Lisa Neigut | Connect the World

What was your first experience with the lightning network and do you still know what you did/saw at that moment?

I first heard about it when I worked at CashApp! It was this cool hot new thing that everyone on the bitcoin backend team was really excited about. I have no idea what I did with this info, probably looked up roasbeef’s twitter. 
Later I spent time trying to set up a lightning node. I couldn’t figure out how to get LND examples to run, they all used Docker, but CLN you could just build and run and that Just Worked.
The first thing I did when I got my node up and running was buy a little lightning bolt lamp from a seller in China who was running a lightning only store.

Is it possible to pay with Lightning in your neighborhood?

Nope! Not at all. But I spend most of my money online these days, that seems like a lot bigger market.

You posted a tweet about “virtual” channels and stated that the Blockchain isn’t just a security feature, that it’s also a broadcast network for information.
What happens with Lightning if you remove the Blockchain layer? And is this desirable?

This is kind of a longer conversation that we could get into. The blockchain is a couple of things. The broadcast of information is a nice one, it’s a well known way of sending information to other parties, kinda like “multicast” (send one time, multiple recipients). 
The other thing a blockchain gets you is something I’d call “instant settlement” which is incredibly new and novel. There’s no other payment rails that works like this. Zero. Basically as soon as the information about wanting to pay someone happens, the money *also* moves.
Compare this to something like a credit card where they “authorize” to take money out of your account, but you don’t actually have to come up with the money until a month or so later. And the merchant gets paid a few days, or a week later by a different bank (who then goes and asks your bank for money). That’s not instant settlement. It’s instant “I owe you one bro”.
Getting rid of the blockchain gets rid of this instant settlement, and puts you back into fiat land of “I promise I’m good for it”.. or something else? Virtual channels mean you have to figure out where the settlement is happening independent of the “I owe you X” accounting. Maybe that’s good? I don’t know. Why do that when you can have a fully backed lightning channel? 
Well probably because you don’t want to have to commit the capital…

In Episode 39 Bastien Teinturier answered a question from Christian Keroles of BTC Inc. It was about how Lightning will work together in the future with the underlying on-chain layer.
He then told us that he was having a talk with you over some beers and that you both came up with the quote: “Lightning is the Blockspace buyer of the last resort”.
Can you explain this will mean in the future?

Lightning channels are batch payments. At some point you may want to close up a batch and re-deploy your capital elsewhere, or consolidate your batches etc. But typically you won’t need to do this immediately, you’ll do it as soon as it’s convenient to (when blockspace is cheap).
With enough capital deployed onto lightning, at some point you’ll have people who always have a usecase for blockspace, provided it’s cheap enough. So lightning in a way becomes a “buyer of last resort”, where when no one else is buying it, it’ll make sense for lightning node operators to use the blockspace to move around/redeploy their channel capital.

Transactions on the Lightning Network are more private than with on-chain transactions. When thinking of mass adoption: Will this eventually be a problem for governments in the future to allow Lightning adoption to go any further?

Who’s asking the government their opinion man?
But it’s a good point – all the major card networks are setup so the only entity that’s allowed to transact on a card network *must* be a bank. Banks are all strongly regulated by the federal government. (Fun fact: JC Penney in the USA was the only non-bank entity that ever got permission to connect directly to the VISA network, every other “node” in the VISA network has a banking license).
Lightning’s just data though right? And bitcoin’s an asset? Why would the government care about information flows about an asset? 😉
In the Lightning Network we can view the Node and channel count and the channel capacity, but what is more important is the amount of volume being done inside the network. This data is private, which is a big advantage.

What is the best way to measure the adoption of the Lightning Network without knowing the exact volume?

Volume only matters when you’re trying to make yourself sound cool to an external audience, when you’re trying to impress someone.
I think the numbers that speak volumes are ones that come from merchant stories about how lightning is helping them save on credit card processing fees and reach new markets. If we don’t have enough stories like that, maybe we should work on building them, as a community. Let’s make stories that speak volumes, not numbers that win beauty contests. It’s ok if small scale commerce doesn’t impress anyone. Let’s not try to be impressive for its own sake.

