Hacker News Reader

Comment by FreeTrade | Hacker News Reader

Sigh. Bitcoin BTC, at 3 transactions per second, doesn't have the capacity to support El Salvador's economy. Even if Lightning worked (it doesn't), BTC doesn't have enough capacity to open channels for users at a reasonable cost. Salvadorians will end up forced to use proprietary payment networks denominated in BTC.

BTC is not cash anymore, it won't work as such. There are other cryptos that could work, we'll have to see how broadly 'bitcoin' is interpreted.

The CGT exemption will be of interest to crypto whales and entrepreneurs and may attract some, but organized crime is a concern given the nature of the asset.

View on HN

Lightning does work. I've used it myself as have many others. You should really work on explaining what it is you actually mean by that statement instead.

Apparently they are using Lightning and a service LN Strike that recently launched.

https://twitter.com/starkness/status/1402635954653904902?s=2...

Lightning is not decentralised. You have up keep your node online 24/7 to prevent theft. Most people can't do that. So they will end up keeping their coins in the custody of nodes that run 24/7. Leading to centralisation. Lightning is just like a MySQL database extra steps.

I read that you put a "bounty" in your transaction so that if your node is offline, another node will have an incentive to prevent the theft for you.

You don't have to be on exactly 24/7. The default delay for spending coins from a commitment txn is 144 blocks in some clients and is usually scaled with channel size. Most clients let you configure it manually too.

That means you'd have to be online once within 24 hours of your counterparty trying to steal your channel balance to stop them. IMO this is pretty reasonable for a cellphone user.

It doesn't follow that this would lead to centralization of custody either. If you want to pay for better liveness you could pay someone to run a watchtower to watch your channels. Basically, it watches the blockchain for bad channel closes and posts the justice txn with the correct settlement for you in case it sees it. I'd be surprised if a service like LN_Strike didn't start using these.

https://github.com/lightningnetwork/lnd/blob/master/docs/wat...

So all anyone one needs to do to steal your money is DDoS you?

According to Bitcoin maximalists 2mb blocks are too much for our infrastructure, and will hinder everyone from running their own node. But apparently keeping a node online 24 hours is not a big deal.

LN requires an on-chain transaction for both opening and closing a channel and requires the node stays online 24/7 to receive payments. That's garbage.

Because Visa and other payment processors don't require opening an account? (e.g. a credit card, which is a much more involved process than opening a channel). Or 24 hour servers?

It's for sure a step backwards from regular cryptocurrency where ownership of private keys is enough to receive or spend. No servers, no "opening accounts" etc.

You can compete with Visa and other payment processors AND not have these arbitrary requirements through on-chain scaling and larger block sizes.

3 transactions a second is 100m transactions per year. So everybody on the planet gets one transaction per 70 years. So everybody gets one transaction in their lifetime to open a channel to a lightning network provider and that's it.

You can do far more than 3 tx per second in mainchain. Look into channel factories. 100s of trustless lightning network channels in 1 onchain transaction. Layer approach is the only viable scaling solution.

How does it work? Don’t you still need O(n) bytes for the payload?

With Lightning channel factories:

- 2 people can make 10,000,000 transactions between each other in seconds

- 10,000,000 people can make 2 transactions between each other in 46 days

Channel factories are not a solution to the user scaling problem.

Are there any videos of Lightning in action in the real world (I.e. not abstract explainers)? Genuine question, I don’t know enough about it and would love to learn.

They're literally using Lightning for payments right now in El Salvador. They didn't do this without having local projects already. Bitcoin Beach (https://twitter.com/Bitcoinbeach) made a lot of this happen.

The cost and complexity of opening channels for users is prohibitive. Any widely adopted solutions will therefore be custodial. Might as well use PayPal.

There are easy to use non-custodial wallets (try Phoenix wallet).

Sure, I just installed it and its saying some impenetrable stuff about a swap address that is not controlled by my wallet and deposits will be converted to Lightning channels.

It's prohibitively complex.

