Is Lightning network truly centralized?

[OmegaStarScream]

I just started reading this article[1] (didn't fully understand the technical details) where it says and proves that this Lightning network will actually be centralized and also a blog post by CoinTelegraph[1] where Gavin Anderson said the same thing, I know that the community doesn't like him but I believe that when he is right, he is right. If its actually what these people claim, is it something to be worried about? I mean what's the worst case scenario that could happen overtime? (a few years from now)



[1] https://medium.com/@jonaldfyookball/mathematical-proof-that-the-lightning-network-cannot-be-a-decentralized-bitcoin-scaling-solution-1b8147650800
[2] https://cointelegraph.com/news/lightning-network-will-be-highly-centralized-gavin-andresen

[Razick]

I have heard many rumors about the Lightning Network being at least partially centralized. I cannot say if it is 100% though as I am not an expert in it. In fact, I watched a video by one of the main developers and find it quite complex and hard to understand, technically speaking. Heck, I don't even understand https all that well which is the equivalent tech for the internet. At the end of the day, as long as it doesn't interfere with my privacy or ability to control my own money and most importantly improves Bitcoin, then I am all for it.

[thecodebear]

From what I understand it will work like the middle picture. There will be multiple LN providers, just meaning that they have created their own implementation of the LN (several companies have already been working on their own LN implementations). So yes the LN will be routed through these centralized providers who run the LN. There will likely be at least several of them though, creating the middle image where there are multiple LN provider nodes in the graph that route LN transactions from bitcoin users. I don't know if they will all communicate to create one single larger LN network like in the picture or just each LN provider will have their own set of clients (wallets that use their LN), created multiple smaller completely centralized Lightning Networks with only one provider in each network.


I don't really see a huge problem in centralization of LN as long as the different provider nodes all connect together to form one big LN network. Bitcoin itself will still be decentralized, and LN will be somewhat decentralized with major LN centralized hubs.

[gentlemand]

Why would anyone take anything written by Mr Fyookball with anything other than a boulder-sized pinch of salt? He's probably the most hollowed out big blocker of them all.

Let's see an actual LN in operation and how it operates it before deciding anything. Much has been theorised. Nothing has been proven.

[Rinaze]

I had a feeling (take it with a pinch of salt) that LTC will probably implement LN faster than BTC. You will have your case study by then!

[Sevvero]

LN is centralized in small banker hubs. You don't have to use them. If you pay 50$ fee you can send via on chain.

[OmegaStarScream]

LN is centralized in small banker hubs. You don't have to use them. If you pay 50$ fee you can send via on chain.

If you don't like bitcoin, its fine, you can use something else but stop spreading FUD. I've been using bitcoin for three years now and I never came across for such a high fees.

[HeRetiK]

Just because LN isn't distributed doesn't mean that it's centralized. Putting the word decentralized under air quotes doesn't change any of that. A decentralized network by its very definition consists of a network of hubs.


Semantics aside, there's a couple of implicit (and explicit) assumptions in Fyookball's argument that in my opinion are rather contentious:

1) The necessity of a fully distributed network to allow for permissionless transactions.

2) That too few people would be able to hold enough coins in reserve to act as a LN hub.

3) That you won't be able to do on-chain transactions to sidestep LN.

4) That LN hubs are implemented with a naive algorithm that doesn't try to actively counteract balance depletion.

5) That on-chain security would suffer due to most transactions happening on a second protocol layer.

6) That throwing math at a problem automatically makes you correct even though your underlying assumptions may be on shaky grounds.

[figmentofmyass]

Let's see an actual LN in operation and how it operates it before deciding anything. Much has been theorised. Nothing has been proven.

that's my take. and it goes both ways; we can't hold the lightning network up to be the answer to all scalability questions, because we haven't seen it work in practice. some tests, sure, but nothing on mainnet, and nothing at the scale that we need to make it useful for bitcoin.

i'm not concerned about centralization and hub-and-spoke as long as the trust model is secure, and there is no custodial trust involved. as far as i can tell, there isn't, so i fully support LN. how much value it will provide remains to be seen.

[futile-resistance]

LN is centralized in small banker hubs. You don't have to use them. If you pay 50$ fee you can send via on chain.

If you don't like bitcoin, its fine, you can use something else but stop spreading FUD. I've been using bitcoin for three years now and I never came across for such a high fees.

