Fundamental problem: Lightning Network leads to centralization and less security

[Carlton Banks]

Feel free to use the ignore button if i waste your time. Just like everybody else can do who feels the same. In the meantime i will try to have an productive discussion. I'm always glad to learn something new or be corrected, but expect that this happens in a civilized and helpful manner.

You have little idea how Lightning works, and are incapable of either understanding or even reading very simplistic explanations to aid your broken understanding. Please stop wasting our time until you can demonstrate you are willing to learn

[d5000]

The LN protocol can only lead to centralization of the LN network.
If we find it too centralized, we can use another network based on another protocol, such as Tumblebit, which also provides payment hubs and payment channels.

That is true, but there is a possible drawback: It wouldn't help adoption if users must choose between different payment methods, all with different potential security models. The on-chain mechanism is still fairly easy to understand, LN adds a lot of complexity - which can be hidden, obviously, but the users must understand the underlying ideas at least a bit to be able to calculate risks. The more methods that are competing, the higher the barrier to enter the system for new users.

That doesn't mean I only would approve on-chain methods. For me, (pegged) sidechains are the most interesting way to scale because the on-blockchain transaction method is only slightly modified and the concept is easy to understand, and so are the risks (51% attack on weak sidechains). LN in my opinion is more a kind of prepaid card replacement for very small micropayments - I would use it this way even if it lead to more centralization, as it would be still better than a fully centralized third-party-service.

[Senor.Bla]

The LN protocol can only lead to centralization of the LN network.
If we find it too centralized, we can use another network based on another protocol, such as Tumblebit, which also provides payment hubs and payment channels.

That is true, but there is a possible drawback: It wouldn't help adoption if users must choose between different payment methods, all with different potential security models. The on-chain mechanism is still fairly easy to understand, LN adds a lot of complexity - which can be hidden, obviously, but the users must understand the underlying ideas at least a bit to be able to calculate risks. The more methods that are competing, the higher the barrier to enter the system for new users.

If we find LN to centralized, then we could change to another service provider. At least in theory. First of there need to be other services we could use and secondly they should be on more or less on the same level. If there are alternatives but they are not sophisticated enough, then they are no help.
We can often see that we end up with two or three market leaders. For credit cards it Visa, MasterCard and American Express. How about smartphone OS? We only use Apple or Android.
I don't see more then three big 2nd layer solutions. What if they all centralize? Once they are established it'll be even hard for a good product to enter the market. Why? I think it will be the network effect. If you have all the users they will be unhappy but still stay with you. Look at Facebook. There are alternatives, but nobody uses them. Why? Because all my friends are at FB.
If all my merchants and friends use LN then i will not switch to an alternative where i can not pay them, because they all just use LN.

And yes, this might cause just a little drawback in Bitcoin adoption. If people have to choose it's always a tiny bit harder then only having one option and deciding to join. But right now we have a whole bunch of Wallets to chose from and it is not stopping people from using Bitcoin. Also i don't think people will chose because of the technical underlining or security. They will chose depending on what they friends use, the costs and maybe some convenient features, which tend to be less safe. They won't care about centralization.

[Carlton Banks]

The LN protocol can only lead to centralization of the LN network.

That is true

No it isn't


Please either bring articulable knowledge to the discussion about the topic, or do us all a favour and be quiet. You're not backing your statement up with any facts, and you're saying things that aren't true

[DooMAD]

The LN protocol can only lead to centralization of the LN network.

That is true

No it isn't


Please either bring articulable knowledge to the discussion about the topic, or do us all a favour and be quiet. You're not backing your statement up with any facts, and you're saying things that aren't true

Either you're trolling again, Carlton, or your bias is now so strong that you're selectively quoting out of context without even realising it.  Stop jumping down someone's throat over your own failure at reading comprehension.  Let's try that quote you replied to again without the twisted bias on your part:

If we find it too centralized, we can use another network based on another protocol, such as Tumblebit, which also provides payment hubs and payment channels.

That is true, but there is a possible drawback: It wouldn't help adoption if users must choose between different payment methods, all with different potential security models (...)

That's how anyone with an open mind will be reading d5000's reply.  But keep being the mindless, rabid attack dog that you are, Carlton, frenzied at the slightest misunderstanding.  It continues to do nothing for your credibility.

[Carlton Banks]

Either you're trolling again, Carlton, or your bias is now so strong that you're selectively quoting out of context without even realising it.  Stop jumping down someone's throat over your own failure at reading comprehension.  Let's try that quote you replied to again without the twisted bias on your part:

But keep being the mindless, rabid attack dog that you are, Carlton, frenzied at the slightest misunderstanding.  It continues to do nothing for your credibility.

