Topic: Would it cost anything to open a Channel on Lightning Network? (Read 522 times)

I made a basic supply and demand argument. Lightning channels adds several orders of magnitude more transaction capacity for sending BTC between users. A list of conditions under which channels close is not very meaningful as a response.


Philosophy involves reasoning. Feelings do not.

Although philosophical, it doesn't feel right.

Isn't "closing channels" very easy to abuse? Say a LN hub pays me 0.0000001 Bitcoin, if I understand correctly I can close that channel whenever I want. I won't have enough funds to pay the fee, but the hub does. Does this mean the hub has to pay many times more than what it paid me in fees?
Or is there a safeguard in place that I don't know about? If it is like I just described, a spam attack can easily bankrupt LN hubs.

Users will have to periodically close/settle LN channels for a variety of reasons even if fees are high. If participants of LN channels start to act maliciously, channels will be forced to close, when channels have nearly all of the balances on one side, the other side will want to close the channel to reduce risk, channels will be opened and closed as users start/stop using bitcoin, businesses will close channels as they accumulate large amounts of bitcoin so they can move coins to cold storage, etc.

Why not?

I thought about that, but ignored it: even if you do the math with 4 MB blocks, it's just 4 times more users or 4 times more transactions.

That would mean many users would never settle their transactions on-chain anymore. That means the small amounts they have in their future Lightning Wallet have the potential to be settled on-chain, but they never do it because it's easier/cheaper not to? Can you even call it Bitcoin if they never go on-chain?

Your math is right. Your figures are wrong: 1MB is not the blocksize any longer.

And your logic is wrong too. When millions of transactions per day are being made over Lightning, the price users are willing to pay on-chain is going to be less. Why wait minutes or hours when most transactions can be made in less than a second? Why pay 100,000 times the fees? Demand will follow supply, and fees will adjust commensurately.

Really ? Anyone ? Are they kind of newbie questions for you ?

I highly doubt that. If 10 million people make only one on-chain transaction per month, 1 MB blocks are full already.

I can't wait for it to happen!

Has it already been specified what is necessary to run your own Lightning Network node?
I guess that you would need one of the Lightning Network clients alongside with enough
locked Bitcoin to provide liquidity. Besides, it is probably necessary to be connected to the internet
in order to route transactions.

Which other requirements do I miss?

Yep, that's the idea. But not really for 500 bits (as in your example), if Lightning is successful enough then 500 bits could be economic to send on-chain again. Lighting really unlocks amounts less than that, 100's of satoshi, or even less than 1 satoshi (i.e. milli-satoshis. Lightning expresses BTC amounts in millisatoshis I believe)

I've been wondering about the promise of 100,000 transactions per second with the Lightning Network. Bitcoin can still handle only about 5 transactions per second. That means, for every Bitcoin transaction, there can be 20,000 Lightning transactions! Even if I use LN for everything I do, it will take me years to make that many transactions.
That made me realize something else: someone can receive a Lightning transaction without every making an on-chain transaction, am I right? If that's the case, I can indeed imagine LN having 20,000 times more off-chain transactions than Bitcoin has. And if that's right, many users will keep their payments offline forever: if you only receive 0.00005BTC through LN, it's not even enough to pay for a Bitcoin fee, so all Bitcoin dust stays off-chain.

No, you only pay a fee when opening and closing a payment channel. Once the channel is open, you can transact freely (i.e. as many times as you want) with the other person in the channel at no cost. If you want to transact with someone outside of the channel but there is some route to them through LN channels, then you can use your same channel to pay that other person by routing a payment through other channels. This routing of payments will either be extremely cheap or have no cost, depending on whether node operators in the route charge a fee for being routed through.

If you are just going to make a single transaction with someone and you don't plan on using them to route payments to other people, then you should not create a channel for that. Just make a normal Bitcoin transaction.

   So back to square 1, we need to pay fees to open the channel each transaction?  Or we keep our own channel open (tieing wallet balance) hope our recipient is on LN with an open channel? 

Like the idea of running a node though, have my old machine for mining that i keep meaning to give a new project.

Nope.

Alot of people make this mistake. There won't be much incentive to attempt to make hubs of payment channels. The resources needed to run a Lightning node are little more than running a regular Bitcoin node, i.e. not much.

So the routing algorithms can use as many node hops as they like, as charging a discernible fee will only hurt how well your node is connected within the Lightning network. No-one's going to choose to pay high fees with this model, as it's going to be far too easy to find a cheaper alternative. Making a business out of that is going to be tough, it's real race-to-the-bottom economics.

Thus conflicting one the foundations of Bitcoin, decentalisation?  After 10 years, we've reinvented the Bank.      

For just one transaction, you'll end up paying more in fees than without using LN.

Correct.

It's very likely (centralized) hubs will emerge: you open 1 channel with the hub, and the hub has many other open channels. Say Bob <> hub <> Alice, or Bob <> hub <> exchange <> Alice.
As long as you don't close the channel, you can keep using it without paying on-chain fees.

If there's a fee to open/close the channel, and I only make a single transaction with the other party, how is that different to the existing fee structure?  If i make regular transfers, do i simply keep the channel open for all future payments (say with exchange)?  As far as i can see, LN makes it cheap for bulk tx but not really helping the user with individual tx.  I hope i'm missing something.

The costs associated with opening and closing LN channels will likely be significant.

The cost to get a transaction confirmed on the blockchain will likely go up from current levels (or else there would not be incentives to use LN in masse), and the amount of transaction space required to open and close a LN channel is much greater than getting a "normal" (eg a payment from one person to another, or a withdrawal from a business to several of its customers) transaction.

Thanks for the answers. Good to know. It looks like very promising as they are all 'yes'.

Point 1. and point 3. are kind of part of the points that confuses me on LN stuff, if I may I would like to detail it more.

If I have good understood, once LN will be avaible for mass adoption, as a user you have to make a multi-sig special transaction on-chain
to create the channel, and if you or your counterpart will link to other users with other channels, and so on, in that case you're creating the famous Lightning Network (LN), off-chain, to transact between multiple users.
Then each transaction on the channel will be off-chain until channel will not be closed, but transactions on-channel, thus off-chain, will not be feeless, and each transaction will cost a fee to pay LN nodes between you and the user you want to pay and viceversa.

Thus there will be on-chain fees, collected by miners as of today, and off-chain fees, collected by LN users.

My questions are:
- Who pays on-chain fees? A and B I suppose, right? In equal shares? 30% 70%? 70% 30%? How will be calculated?

- How will be calculated off-chain fees ?

- In my view, off-chain fees will be collected by users who can afford running a full-node running 24h 7d with multiple opened channels, that could be   kind of an interesting business not for miners but for simple users also, or not?

- Once you will have a LN with multiple channels, how do you balance between them once a channel between two users (A and B), and hence the bitcoin encumbered on-chain within that transaction, will be released?
It's highly unlikely they will be still entirely between A and B wallets, quite the contrary, some part of them will be paid as off-chain fees, others will be sent to other users (let's say C and D), which though remain with opened channels. What will happen in that event to those bitcoins? I suppose they will be avaiable on-chain to C and D wallets, but it seems kind of risky to me, because you could create negative balances on the Lighting Network, how do you manage that?
Let's say C has a 0,5 BTC balance off-chain, A and B close their channel, doing that C "earns" 1 BTC in his on-chain wallet and "lose" 1 BTC in his LN off-chain, hence with a balance off-chain in the LN of -0,5 BTC. What will happen if he closes all his channels?

Thanks in advance