5: Do Channel Factories make sense in the future or it's only another optimalization resolve?

Q6: Please, specify, where exactly the Lightning layer is (internet, blockchain etc.) and how exactly the nodes communicate (and in which layer, too).

Q7: Do you think that understanding LN is much harder than understanding Bitcoin?

[WhyFhy]

anyway. the whole point of bitcoin is once confirmed the payment is done, irreversible, fully secure, no counterparty needed, no trust. its just done,

None of that changes if people use the Lightning Network. I don't really get why you're so against it. Who cares if people use something they like?

unit of account, LN=msats bitcoin=sats

I've seen this argument before, and I don't know how LN handles the rounding on settlement, but honestly I couldn't care less about possibly losing 0.999 milli-satoshi.

I've played with it, and I like it! I know it's immature, experimental or even high risk to use, but considering the possibilities, I'm totally fine with risking a few bucks on it.

That's the spirit! https://discord.gg/YCqQZqq <-LN Discord, Full of devs if you have any problems its pretty much so an LN ecosphere.

[franky1]

windfury. can i just give you some advice
try not to get involved in LN discussions until you understand it and atleast used it.

I will stop to get involved in Lightning discussions, if you stop with your misinformation rampage about the Lightning Network.

not misinformation, just not positive unicorn rainbow dreams and empty promises of LN PR

Plus for the benefit of the people discussing in the topic, do you, or anyone else believe this person is correct in his opinion of the Lightning Network?

https://twitter.com/rogerkver/status/1135567225102905345

The #LightningNetwork requires trust for 99.99% of users because they are using custodial non-wallets similar to PayPal.

now your derailing into social drama. funny how you go full on time waste to try to drag in the bitcoin cash derailments but dont spend time learning LN

but anyways. the whole "what is a custodian" is the main point
custodians do not have to be the single owner/holder/manager.
they can just be a key holder with partial responsibility to maintain something.

look at schools, residential buildings, even police evidence.
a custodian does not need to fully own(building lease/school head) they can be the janitor or officer handed keys and given some privelidges/authority of what they are allowed to handle.

so yes LN is custodial as its counter party. you and another party are custodians
(i know you previously tried to think custodians had to be a centralised powerhouse institution. but thats not the case)

[LoyceV]

To get back on topic, I'll share more of my own experience for the FAQ.

First, Eclair syncs very fast since yesterday's update! The sync time was annoying earlier, so I like it.

What happens if a funding transaction doesn't confirm in time?
I opened a channel with a low fee, which didn't confirm within a day. This is the current state:

• Status: "CLOSING (uncooperative)"
• Closing type: "LOCAL (initiated by your node)"
  
• Cause of closing: "sync error"
• "Refund in 58 blocks"

Until now, it wasn't clear to me how this would work, and it's still unclear to me how the transaction fee is determined.
The closing fee (0.2+ mBTC) was high enough to almost instantly confirm my funding fee (CPFP).
This fee moved my funds from one Bech32 address to another, and I assume it will move to my own address once it closes. That means another fee, which makes it close to 20% of the total channel balance. At this moment it costs me just over $3 to close the channel (even though I haven't been able to use it at all). Bitcoin fees have been more than 10 times higher than they are now, which would put the total fee to just close a channel around 5 mBTC ($100 when Bitcoin was at it's peak).
Channels with a low balance (the minimum is 1 mBTC) won't even hold enough funds to pay a high fee. That means the channel can't be closed, and whoever has funds in them more or less loses them.

Those high fees will be a serious limitation even for a second layer channel. Instead of closing a channel, it now becomes much more interesting to send the entire balance to a LN-service in exchange for an altcoin. But I'm not sure if that's possible, until now the channel doesn't allow me to send my full balance. It varies a bit, so I'm not sure what the minimum reserve is.

Another disappointment is that I have to wait for the channel to close. If there have been no transactions at all on the channel, I'd like to see it instantly returned to my own wallet. Or even better: if the funding transaction doens't get confirmed, it would be much better to just let it drop out off mempool and never open the channel.
I'm now at the rediculous situation that my transaction was only confirmed because the closing transaction paid for it through CPFP.

