Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Lightning Network Search and Analysis Engine

1ml.com

Post reply on HN
Print view View original

71–80 of 104 posts

Re: Show HN: Lightning Network Search and Analysis Engine

#71

Post by dx034 »

nosuchthing wrote: post #37
imaginenore wrote: post #21

Have they solved the distributed routing problem? I watched many videos on the LN, and it's never addressed, it's always this hand-wavy explanation that a route is somehow found, on a network that changes, potentially, thousands of times per seconds, and has tens of millions of nodes.

No. The LN white paper doesn't even explain how they plan to solve routing. https://www.youtube.com/watch?v=Ug8NH67_EfE&feature=youtu.be... Without decentralized routing, It seems like they're simply prioritizing nodes with large sums of BTC as the central operator hubs. Centralized payment processing hubs?

So like a Visa/Mastercard for Bitcoin? Probably not much to gain from that, I doubt it'll be more efficient than established networks (unless you want to deal in transactions that aren't allowed on those networks).

Re: Show HN: Lightning Network Search and Analysis Engine

#72

Post by polyomino »

keymone wrote: post #67
thisisit wrote: post #63

I have watched couple of introductions on LN and had some questions, if someone can help: a. What effect does this have on transaction sizing? b. Does the the node owner charge fees for each transaction? c. How are the channel amounts calculated? Do they include mining and node fees?

a. LN transactions and Bitcoin transactions are different things, i will assume you mean "how does LN affect Bitcoin's chain capacity?" - to use LN you have to make at least one Bitcoin transaction - so called "funding" transaction which opens or refills a channel, afterwards you can make millions of LN micropayments - none of them will affect Bitcoin tx capacity. b. Yes, it's part of node configuration, you can set…

LN transactions are bitcoin transactions though they aren’t broadcast to the blockchain most of the time.

Re: Show HN: Lightning Network Search and Analysis Engine

#73

Post by thisisit »

keymone wrote: post #67
thisisit wrote: post #63

I have watched couple of introductions on LN and had some questions, if someone can help: a. What effect does this have on transaction sizing? b. Does the the node owner charge fees for each transaction? c. How are the channel amounts calculated? Do they include mining and node fees?

a. LN transactions and Bitcoin transactions are different things, i will assume you mean "how does LN affect Bitcoin's chain capacity?" - to use LN you have to make at least one Bitcoin transaction - so called "funding" transaction which opens or refills a channel, afterwards you can make millions of LN micropayments - none of them will affect Bitcoin tx capacity. b. Yes, it's part of node configuration, you can set…

a. I meant bitcoin transaction sizing. Bitcoin fees are driven by transaction sizes and my understanding (edited) is for a LN channel to be opened and closed, the transaction needs to committed to the blockchain. What are the expected size of these transactions? And how does Segwit work for these transactions?

b. Thanks for the answer. (Edited): So, it seems the current argument there will be too much competition for nodes hence people will automatically set low or no fees.

c. This is unclear to me. Normally, when I am paying for stuff in bitcoin sellers tend to attach an estimated fee for the transaction. Lets take an example, I need to pay 0.1 BTC they will charge 0.101 BTC. The 0.001 BTC is the transaction fee.

With LN nodes charging fees on top of it, businesses now will have two fees to account for?

Additionally, my understanding now is that there are three fees now:

a. Channel Opening fees to miners

b. Node fees which might be zero

c. Channel Closing fees to miners

Re: Show HN: Lightning Network Search and Analysis Engine

#74

Post by keymone »

polyomino wrote: post #72
keymone wrote: post #67

Earlier quoted context omitted.

a. LN transactions and Bitcoin transactions are different things, i will assume you mean "how does LN affect Bitcoin's chain capacity?" - to use LN you have to make at least one Bitcoin transaction - so called "funding" transaction which opens or refills a channel, afterwards you can make millions of LN micropayments - none of them will affect Bitcoin tx capacity. b. Yes, it's part of node configuration, you can set…

LN transactions are bitcoin transactions though they aren’t broadcast to the blockchain most of the time.

you could argue either way depending on your definition of what bitcoin transaction is. if it's a separate entry in the blockchain then no, LN transactions aren't bitcoin transactions. if it's transfer of value within bitcoin's network with same security characteristics, then yes, LN transactions are bitcoin transactions.

