Can Lightning Network Hubs be owned by banks?

[Peter11kz]

The following 2017 youtube video says the lightning network hubs will charge money similar to banks (about 6.5 minutes into the video):

https://www.youtube.com/watch?v=UYHFrf5ci_g

1. Is it true and may the banks start owning hubs?

2. The same video later-on (6.5 minutes into the video)says the block was limited in size to create a monopoly: is it true or will larger blocks produce a problem such as more processing power and therefore even longer transaction time/fees?

[DannyHamilton]

The following 2017 youtube video says the lightning network hubs will charge money . . .
1. Is it true

It is possible.  However, users of the Lightning Network will configure their software to search for the cheapest path available. As such, competition will keep those fees low. If someone chooses to run a node with a high fee, then nobody will create a path using that node, and the node owner won't generate any revenue.

and may the banks start owning hubs?

Sure. If they want to, and if the laws in their jurisdiction allow them to. Anyone can run a hub, even you.

2. The same video later-on (6.5 minutes into the video)says the block was limited in size to create a monopoly:

That is not true. The block is limited in size to PREVENT a monopoly.

[Peter11kz]

will larger blocks produce a problem such as more processing power and therefore even longer transaction time/fees?

There's no correlation between larger block and processing power (mining hashrate). With larger blocks, average transaction time/fees will be reduced (with assumption all other variable such as average transaction created per second and average block time remains same).

If all of that is true then how come Bitcoin doesn't have a larger block size?

[DannyHamilton]

how come Bitcoin doesn't have a larger block size?

The block is limited in size to PREVENT a monopoly.

More importantly, there isn't a specific size that is the "right size".  It could be smaller, it could be bigger.

It already IS bigger than it would be if it was smaller. You might want it bigger than that, but no matter what size you pick, there will probably be someone else that wants it even bigger than that.

In the end, the issue of transaction volume isn't solvable by increasing block size, so why bother increasing it at all? It's already working just fine at the current size.  To scale the quantity of transaction it is better to look to alternative solutions such as Lightning Network and/or off-chain transactions for those cases where such transactions make sense.

[TangentC]

how come Bitcoin doesn't have a larger block size?

The block is limited in size to PREVENT a monopoly.

More importantly, there isn't a specific size that is the "right size".  It could be smaller, it could be bigger.

It already IS bigger than it would be if it was smaller. You might want it bigger than that, but no matter what size you pick, there will probably be someone else that wants it even bigger than that.

In the end, the issue of transaction volume isn't solvable by increasing block size, so why bother increasing it at all? It's already working just fine at the current size.  To scale the quantity of transaction it is better to look to alternative solutions such as Lightning Network and/or off-chain transactions for those cases where such transactions make sense.

My personal opinion,
You increase onchain transaction capacity
1. To keep the blockchain fees affordable  
2. To keep transaction process timely so you are not waiting days for a transaction to complete

Small point, no matter how fast an offchain solution is, if the onchain capacity fails so too will the offchain.
As their are known dangers to offchain , if the onchain does not broadcast withdrawal info in a timely fashion.
So at some point onchain must be improved or even offchain faces failure.

Also using LN only makes sense if you are transacting multiple times between the same people.
The normal way people pay for gas or food or rent, is weekly, daily, or monthly, but not the 100s of times needed to make LN worth it.
Only Banks process thousands to millions of transactions per day between other banks that would really be advantageous to use LN.
 

[DannyHamilton]

My personal opinion,
You increase onchain transaction capacity
1. To keep the blockchain fees affordable

Affordable is a relative term.  If I'm sending the equivalent of a few million dollars, then a few hundred dollars in fees might be "affordable".  If I'm trying to buy a piece of candy that's worth about $0.10, then a $0.05 fee is ridiculous. If I've worked hard and saved for decades so I can finally buy that "thing" that I've always wanted, an extra few dollars might be more than I can "afford" at that moment.

2. To keep transaction process timely so you are not waiting days for a transaction to complete

Pay a competitive fee and your transaction will be timely.

Small point, no matter how fast an offchain solution is, if the onchain capacity fails so too will the offchain.
As their are known dangers to offchain , if the onchain does not broadcast withdrawal info in a timely fashion.
So at some point onchain must be improved or even offchain faces failure.