[d5000]

Those high fees will be a serious limitation even for a second layer channel.

Thank you for your report. I fully agree, the fees are one of the main limitations of LN.

In the current situation, however, it would be enough if a higher percentage of all low-value transactions moved to LN. I've just looked ad Bitinfocharts: the median transaction value is currently at 0.053 BTC ($406.46 USD). If only half of the transactions with a lower value (roughly, these lower than 100 USD, I guess) could move to LN then we would have much less congestion and much lower fees - basically, we would have the same fee level at weekdays that we now have at weekends. Then you could open and close channels for less than 1 USD.

The problem is that even a growth in 25% of transaction blockchain space usage (in kB) or "adoption" in other words would bring us again to current fee levels of single-digit dollar values.

Channel factories could surely help, but I think they'll be not enough.  That's also why I'm also interested in side-chain and pegged-chain technologies. Those could complement LN and make really massive scaling possible.

[Wind_FURY]

windfury. can i just give you some advice
try not to get involved in LN discussions until you understand it and atleast used it.

I will stop to get involved in Lightning discussions, if you stop with your misinformation rampage about the Lightning Network.

not misinformation, just not positive unicorn rainbow dreams and empty promises of LN PR

Yes they are. Read your posts.

Plus for the benefit of the people discussing in the topic, do you, or anyone else believe this person is correct in his opinion of the Lightning Network?

https://twitter.com/rogerkver/status/1135567225102905345

The #LightningNetwork requires trust for 99.99% of users because they are using custodial non-wallets similar to PayPal.

Words

Stop meandering, and making confusing posts that doesn't answer anything. You're not helping the topic. I want an explanation why everyone on Twitter replied that Roger Ver is wrong. Do you agree with "they are using non-wallets similar to Paypal", or not? Why?

[LoyceV]

Until now, it wasn't clear to me how this would work, and it's still unclear to me how the transaction fee is determined.
The closing fee (0.2+ mBTC) was high enough to almost instantly confirm my funding fee (CPFP).
This fee moved my funds from one Bech32 address to another, and I assume it will move to my own address once it closes.

The funds were indeed sent to my own address with a very high fee per byte (total 0.18 mBTC, slightly lower than the previous fee). After this, I consolidated my funds with a 50 times lower fee, and this took 100 minutes to confirm. I can't help but feel like LN is wasting a rediculous amount on fees. I assume it's set by the node, but I'm not sure what it's based on.

Eclair doesn't allow coin control. Instead, it uses all available inputs (including dust) and consolidates them. That also leads to higher fees than needed. You can reduce the total fees by consolidating inputs with a low fee before funding a channel.

I'm funding another channel now, with a bit higher fee this time (currently 9 times more than what Bitcoin Core thinks is needed to confirm within 2 hours).

https://1ml.com gives an extensive overview of LN nodes, including an overview of the base fees. I'm not sure what everthing means though, there's a lot to take in for new LN users.
I'm now bookmarking the nodes on my laptop, so I can find them back on Android.

What I really like about LN, is that many inputs no longer lead to a higher fee. "Bitcoin dust" isn't a problem anymore as long as it's within LN.
LN doesn't allow to send a payment without making a payment request, but I read this might be included in a future upgrade. I can imagine this will be great for advertising: send a fraction of a satoshi to all channels you can find, and add a message. Without some sort of spam protection this might become a problem in the future.

[WhyFhy]

Hey LoyceV , as far as swapping to alts there's zigzag.io  Dash,LTC and ETH are on the list
Fee's are high though.

I heard submarineswaps is doing LTC I have not used it yet though

[LoyceV]

Hey LoyceV , as far as swapping to alts there's zigzag.io  Dash,LTC and ETH are on the list
Fee's are high though.

I haven't tried zigzag yet. I did try coinplaza.it, which has surprisingly low fees for a small transaction (even to onchain Bitcoin).
There's also fixedfloat.com, which I haven't tried yet.

