Bitcoin may need to be HIGH-VOLUME digital cash to survive

[ebliever]

Those are the objective facts. my opinions

FTFY

If you are going to dispute what he says, it would be helpful if you could show how he is wrong. It rather looks to me like the facts are on his side, and your opinion is only that. Given storage limitations for the foreseeable future how can you believe a cryptocurrency can keep transactions on-chain while scaling up massively from the current state?

[1714858963]

1714858963

[1714858963]

1714858963

[Carlton Banks]

Also, if we can reach high volumes of transactions with a layer on top of the existing network

That's called a banking network.  It can only work through centralization.  LN hubs are banks, knowing you, knowing what you do, and blocking what you want to do eventually, reporting you, asking you to pay for their services.... like normal banks.  If you are fine with that, why not use normal banks and fiat ?  It works since ages.

No, not like banking at all. Hubs may well develop naturally, but they're not the network model for Lightning, one can just as easily find one's self connected to a very large Bitcoin node, that does not mean Bitcoin is centralised around 1 well connected node. Lightning is also peer-2-peer, no different to Bitcoin.


If there is no-one forcing anyone to use certain LN nodes, and the only barrier to entry is running a Bitcoin full node, I'm failing to see any loss of control, or any centralisation in that.

[Matthias Green]

Well powers of individual gadgets are growing, so if any device, for example, contains a dedicated portion of calculation power to permanently maintain hashing, and with the development of wireless energy, it may look possible. But the face of the entire industry will change with the help of such means.  

[dinofelis]

No, not like banking at all. Hubs may well develop naturally, but they're not the network model for Lightning, one can just as easily find one's self connected to a very large Bitcoin node, that does not mean Bitcoin is centralised around 1 well connected node. Lightning is also peer-2-peer, no different to Bitcoin.

Using LN has only a meaning if you send many transactions through the same channel.  If you do this P2P, then this is just the old "micropayments".  LN only has a meaning if you get further away than one hop.  If you are not immensely rich in BTC, you cannot keep open a lot of channels with enough coins to a lot of peers, without having to settle after a few transactions.  If off-chain transactions are orders of magnitude more frequent than on-chain transactions (with high fees), your channels should, on average, transmit that number of transactions before settling, otherwise you will lose out.  Your settling fees per channel will be divided by the number of transactions you can get through a channel to give you the cost per transaction to you.  You see the huge economies of scale here.  The bigger the hubs are, the better fees they can offer, the better they can guarantee that they will be able to transact.  Also, the better the connections are between the hubs your transaction will use, the better it works.  In fact, ONE CENTRAL BIG HUB with all people connected to it would be the optimal LN.

If there is no-one forcing anyone to use certain LN nodes, and the only barrier to entry is running a Bitcoin full node, I'm failing to see any loss of control, or any centralisation in that.

Because opening a channel is locking in your coins until settlement, which means paying a (big) fee on-chain, and transmitting a transaction can induce a settlement on another channel (and its inherent fee risk).  You can also not use the same coins on different channels.

If you, as a small fish, set up, say, 8 channels to 8 other nodes, you can only lock in at most 1/8 of your bitcoin holdings into each of the channels, which will settle each time your unknown other node settles, and which will settle each time the sum of transactions you treated go beyond 1/8 of your holdings.

[RodeoX]

I used BTC as cash when it traded for pennies and I use it at it's current valuation. Works just the same. It's money because that's how I use it. I can't see anyone convincing me otherwise while this is still true.

[jonald_fyookball]

Those are the objective facts. my opinions

FTFY

If you are going to dispute what he says, it would be helpful if you could show how he is wrong. It rather looks to me like the facts are on his side, and your opinion is only that. Given storage limitations for the foreseeable future how can you believe a cryptocurrency can keep transactions on-chain while scaling up massively from the current state?

We don't know all the ways that a cryptocurrency can scale, how long it will take to reach 'world wide levels' , what kinds of bandwidth technology will be available in the future.  Heck we can't even agree on what constitutes node centralization. So to make a blanket statement that "it cant be this or that" is presumptive.

[dinofelis]

We don't know all the ways that a cryptocurrency can scale, how long it will take to reach 'world wide levels' , what kinds of bandwidth technology will be available in the future.

If your statement is that one doesn't yet know all ways, beyond "writing all transactions in a single block chain", to make a crypto currency, *you are perfectly right*.  This "writing all transactions in a single big block chain" was a first idea, introduced by Satoshi, on which most alt coins today are based too, and THIS kind of crypto currency, CLEARLY HAS SCALING PROBLEMS.