Just like the rest of your ilk, you're still obsessed with individuals (mostly me) and personalities, incapable of arguing straightly or honestly about technical points.


You're wrong, that little exchange between d5000 and Boussac was not clear at all. I provided clarity.


Is there something wrong with the content of what I said? There is not, which is why, ironically, you can only attack my character. Who's the real troll

[DooMAD]

Either you're trolling again, Carlton, or your bias is now so strong that you're selectively quoting out of context without even realising it.  Stop jumping down someone's throat over your own failure at reading comprehension.  Let's try that quote you replied to again without the twisted bias on your part:

But keep being the mindless, rabid attack dog that you are, Carlton, frenzied at the slightest misunderstanding.  It continues to do nothing for your credibility.

Just like the rest of your ilk, you're still obsessed with individuals (mostly me) and personalities, incapable of arguing straightly or honestly about technical points.

You were warned that you were now on my shitlist and that you would now be singularised for the brunt of my wrath because you insist on constantly being a pernicious dick.  I'm afraid you've brought this on yourself.  People were trying to discuss technical points before you stormed in to go for the throat.

That little exchange between d5000 and Boussac was not clear at all. I provided clarity.

You jumping to the wrong conclusion isn't clarity.  It might look a bit clearer if bias wasn't clouding your vision.

Is there something wrong with the content of what I said? There is not, which is why, ironically, you can only attack my character. Who's the real troll

You insulted someone because you misunderstood something.  You literally just attacked d5000's character and intelligence for no reason.  So yes, there's something wrong with the content of what you said.  There's evident justification for attacking your character.  You are the troll.

Back to the topic:

If we find it too centralized, we can use another network based on another protocol, such as Tumblebit, which also provides payment hubs and payment channels.

That is true, but there is a possible drawback: It wouldn't help adoption if users must choose between different payment methods, all with different potential security models. The on-chain mechanism is still fairly easy to understand, LN adds a lot of complexity - which can be hidden, obviously, but the users must understand the underlying ideas at least a bit to be able to calculate risks. The more methods that are competing, the higher the barrier to enter the system for new users.

If we find LN to centralized, then we could change to another service provider. At least in theory. First of there need to be other services we could use and secondly they should be on more or less on the same level. If there are alternatives but they are not sophisticated enough, then they are no help.
We can often see that we end up with two or three market leaders. For credit cards it Visa, MasterCard and American Express. How about smartphone OS? We only use Apple or Android.
I don't see more then three big 2nd layer solutions. What if they all centralize? Once they are established it'll be even hard for a good product to enter the market. Why? I think it will be the network effect. If you have all the users they will be unhappy but still stay with you. Look at Facebook. There are alternatives, but nobody uses them. Why? Because all my friends are at FB.
If all my merchants and friends use LN then i will not switch to an alternative where i can not pay them, because they all just use LN.

And yes, this might cause just a little drawback in Bitcoin adoption. If people have to choose it's always a tiny bit harder then only having one option and deciding to join. But right now we have a whole bunch of Wallets to chose from and it is not stopping people from using Bitcoin. Also i don't think people will chose because of the technical underlining or security. They will chose depending on what they friends use, the costs and maybe some convenient features, which tend to be less safe. They won't care about centralization.

One would hope that the way LN is implemented would mean that it's not an either/or style choice.  It should just be another part of Bitcoin.  An option to open a channel, send your transactions back and forth without any delays, and then settle when you're done.  It's also best to not think of LN as a "service provider" as such.  There may well be service providers that utilise LN and those services would obviously be fairly centralised, but LN itself is closer to a protocol and should hopefully remain as neutral and impartial as Bitcoin's existing protocol.  But we need to be careful how it's implemented to ensure that happens.

[classicsucks]

Somebody correct me if i'm wrong, but LN does not effect Bitcoin directly. Nobody is forced to use LN. Even with LN we can all just use Bitcoin. So LN is not a risk to Bitcoin.

Some are saying that Blockstream is deliberately keeping the blocksize small to force people onto LN. I haven't seen proof of this theory, but they are the primary developers of LN, and they've received $75 million based on their assertion that people will need to use LN. If it's true, it would be a major risk for bitcoin.

[Carlton Banks]

Somebody correct me if i'm wrong, but LN does not effect Bitcoin directly. Nobody is forced to use LN. Even with LN we can all just use Bitcoin. So LN is not a risk to Bitcoin.