Re: Show HN: Lightning Network Search and Analysis Engine

#75

Post by keymone »

thisisit wrote: post #73
keymone wrote: post #67

Earlier quoted context omitted.

a. LN transactions and Bitcoin transactions are different things, i will assume you mean "how does LN affect Bitcoin's chain capacity?" - to use LN you have to make at least one Bitcoin transaction - so called "funding" transaction which opens or refills a channel, afterwards you can make millions of LN micropayments - none of them will affect Bitcoin tx capacity. b. Yes, it's part of node configuration, you can set…

a. I meant bitcoin transaction sizing. Bitcoin fees are driven by transaction sizes and my understanding (edited) is for a LN channel to be opened and closed, the transaction needs to committed to the blockchain. What are the expected size of these transactions? And how does Segwit work for these transactions? b. Thanks for the answer. (Edited): So, it seems the current argument there will be too much competition for…

> What are the expected size of these transactions?

LN deposit/close transactions don't carry any information about LN payments made on the channel, so their size doesn't grow with usage. They are simply 2-2 multisig transactions, so somewhat larger than the minimal 1-in-1-out tx, but probably (didn't check this) not larger than median bitcoin transaction.

> And how does Segwit work for these transactions

those are always pay-to-witness-script-hash transactions, so they don't even work without segwit: https://github.com/lightningnetwork/lightning-rfc/blob/maste...

> current argument there will be too much competition for nodes hence people will automatically set low or no fees

it's not as easy, there are multiple incentives and discentives working here. lowering fees will bring more volume but increase cpu/io load. higher volume also means you need to have more capacity on your channels, which essentially means lots of funds in a hot wallet - a target for hackers. high volume also means you might become target for authorities wanting to track payments in LN. et cetera.

> With LN nodes charging fees on top of it, businesses now will have two fees to account for?

technically yes, but the fee for deposit transaction is paid only once per channel and merchants (usually) won't be involved with it. it's reasonable to assume that users will mostly fund channels with larger nodes because that brings better connectivity (you don't want to have a channel with starbucks, mcdonalds and burger king, you want single channel with a node that invested in having a lots of channels).

as for the three fees you listed - yes (but not necessarily to miners?). the assumption is that you will rarely be opening and closing channels, just like you're rarely opening and closing a new debit card. you'll open with a reliable node that makes you routable, you'll once in a while deposit more funds and eventually you will close the channel. in between you can have bajillion transactions for which you will pay close to zero fees.

Re: Show HN: Lightning Network Search and Analysis Engine

#76

Post by utunga »

zaarn wrote: post #58
ExactoKnight wrote: post #55

Earlier quoted context omitted.

The same way Bitcoin can't realistically be stopped by any government. (Short of seizing network hashing power..)

Or by ruling that it is illegal to buy, sell or own bitcoin. While bitcoin might continue, most citizens will stop bothering at that point because buying or selling bitcoin shows up in tax audits. The government can also make it illegal to operate a bitcoin node or transmit bitcoin transactions, including for ISPs, which would get the protocol quickly blocked. Additionally they could rule Bitcoin to be an illegal for…

> While bitcoin might continue, most citizens will stop bothering at that point because buying or selling bitcoin shows up in tax audits.

But they'd only show up for a short period of time (in this scenario) because after a while bitcoin wouldn't really be worth anything. Lol.

Re: Show HN: Lightning Network Search and Analysis Engine

#77

Post by cup-of-tea »

tlrobinson wrote: post #54
colordrops wrote: post #42

Earlier quoted context omitted.

The whole thing is fishy. The lightning network is pushed by Blockstream, a company which has a large number of devs working on bitcoin core. They were responsible for all the propaganda and talking points against bumping the block size leading to the BCH fork, because they wanted the "correct" solution of their lightning network. The fact that it is so complex and ill defined is dangerous and an antithesis of decent…

This is pure FUD. Satoshi himself proposed the first version of payment channels [1]. The Lightning Network paper has zero authors from Blockstream [2]. At least two other entirely independent companies are working on Lightning Network implementations [3][4]. It's well understood how trustless LN transactions work [5]. The spec [6] and code are open source, feel free to look for a "backdoor". 1. https://en.bitcoin.it…

Whenever I see the term "FUD" nowadays I think "ah, someone who is invested in a cryptocurrency, this will definitely be an unbiased view".

