FWIW that is the goal of the Lightning Network on Bitcoin: https://lightning.network/

The lightning network will be far less effective than people are hoping. You have to a) pay to open up a channel with each person you wish to use it, and b) each stake capital in that channel for it to be functional. This is useful only in specific scenarios like exchange-to-exchange transfers. Exchanges are going to be very unlikely to ever set up channels to customers. They haven't even implemented segwit yet, there's no way they are going to bother setting up the infrastructure needed to charge for and activate lightning channels.

> You have to a) pay to open up a channel with each person you wish to use it

This simply isn't true. You open one channel and then payments are routed over channels to your destination. It's no different to TCP/IP in this way and the reason why that analogy is used a lot in describing the network

Having a one channel single destination can be done with Bitcoin alone and the new timelock transactions - the purpose of lightning is automating these transacations and the routing between channels

Opening a channel in lightning is no different to having a balance in a wallet or a bank account - you don't notice that it is "locked". Channels can also create other channels using channel factories[0]. This all happens under the hood much in the same way somebody accessing YouTube knows little about routing infrastructure.

There really is a lot of FUD about Lightning and they get repeated a lot by people who should know better. There is now a series of simple infographics that explain what Lightning is and how it works so there shouldn't be any more excuses for such simple misinformation being repeated[1][2][3][4][5]

Your specific claim about one channel = one receiver is addressed in [5]

I think the HN audience should go out and try it themselves and read the paper. It's the most eye-opening tech experience i've had since reading the Bitcoin white paper.

If you've been kicking yourself about not getting into Bitcoin early - you have a new opportunity with Lightning now. There are only hundreds of nodes at the moment and only a couple of vendors accepting payments on the testnet and a ton of opportunity to build an entirely new infrastructure around it.

[0] https://bitcoin.stackexchange.com/questions/67158/what-are-c...

[1] https://i.imgur.com/L10n4ET.png

[2] https://s3.amazonaws.com/bitcoindesigned-prod/media/what-are...

[3] https://s3.amazonaws.com/bitcoindesigned-prod/media/lightnin...

[4] https://s3.amazonaws.com/bitcoindesigned-prod/media/lightnin...

[5] https://s3.amazonaws.com/bitcoindesigned-prod/media/lightnin...

> If you've been kicking yourself about not getting into Bitcoin early - you have a new opportunity with Lightning now.

Why not just delete the rest of your post and just leave this? This once sentence is what really drives all of the rest of your breathless hype. Naked greed. It's the only thing the entire crypto "space" has.

Lightning network is pure vaporware. It was "coming any day now" when Mt Gox collapsed. It's still coming soon. It will always be coming soon.

You'll be happy to hear then that a lot of progress has been made on Lightning since MtGox imploded

I probably should have linked to these from my original comment - rather than removing anything

Specifically with lnd the lightning daemon[0], BOLT's (the equivalent of Bitcoin BIP's) 1 through to 11 have been implemented in the core daemon[1]

There is also the Neutrino light client[2] (screenshot [3])

Far from being vaporware - you're only a few docker commands away from trying it out yourself[4]

[0] https://github.com/lightningnetwork/lnd

[1] https://github.com/lightningnetwork/lightning-rfc

[2] https://github.com/lightninglabs/neutrino

[3] https://blog.lightning.engineering/assets/images/app-chans-g...

[4] https://github.com/lightningnetwork/lnd/tree/master/docker#c...

This sounds just like a 15-year-old ad for Perl 6.

Bitcoin is the multi-level marketing of millennials.

Spot on!

Personal attacks are not OK on Hacker News, no matter what you're replying to.

https://news.ycombinator.com/newsguidelines.html

It's somewhat reductive but still pretty accurate.

Thanks for posting those infographics; they answer some lingering questions that have lain unanswered for years. I still see one problem, though, which is that the few on-chain transactions that are required (such as topping up a channel) will be far more expensive than in those other blockchains. Like $20 vs. 20 cents. So there are plenty of scenarios where LN fees will still be higher overall.

Other chains are cheaper because they have fewer users or higher trust/censorability.

With LN there are less transactions on-chain which reduces fees. Also, finally the temper-tantrum throwing core-supporters are quickly realizing that a block size increase is absolutely necessary going forward

"Getting into Lightning" won't make you money as you might think.

The network isn't a coin as such - merely a way of sending other kinds of coins. Hence it doesn't have a value which could skyrocket making one rich.

I worry that could hamper it's uptake.

> a) pay to open up a channel with each person you wish to use it

I think you only need a path to the person you want to pay, not a direct channel. eg if Alice wants to pay Bob, and they both have a channel to Greg, then Alice can send money to Bob through Greg.

> They haven't even implemented segwit yet,

The wikipedia[1] page seems to say that most miners implement segwit now.

* https://en.wikipedia.org/wiki/SegWit

Segwit is still not fully supported in the Core wallet [1] - change goes to P2PKH addresses and there's no GUI support for generating (or even sending to AFAIK) bech32 addresses.

While certain charts regulary show 10% usage [2], when analyzing UTXOs, the numbers are a miniscule 0.03% of transactions. [3]

[1] https://github.com/bitcoin/bitcoin/pull/11403

[2] http://segwit.party/charts/

[3] https://twitter.com/dukeleto/status/954759096544124928

It's named network because you do not need a channel with every individual user. You have a channel with one or more LN servers, they in turn have connections with yet more servers.

