Topic: Share your opinions of Lightning (Read 2031 times)

For me, Lightning is the equivalent of a prepaid card in the "fiat" world. It works fast and without much hassle, but it is a little bit less secure than the blockchain. So I would use it for payments up to $50, maybe $100 if everything works well. But I wouldn't like to get my salary via it.

Two things I didn't fully understand, though - maybe someone here can answer them:
1) Let's say there are some very big hubs that have hundreds of thousands or even millions of channels open. What happens if one of these hubs decides to scam his clients and resets the channels? Is the blockchain capacity high enough to allow a majority of its clients to close the channel in time?

2) What happens if one node A gets used as a "hub" mostly in one direction (X -> A -> Y, with very few payments flowing from Y -> A -> X)? An example: X is a group of miners and Y a group of exchanges, and A is the biggest "connector" between them. The flow to the miners would be less important than the flow to the exchanges. Could this be a problem or is it a non-issue?

I am sure that Lightning will someday be lost and unconsciously it will give negative effect to bitcoin players or other digital currency

The Lightning network is vital for Bitcoin IMHO. The idea that every micro-transaction has to be verified by the entire network makes no sense, it's like saying if I want to buy a sweet from my local market I need the worlds banking network to verify and approve it.

Lightning Network finally makes Bitcoin scalable in terms of transactions per second. It also makes it possible to obfuscate transactions off-chain particularly when combined with Sphinx and Hornet.


The simplest way to think about Lightning Network is to think of how you make transactions in daily life.
If you want to pay me $100,000 then you want some evidence that you've paid me and so you would want to record that transaction to the Blockchain. The transaction is broadcast, recorded and becomes immutable and proveable from that point forward. There is a cost associated to that however which you cover with fees.
If you want to pay me $1 then you would likely be happy to just hand me $1 and forego the fees. The Lightning Network creates a payment channel between you and me so your wallet makes a note that its given me $1 and my wallet makes a note that you've paid me $1. You and I perhaps make regular payments between us so I pay you $2 for something then I pay you $3 for something else and you pay me $2.50 for something. The balance is I owe you $1.50. You or I decide we want to persist that to the blockchain so we instruct the wallet to write a transaction for $1.50 to the blockchain.

Several important benefits just occurred.

1. We just skipped all those micro-transaction fees and just had one final fee so we saved money.
2. An observer of the blockchain can't see what micro-transactions happened so there's some obfuscation of transactions there.

Now you can extract that by considering a scenario where I want to pay someone called George who I don't know and so I don't have a payment channel to. But I do know and trust you with our payment channel and you know and trust someone called Jane with a payment channel and Jane knows and trusts George with a payment channel so if I give you $1 and you give that to Jane and she gives it to George then the net result is that George has been paid by me but there's still no fees and the transaction is still obfuscated. Each channel may be written out at different times so from an observer it becomes difficult to unpick payments moving about. Not impossible and if I'm a node in that channel I can see where the money is coming from and going to.

So this is where protocols like Sphinx and Hornet come in by creating encrypted shells similar to how TOR works to hide the flow off money as it happens off-chain.

So Lightning is just moving transactions off-chain and into a peer to peer network.

To be clear, I think LNs are a solution to the payment problems of 99% of normal people. The only UseCases that fall outside of the scenario where LNs exist is one where there would be no fees to pay money. This would require true decentralisation which is something that has not been succesfully done before.

You cannot deny that the economical and practical realities of this scenario, if you would for a second go with me in my thoughts, would look a lot like an improved legacy payment system. Maybe you could try and convice me with arguments about how it will not be centralised in an economical sense? I get that it is a genius technological solution, I am just really worried about the economical and practical implications.

Kind regards

Proposed solutions can flow from action, not necessarily words. Besides, English is not my native language so I am sorry if I did not make myself clear but are LN the solution core wants to go with or are they not?

On your point of the danger of blocksize increases: first of all there are many documented experiments and experienced coders that could assist in such a task since there are chains with bigger blocks based on BTC. Heck, BTC even was BTC unlimited at one point. And you are telling me that the way they are testing LN of the testnet and the way the advertise it as being the perfect solution is rational and safe? They did not only outright refuse to increase the blocksize at the time when this was really needed (and which could have helped a lot actually with market depth, who knows how much), the DoS protection that the blocksize limit was has become obsolete. You need a blocksize that is just above what is needed with normal transactions in the network. That is how Satoshi saw it, how it is logical and how you account for all factors keeping the current Bitcoin economy intact.

