You don't get to just assume that the Lightning Network will grow into a massive, robust worldwide network. Maybe it will, but nobody knows what will happen.

Bcash is pumping because of manipulation. They are trying to start some momentum ahead of the big conference at the end of the month. The longer you are in crypto the more you realize how much manipulation goes on in every market.

BCash always pumps shortly before TV appearances and such so roger can say BCash has outperformed bitcoin over a week, a month, 3 months, etc, despite being in a general downward trend the entire time

Correct on all points.

Most people that have actually studied the routing problem associated with LN realize that LN actually scales worse than the bitcoin base layer. As such, it has no real opportunity for widespread adoption and has no chance to reverse the trend of shrinking merchant adoption of Bitcoin.

BTrash enthusiasts are excited about LN going to main chain beta because it means that some of the uneducated folks out there might start to realize that LN is a complete goat rodeo.

The soon people realize this, the sooner we can all move on to better scaling options.

What is the "routing problem associated with LN"? I was unable to find any information after a google search, which seemed like something that would be easy to do given your appeal to the authority of a consensus.

Think about how bitcoin routes it's transactions: a data package containing some inputs addresses, some output addresses, and a signature must be broadcast to all of the full nodes, which is a few thousand nodes x an average of 255 bytes of data per transaction.

With simple, single hop LN transactions, about that same amount of data only needs to be broadcast to one peer. Yeah!! We just saved tons of data. That's the appeal of LN.

Now let's look at multi-hop payments between unrelated parties.

Let's say Dave at work is running an March Madness office pool for .01 BTC, and you don't know Dave that well, so you don't have a LN connection with him. How are you going to send him the money?

You could ask Dave, "Hey, tell me all your LN connections, so I can find a route." But he's probably not going to want to do that for privacy reasons (i.e. he doesn't want to disclose how many $100 channels he has open, duh) . So instead, you have to ask send a request to all your connections "Does anyone have a channel with Dave for .01 BTC?" Let's say you have 20 LN channels, that will probably fail. So now you need to send a request to your 20 channels "Do you know anyone that has a channel with Dave for .01 BTC?" and then they have to ask their friends the same question, and so on until you've sent and received thousands of request. That stacks up to a lot of data fast.

But that's the stupid way, right? We really just need our major nodes like Coinbase, Gemini, Bitfinex, etc. to maintain a "state map" of all the existing routes, right?

Set aside the centralization, AML/KYC and other crap, let's just look at the data problem.

With Bitcoin, the "state" of the network only updates every 10 minutes and is computed by the nodes using the smaller transaction data. The "state" of the LN changes every second (or millisecond), and the only way to know it is to have everyone in the network broadcast their new state with every change they make. So more data actually has to be pushed through the network, and there's not the reliable timetable for it to happen like with Bitcoin (10 minutes).

The TL;DR version: LN scales much more like like Ethereum, where network "states" must be broadcast and maintained at super frequent intervals. We can already see that Ethereum struggles much more with capacity than Bitcoin. Eth chokes on large transaction volumes much faster than BTC by hitting real resource limits, not artificial ones like the 1MB cap.

head asplode

Some of these comments flat out make my head hurt.

Not sure how long some of you guys have been following bitcoin but comments like these (and btrash) make me think six months tops.

Bitcoins whole premise and appeal to start was decentralized digital currency. Low fees and fast transactions are pretty important don't ya think???

"Btrash" is good old fashioned bitcoin scaled the way it was originally intended, plain and simple. If you don't agree with a block size increase being the best way to scale bitcoin, okay cool. But calling bitcoin cash trash is simply idiotic.

hi roger

vikthunder,

All channels are publicly known.

You don't need to know the LN transaction history of a channel to route a payment through it.

https://github.com/lightningnetwork/...ntroduction.md

Hope this helps.

Cheers,
b

Channel updates are not sent out with each payment, nodes are also source routed so each node can use whatever path selection algorithm it wants. The forwarding nodes will also be able to choose less congested paths. This ignores further optimization and development that will inevitably happen over time with LN (channel factories and side chains). Another point over looked is that the person sending to dave can still use onchain transactions instead on LN (if we assume there are the issues as you stated).

If/When automatic channels can be opened etc almost all the problems here go away.

The funny thing about bcash proponents, they always hawk back to the whitepaper. The BTC whitepaper is a total of 9 pages long. Do you think a whitepaper 9 pages long is adequate enough to describe all the design parts of a system that wants to scale so anyone in the world can use bitcoin? How could Satoshi envisage what BTC would become in 5 years, 10 years, 50 years time? Increasing the blocksize means that miners with the best hardware get to propagate blocks quicker and this leads to the big mining pools having even more of the hashrate than they do now.

All the Lightning network does is give people a choice if they want to send a payment on-chain or using a side chain. It also comes with the added benefits of large scaling and cheaper transactions than sending them onchain but I geuss they don't like to mention than in r/btc.

