Opening LN channel with zero balance?

[pawel7777]

As in subject: will Lightning Network allow opening channels with zero funds and without initial 'deposit' transaction (by just generating address and associating it with the channel)?

ie. If I wanted to send btc to other person (new to bitcoin) would that person need to have any own funds beforehand (on-chain)?

I'm guessing it's possible, but would like to confirm this with someone more knowledgeable.

[Quickseller]

When a LN channel is opened the total amount in a channel is sent to a 2-of-2 multisig address when each party has one of the two keys necessary to spend the BTC.

If someone had settled funds, then nothing from a technical standpoint would stop him from sending his BTC to a new LN channel with no when you have an opening "balance" of 0.

From an economic standpoint, this would make little sense because the person with settled funds would have to cover the entire tx fee (which is likely to be very expensive), and would have slightly less security because the other party to the LN channel would have no "deposit" they would lose in the event that he tried to broadcast an "old" transaction to close the LN channel.

[stdset]

Alice buys a car from a car dealer Bob.
She has yet to join LN so she opens a channel with Bob contributing a significant sum (price of the car), Bob doesn't contribute any funds to the channel. Entire channel balance immediately transferred to Bob with the first commitment transaction.
Alice plans to use this channel later to receive salary from her employer.

[achow101]

Alice buys a car from a car dealer Bob.
She has yet to join LN so she opens a channel with Bob contributing a significant sum (price of the car), Bob doesn't contribute any funds to the channel. Entire channel balance immediately transferred to Bob with the first commitment transaction.
Alice plans to use this channel later to receive salary from her employer.

It would be far more efficient for Alice to simply make an on chain payment for the car, and then open up a direct channel with her employer. This would reduce the number of transactions that Alice would need to make in order for her to send and receive her payments.

[stdset]

Alice buys a car from a car dealer Bob.
She has yet to join LN so she opens a channel with Bob contributing a significant sum (price of the car), Bob doesn't contribute any funds to the channel. Entire channel balance immediately transferred to Bob with the first commitment transaction.
Alice plans to use this channel later to receive salary from her employer.

It would be far more efficient for Alice to simply make an on chain payment for the car, and then open up a direct channel with her employer. This would reduce the number of transactions that Alice would need to make in order for her to send and receive her payments.

My scenario, onchain txses: channel opening.
Your scenario, onchain txses: car purchase, channel opening.
Possible drawbacks in my scenario:
1) Alice can't pay for anything with the new channel before she receives the first (after the purchase) salary.
2) If Bob doesn't offer intermediary services or if he charges incompetitive fee for conducting payments, it's better to open a channel not with Bob, but with a competitive hub and route the payment to Bob through the hub. If Bob's channel capacity with the network isn't enough, he can open another channel with a hub.
So, what do you mean?

[achow101]

My scenario, onchain txses: channel opening.
Your scenario, onchain txses: car purchase, channel opening.
Possible drawbacks in my scenario:
1) Alice can't pay for anything with the new channel before she receives the first (after the purchase) salary.
2) If Bob doesn't offer intermediary services or if he charges incompetitive fee for conducting payments, it's better to open a channel not with Bob, but with a competitive hub and route the payment to Bob through the hub. If Bob's channel capacity with the network isn't enough, he can open another channel with a hub.
So, what do you mean?

Over the long term, there will be more on chain transactions and more costs with your scenario.

In the US, the average car is ~$25,000 and the median household income is ~$50,000 per year.

Under your scenario:
Alice opens the channel with the Car dealership (1 on chain tx). In order for her employer to pay Alice through the car dealer, they would need to have a payment channel with more than $50,000 to pay Alice for the year (which I find unlikely), or they will need to load their payment channel to match Alice's salary (another on chain tx). Now in order for Alice to receive her entire $50,000 salary, she (or the car dealer) will need to close a payment channel (another on chain tx), load up Alice's channel (another on chain tx), and probably open up another channel with whoever they closed with (another on chain tx) In one year, Alice will have made or required to be made 5 on chain txs. Furthermore, the car dealer is likely to charge a fee for being a third party between Alice and the employer.

Under my scenario:
Alice sends 1 on chain transaction for the car. She opens a payment channel with her employer, who loads into the channel however much they expect to pay her that year, so $50,000 (another on chain transaction). At the end of the year, the channel is loaded with her next year's salary (another on chain transaction). In total, she makes only 3 on chain transactions, and does not need to pay a third party for being a hub. Like with the Car dealer, Alice can buy things by using her employer as a hub, but because the channel already has her entire yearly salary there, no on chain txs need to be made to load any more money into that channel (unless she is spending from here savings).