Small point, if everyone in the world were to start to use bitcoin for EVERY transaction they engage in every day, you won't be able to create a block big enough for them all to be confirmed "quickly".  Some transactions MUST be off-chain. It simply won't work any other way.  The more transactions that are moved off chain, the faster (and cheaper) the remaining on-chain transactions will be (for ANY given block size).

Also using LN only makes sense if you are transacting multiple times between the same people.
The normal way people pay for gas or food or rent, is weekly, daily, or monthly, but not the 100s of times needed to make LN worth it.
Only Banks process thousands to millions of transactions per day between other banks that would really be advantageous to use LN.

I don't think LN works the way you seem to think it works. Perhaps I'm the one misunderstanding it, but shouldn't you be able to send bitcoins to someone over LN that you don't have a direct channel with?  I haven't kept up with the developments over the past 18 months or so, but the last time I looked at it there's no need to "transact multiple times between the same people" for LN to work well.

It really surprises me that Exchanges (such as Coinbase and Kraken) don't set up LN and strongly encourage their customers to receive most of their withdrawals in the form of a LN channel.  Wouldn't that effectively turn the exchange into a hub?  If I want to transact with ANYONE that ALSO has an account at the same exchange, then my transaction would just need 2 hops (one from me to the exchange, and a second from the exchange to the person I'm transacting with).  If the big exchanges then set up a few large channels between each other, I could transact with any of the millions of customers that acquire most of their bitcoins at exchanges.

Same thing with mining pools.  If each of the largest pools were to set up LN and strongly encourage their miners to all take their withdrawals over LN, then the pools would become yet another hub. The pools can even set up their channels in the blocks that they create (which reduces the "cost" of setting up the channels to the lost opportunity of including some other fee-paying transaction in that block instead).  If they try to do it mostly when the blocks aren't full, then they can effectively open and/or close their own channels for free.

[NeuroticFish]

1. Is it true and may the banks start owning hubs?

Yes, it's very possible.
I don't know LN very well, but it has occurred to me that banks and other institutions may actually set low or no fee to route as many transactions as possible and analyze them.
User info may bring them more money (in a way or another) than the fees you'd save.

It really surprises me that Exchanges (such as Coinbase and Kraken) don't set up LN and strongly encourage their customers to receive most of their withdrawals in the form of a LN channel.  Wouldn't that effectively turn the exchange into a hub?

That's probably (in) the future. But since they are leading exchanges, they won't lose easily their chance to become hubs. And until then - as late as possible - they can still ask for hefty withdrawal fees to discourage people from moving their money out.
I think that only strong (and many) voices from the community and their customers could make them change their current practices.

[Pmalek]

I don't think LN works the way you seem to think it works. Perhaps I'm the one misunderstanding it, but shouldn't you be able to send bitcoins to someone over LN that you don't have a direct channel with?  I haven't kept up with the developments over the past 18 months or so, but the last time I looked at it there's no need to "transact multiple times between the same people" for LN to work well.

You don't need a direct channel with the person you send your coins to. The network will find a way over 1 or more intermediaries (let's call them that way) before it reaches you. But the longer the route, the bigger the transaction fees will be. Each node charges a base fee and a fee rate (percentage of the transaction being routed). These fees are small compared to on on-chain transactions, but still.

I think what TangentC was trying to say is that it doesn't make sense opening a payment channel with someone just to send 1 or 2 transactions. You pay regular fees for opening and closing the channel, so if you send fees only once, you would actually pay more than for an on-chain transaction. There are custodial wallets though where you don't pay any opening/closing channel fees because you will be using their own payment channels. 

[DaveF]

It really surprises me that Exchanges (such as Coinbase and Kraken) don't set up LN and strongly encourage their customers to receive most of their withdrawals in the form of a LN channel.  Wouldn't that effectively turn the exchange into a hub?  If I want to transact with ANYONE that ALSO has an account at the same exchange, then my transaction would just need 2 hops (one from me to the exchange, and a second from the exchange to the person I'm transacting with).  If the big exchanges then set up a few large channels between each other, I could transact with any of the millions of customers that acquire most of their bitcoins at exchanges.

Same thing with mining pools.  If each of the largest pools were to set up LN and strongly encourage their miners to all take their withdrawals over LN, then the pools would become yet another hub. The pools can even set up their channels in the blocks that they create (which reduces the "cost" of setting up the channels to the lost opportunity of including some other fee-paying transaction in that block instead).  If they try to do it mostly when the blocks aren't full, then they can effectively open and/or close their own channels for free.

