Topic: Jihan blocks segwit on LTC: price crashes (Read 6751 times)

Yes.  It's called on-chain scaling. lol

No. I thinking outside the box of using the current full nodes as a means to do multiple transactions before settling on the blockchain. My computer uses no more than 2% of cpu power and that's when a new block is propagated.

You have just explained why i don't like LN. Hence seeing if there is another solution.

I think theoretically an SPV client could be an LN hub.  Why not?

A full node with open LN channels is what is called an LN hub, no ?

The problem is that such a thing needs you, as a full node owner, to lock up quite a lot of BTC into the channels with your neighbours, and not settle before you've done quite a lot of transactions through those channels, because you only get small fees for the LN transactions, and you will need to pay a big fee to settle the channels on chain.

So the higher the amounts of BTC that you can lock up in channels, and the higher the amounts of BTC that your partner nodes you link to can lock up in channels, the longer your channels will "live" before being exhausted in one direction or another, and the more profits you can make on the LN as an active BTC locking hub.

This means that the benefits essentially will go to those whale hubs that can lock up a lot of BTC into channels with one another, and with "their customers", which are the small fish, that can never hope to set up LN links that will last long enough to reimburse the fee of the settling. 

So you essentially get the whale banks that are linked with big LN links to one another (replacing the central bank settlements in the fiat system), and customers, linked to them with all their holdings, not able to settle because of the high fees, and completely at the mercy of their bank/hub to which they are connected.

Note that on the LN, you are at the mercy of your link partner, because any of you can decide to settle (and hence have both of you pay a big on chain fee).  So a LN hub has all interest to be only connected to "secure" partners that will not easily settle: other big hubs, or dependent customers.  Most probably, with other big hubs, the fees are 50/50, and with the customers, 100% of the settling fee is for their account.  The central hubs can then also decide to whom you can pay, can report it, can help collect taxes on it, can stop transactions if they don't like them...

If the LN is processing about 1000 times more transactions than the chain can handle, you figure that settling on chain is a "one in a thousand" action, that is not easily done by a small fish, especially if the miners have already pre-sold the block chain room to the big hubs and "savage" transactions have almost no room.

pretty much.

That what i am trying to figure out. When A send btc to B, it is broadcasted to all nodes. The full node/miner node validate the unconfirmed tx. What else exactly does the full node do? Check for double spent?

If I understand you correctly then yes you are right, you need transactions to open and close the channel, which is dumb if you only want to do 1 tx.  That's why a p2p implementation of LN doesnt make a whole lot of sense because almost everyone has to have 2 channels open with everyone else all the time...and you still have to open and close those channels as well if you want to settle...so it makes a lot more sense with centralized hubs...

why not just go for on chain scaling? 

The aim of scaling should be decentralised, if there is a way. My thinking is that there should be a minimum fee per tx of say 1000 sats or lower that goes to the full noder and the usual tx fee that goes to the miner. (more on that later)

In LN, as i understand it, so correct me if i am wrong. A wants to send a 1BTC to B - both send fund (2BTC) to Funding tx (1 tx on blockchain). A send 1BTC to B, B now has 2BTC. Both broadcast to blockchain, total 2 txs. If A send 1BTC to B using bitcoin wallet - only 1 tx on the blockchain. Therefore LN only works when it comes to 3 or more txs.

Ok I get what you are saying now.

Yes, that would be the ultimate goal is to have ordinary users just be hubs.

The problem is routing... i've thought a little bit about this but haven't really seen anyone write about how to do in a really decentralized mesh-network type way.  In order to route your payment to someone without centralized hubs, every node would have to be connected to at least 2 other 'unrelated' nodes and have a channel with them, then there would have to be a routing mechanism, and you still run into the issue that the channels need to be open all the time and be of sufficient 'width'.  Probably it would be possible to do but unwieldy.  I could imagine a future where such networks are possible but a hassle to use and most people just defer to the big companies that run the massive hubs.

A LN hub/node will need to have access to a trustworthy Bitcoin full node because it will need to monitor for closing transactions that are not the most recent transaction.

Probably not. A centralized exchange will do a better job of price discovery because their order books would logically be more liquid. With a pair of BTC/LTC LN hubs, there would be a more limited amount of BTC/LTC the hub could exchange. I also suspect that those that employ BTC/LTC LN hubs in order to exchange BTC for LTC (and vice versa) will start out with a fixed amount of each of LTC and BTC, and will exchange BTC for LTC on an exchange whenever someone wants to use their hub to buy LTC with BTC, and would charge a spread over the price on the exchange. 

I know that.

A LN hub works using a software. My computer is capable of running more than one software at a time you know. Thus replace hub with full noder. I.e i run both Bitcoin wallet and LN software.

You seem to not really understand how LN works.   You can't simply say "use a regular bitcoin node instead of an LN hub" because they perform different functions. 


A non-mining node relays transactions, can verify if miners are doing their job (iow validate blocks), and can give information to SPV clients.  It doesn't provide any network security in the sense that miners do (compete with hashpower to extend and secure the ledger).

A LN hub is an entity that opens many offchain bi-directional payment channels with a large number of different users.  The LN hub does not require any additional security (perhaps that is where your misunderstanding comes from).  Its usefulness comes from the fact that many people are connected to it.  The hub may ask for a fee (or not) for its service but the transactions still need to be eventually settled to the blockchain, which is where the real security takes place.

Bi-directional payments can be done using Full Node and no need for LN hubs. The principles of LN can be applied to Bitcoin and therefore tx fees would go to miners and Full Noders. Thus securing the network in the future when POW reward/subsidy goes down.

Maybe I didn't... please elaborate.

Not sure what you mean or you misunderstood what i meant.

Exactly.  You can't do off chain by doing it on chain. 

LN hub, outside Bitcoin as in off-chain.
Full Node, inside and part of Bitcoin already.

There's no reason a person running a non-mining bitcoin node couldn't also run a LN hub... but the two things are different because they operate on
different levels.  

There are still lots of jobs horse related.

So i will ask this question again as i've never had a satisfactory answer. Why LN with hubs and not LN using Bitcoin Full Nodes (non miners)?

The whole idea to take txs fees away from the main chain would pose a risk when the rewards goes low in 20 years time. Txs fees are suppose to cover the miners' costs in the future. If there is lack of fees then who will secure Bitcoin blockchain that the value depends on.

Centralized cryptocurrency-to-cryptocurrency exchanges would be made obsolete. There would be little advantage to using a centralized exchange. Thats just better technology for you, improvements in technology make certain business models obsolete. There were lots of jobs for horses until the car came along.

They could change their business model and become LN hubs that support both BTC and LTC and thus be nodes that take orders for exchanging coins and charge tx fees, but their profit margins would probably be a lot lower. They could also become fiat <> cryptocurrency exchanges too.

It's not really any different than I described. The coins are trustlessly "swapped", so there is no trust or escrow, and the exchange happens almost instantly. All the user has to do is type in a BTC address into their LTC wallet and the LN will work out the details of which coins to swap with who to get the best deal etc and then pay those coins to the receipient. A clever UI will make this a very seemless process.