Lightning is fairly useless for big one time purchases. It is far more efficient to use just a normal on chain transaction to make a big purchase like a car or a house.

A far more likely scenario for one party to put in 0 funds into a channel is for an employer to be paying an employee.

[stdset]

Over the long term, there will be more on chain transactions and more costs with your scenario.

In the US, the average car is ~$25,000 and the median household income is ~$50,000 per year.

Under your scenario:
Alice opens the channel with the Car dealership (1 on chain tx). In order for her employer to pay Alice through the car dealer, they would need to have a payment channel with more than $50,000 to pay Alice for the year (which I find unlikely), or they will need to load their payment channel to match Alice's salary (another on chain tx). Now in order for Alice to receive her entire $50,000 salary, she (or the car dealer) will need to close a payment channel (another on chain tx), load up Alice's channel (another on chain tx), and probably open up another channel with whoever they closed with (another on chain tx) In one year, Alice will have made or required to be made 5 on chain txs. Furthermore, the car dealer is likely to charge a fee for being a third party between Alice and the employer.

Under my scenario:
Alice sends 1 on chain transaction for the car. She opens a payment channel with her employer, who loads into the channel however much they expect to pay her that year, so $50,000 (another on chain transaction). At the end of the year, the channel is loaded with her next year's salary (another on chain transaction). In total, she makes only 3 on chain transactions, and does not need to pay a third party for being a hub. Like with the Car dealer, Alice can buy things by using her employer as a hub, but because the channel already has her entire yearly salary there, no on chain txs need to be made to load any more money into that channel (unless she is spending from here savings).


Lightning is fairly useless for big one time purchases. It is far more efficient to use just a normal on chain transaction to make a big purchase like a car or a house.

A far more likely scenario for one party to put in 0 funds into a channel is for an employer to be paying an employee.

Alice's employer not only pays salaries, he also receives payments. So there's no need to load entire yearly salary to a personal channel with Alice. There's even no need to have a separate channel for Alice. It's much better to have a channel with a hub, pay salaries through that channel and receive payments through the same channel. Alice can receive many monthly salaries through a channel created with a car purchase, it has very high receiving capacity, it will save many onchain txses.

[Quickseller]

Alice buys a car from a car dealer Bob.
She has yet to join LN so she opens a channel with Bob contributing a significant sum (price of the car), Bob doesn't contribute any funds to the channel. Entire channel balance immediately transferred to Bob with the first commitment transaction.

This would probably not be a good idea for either party. Both parties would need to pay higher fees than if they had covered the tx fee for an on-chain transaction. There is also no guarantee that either party will facilitate LN transfers in the futures and be willing to pay/accept what the other party deems to be a reasonable LN fee.

It would also not be unreasonable to expect you to receive a LN payment via this channel, and subsequently spend the majority of said payment in a subsequent LN transaction. If this were to happen, then you could attempt to broadcast a settlement transaction that is not the most recent, that would result in you receiving more BTC than what you and the dealership have most recently agreed to -- in this situation, the dealership should normally be able recover then entire balance within the channel before you can spend the portion spent to you, however doing so would incur an additional tx fee that the dealership would need to bear in it's entirety in order to claim an amount that they are due anyway....and you are not risking very much/anything when you attempt to steal these funds (at most a nominal amount).  

Alice plans to use this channel later to receive salary from her employer.

If you do this, then the dealership might see that your portion of the LN channel is receiving the exact same amount on a very regular basis, they might be able to reasonably conclude the amount of your bi-weekly (or whatever frequency) paycheck, harming your privacy (the exact information they will know is that the balance of your channel is increasing every other Friday by the same amount, they will not have proof if you are receiving this payment, or if you are acting as an intermediary. They will also know the  channel that is sending the payment to you, they wont know if that channel is sending the payment directly or if they are acting as an intermediary).

Also, if you only have one LN channel open, then you will need to accept whatever LN fee the dealership is charging (they can change the LN fee at anytime without notice), or else close/settle the LN channel, delaying your access to your money AND be forced to pay a very high tx fee.

[stdset]

There is also no guarantee that either party will facilitate LN transfers in the futures and be willing to pay/accept what the other party deems to be a reasonable LN fee.

If Bob doesn't offer intermediary services or if he charges incompetitive fee for conducting payments, it's better to open a channel not with Bob, but with a competitive hub and route the payment to Bob through the hub. If Bob's channel capacity with the network isn't enough, he can open another channel with a hub.

then you could attempt to broadcast a settlement transaction that is not the most recent, that would result in you receiving more BTC than what you and the dealership have most recently agreed to -- in this situation, the dealership should normally be able recover then entire balance within the channel before you can spend the portion spent to you, however doing so would incur an additional tx fee that the dealership would need to bear in it's entirety in order to claim an amount that they are due anyway....and you are not risking very much/anything when you attempt to steal these funds (at most a nominal amount). 