I posted about this a while ago. I mentioned it being a support issue. Coinbase / Kraken move a ton of BTC each and every day.
The capacity of the ENTIRE lightning network at the moment is a bit under 1200 BTC

It's a loop. Till they know there are enough channels with enough capacity to move the amount of BTC to and from their customers it's pointless to do it since it's just going to generate more support issues. Till we get more big places on lightning it's pointless to have large channels open since it's locking funds and they may or may not be used.

For mining pools, at least the larger ones, lightning would just cost them money in terms of needing more back end servers and support. A lot of them send their payments only in their own blocks (or at least they used to) so although there is a cost in "we could have had the fees from someone else" it's really not a cost to them.

-Dave

[DannyHamilton]

I think what TangentC was trying to say is that it doesn't make sense opening a payment channel with someone just to send 1 or 2 transactions. You pay regular fees for opening and closing the channel, so if you send fees only once, you would actually pay more than for an on-chain transaction. There are custodial wallets though where you don't pay any opening/closing channel fees because you will be using their own payment channels. 

My assumption is that there will eventually mostly be channels established with the main sources of acquiring bitcoin for the average person.  Places such as exchanges, mining pools, or payroll companies.  In that way, the fee will be paid once to set up the channel to get the bitcoins to you in the first place.  You then, as you use your bitcoin with people that you don't have a direct channel with, they'll typically only require 1 or 2 intermediaries.  Additionally, the competition will drive down the LN fees. If I don't like my exchange's (or payroll company's) LN fees, I'll switch to an exchange (or payroll company) that has lower fees. If thee fees are too high everywhere, then it creates an opportunity for someone else to start a service that with lower fees and still earn a living.

[TangentC]

I think what TangentC was trying to say is that it doesn't make sense opening a payment channel with someone just to send 1 or 2 transactions. You pay regular fees for opening and closing the channel, so if you send fees only once, you would actually pay more than for an on-chain transaction. There are custodial wallets though where you don't pay any opening/closing channel fees because you will be using their own payment channels.  

My assumption is that there will eventually mostly be channels established with the main sources of acquiring bitcoin for the average person.  Places such as exchanges, mining pools, or payroll companies.  In that way, the fee will be paid once to set up the channel to get the bitcoins to you in the first place.  You then, as you use your bitcoin with people that you don't have a direct channel with, they'll typically only require 1 or 2 intermediaries.  Additionally, the competition will drive down the LN fees. If I don't like my exchange's (or payroll company's) LN fees, I'll switch to an exchange (or payroll company) that has lower fees. If thee fees are too high everywhere, then it creates an opportunity for someone else to start a service that with lower fees and still earn a living.

Pmalek is correct, my thoughts was exactly that.

IMO, you need 100s or more of transaction between you and whomever you are transacting with on LN, to make it worthwhile to use LN instead of just sending it normally.

You mentioned earlier, why does coinbase not use LN hubs,
Since they are an exchange, they use their internal accounting system to perform offchain transactions without needing to bother with the complexity of runing LN or Liquid 3rd party software.
https://help.coinbase.com/en/coinbase/trading-and-funding/cryptocurrency-trading-pairs/how-to-send-and-receive-cryptocurrency

If you’re sending to a crypto address that belongs to another Coinbase user who has opted into Instant sends, you can use off-chain sends.
Off-chain sends are instant and incur no transaction fees

So any coinbase user can send funds offchain instantly to another coinbase user and incur no transaction fees at all,
while LN or Liquid users, pay the initial onchain transaction fee, LN hubs fees, and final onchain withdrawal transaction fee.


  

[Pmalek]

Additionally, the competition will drive down the LN fees. If I don't like my exchange's (or payroll company's) LN fees, I'll switch to an exchange (or payroll company) that has lower fees.

The way the Lightning Network is set up, I think that is already happening. Your wallet will try to find a route from point A to point B which is the cheapest one. Therefore If I offer a cheaper routing fee to the same destination compared to you, there is no reason why someone would use your route instead of mine. Actually, one reason would be if I didn't have enough liquidity in my channel. Someone who has used LN can correct me, but I think the payment routes are chosen automatically by the wallet, based on what is the most cost-effective choice. The sender can't select the route his payment will take or can it? 