Some are saying that Blockstream is deliberately keeping the blocksize small to force people onto LN. I haven't seen proof of this theory, but they are the primary developers of LN, and they've received $75 million based on their assertion that people will need to use LN. If it's true, it would be a major risk for bitcoin.

There are multiple implementations of LN in ongoing development, 5 the last time I checked. Blockstream are only 1 of that 5.

You cannot qualify your statement that 5+ organisations developing their own form of Lightning protocol is a risk to Bitcoin.

The truth is the opposite: healthy competition in LN implementations can only help to improve Bitcoin's transaction capacity by orders of magnitude higher than any workable blocksize increase ever could, and provide strength in diversity should one of the many designs develops a problem.

[Senor.Bla]

...
Back to the topic:

If we find it too centralized, we can use another network based on another protocol, such as Tumblebit, which also provides payment hubs and payment channels.

That is true, but there is a possible drawback: It wouldn't help adoption if users must choose between different payment methods, all with different potential security models. The on-chain mechanism is still fairly easy to understand, LN adds a lot of complexity - which can be hidden, obviously, but the users must understand the underlying ideas at least a bit to be able to calculate risks. The more methods that are competing, the higher the barrier to enter the system for new users.

If we find LN to centralized, then we could change to another service provider. At least in theory. First of there need to be other services we could use and secondly they should be on more or less on the same level. If there are alternatives but they are not sophisticated enough, then they are no help.
We can often see that we end up with two or three market leaders. For credit cards it Visa, MasterCard and American Express. How about smartphone OS? We only use Apple or Android.
I don't see more then three big 2nd layer solutions. What if they all centralize? Once they are established it'll be even hard for a good product to enter the market. Why? I think it will be the network effect. If you have all the users they will be unhappy but still stay with you. Look at Facebook. There are alternatives, but nobody uses them. Why? Because all my friends are at FB.
If all my merchants and friends use LN then i will not switch to an alternative where i can not pay them, because they all just use LN.

And yes, this might cause just a little drawback in Bitcoin adoption. If people have to choose it's always a tiny bit harder then only having one option and deciding to join. But right now we have a whole bunch of Wallets to chose from and it is not stopping people from using Bitcoin. Also i don't think people will chose because of the technical underlining or security. They will chose depending on what they friends use, the costs and maybe some convenient features, which tend to be less safe. They won't care about centralization.

One would hope that the way LN is implemented would mean that it's not an either/or style choice.  It should just be another part of Bitcoin.  An option to open a channel, send your transactions back and forth without any delays, and then settle when you're done.  It's also best to not think of LN as a "service provider" as such.  There may well be service providers that utilise LN and those services would obviously be fairly centralised, but LN itself is closer to a protocol and should hopefully remain as neutral and impartial as Bitcoin's existing protocol.  But we need to be careful how it's implemented to ensure that happens.

That is a good point. I tend to have a picture in my mind where LN will end up controlled by a few hubs. Opening a payment channel between two parties is already possible with Bitcoin as far as i know.
The more interesting part is that all participants of a channel can interact. So everybody using Amazon or Apple would be potentially connected. I would expect Amazon and Apple to run there own Hubs. Which would them give certainly a lot of power and they might even decide to charge a fee. I also can not see how google is missing out on collecting a huge amount of data running own Hubs. Wouldn't that be a further compromise to privacy?

[Carlton Banks]

The more interesting part is that all participants of a channel can interact. So everybody using Amazon or Apple would be potentially connected. I would expect Amazon and Apple to run there own Hubs. Which would them give certainly a lot of power and they might even decide to charge a fee. I also can not see how google is missing out on collecting a huge amount of data running own Hubs. Wouldn't that be a further compromise to privacy?

Lightning is peer to peer in essence. That means there is no hub/node distinction, everyone is a node, anyone can start a node.

Sure, some nodes will be more popular than others, but everyone can decide for themselves whether to use a cheap high volume channel (suitable when you don't care about privacy), or a more expensive low volume channel (suitable for privacy sensitive payments), or anything in between as you choose.

The main point is that because Lightning is p2p, it's a platform for a diversity of needs. Choose your own compromise/trade-off.

[mezzomix]

Some are saying that Blockstream is deliberately keeping the blocksize small to force people onto LN. I haven't seen proof of this theory, ...

There can't be a proof because this theory is wrong. segwit allows more transactions per block than the current system.

[Senor.Bla]

The more interesting part is that all participants of a channel can interact. So everybody using Amazon or Apple would be potentially connected. I would expect Amazon and Apple to run there own Hubs. Which would them give certainly a lot of power and they might even decide to charge a fee. I also can not see how google is missing out on collecting a huge amount of data running own Hubs. Wouldn't that be a further compromise to privacy?