Now consider LN Hubs will be processing offchain transactions in microseconds ,
So LN hubs hardware requirements/cost will grow at a much higher rate, especially the Main Hubs.
Unlike Bitcoin, that only charges to include a transaction in a block,
LN will impose variable fees on you for each hub needed to complete a LN transaction plus the 2 onchain bitcoin fees.

So how in your mind do you see LN fees staying cheap,
when there hardware requirements/cost for nodes are greater, and they have the ability to set their own LN fees.  

I don't worry about this for the following reasons:
Bitcoin fees spiraled up because users have no choice. Miners only compete against each other for blocks, but that doesn't increase block space.
In LN, there will be many different routes, and hub owners compete against each other on fees. If fees are high and hub owners make large profits, new actors will join the hub-market until fees are lower again.
More transactions obviously increase hardware requirements, but more transactions also means there are more fees to be earned. At the moment I don't think hub owners make a profit from the few small fees. They actually need much more transactions to earn money. If there would be a million times more transactions, I don't think the hardware cost per transaction will be very high.

Update on my little experiment: now that I have a second channel, a 5 sat transaction takes only 5 msat fee, instead of 1 sat. That makes sense, as the other channel has a lower fee.
However, a 100 sat transaction now takes over 3 sat fee! It uses the new channel, while the other channel would have charged only 1 sat. So I guess routing still needs some work.
For small transactions, 3% is not a problem. However, if I would send (say) 0.1 BTC, I would like to see the fee upfront if it's going to be around 3%. I'm not sure if LN even supports that before completing the channel.

Another update: Despite having 2.5 mBTC "Total can send", I can't send 2 mBTC ("Your Lightning channels do not have enough balance"). When I try to send 1 mBTC, I get: "Payment failed, Route not found". After trying a few times, Coinplaza banned my IP for a few minutes.
I've read before that "larger" payments often can't be routed, now I found out by myself

I made several smaller payments to my account at tippin.me, which all worked. The total was >1 mBTC now. But: fees were just over 3%, which is the maximum Eclair allows (and the only thing about it in settings). I've recently made cheaper on-chain payments!
From tippin.me, I sent 101010.10 sat to coinplaza.it. This got me 1 mBTC on-chain to my Bitcoin address. And although 1 sat/byte is enough, coinplaza.it paid 128 sat/byte, which is about 31 times more than the margin they had on this transaction.

I pay much less fee through tippin.me than through the last channel I opened in Eclair. I'll try opening a channel to tippin.me's node next time.

After many transactions, Eclair doesn't show my full Payments history anymore. I can only see the last 5 days.
So far, I've lost less than 1 mBTC in total on fees. Not bad for a week and a half of LN-experimenting.

I think I picked a LN node without enough cheap connections.
I'm closing the "expensive" channel now, see how that works. It says "CLOSING  (cooperative"). It should be able to do this at very low fee. It looks like my 1.42 mBTC LN-balance turned into 1.32 mBTC on-chain. The total fee wasn't that high, but still 10-20 times higher than needed for a fast confirmation.

Question
How do I choose which node to open a channel to?
And: I expected payments to use both channels when possible, but until now that hasn't happened.

[franky1]

Here is a thought experiment for you.
Bitcoin Devs refused to increase blocksize because of the excuse non-mining nodes would fail to keep up,
even thru block time is only 10 minutes.

love it. i can just imagine that conversation

bitcoin dev salaried by corporation:
     'bitcoin cant scale because you cant receive, validate and broadcast 3000 tx in 10 minutes'
bitcoin dev salaried by corporation:
     'LN is where wveryone should be because you cant recieve, validate and broadcast thousands/unlimited tx in seconds'
community:
     'why cant bitcoin?'
bitcoin dev salaried by corporation:
     'coz hardware'
community:(in sarcastic tone)
     'ohhh the same hardware LN needs to receive, validate, sign, revoke old sig and broadcast thousands tx in seconds'

[Wind_FURY]

Until now, it wasn't clear to me how this would work, and it's still unclear to me how the transaction fee is determined.
The closing fee (0.2+ mBTC) was high enough to almost instantly confirm my funding fee (CPFP).
This fee moved my funds from one Bech32 address to another, and I assume it will move to my own address once it closes.