I think that Moore's law is a good indication of future technological growth at best.  In order for such a crypto currency to be able to swiftly deal with most daily transactions in the world, we need about 5 orders of magnitude.  Because VISA and Mastercard are NOT handling all the transactions in the world either, and are 4 orders of magnitude higher in transactions per second.  But there are a lot of other payments, coffee is typically done with cash. 

5 orders of magnitude with Moore's law requires about 25-30 years.  So, 30 years from now, block chains where we write all transactions like with bitcoin, are going to be able to deal with all transactions world wide.  Over 30 years, most of the currently existing crypto currencies will have run in fundamental problems, and be entirely centralized.

But it will not take 30 years to discover smarter ways to do crypto payments without having this burden.  There ARE schemes in development which can do the essential function of crypto in much leaner ways.

[Carlton Banks]

ONE CENTRAL BIG HUB with all people connected to it would be the optimal LN.

All good, except this part. That's not "optimal", that's "cheapest", which are separate concepts.

You will have a choice. Don't need super high levels of privacy to buy a packet of chewing gum? Then use a larger channel to save on fees (and who says the street vendor will let you use that large channel without sending it through THEIR node first? Everyone competes on a level playing field, as there is no real barrier to be a Lightning network node)

Because opening a channel is locking in your coins until settlement, which means paying a (big) fee on-chain, and transmitting a transaction can induce a settlement on another channel (and its inherent fee risk).  You can also not use the same coins on different channels.

If you, as a small fish, set up, say, 8 channels to 8 other nodes, you can only lock in at most 1/8 of your bitcoin holdings into each of the channels, which will settle each time your unknown other node settles, and off which will settle each time the sum of transactions you treated go beyond 1/8 of your holdings.

No, you've already accepted that aggregating multiple LN transactions into 1 on-chain transaction will allow far more transaction capacity, which will shift the more meager capacity away from the on-chain Bitcoin network. Settlement fees could end up cheaper than today's transaction fees no problem. Settlement transactions are hardly highest priority, and that together with the effects of some capacity shifted onto LN will create positive and immutable incentives to keep on-chain AND off-chain transaction relatively cheap. Fewer people needing on-chain tx's naturally frees up on-chain capacity.

The positive incentive to settle to one's detriment doesn't exist, all honest participants will want to save money, not cost themselves and others unnecessary fees. Of course, people could use malign incentives to deliberately close channels earlier than makes economic sense, but it's not like Bitcoin miners are just nice people wanting to help everyone out right now, our current 3rd parties are also being uncooperative and obstructive towards the network's best interests. If anything, LN improves that situation, as we can choose who processes our LN transactions, and deny business to those who want to be anti-social or destructive.

[dinofelis]

ONE CENTRAL BIG HUB with all people connected to it would be the optimal LN.

All good, except this part. That's not "optimal", that's "cheapest", which are separate concepts.

Same notion.

You will have a choice. Don't need super high levels of privacy to buy a packet of chewing gum? Then use a larger channel to save on fees (and who says the street vendor will let you use that large channel without sending it through THEIR node first? Everyone competes on a level playing field, as there is no real barrier to be a Lightning network node)

It would be totally ridiculous to have to set up a LN channel for a single payment, because setting up a LN channel will at least involve ONE on-chain payment.  So this is nothing else but an on-chain transaction.

I'm not talking about privacy.  With bitcoin being a transparent ledger, there is no privacy with on-chain transactions.  LN and banking is more private than on-chain bitcoin.  Use monero or something like that if you need privacy in crypto.  Don't use bitcoin.  So this is not the issue.  The point with banking is that banks impose rules, impose fees, impose restrictions, impose reporting, impose a lot of things.  If you are OK with all these things, FIAT BANKS are better than "guy on the internet" banks.  There's nothing that LN can offer, that normal fiat banks can't offer.

Crypto, with all its burden, has an edge over fiat in those cases where fiat cannot be used.  But this edge is entirely lost with a LN banking layer.  The fiat banking layer works better than the LN banking layer ever can.

No, you've already accepted that aggregating multiple LN transactions into 1 on-chain transaction will allow far more transaction capacity, which will shift the more meager capacity away from the on-chain Bitcoin network. Settlement fees could end up cheaper than today's transaction fees no problem.

If your transaction on the "amateur LN network" needs, say, 20 hops to get to destination, that would mean that ON AVERAGE, every amateur LN channel needs to process at least 20 transactions in order even to BREAK EVEN with on chain transactions.  That means that ON AVERAGE, every amateur LN channel needs to have locked in at least 20 times the average amount of a transaction.  