As the other guy said, there only needs to exist a path of channels between you and the entity you're paying, and the payment will be routed through the network.

> You have to a) pay to open up a channel with each person you wish to use it

Bitcoin's lightning network doesn't require you to open a channel for each person you want to pay https://unlikekinds.com/t/bitcoin-s-lightning-network-doesn-...

I keep seeing this brought up over months and months, but how close are we even to using it? Is the technology a thing yet? Or is it all white papers.

I demoed [0] receiving one tenths of a cent's worth of bitcoin (0.00000010 BTC) instantly, and on Bitcoin's mainnet (not a test network).

I also demoed [1] purchasing swag from Blockstream's store [2] which is powered by their newly released Wordpress e-commerce plugin [3] that enables Lightning payments.

I did not screencapture the videos because Qubes dom0 does not come OOTB with screenrecording software, and I don't want to enable networking in dom0 to install software that does.

[0]https://twitter.com/notgrubles/status/955611467889422337

[1]https://twitter.com/notgrubles/status/954933507863863296

[2]https://store.blockstream.com/

[3]https://blockstream.com/2018/01/16/lightning-charge.html

It exists in its early stages, but it's terribly impractical, hence no exchange or end user is in a rush to adopt it. Its design is almost like someone took a hard look at Bitcoin and concluded "Gee, this is not cumbersome enough for people, we need to make transacting a worse and more confusing experience".

Here is a graph showing the number of lightning network payment channels on the bitcoin mainnet: https://p2sh.info/dashboard/db/lightning-network?orgId=1

Working implementations exist and are usable, but they are being very cautious before releasing to the general public.

it's not useful for speculating on the price of bitcoin so few people will use it

You can buy coffee https://twitter.com/alexbosworth/status/955870434230132736

There's a few different ones that have made huge jumps forward; RaiBlocks can apparently do around 7,000tx/s.

Someone did a stress test the other day. Nowhere near 7ktx/s and a lot of nodes started to crash. https://medium.com/@bnp117/stress-testing-the-raiblocks-netw...

Honestly it is scary to see how easy is to launch a DoS attack just by precomputing the PoW.

Edit: Here are some people saying their nodes went out of sync https://www.reddit.com/r/RaiBlocks/comments/7rzlid/minor_str...

Radix is publishing some eye popping numbers on their DAG-based testnet: https://twitter.com/radixdlt/status/956087291696566272

We'll have to see how it functions in the real world later this year...

Meanwhile, your payment card provider is probably doing that many transactions per second just in the small city I live in.

That's 25 million per hour, we must have very different definitions of what a small city is.

[citation needed] A quick google says Visa peaks at 4k txs/sec

[0]: VisaNet handles an average of 150 million transactions every day and is capable of handling more than 24,000 transactions per second.

[1]: 65,000 Transaction messages per second

[0] https://usa.visa.com/run-your-business/small-business-tools/...

[1] https://usa.visa.com/dam/VCOM/global/about-visa/documents/vi...

Also, consider there are other networks like MasterCard:

[2]: MasterCard’s 2012 annual report asserts that its network can handle more than 160 million transactions per hour with an average network response time of 130 milliseconds. Operating 24 hours a day every day, MasterCard reports that its processing systems have consistently maintained availability 99.9 percent of the time.

[2] https://www.philadelphiafed.org/-/media/consumer-finance-ins...

Thats around the world, not in one town/city

A currency with 'decentralization' as a goal should probably aim higher than satisfying the transaction rate requirements of a single town/city.

Id like to also add that there are an order of magnitude more transaction options provided by cryptocurrencies than there are for fiat payment systems. Id bet money that, collectively, the cryptocurrency ecosystem is able to process more transactions per second than all of the leading fiat processors combined.

Probably, but only because crypto doesn't scan for fraud and abuse among other things. Monero is happy to let someone completely empty your life savings.

Cc processors on the other hand are on the hook for fraud, so they are checking every transaction to make sure it doesn't match certain hueristics that could indicate theft or fraud.

I'm optimistic crypocurrencies (not all, but some) will achieve global adoption as a payment option. I think its possible to build a cryptocurrency that could emulate the fraud / refund detection features of cc processors. If there isn't one out there, I'm sure it will exist in the future. I certainly don't see monero being used in my local bodega or as a mainstream currency because thats not really what monero is trying to achieve.

I'd take that bet in a hot second (not that we could ever get verifiable numbers to prove it). There's just no way sorry

Is it doing that in production at decentralized scale?

Steem regularly processes 1.5 million transactions per day and is at 0.13% capacity[1]. If my math is right, that comes out to the ability to process 14k tx/s.

[1]: http://www.blocktivity.info

It's important to note that that is theoretical capacity. We've no idea what it's actual performance would be.

That's a great observation. I was struggling to put my view into words. That captures it perfectly.

Edit: Yes, against the guidelines to talk about downvotes, but really? This is more than a "me too". Parent was unusually on point for me, and felt compelled to offer thanks for the succinct summary

I had a similar thought recently, only I thought MS-DOS. Same difference.