[pooya87]

It really surprises me that Exchanges (such as Coinbase and Kraken) don't set up LN and strongly encourage their customers to receive most of their withdrawals in the form of a LN channel.

I think it is the fear of losing money to a bug. I don't follow LN closely but I hear the LN devs warn people about possibility of bugs and how they shouldn't use it with large amounts. It is understandable that an exchange doesn't want to bother with the technology until it is matured.
I only know one exchange (Bitfinex) that has an LN channel and they started it with a very small capacity for this very risk, then increased it after some time passed but it is still a small capacity.

Same thing with mining pools.

I don't think it works for pools unless they change their withdrawal setup and pay every time they find a block. Not to mention that the miners may not want to keep their coins in a channel (maybe want to sell it on an exchange, move to cold storage,...) which means they would have to pay another on-chain fee themselves to move the coins to their final destination.

[Wind_FURY]

The following 2017 youtube video says the lightning network hubs will charge money . . .
1. Is it true

It is possible.  However, users of the Lightning Network will configure their software to search for the cheapest path available. As such, competition will keep those fees low. If someone chooses to run a node with a high fee, then nobody will create a path using that node, and the node owner won't generate any revenue.

Although it might be debatable if Lightning Node Operators will keep fees very low. To provide the Lightning Network with liquidity, it needs a finite “commodity”, which also is your “capital”, Bitcoin. I don’t believe the more specialized/connected and more “professional” nodes will provide the lowest close-to-zero/almost-free fees.

[Timelord2067]

I think most people are over thinking the issue of banks setting up their own hubs within the Lightning Network Framework...

Anyone (including banks) can buy bitcoins, install a Lightning Network program such as Eclair or Zap (or others) and create channels to any-other node they want to without anyone's permission.

Elon Musk could set up channels worth one bitcoin to each of about the first thirty-thousand nodes on the 1ml.com list as a means of perma-storing his bitcoins into perpetuity.  Need a bitcoin to pay a bill? close the channel at the bottom of the list and one bitcoin awaits in a few short hours of confirmation time.

Even BTC 0.01 (worth about $ 500 at the moment) would be enough of a channel to capture traffic and would enable Musk to create channels to close to Three Million other nodes.  (Cornering the traffic/market?)

[Pmalek]

I think it is the fear of losing money to a bug. I don't follow LN closely but I hear the LN devs warn people about possibility of bugs and how they shouldn't use it with large amounts. It is understandable that an exchange doesn't want to bother with the technology until it is matured.

There is a great article on the topic of LN vulnerabilities. I don't know if you have read it, but if you are interested, here is the link > https://www.coindesk.com/bitcoin-lightning-network-vulnerabilities-not-exploited-yet.

One of the most severe threats is called a Time-dilation eclipse. The article mentions that light clients, such as Blue Wallet or Phoenix Wallet are particularly vulnerable to this attack vector. To carry out such an attack the scammer would have to create a whole network of nodes that the victim would be connected to. The victim is then only connected to malicious nodes. The scammer can deliberately slow down the process of sharing transaction data with his victim. He can close the channel and steal the coins before this data is received by the victim. That's the idea anyway.       

I only know one exchange (Bitfinex) that has an LN channel and they started it with a very small capacity for this very risk, then increased it after some time passed but it is still a small capacity.

If this article can be trusted, and it seems to be up-to-date, the list of exchanges that accept LN payments is slowly on the rise. Besides Bitfinex, there is OKCoin and OKEx. Paxful has declared intention to integrate the LN...etc. 

[acquafredda]

Anyone (including banks) can buy bitcoins, install a Lightning Network program such as Eclair or Zap (or others) and create channels to any-other node they want to without anyone's permission.

Sorry I cannot help myself stop smiling on this one as it would be awesomely funny to see a banker connected to the Lightning Network via Eclair or Zap.
Jokes apart, I agree with you and I believe banks could be the ones benefiting the most from a widespread and increasingly adopted Lightning Network.
They first offer bitcoin to their clients (eating on fees like exchanges do) and then they could re-direct their clients on their hubs to offer them both their own services and access to the network to freely transact with other peers.
Bank hubs will be the future of LN, whether we like it or not.
Look at what happened with the bitcoin treasuries, it is only about time.

[Timelord2067]

Anyone (including banks) can buy bitcoins, install a Lightning Network program such as Eclair or Zap (or others) and create channels to any-other node they want to without anyone's permission.