The funds were indeed sent to my own address with a very high fee per byte (total 0.18 mBTC, slightly lower than the previous fee). After this, I consolidated my funds with a 50 times lower fee, and this took 100 minutes to confirm. I can't help but feel like LN is wasting a ridiculous amount on fees. I assume it's set by the node, but I'm not sure what it's based on.

Here is a thought experiment for you.
Bitcoin Devs refused to increase blocksize because of the excuse non-mining nodes would fail to keep up,
even thru block time is only 10 minutes.

There are externality costs. Increase the block size, and the costs are transferred, and subsidized somewhere else. It would also sever the network of full nodes, centralizing it futher. It's best to overshoot security in my opinion.

Plus it would set a bad precedent. The best example, Bitcoin Cash ABC and SV.

Now consider LN Hubs will be processing offchain transactions in microseconds ,
So LN hubs hardware requirements/cost will grow at a much higher rate, especially the Main Hubs.
Unlike Bitcoin, that only charges to include a transaction in a block,
LN will impose variable fees on you for each hub needed to complete a LN transaction plus the 2 onchain bitcoin fees.

So how in your mind do you see LN fees staying cheap,
when there hardware requirements/cost for nodes are greater, and they have the ability to set their own LN fees.  

Enjoy the answer to that, as you follow it thru it's natural evolution.  

It's a theory at best, and I agree that the Lightning Network can fail, the same as Bitcoin as a software experiment can also fail, but why stop its development if it doesn't affect, or make contentious changes on the conensus layer?

FYI:
@Windfury, I know you are unable to think , so don't worry yourself with the thought experiment.
Just keep playing with your crayons.  

Scared to be memefied?

Now consider LN Hubs will be processing offchain transactions in microseconds ,
So LN hubs hardware requirements/cost will grow at a much higher rate, especially the Main Hubs.
Unlike Bitcoin, that only charges to include a transaction in a block,
LN will impose variable fees on you for each hub needed to complete a LN transaction plus the 2 onchain bitcoin fees.

So how in your mind do you see LN fees staying cheap,
when there hardware requirements/cost for nodes are greater, and they have the ability to set their own LN fees.  

I don't worry about his for the following reasons:

Bitcoin fees spiraled up because users have no choice. Miners only compete against each other for blocks, but that doesn't increase block space.

In LN, there will be many different routes, and hub owners compete against each other on fees. If fees are high and hub owners make large profits, new actors will join the hub-market until fees are lower again.

More transactions obviously increase hardware requirements, but more transactions also means there are more fees to be earned. At the moment I don't think hub owners make a profit from the few small fees. They actually need much more transactions to earn money. If there would be a million times more transactions, I don't think the hardware cost per transaction will be very high.

Shower thought. I believe if Lightning is used in a massive scale, its fees will be higher than most altcoins because routing hubs are staking a valuable asset in Bitcoin to maintain their channels.

[franky1]

Shower thought. I believe if Lightning is used in a massive scale, its fees will be higher than most altcoins because routing hubs are staking a valuable asset in Bitcoin to maintain their channels.

towel thought:
when was the last time someone went to a bank with a bank note(promise/iou) and exited the bank with gold
when was the last time someone went to a bank with a bank note(promise/iou) and exited the bank with less precious metals

if you say the 1950's
in the 1950's
how many went to a bank with a bank note(promise/iou) and exited the bank with gold
how many went to a bank with a bank note(promise/iou) and exited the bank with less precious metals

the answer is more people deposited gold in but ended up taking silly small coins out

big mega hint:
LN is not a sole bitcoin feature of only bitcoin design and function
LN REQUIRED bitcoin to change to give bitcoin an access method to use LN
LN is a separate network
LN is a network of independant/guaranteed payment. but a network of requiring users to be funded to join, online to be paid and requires another entities authorisation just to attempt to get funds moved