Inbound liquidity is one of the biggest challenges for onboarding new Node owners. Luckily there are many more options to have inbound liquidity for example participating in our Rings of Fire community, using Magma of Amboss.space or Flow from Voltage. You talked and wrote a lot about dual funded channels on demand and about Liquidity Ads. This can help advertise requests from a peer directly over the Lightning gossip network.
What possibilities are there for a Core Lightning Node owner to start with balanced channels and maintain the liquidity in their channels in the best way?

This is a great question. I would seem to be the person who would know the most about this, but I am not super experienced at running a node to be honest.
Starting with a balanced channel is theoretically the best for getting re-use of the sats in your channel; the reality of that I think really depends on where that channel is, the network of capital around you, and the payments people want to make.
I’m excited about the fee ratecard spec project and splicing; I think these two additional improvements, along with liquidity ads and dual-funding, are gonna give lightning node operators some really slick tools for liquidity management and (re)deployment.
I think right now the state of the art in liquidity management is some manner of “rebalancing”; I think we can get to a place where it’ll be “be the best payment router on the network”, if we can align the goals of a routing node’s capital placement with payment success, that’ll be really big for us as a network I think.

Rusty Russell posted a proposal for a new set of Lightning gossip messages. One peer can then generate a minisketch from its complete graph of the public network and send that to all of its peers, who can use the minisketch to find any updates to the network since its last reconciliation. The goal is to maximize efficiency.
Is the gossip problem the result of a too crowded Lightning Network? ANd how important is it to maintain maximum efficiency within the network?

There’s a lot of things about gossip currently that are ripe for improvement. Something like minisketch is great because it will reduce the amount of data you need to exchange between nodes in order to get everyone on the same page.
But if the page is changing all the time, then you’re still gonna need to do a lot of syncing to get everyone on the same page.
There’s a couple of things that make gossip changes “bad”: first is having lots of channels, which I’d say having “multiple channels per peer” is a big contributor to this. Every extra lightning channel creates 3 new gossip messages, two of which change a lot (which we can get into next). So more channels is terrible for creating more gossip traffic.
The other thing is that every time any single channel side changes the rate they charge for routing a payment, that’s more gossip that has to be propagated to every node. So you make multiple channels between two peers and then you update the rates on them every time your capacity changes and all the sudden you have tons of gossip changes that need to go out.
Two ways we can help with this: splicing, so making multiple channels per peer less attractive (dual-funding was an early precursor to this); and the fee rate card idea, which allows you to dynamically price your liquidity with a single gossip update (and even allows you to set negative fees without spamming the network with gossip).

Most Lightning Nodes run on a Raspberry Pi. Do you think it will become a problem that the majority of the nodes are running on such a device with minimum specifications?

Problem for the network? No. Problem for themselves if the memory card dies? Yes definitely. Bigger problem “for the network” is how unreliable Tor is for long running connections – it slows down payments, or payments get stuck/go to chain.
I like to think the eltoo upgrade will help a lot with small capacity computers, since the backup storage requirements are much smaller; I’m also bullish on projects like Greenlight + VLS that will enable you to keep your keys at home on a Raspi and run the node infra (channel state, watchtowers) via a cloud service.

Is it more important to work on the security level of the Lightning Network or focusing on adopting with introducing more use cases for Lightning in the field?

Personally, I think we need more people working hands on as “acquirers” for lightning, that’s businesses who are in the business of getting vendors and shops to offer lightning as a form of payment. Credit card networks had these in the early days and they were really successful.
Also fun fact about credit card networks is they used to do “airdrops” of cards to entire neighborhoods (everyone now has a credit card!) until Congress banned the practice in 1971 because it was basically mass chaos.

We have now already more than 4500 BTC on the second layer.
Imagine the theoretical total amount of BTC on the LN of 21M is reached. So all BTC is on the LN. What will happen with layer 1: miners…. What possible problems could the network face?

Lightning is “just” batched on-chain payments. On chain payments aren’t going to go away. They make sense for small payments, they stop making as much sense for larger payments (because you pay per satoshi of volume transferred, unlike onchain where you pay per byte of blockspace used).
So lightning will always be needing to “process batches” by making onchain payments (which will pay miners) and larger transactors will always want to pay on-chain (also paying miners).

How big should lightning channels be, to guarantee big amounts transfers  over the LN?

What a great question! I have no idea. But again, that batching idea is pretty key here. You gotta take into account how much you’d need to earn in routing for each of those payments to cover the onchain cost of making the channel in the first place; this equation naturally favors larger channels!