Again, if the dealership doesn't offer LN intermediary services, they can use their respective hubs. Also, if one party tries to cheat and broadcasts an obsolete commitment tx, another party can take all channel funds, that will probably not only cover all necessary fees, but much more. Also, since all parties know, that broadcasting obsolete commitment txes is disadvantageous, they won't even attempt doing so.

[Quickseller]

There is also no guarantee that either party will facilitate LN transfers in the futures and be willing to pay/accept what the other party deems to be a reasonable LN fee.

If Bob doesn't offer intermediary services or if he charges incompetitive fee for conducting payments, it's better to open a channel not with Bob, but with a competitive hub and route the payment to Bob through the hub. If Bob's channel capacity with the network isn't enough, he can open another channel with a hub.

then you could attempt to broadcast a settlement transaction that is not the most recent, that would result in you receiving more BTC than what you and the dealership have most recently agreed to -- in this situation, the dealership should normally be able recover then entire balance within the channel before you can spend the portion spent to you, however doing so would incur an additional tx fee that the dealership would need to bear in it's entirety in order to claim an amount that they are due anyway....and you are not risking very much/anything when you attempt to steal these funds (at most a nominal amount). 

Again, if the dealership doesn't offer LN intermediary services, they can use their respective hubs. Also, if one party tries to cheat and broadcasts an obsolete commitment tx, another party can take all channel funds, that will probably not only cover all necessary fees, but much more. Also, since all parties know, that broadcasting obsolete commitment txes is disadvantageous, they won't even attempt doing so.

1 - if I have a LN channel open with the dealership, and the dealership does not offer intermediary services, then the only use of that LN channel is to transact with that dealership. If the dealership is connected to a LN hub that I am not connected to, then I cannot transaction with that LN hub without the cooperation of the dealership, or I would need to route payments via other means to that LN hub via other cooperating LN channels.

2 - Say for example that you buy a $25,000 car from a local car dealership. You open a LN channel in which you send $25,000 and the dealership sends $0.00, and you subsequently sign transactions changing the balance of the LN channel so that the dealership has $25,000 and you have $0.00 -- now the car is paid for. The following Friday, you get paid by your employer via the LN channel you have open with the car dealership, and you get paid $3,000, now the balance of the LN channel is you have $3,000 and the dealership has $22,000. You subsequently use that same LN channel to pay $1,500 to your landlord, and $1,490 towards your other monthly expenses -- now the LN channel has a balance of $10 for you, and $24,990 for the dealership. If you were to broadcast a closing transaction that resulted in you receiving $3,000, then the dealership would be able to claim the full $25,000 in the LN channel, however the tx fee would likely be greater than $10, so they are loosing money by doing this, and there is the potential that for whatever reason, the dealership does not realize that you broadcast an old closing transaction and/or the dealership is not able to claim the entire balance in the LN channel because of on-chain transaction backlogs, so there is no financial risk to you broadcasting an old transaction, because after tx fees, you would have gotten nothing after the channel is closed anyway.   

[stdset]

1 - if I have a LN channel open with the dealership, and the dealership does not offer intermediary services, then the only use of that LN channel is to transact with that dealership. If the dealership is connected to a LN hub that I am not connected to, then I cannot transaction with that LN hub without the cooperation of the dealership, or I would need to route payments via other means to that LN hub via other cooperating LN channels.

That's why I suggest opening a channel not with the dealership, but with a hub, in this case.

then the dealership would be able to claim the full $25,000 in the LN channel, however the tx fee would likely be greater than $10, so they are loosing money by doing this, and there is the potential that for whatever reason, the dealership does not realize that you broadcast an old closing transaction and/or the dealership is not able to claim the entire balance in the LN channel because of on-chain transaction backlogs, so there is no financial risk to you broadcasting an old transaction, because after tx fees, you would have gotten nothing after the channel is closed anyway.   

If tx fees are so significant, the dealership, as well as other intermediaries should abstain from completely emptying balances of their counterparties in their respective channels. Always leave enough in a channel to cover possible onchain fees. Also if onchain fees are so high, it's an additional incentive to use LN.
Every LN participant and especially intermediaries, should be online often enough to check for old commitment txes, before their outputs become spendable for their counterparties. Transactions backlog may be a problem. But provided that enough funds are left in a channel, it should be possible to offer miners enough fee to get ahead of queue without suffering even a small loss. Also there are solutions proposed for fast dynamic blocksize: https://bitcointalk.org/index.php?topic=1713807.0

[dinofelis]

1 - if I have a LN channel open with the dealership, and the dealership does not offer intermediary services, then the only use of that LN channel is to transact with that dealership. If the dealership is connected to a LN hub that I am not connected to, then I cannot transaction with that LN hub without the cooperation of the dealership, or I would need to route payments via other means to that LN hub via other cooperating LN channels.