Lol, I saw that too. Now someone is trying to spoil the show, it will be better they leave Bitcoin for good and get into another crypto currency like Litecoin, Ethereum etc. By the way, I was expecting LN (Lightening Network) to be distributed and not centralized as the post claims it would. So I don’t still believe this until see for myself.

[n4poleon]

It has a degree of centralization, in terms of hubs (payment channels) as hubs will take a role of intermediaries, in a sense that, it is an intermediary of your transaction to the chain. In my opinion, it is a trade-off if you want efficiency. However, those channels are trustless in theory.

[Carlton Banks]

Let's see an actual LN in operation and how it operates it before deciding anything. Much has been theorised. Nothing has been proven.

that's my take. and it goes both ways; we can't hold the lightning network up to be the answer to all scalability questions, because we haven't seen it work in practice. some tests, sure, but nothing on mainnet, and nothing at the scale that we need to make it useful for bitcoin.

i'm not concerned about centralization and hub-and-spoke as long as the trust model is secure, and there is no custodial trust involved. as far as i can tell, there isn't, so i fully support LN. how much value it will provide remains to be seen.

Concurred.

There will be bigger and smaller nodes in the network, which is no different to any network where the participants can choose the amount of resources they use in order to participate, this happens with both the Bitcoin network, the cell phone network or the internet itself. Lightning node operators will necessarily need more resources than Bitcoin network participants do, but those resources are..... some BTC. And everyone on the Bitcoin network probably has some BTC anyway, it's just a common sense reason to use the Bitcoin network, not a requirement. So the on-chain "scaling" proponents can exhort us all to join them crying themselves to sleep like this if they want to, but it just is what it is, sorry gentlemen.

Lightning is not the complete answer to the scalability problem, but it sure as hell has huge potential to be either one technology in line-up of scaling solutions, or a stepping stone to whatever scaling tech predominates in the end (and it will very likely be a 2nd layer network, leveraging the combination of high security and user led consensus that the blockchain gives us the way the current balance of incentives is structured on the Bitcoin network). The only route that's not going anywhere much is the so-called "on-chain scaling" ideas. Sure, we can aggregate signatures and other such tricks to encode blockchain transactions more efficiently, but it's never going to create the magnitude of transaction increases needed to allow Bitcoin to fully compete against the euro, dollar or yuan.

[DooMAD]

While it's true that the Lightning Network certainly involves a slightly different trust model to the on-chain transactions we've grown accustomed to, you can't make the argument that it's centralised.  It's reductionist at best and outright misleading at worst.  Lightning isn't something we should be overly fearful of, providing its use is purely optional and there is always a viable alternative for those who don't wish to utilise it.  As events unfold, it seems likely we're going to see Lightning operating live on certain Altcoins before it's implemented in Bitcoin, so everyone will have a chance to see it in action and give us time to analyse and assess it.

[lottery248]

LN is centralized in small banker hubs. You don't have to use them. If you pay 50$ fee you can send via on chain.

but in the current bitcoin blockchain is suffering from the high amount of fee due to the so far major amount of unconfirmed transactions. i don't mean i don't like bitcoin doing so, it is similar to bitcoin network of now. i don;t sure other than that.

[Carlton Banks]

As events unfold, it seems likely we're going to see Lightning operating live on certain Altcoins before it's implemented in Bitcoin, so everyone will have a chance to see it in action and give us time to analyse and assess it.

Well, the global Bitcoin testnet is a better model for how Lightning will behave on the Bitcoin network itself, and Lightning channels (and transactions) have been confirmed/transmitted (respectively) on Bitcoin testnet for several months now. The testnet is just a Bitcoin blockchain without any hashing difficulty re-targeting, every other aspect of the Bitcoin testnet is identical to Bitcoin (with the exception that softforks are easily activated, none of the political nonsense that gets you of out of bed in the morning )

[hv_]

As events unfold, it seems likely we're going to see Lightning operating live on certain Altcoins before it's implemented in Bitcoin, so everyone will have a chance to see it in action and give us time to analyse and assess it.

Well, the global Bitcoin testnet is a better model for how Lightning will behave on the Bitcoin network itself, and Lightning channels (and transactions) have been confirmed/transmitted (respectively) on Bitcoin testnet for several months now. The testnet is just a Bitcoin blockchain without any hashing difficulty re-targeting, every other aspect of the Bitcoin testnet is identical to Bitcoin (with the exception that softforks are easily activated, none of the political nonsense that gets you of out of bed in the morning )

Aha. And the testnets have same bad actors and attacks as productive bitcoin?

I see you try selling shit to innocent again.


Back to ignore.