Lightning is peer to peer in essence. That means there is no hub/node distinction, everyone is a node, anyone can start a node.

Sure, some nodes will be more popular than others, but everyone can decide for themselves whether to use a cheap high volume channel (suitable when you don't care about privacy), or a more expensive low volume channel (suitable for privacy sensitive payments), or anything in between as you choose.

The main point is that because Lightning is p2p, it's a platform for a diversity of needs. Choose your own compromise/trade-off.

I need one more thing to be clarified.
Alice and Bob opened a LN channel between them. So did Carol and Dan.
Know i was under the impression that Alice could not send funds to Dan, since the two channels are not connected. Just because they all use LN doesn't mean that they are connected or is this assumption wrong?
But if for example Bob and Carol would also have an open channel, then this would link the two channels (A-B and C-D), so that Alice could finally send funds to Bob.
Alice would send funds through Bob and then Carol to Dan. As i heard they all would have to be online to do this. Is this correct?

[Carlton Banks]

if for example Bob and Carol would also have an open channel, then this would link the two channels (A-B and C-D), so that Alice could finally send funds to Bob.
Alice would send funds through Bob and then Carol to Dan. As i heard they all would have to be online to do this. Is this correct?

Yes.

Routing through others connected to your LN node is part of what makes Lightning attractive, a high quality of privacy can be established that way (better than on-chain Bitcoin, only the parties directly connected to you are aware of your transactions, unlike when published on the public blockchain)

They don't have to be permanently online necessarily, but it's ideal if they are. There is at least one way to exploit channels to steal from others when one party goes offline, but not if they stay online. This is a significant flaw IMO, but it's not impossible that a p2p solution could be found for this also (a 3rd party solution has been devised). It's all still a work in progress.

[classicsucks]

Somebody correct me if i'm wrong, but LN does not effect Bitcoin directly. Nobody is forced to use LN. Even with LN we can all just use Bitcoin. So LN is not a risk to Bitcoin.

Some are saying that Blockstream is deliberately keeping the blocksize small to force people onto LN. I haven't seen proof of this theory, but they are the primary developers of LN, and they've received $75 million based on their assertion that people will need to use LN. If it's true, it would be a major risk for bitcoin.

There are multiple implementations of LN in ongoing development, 5 the last time I checked. Blockstream are only 1 of that 5.

You cannot qualify your statement that 5+ organisations developing their own form of Lightning protocol is a risk to Bitcoin.

The truth is the opposite: healthy competition in LN implementations can only help to improve Bitcoin's transaction capacity by orders of magnitude higher than any workable blocksize increase ever could, and provide strength in diversity should one of the many designs develops a problem.

I stand corrected about Blockstream's LN implementation. Is it fair to say that their LN code is the most popular version given that they are a $75 million corporation and the prime mover for the technology?

LN could be a threat to bitcoin in that the main cheerleaders for it are also blocking bitcoin from working properly. If LN worked and solved all of the problems, I'd be happy. As I said before, I doubt LN will ever see widespread use, and altcoins will simply take more and more market share until if/when core wakes up and smells the coffee.  Isn't it obvious that the majority isn't interested in LN based on the massive resistance to Segwit? LN will likely go the way of Segwit - another over-complicated dorky solution that never got off the ground...

Lightning is peer to peer in essence. That means there is no hub/node distinction, everyone is a node, anyone can start a node.

"Anyone" can run a bitcoin node as well, but only ~5000 do... You're forgetting that people need to be tech-savvy and security-conscious, which eliminates 99.9% of the population off the bat. Also they need to be prepared to spend money on hosting and spend time adminstering the node.


Finally, I don't want to open a payment channel with anyone, I just want to pay them and get paid by them, and that technology exists today and is(was) working until recently. As others have said, with LN we could well be headed toward using the mega-corporate channels with the corporations data mining and spying. All sounds very familiar to me...

[Carlton Banks]

Isn't it obvious that the majority isn't interested in LN based on the massive resistance to Segwit?

Segwit compatible nodes are 51% of the network. And rising every day. If that's the majority of users rejecting it, I don't know what would convince you. 101% segwit nodes, maybe?

"Anyone" can run a bitcoin node as well, but only ~5000 do...

Bitcoin nodes are ~ 6000. And rising every day. You're in favour of changes that endanger that trend.

Finally, I don't want to open a payment channel with anyone, I just want to pay them and get paid by them, and that technology exists today and is(was) working until recently. As others have said, with LN we could well be headed toward using the mega-corporate channels with the corporations data mining and spying. All sounds very familiar to me...