Sorry I cannot help myself stop smiling on this one as it would be awesomely funny to see a banker connected to the Lightning Network via Eclair or Zap.

Those two (Eclair and Zap - and to a lesser extent Blue LN are the only wallets (as an end user) that I am familiar with.  I'm sure there are some/many LN wallets for back of house applications.

Jokes apart, I agree with you and I believe banks could be the ones benefiting the most from a widespread and increasingly adopted Lightning Network.
They first offer bitcoin to their clients (eating on fees like exchanges do) and then they could re-direct their clients on their hubs to offer them both their own services and access to the network to freely transact with other peers.
Bank hubs will be the future of LN, whether we like it or not.
Look at what happened with the bitcoin treasuries, it is only about time.

Banks may have been slowly buying up bitcoin - a little at a time (as low as one bitcoin per week) so as not to draw attention to themselves.

After writing the above, it also occurred that multiple nodes/hubs could be set up by a person or organisation with deep enough pockets in various parts of the globe to have not just one mammoth node/hub, but a dozen/dozens of smaller niche nodes interconnected between themselves, and also to a slew of much smaller nodes/hubs with minute channels, thus having two or more steps in the link from sender to receiver.

[Wind_FURY]

Anyone (including banks) can buy bitcoins, install a Lightning Network program such as Eclair or Zap (or others) and create channels to any-other node they want to without anyone's permission.

 
Sorry I cannot help myself stop smiling on this one as it would be awesomely funny to see a banker connected to the Lightning Network via Eclair or Zap.
Jokes apart, I agree with you and I believe banks could be the ones benefiting the most from a widespread and increasingly adopted Lightning Network.
They first offer bitcoin to their clients (eating on fees like exchanges do) and then they could re-direct their clients on their hubs to offer them both their own services and access to the network to freely transact with other peers.
Bank hubs will be the future of LN, whether we like it or not.
Look at what happened with the bitcoin treasuries, it is only about time.

What’s funnier are the people who spread FUD that the government will monitor all Lightning transactions, just as they’re doing with onchain transactions. OK, then government will run Lightning nodes, buy Bitcoin, and then open payment channels? Thanks for the liquidity.

[PrimeNumber7]

It really surprises me that Exchanges (such as Coinbase and Kraken) don't set up LN and strongly encourage their customers to receive most of their withdrawals in the form of a LN channel.  Wouldn't that effectively turn the exchange into a hub?  If I want to transact with ANYONE that ALSO has an account at the same exchange, then my transaction would just need 2 hops (one from me to the exchange, and a second from the exchange to the person I'm transacting with).  If the big exchanges then set up a few large channels between each other, I could transact with any of the millions of customers that acquire most of their bitcoins at exchanges.

You make a compelling argument, but I suspect many exchanges are hesitant to do this because it would create a customer service/support nightmare. When a LN transaction fails, it could be because of the current topography of the network, or it could be a problem with one or both of the participants.

The exchanges also would not be able to open channels with all of their customers for their entire balance because doing so would prevent the exchange from having any of their coin in cold storage. If you have 1 BTC on an exchange, and you want to receive a LN transaction from the exchange in the amount of 1 BTC, the exchange will need to either have an open channel with you with sending capacity of at least 1 BTC, or for there to be a path of channels between the exchange and you with capacity of the same amount. An exchange could have a very well-connected node that many customers can receive withdrawals from and make deposits to, however, some deposits/withdrawals may fail.

Same thing with mining pools.  If each of the largest pools were to set up LN and strongly encourage their miners to all take their withdrawals over LN, then the pools would become yet another hub. The pools can even set up their channels in the blocks that they create (which reduces the "cost" of setting up the channels to the lost opportunity of including some other fee-paying transaction in that block instead).  If they try to do it mostly when the blocks aren't full, then they can effectively open and/or close their own channels for free.

Pools could fund a new LN channel with coinbase transactions. This channel would generally immediately be spent in its entirety in order to pay the miners, and the miners have bills to pay, so they would generally need to send the entire amount they receive to an exchange. This means there is a lot of one-way transactions on LN, instead of back and forth transactions that LN is really designed for.

I think it would be more likely for exchanges to operate very well connected LN nodes, and act somewhat like "banks". I can see major exchanges opening large LN channels amongst themselves in order to facilitate transfers between exchanges.