To be honest, I was really rather surprised to find a member that has been around for so long brush off this problem of a new player in the economy. Do you really think people who don't use Bitcoin at this moment and will start using it because LNs become SEC regulated money transferring businesses that are way cheaper and faster and better in every and any sense, will stop using it if a ''hub is being a dick''? I find this line of thought does not take into account economical reality. If any LN that is well implemented becomes a sort of PoS for Bitcoin in the sense that liquidity providers van get a fee in BTC for their BTC, people will naturally gravitate towards the best and cheapest and most liquid option. How many times a day do you search something on the internet with Yandex? Not often? Maybe because Google is better? But in your line of reasoning, Google here has no power because you can just choose another search engine. This is in my opinion extremely flawed, because people do not choose a long term prospect of decentralisation or lubby-dubby ''the next hub is cheaper but I'll just stay rght hurr''

Realistically speaking people will gravitate towards large LNs. This provides a new source of power within the Bitcoin economy. 

The lightning network is not something that was proposed by or being worked on by Bitcoin Core developers.

Because we don't know how Segwit and the deployment of LN is going to effect the economy and the transactional capacity. There are consequences to increasing the block size (and there are positives to having full blocks and a mempool that is not emptied on every block) which we do not think are worth having for a larger block size limit. The lightning network is not yet deployed and increasing the block size limit further may be more harmful than not.

Not really. The people who are connected to those hubs can close channels any time that they want. So if a hub is being a dick and refusing to close a channel, then you can close it yourself, albeit you will need to wait a little bit before you can access your coins. Likewise you can do the same the the hub. Furthermore, if a hub were to be engaging in that behavior, many people would likely just stop using them. They would close their channels with that hub and route payments through someone else. Anyone can be a hub, so there likely won't be a shortage of alternative hubs to use and that people are simultaneously connected to.

What makes you think that? There are at least 2 other organizations besides Blockstream who are building software for LN. They employ more people than Blockstream does and all of them are collaborating on making their software compatible via the Lightning RFC spec.

You overestimate the manpower and influence of Blockstream.

First of all, as I said earlier, anyone can be a LN hub. You don't and shouldn't be connected to only one hub. There will certainly be other hubs set up by the other LN developers for people to use.

Secondly, Blockstream is not behind Bitcoin Core nor do they control Bitcoin Core in any way, shape, or form.

I would like to point out that I am not opposed to any solution that would scale the transaction capacity of the Bitcoin network. If we are going to go for market depth with the people that have heard of Bitcoin at this point, we need way more transactions per second that the current network allows for.

That being said, after some research into the proposed solution of the Bitcoin core team, a couple of questions arose that I feel they did not answer with sufficient clarity. For example, even though in the whitepaper for Lightning Networks (LNs) the development team specified that a blocksize increase is still needed, why is there so much resistance to using it now to mitigate full blocks? It is not a problem as of writing this post, but we have seen in recent months that it could very well be and it therefore cannot be used as a reason to not research and implement bigger blocks.
      Secondly, as they described their version of the second layer network, it contained certain technical aspects that required larger ''hubs'' to exist within the network to provide the liquidity to be able to settle the contracts. These providers of liquidity would be incentivised by a fee that is internal to the network. This creates a new and powerfull actor in the Bitcoin economy: these liquidity hubs. They can broadcast, or not broadcast when they want to the main chain. This allows them to maximise their profit, whilst reducing that of the main chain miners. It also presents us with the prospect of a system that feels more like a better legacy payment system than anything revolutionary. It would still not be censorable or anything like that, but we would all be paying a bunch of rich people for using a payment system. Furthermore, the most experienced and trusted source of LNs will become Blockstream who can therefore probably build the best LN. Big whales will naturally gravitate towards their LN, making them the most profitable operation. Their LN would be the best and probably the cheapest on all LNs so acounting for rational actors they will receive all new Bitcoins users and vendors. In the best case scenario there will be a handfull of these LN providers, is this desirable? Blockstream would be behind core, the main LN and could have influence over miners through non-broadcasting, is this desirable?