Which part of "anyone can start a p2p Lightning node" don't you understand?

[Wind_FURY]

If you check who wrote the article, he is a big blocker. Then of course his opinions will be biased against Segwit and the Lightning Network. He is more willing to take the risks of a hard fork to Bitcoin Unlimited than take a safer path in Segwit. What does that tell you about him? This is more politics than a technical argument.

[classicsucks]

Isn't it obvious that the majority isn't interested in LN based on the massive resistance to Segwit?

Segwit compatible nodes are 51% of the network. And rising every day. If that's the majority of users rejecting it, I don't know what would convince you. 101% segwit nodes, maybe?

Nice try, miner support for Segwit is 25.6%  Keep on wishing for that 101% and keep on ignoring the giant mempool and borderline-unusable-for-normal-transactions bitcoin network.

"Anyone" can run a bitcoin node as well, but only ~5000 do...

Bitcoin nodes are ~ 6000. And rising every day. You're in favour of changes that endanger that trend.

First of all, don't put words in my mouth. If I'm a critic of Core/Segwit/LN, that doesn't automatically make me a BU/Hard Fork fanboi. BTW the Chinese introduced a synthetic fork proposal that was completely ignored here on this echo-chamber forum. Study, learn, and think for yourself.

Second, if you're claiming 6000 bitcoin nodes then you're stretching a bit, and you're including the 700+ Unlimited nodes, LOL. I've heard anectodally that the total number of bitcoin nodes was at 20k at one point, but fell as the blockchain size increased dramatically.  Do you think that 6000 LN nodes would make the network secure and allow payments to be process at the claimed rates?

Finally, I don't want to open a payment channel with anyone, I just want to pay them and get paid by them, and that technology exists today and is(was) working until recently. As others have said, with LN we could well be headed toward using the mega-corporate channels with the corporations data mining and spying. All sounds very familiar to me...

Which part of "anyone can start a p2p Lightning node" don't you understand?

"Anyone can swap the engine in their Honda Civic".
"Anyone can become an astronaut"

I find it frightening that you're proposing that random fools should run a node in this secure network. Imagine the security implications...

[Carlton Banks]

Nice try, miner support for Segwit is 25.6%  Keep on wishing for that 101% and keep on ignoring the giant mempool and borderline-unusable-for-normal-transactions bitcoin network.

Segwit compatible nodes are 51% of the network. You didn't ask about miners, you stated that a majority of users had rejected it.


If you really think moving the goalposts is how to win the game, you must be more stupid than the audience you believe you can fool

if you're claiming 6000 bitcoin nodes then you're stretching a bit, and you're including the 700+ Unlimited nodes, LOL. I've heard anectodally that the total number of bitcoin nodes was at 20k at one point, but fell as the blockchain size increased dramatically.  Do you think that 6000 LN nodes would make the network secure and allow payments to be process at the claimed rates?

They're all Bitcoin nodes, because they have to behave like Bitcoin nodes to participate in the network, by definition. Anything that doesn't observe the rules is rejected, you're arguing non-pertinent semantics, not technical details

Which part of "anyone can start a p2p Lightning node" don't you understand?

"Anyone can swap the engine in their Honda Civic".
"Anyone can become an astronaut"

I find it frightening that you're proposing that random fools should run a node in this secure network. Imagine the security implications...

More concerned that you believe using the Bitcoin software "isn't for everyone". There are no security implications, it only strengthens the network. Stop using unqualified statements to discourage people to run a node, it's a highly irresponsible falsehood

[classicsucks]

Segwit compatible nodes are 51% of the network. You didn't ask about miners, you stated that a majority of users had rejected it.

If you really think moving the goalposts is how to win the game, you must be more stupid than the audience you believe you can fool

Sorry, "the game" is already lost for you and Core as well. Segwit requires 95% consensus by miners and nodes and therefore won't happen. No need to move any goalposts. No need to split hairs about who supports what - it's over and neither Segwit nor BU will be adopted. However, it's the first time a Core release has been soundly ignored. For some reason they have become more shrill, rather than more humble.

I find it frightening that you're proposing that random fools should run a node in this secure network. Imagine the security implications...

More concerned that you believe using the Bitcoin software "isn't for everyone". There are no security implications, it only strengthens the network. Stop using unqualified statements to discourage people to run a node, it's a highly irresponsible falsehood

Now you're just getting ridiculous, my friend - can you not see that there is a distinction between bitcoin users and bitcoin node operators?

Paying for and running a server 24/7 that processes financial tranasactions isn't for everyone. Can you at least agree with that?