That's why I suggest opening a channel not with the dealership, but with a hub, in this case.

This is why LN is banking.  A few big hubs are the banks.  Everyone will want to connect to them, and they set their conditions.

[stdset]

1 - if I have a LN channel open with the dealership, and the dealership does not offer intermediary services, then the only use of that LN channel is to transact with that dealership. If the dealership is connected to a LN hub that I am not connected to, then I cannot transaction with that LN hub without the cooperation of the dealership, or I would need to route payments via other means to that LN hub via other cooperating LN channels.

That's why I suggest opening a channel not with the dealership, but with a hub, in this case.

This is why LN is banking.  A few big hubs are the banks.  Everyone will want to connect to them, and they set their conditions.

That's may be true to some extent. In the sense that hubs and spokes architecture is simpler and cheaper than mesh-like. But what power will hubs have? Rising fees? There will always be enough competition to keep fees low. What else?

[dwieyani]

zero balance what is the reason why I do not understand please love feedback for me.
whether this channel LN zero balance?

[fathur.aza]

do not let all no problem if they are less guarantees about LN.

[Carlton Banks]

1 - if I have a LN channel open with the dealership, and the dealership does not offer intermediary services, then the only use of that LN channel is to transact with that dealership. If the dealership is connected to a LN hub that I am not connected to, then I cannot transaction with that LN hub without the cooperation of the dealership, or I would need to route payments via other means to that LN hub via other cooperating LN channels.

That's why I suggest opening a channel not with the dealership, but with a hub, in this case.

This is why LN is banking.  A few big hubs are the banks.  Everyone will want to connect to them, and they set their conditions.

Except one crucial difference.

Not just anyone can open a bank, there are very significant real-world political obstacles to get past, very few genuinely new banks open as a result.

Lightning can behave like banking, but because people have the freedom to open their own banks (really more like a mixture of a bank and a VISA/Amex type payment processor), they will do that.

Some people will try, and fail. Regular everyday people paying each other small amounts through channels might decide the price is too high. Or, they might decide that price is worth it, for the enhanced privacy of doing business in a private channel.

But even small businesses will be able to compete on costs with channels. Maybe an ecosystem of small payment processors might arise, servicing a small geographical area (for instance, a landowner leasing shop units would do well to make using their channel a part of the contract for all the shop leases they offer).

Or maybe the economics will be such that very small shops can afford to run their own channel. If a small shop has goods of sufficient quality, the clientele may prefer to pay the higher channel fees in order to get the high quality product that our small shop sells.



So no, dino, it's not "just like banking". It resembles it in some ways, but in fact it resembles Hayek's "free banking concept " most of all. Now do be quiet, people are interested in having a productive conversation about Lightning.

[goregrind]

Except one crucial difference.

The strongest point of Bitcoin is its decentralization and this is why the scaling debate is such a big issue.
Why is it that we're now ok with centralization ? If lightning leads to a few centralized hubs then we need
to design it in such a way that this won't cause issues further down the road.
We're taking this too lightly.

[stdset]

The strongest point of Bitcoin is its decentralization and this is why the scaling debate is such a big issue.
Why is it that we're now ok with centralization ? If lightning leads to a few centralized hubs then we need
to design it in such a way that this won't cause issues further down the road.

What do you mean by 'few'? More than likely LN hubs will be orders of magnitude more numerous than mining pools are now.

[pinkflower]

The explanations have gotten way too complicated for what pawel777 was asking. Im interested in the answer too. Allow me to rephrase the question. In LN, do we need some kind of show money to open a channel? Is it possible to only open a channel now without the show money for future use if I already have the show money?

[achow101]

The explanations have gotten way too complicated for what pawel777 was asking. Im interested in the answer too. Allow me to rephrase the question. In LN, do we need some kind of show money to open a channel? Is it possible to only open a channel now without the show money for future use if I already have the show money?

With the current lightning spec document, both parties need to put in some amount of money so that both parties have something to lose so there is an incentive for both parties to not scam the other person.

Edit: You can open a channel where only one party funds it. I read this part of the spec incorrectly. See https://bitcointalk.org/index.php?topic=1848861.msg18444685#msg18444685