Re: Show HN: Lightning Network Search and Analysis Engine

#78

Post by thisisit »

keymone wrote: post #75
thisisit wrote: post #73

Earlier quoted context omitted.

a. I meant bitcoin transaction sizing. Bitcoin fees are driven by transaction sizes and my understanding (edited) is for a LN channel to be opened and closed, the transaction needs to committed to the blockchain. What are the expected size of these transactions? And how does Segwit work for these transactions? b. Thanks for the answer. (Edited): So, it seems the current argument there will be too much competition for…

> What are the expected size of these transactions? LN deposit/close transactions don't carry any information about LN payments made on the channel, so their size doesn't grow with usage. They are simply 2-2 multisig transactions, so somewhat larger than the minimal 1-in-1-out tx, but probably (didn't check this) not larger than median bitcoin transaction. > And how does Segwit work for these transactions those are a…

Thanks for the other clarifications but this part leaves me confused.

> mostly fund channels with larger nodes because that brings better connectivity (you don't want to have a channel with starbucks, mcdonalds and burger king, you want single channel with a node that invested in having a lots of channels).

Lets take an example. Bob wants to pay Starbucks for his coffee. Alice is running a node which has a channel open to Starbucks. Now he can open a channel to Starbucks directly and pay for his coffee. But that is time/fees consuming so he will pay Alice, who will then pay to Starbucks on his behalf?

If that is true, doesn't that put too much power in Alice's hand? And if there are some nodes which have heavily invested in cpu/io and coins to their channels, wont they effectively control a large part of the transaction volume and hence make the network more decentralized?

> you will rarely be opening and closing channels, just like you're rarely opening and closing a new debit card.

I get confused every time this analogy is brought up. Sure, I don't need to open/close debit cards but those cards are loaded automatically when I get my salary/income. And the money is not tied, I can move it around as much as I want.

On the other hand, again, as per my understanding LN requires me to commit an amount and I have to make do with it until the amount runs out. So, its more like a prepaid gift card with fixed commitments and a reload feature.

Re: Show HN: Lightning Network Search and Analysis Engine

#79

Post by keymone »

thisisit wrote: post #78
keymone wrote: post #75

Earlier quoted context omitted.

> What are the expected size of these transactions? LN deposit/close transactions don't carry any information about LN payments made on the channel, so their size doesn't grow with usage. They are simply 2-2 multisig transactions, so somewhat larger than the minimal 1-in-1-out tx, but probably (didn't check this) not larger than median bitcoin transaction. > And how does Segwit work for these transactions those are a…

Thanks for the other clarifications but this part leaves me confused. > mostly fund channels with larger nodes because that brings better connectivity (you don't want to have a channel with starbucks, mcdonalds and burger king, you want single channel with a node that invested in having a lots of channels). Lets take an example. Bob wants to pay Starbucks for his coffee. Alice is running a node which has a channel op…

> If that is true, doesn't that put too much power in Alice's hand? And if there are some nodes which have heavily invested in cpu/io and coins to their channels, wont they effectively control a large part of the transaction volume and hence make the network more decentralized?

what do you mean by "power"? yes, there will be nodes handling more transactions than other nodes, but that in no way prevents you from using other nodes or even investing into opening channels yourself and competing.

> those cards are loaded automatically when I get my salary/income. And the money is not tied, I can move it around as much as I want.

not everybody carries a debit card connected to their main bank account. and those that do usually have daily limits on how much they can spend, so you can't really move it around as much as you want. similar to that, not every BTC holder stores their BTC in hot wallet. funding your debit card is similar to funding your hot wallet which is similar to funding your LN channels.

don't take the analogy too far though. point is that LN allows you to do is micropayments without flooding the blockchain. there are tradeoffs involved. that's it.

Re: Show HN: Lightning Network Search and Analysis Engine

#80

Post by Klathmon »

AlexCoventry wrote: post #11

What are the regulatory hurdles to operating a lightning node in the US? It makes you a money transmitter, right?

I'm not a lawyer, but this is my understanding:

A LN node is no more a "money transmitter" than any internet router that a bank transaction flows through is a "money transmitter".

The node or router can't "steal" or modify the amount or destination in any way, and there is no counterparty risk with either of them, so they both aren't "money transmitters".

Post reply on HN