"Might as well use PayPal." for me that applies to 99% of crypto currency projects!

There's a big difference between a solution which requires you to use a custodial service, and one which gives you the option of non-custodial service - the latter allows for a much greater ability to hold custodial services accountable, by opting to self-custody, without forcing you to change your payment medium in the process.

Bitcoin beach is run by a few college students. Not exactly something to depend on.

Funny saying that on HN.

I don't think it is. Startups fail at tremendous rates and are well known for playing it really fast and loose, even if some of them are wildly successful. Not exactly something I'm interested in relying on for finances until there is a much longer track record.

Yep, it's survivor bias.

I'm not sure how you can say lightning doesn't work when it in fact does, 1ml.com shows you growing adoption and there are many private channels that won't even be known. There are many easy to use wallets, etc.

Channel factories will bundle channel open/closes into a single transaction, so it won't be burdensome on the main chain.

It is not hard to scale Bitcoin for millions of people for everyday transactions. You give people a custodial wallet and track the balances off-chain using a trusted third party. As long as a government is willing to allow such a service to legally operate:

1. You can easily support billions of bitcoin transactions a second since the transactions only change centralized database entries.

2. Users can send and receive funds using a website. You could even issue tokens that user could use to authorize fund transfers at Point of Sale (see Credit Cards).

If you want to ensure that this trusted party isn't hacked, you can have five trusted parties instead of one and use multisig so that 4 of 5 of them must consent and sign to do an on-chain transaction. You can something similar for the internal ledger maintained by the trusted parties to ensure that one of them doesn't maliciously alter user balances.

The typical rejoiner to this is: Why do you need Bitcoin to do this? Why not do it with fiat? They aren't wrong per se, you could do this with fiat and it would work pretty well. However building a system like this with Bitcoin does have some advantages:

1. User's can withdraw funds on-chain,

2. there is a clearly defined on-chain layer for settling large balances between institutions,

3. bitcoin isn't created by fiat from a government bank. This removes the risk of inflation in longterm international contracts, etc...

Why bother with Bitcoin at all if you need a trusted third party? Crypto maximalists have let greed blind them from the initial mission of crypto.

>Why bother with Bitcoin at all if you need a trusted third party?

This is a good question and we should always be asking "Why Bitcoin?" and think about the trade-offs.

For small amounts of bitcoin a trusted third party, especially an institutional trusted third party, makes a lot of sense. Convenience, availability and low transaction fees dominate the user's needs.

For large amounts of bitcoin, you probably don't want to use a trusted third party and you can afford to run your own infrastructure. For instance a national bank should not keep all its bitcoin on coinbase.

Bitcoin allows you to choose between these two trust models. It is very hard to remove a trusted party from a system which fundamentally relies on a trusted party. However it is often easy to inject a trusted party into a system to make it more useable or scalable. Thus, for maximum flexibility you want your base layer to be as trustless as possible and then add trust as needed to solve different use cases.

I’m still not seeing the value of Bitcoin for smaller transactions. Just use the existing banking system and if you want to buy Bitcoin then buy Bitcoin.

Thats all great until fractional reserve banking enters the game. In the end we will efficiently trade IoUs and the best outcome would be that they are redeemable and we have a gold-standard-like thing again. That in itself would be a godsend already though probably not what Bitcoin maxis vision looks like

>That in itself would be a godsend already though probably not what Bitcoin maxis vision looks like.

Bitcoin maxis have a very ambitious and grand vision and I don't think they should compromise it. It sets a high bar that drives innovation. I do think that over the short term any nation scale usage of bitcoin for everyday payments will look very different from that vision and that is ok.

What about using a blockchain internally within the country to manage the high throughput and still avoid centralization by relying on a single "trusted" third party?

For e.g., mining operations can be distributed within the country itself to ensure decentralization within the country. Then this internal blockchain network interfaces with the slower and more expensive bitcoin blockchain to do final settlements?

Could this work?

I have a lightning wallet and it works - please don't post uninformed information.

Reply on news.ycombinator.com