These are just a couple of things that spring to mind. I do not intend to sound cynical about Bitcoin, I think there are many good Use Cases for what they described in their whitepaper. In fact, I think it might be 99% of what users realistically want out of a payment system. It is just that I have felt like Bitcoin in the past was more of an anarchistic, give-the-power-to-the-people type of project.

tl;dr: I have two concerns about LNs: they still need a blocksize increase so why the delay/refusal and it requires large liquidity hubs which introduces powerfull players (BTC rich people) into the Bitcoin mining economy and development

I hope that someone will convince me that I am wrong about all of my points.

Kind regards and thanks for reading

Right now I feel like lighting network is a little to un-user-friendly to be deployed straight away. It has been proven that it works but I don't think everyone can make it work straight away. We need some sort of easy to understand UI for the public. In general though lightning is a great idea for the long run but I feel like we must do it right.

I think it's a great thing.

It allows bitcoin to scale using layer 2 solutions. I can't realistically see bitcoin going mainstream and solely using the blockchain, it's just not feasible. I believe 2nd layers were always intended and baked into bitcoin from the get go. Lightning allows bitcoin to scale. Is it finished? is it ready now? will it work without any bugs? probably not to all three, but then again what does work right out of the gate. The point is it's got some good devs who can and will make this work for the masses. Also, it doesn't necessarily need segwit to work, but i also believe segwit to be a great addition to bitcoin and makes things such as lightning network much easier going forward.

LN is meant to offload a large volume of recurring transactions, for example for micro payments to streaming services or internet access, or exchange transactions, that currently utilize much of Bitcoin’s valuable transaction capacity, leaving more room on the blockchain for other transactions. But there are a lot of opinions about what it is impossible to implement in practice.

1. There are no "hubs" and "spokes" distinction in the model, you're misinterpreting what people mean when they say that
2. The basic answer to your question is "yes"


Payments between people without direct channels open between each other can be made by using any number of intermediate channel hops. And so the most used channels will end up very fluid and very hub-like. But that includes anyone whose channel is well connected given a particular pattern of payment routing, all but the most underused channels will play some role in routing at some point.

Who you open channels with will depend on your preferences. Someone you know well or personally (regular businesses or friends) you'll be happy to open direct channels with. The less familiar you are with who you're trading with (a street food vendor on vacation maybe), you might be more inclined to use an intermediary channel that you and your trading partner agree to. But as everyone will be routing payments, calling these intermediaries hubs is just short-hand; the reality will be a continuous scale of hub-ness, from more hub-like to less hub-like.

No one has yet to explain to me fully how hubs and channels will work. Are channels restricted to only one payee or can the hub route them elsewhere? If you have to open and close a channel to pay only one person then I may as well just use the regular blockchain as it's likely I'll only ever be paying that person once. That reduces on chain transactions to just the one rather than two for opening and closing a channel.

If the hubs can route payments elsewhere then it's a whole different ballgame.

I think so too. Lightning is a bit of a hindrance for lovers of its own nodes, but for the rest of users it will be beneficial, undoubtedly.

The more I read, the more I understand that I'm far away of the full understanding of blockchain's fundamentals

Will governments still support the idea that transactions will not be reflected on the Blockchain, but rather on sidechains where they will not have access to that data? Will they create their own LN hubs to track and trace off-chain transactions? < Like the NSA did with Exit Nodes on the Tor network? >

This is going to become very interesting, I should stock up on the Popcorn. ^smile^ 

Thank you for being very specific.  Your answer to my confusion certainly illustrates the OP's premise that more information and explanation for those of us technically less savvy.  Hell, I used to be a "rocket scientist" (kind of a joke there) and I can't keep up, not enough hours in the day.

Its implemented from what I know.  Just the network needs to commence.  Anyone know a date?

Lightning is neither a hard fork nor a soft fork. It will be deployed on the main network because Segwit has already activated. There is nothing to "pass". Lightning is not a consensus rule change; nothing needs to be activated or passed in order for it to be deployed. It just needs to be implemented. By deployed, I mean used.

So, Lightning is not a UAHF, but is going to pass because Segwit already was implemented?  It's a done deal, or....am I totally missing something?

That is misinformation at its finest. More reason to have an information drive to educate the community about the risks of having no replay protection in case of a hard fork that would split the blockchain.