I wonder what would be said if Satoshi weren't anonymous. You wouldn't have this mystical deified entity thats treated as such. Instead you would see a fat nerd or some old asian guy that no one would think is perfect, and his whitepaper wouldn't be gospel but it would certainly go down as revolutionary.

Something something put on the mask

How do you think the "All channels are publicly known" thing happens? In the current version of LN, all of the transactions across the entire LN are broadcast to each and every participant.

This setup is way, way less efficient than regular bitcoin.

In Bitcoin, with 10 million users and 10k nodes, the data for each transaction must be transmitted between 10k computers.

In the current version of the LN with 10 million users, there will be 10 million nodes - because each user must run a node. To maintain this same routing system, each transaction must be transmitted between 10 million computers, therefore, at the 10M user scale, LN is approximately 1000x less efficient than regular bitcoin in terms of world wide bandwidth usage.

For individual nodes, LN is only slightly less efficient than running a bitcoin full node because the data package for a LN transaction is a little larger than a typical bitcoin transaction. However, most people in bitcoin don't need to run a full node.

To maintain the current routing map of the existing LN, updates are sent. But that's not the proposed future state of routing, because that system obviously sucks.

"Forwarding nodes will also be able to choose less congested paths" How?

In order to choose less congested paths, you need a network map. In order to have a network map, you must know the state of all/many of the channels in the network. To know the state of all/many of the channels, you must send/receive/process a ton of data - with all the shortcomings I noted in my OP.

I've read a lot about LN routing and I have yet to see any white paper, youtube video, or even a napkin sketch about how a routing system may develop that is more data efficient than native bitcoin. Everyone, always hand-waves, just like you just did, and says, "there will be optimized algorithms for path finding." But you can't optimize a path without first having all the data. And having all the data is bandwidth and computationally expensive.

It's also a privacy disaster. Routing is only one of the major disasters of LN...there are others.

I love lightning fuds, the worst case scenario still allow some insane scaling, coinbase open a lightning channel with binance, Such a simple connection reduce on chain settlement by a significative % without needing any advanced routing, complex routing add icing on the cake but is not even necessary to make LN usefull.

https://github.com/lightningnetwork/...cement-message

I have no idea where you're getting this idea from.

This is the value proposition for LN, and that level of action should work amazingly well. Exchange-exchange and a small subset of micro-transactions are the use case.

That doesn't fit bitcoin into 1MB and it doesn't solve daily use coffee transactions.

Maybe I'm just stupid and can't figure this out on my own.

How are "all channels publicly known" if you don't transmit that information to all of the participants? Can you explain how that works?

I think a big part of the problem with understanding the LN routing is all the FUD surrounding it.

Some people divert the discussion about the LN at the peer to peer level, me exchanging funds with my neighbor. There the routing is a simple single connection that no one needs to know about.

But in order to execute multi-hop payments, either someone needs to know large portions of the network state (which is resource expensive to maintain), or I need to send out hundreds to thousands of blind requests and hope something pops up.

I think most people that know what they are talking about envision a system where I have a LN channel with a large hub like Coinbase, and then Coinbase maintains a routing network map to thousands of other participants. Unfortunately, this network topology is basically the antithesis of Bitcoin in the first place.

Can that work? Yes. Is it centralized? Extremely. Is it private? Not at all. KYC/AML? You bet. Trust-less? Somewhat -- I'm not trusting Coinbase with my keys, but I'm trusting them to stay on-line as a major participant. If they crash, they crash the LN with it.

If you have another network routing system that's more data efficient than native Bitcoin, I'd love to hear about it.

You do a special kind of transaction on the main Bitcoin chain that funds a Lightning Network channel with some amount of bitcoin. Then on LN, you point to the funding transaction and announce that a channel exists between two LN nodes, with associated Bitcoin keys, funded with some amount of money.

The subsequent LN transactions that happen on that channel between the two nodes are not announced to the network.

Does it look like there may be a bit of a rebound due?

There was supposed to be that G-20 crypto summit in the next couple of days though, so who knows what may happen then?


Hope everybody's having a nice weekend - so pretty out!

The subsequent LN transactions that happen on the channel need to be announced to the network if you want to maintain a network map and allow routing through your channel. (support multi-hop routing)

If you don't announce your updated transaction state, the network has no way to know your current up/down capacity and will not be able to use you to help route other payments.

That's where the routing problem kicks in. If you don't broadcast an update of your channels state, you can't support multi-hop payments (semi-efficiently). The only way you could support muti-hop is if you make yourself available for queries. But querying random nodes to find a route is obviously less efficient than having a network map. Either way, the system is less efficient that base layer Bitcoin.

People are overestimating how many hop will be needed.
For a while most payments will be direct p2p. If you add an intermediary hop that won't be that complex either.
The requirement for super advanced routing require some insane adoption and super slick ui.