If, on average, a LN transaction is, say, $50,- and every amateur node has, say, 8 links to its neighbours, and it takes 20 hops to get to destination, that means that I need to lock in $8000,- on my node, to break even: in that case, a LN transaction will be just as expensive as an on chain transaction.  If I need to offer 10 times smaller fees, I need to lock in $80 000,- and hope I'll not have too many spurious settlements or I'm losing money on fees.

Settlement transactions are hardly highest priority, and that together with the effects of some capacity shifted onto LN will create positive and immutable incentives to keep on-chain AND off-chain transaction relatively cheap. Fewer people needing on-chain tx's naturally frees up on-chain capacity.

Nope.  Settlement transactions are extremely urgent, because they have to be locked in before they expire.  And they lock in my havings.
The chain will be full of settlement transactions, because that's the scaling idea.  

Remember, if on average a LN transaction needs 20 hops, every channel needs to process at least 20 transactions before settlement in order to have a lower rate of settlement transactions on chain than if the transactions were simply executed on chain directly.

You can get your channels much longer open if you have BIG amounts and a lot of traffic in both ways.  Again, this gives a HUGE economy of scale for big hubs.

What fool is going to lock in his bitcoins into channels if he doesn't have mountains of them, with almost no hope to get something out of it, and with a high risk of it costing him money, and not having his coins available when he needs them ?

[dinofelis]

The positive incentive to settle to one's detriment doesn't exist, all honest participants will want to save money, not cost themselves and others unnecessary fees.

Well, if you suddenly see a lovely pair of shoes you want to buy, and your bitcoins are locked in a channel, you settle the channel to be able to buy them.  Too bad for your counterparty.

[jonald_fyookball]

But it will not take 30 years to discover smarter ways to do crypto payments without having this burden.  There ARE schemes in development which can do the essential function of crypto in much leaner ways.

I'd be interested to hear about these schemes.

It's a little disappointing to see defeatist attitudes and limiting beliefs when we should be inspired by Satoshi to think outside of the box.

[dinofelis]

I'd be interested to hear about these schemes.

Well, to give you an idea, Byteball has something else than a block chain.  The system is not entirely well-designed and has a lock-in danger, and it contains other problems.  But it contains a significant conceptual step.   I'm myself also exploring something of the kind, but it is still way way too early to talk about it, as I may be entirely wrong, but there's more than just one way of "keeping accounts", and recording all transactions of everybody is not the only way.

[Kprawn]

Did Satoshi create this to be a Store of Value or a alternative digital P2P payment option? I think his vision was to replace banks & eventually fiat

currencies and to scale this to enable that function. The fees should increase over time to replace the Block reward, and for that to happen... you

will need MASS adoption.  ... I think you made a valid statement. 

[dinofelis]

Did Satoshi create this to be a Store of Value or a alternative digital P2P payment option? I think his vision was to replace banks & eventually fiat

The problem is that the technique he invented can't handle this, and this was pointed out to him already early on.  But as he didn't know anything better, he made bitcoin.  Which was a very good start as a first working crypto.  What was a bigger blunder, was that he didn't fully grasp the consequences of his own invention of trustless consensus, and that he has frozen in a parameter (1MB blocks) without realizing that it was frozen in.

[Carlton Banks]

The positive incentive to settle to one's detriment doesn't exist, all honest participants will want to save money, not cost themselves and others unnecessary fees.

Well, if you suddenly see a lovely pair of shoes you want to buy, and your bitcoins are locked in a channel, you settle the channel to be able to buy them.  Too bad for your counterparty.

It would make much more sense to buy the shoes with the money in the channel? If you have money in a channel, it can be routed to anywhere else on the Lightning network, that's how Lightning is designed. You basically don't get LN. Not my problem, I'm afraid.

[AngryDwarf]

If there is no-one forcing anyone to use certain LN nodes, and the only barrier to entry is running a Bitcoin full node, I'm failing to see any loss of control, or any centralisation in that.

I thought the idea of segwit was that people could use LN without running a full node. Or do we need to shrink full nodes so that they can be run on a mobile phone?

[Carlton Banks]

If there is no-one forcing anyone to use certain LN nodes, and the only barrier to entry is running a Bitcoin full node, I'm failing to see any loss of control, or any centralisation in that.

I thought the idea of segwit was that people could use LN without running a full node. Or do we need to shrink full nodes so that they can be run on a mobile phone?

Just like the difference between full nodes and SPV nodes, you can run Lightning nodes either using a full node or using a lite node. Appropriate trade-offs apply, again, just like they do with lite nodes and full nodes. i.e. there's a desktop version with more capabilities that uses more resources, and a version with stripped down capabilities with stripped down resources to use it.