Lightning Labs lnd Goes Wumbo

Hi and welcome to the Stephan Livera podcast, a show about Bitcoin and Austrian economics. Are you wondering what is the latest with Lightning Network? What does it mean when channels go large sized with Wumbo? My guest today is Laolu, also known as Roasbeef, the CTO of Lightning Labs. This show is brought to you by swanbitcoin dot com, the best place to auto stack your Bitcoin. In the US, with incredibly easy setup and low fees, I personally appreciate that Swan is Bitcoin only and they're dedicated to Bitcoin education. Go to swanbitcoin dot com slash livera to get ten dollars of free Bitcoin when you start stacking with Swan. And Swan has some news to share. They've had massive demand for daily buys since the day they launched the service. One of the big positives of regular recurring buys is smoothing out price volatility, so buying daily will catch those dips even better than buying weekly. There are a limited number of spots in the Swan Daily Buys beater, so head over to swanbitcoin dot com slash dailybuys to get into the beater. That's swanbitcoin dot com slash dailybuys. Are you looking out for your Bitcoin backups and security? Have you got your Bitcoin seed words or recovery phrase backed up? Check out the Cipher Wheel produced by CipherSafe at ciphersafe dot io. It's a wheel-shaped product, and you can put in the words from your seed using the tiles, and it also comes with a padlock tamper evidence. Seal so you know if it's been opened. Make sure that you or your loved ones have access if something happens to you, don't just keep it on that piece of paper. Orders are going out now, go to ciphersafe dot io and use code livera for a discount. Next is Knox. Knox is a Bitcoin custodian dedicated to ensuring their insurance protection covers the full value of their customers' assets. For example, suppose a fiduciary wants to hold two hundred and fifty million dollars worth of Bitcoin with Knox, Knox will seek to obtain two hundred and fifty million dollars of insurance dedicated to exclusively to that account and adjustable to volatility, no fractional coverage or narrow scope, insurance for what it's worth, a tool to transfer risk. Knox is backed by investors such as Fidelity Investments Canada, Initialized Capital, and Inovio. If you are a Bitcoin company, investment fund, trust, or family office, check out Knox for your insured custody. The website is knoxcustody dot com. And finally, Unchained Capital, Bitcoin native financial services. Unchained Capital are making it easy for you to access multi-signature and And they are providing products in a way that respects the "not your keys, not your coins" ethos of Bitcoin. So if you're thinking about your Bitcoin security, why not consider going from zero to multisig with Unchained? They're offering a vault concierge onboarding package where you can get a guided setup call and have hardware devices mailed out to you, and as part of the package, you'll have a thousand dollars in your vault also. So the prices range from fifteen hundred to one thousand dollars, and use code Laverara for a discount. Also, Unchained Capital have an incredible amount Base, which you can find at their website, go to unchained-dashcapital dot com. Alright, so for this interview, we get into a lot of stuff around the Lightning Network, what's the current state of it, what's gonna happen with Wumbo, backups, HTLC interception, we talk about Loop as well, some of the attacks on Lightning that are possible, and what are some of the mitigations as well as protocol upgrades. So make sure you are subscribed using a podcatcher application, and you can make sure you set the speed to a comfortable level for this interview.

What's up? What's up? Yeah, it's good to be back. I think the last time I was here was like two years ago, maybe or something like that. But yeah, it's been, it's been some time. And since then, you know, you've kind of, you've kind of blown up since then, so, you know, definitely good to see you, make some progress on this.

Oh, for sure, thank you. And, you guys have been doing some great work over at Lightning Labs. I was really excited to see some of We're seeing, you know, some real development in the ecosystem. So what are you seeing that you're excited about?

Oh, yeah, I'm so excited about what's going on. I think, I think it's kind of now it's where it's like, you know, I feel like progressively we kind of like slowly matured the network over time, so you know, maybe initially it was a little bit more like just about like, you know, mic payments, we have a search layer, things like that, but I think now we're kind of like seeing like a bunch of like Lightning like growing up a little bit more from kind of like just the little like, maybe like micromine networks and something that's like turning into a little bit more as far as like, you know, like payment size and also kind of like, you know, people being more confident in the software itself. I think also generally like the network matured a lot, like if you, if you'd say, okay, well, you know, maybe things weren't, weren't super, you know, kind of like, stable before, but like I think, I think at

Yeah. So we're seeing things like, as you said, mining payouts, potentially, well, mining payouts are now being done over Lightning in the case of NiceHash. So that's a really interesting development in the ecosystem and, people could say, well, in the past maybe mining, pools had to batch the payout, whereas now they can payout in a more instant or closer to, real time way. And that's a really interesting development, and, and even, you know, we're seeing the growth of some of the These Sats back st-type, businesses, things like Fold or Midrefill, so those are exciting as well. And Strike

Two, which is like, launched, relatively recently.

Yeah, and they're built on top of LND as well, aren't they? So that's also another interesting one, and, yeah, it's really cool to just sort of see the way this thing is growing up, and now payments are starting to become a little easier to route through, and obviously we're gonna get into Wumbo and all of that as well.

Talk about Wumbo. So, you know, the last time we spoke on the show, you were kind of talking about the idea of Wumbo, but now Wumbo is becoming reality. So, tell us a little bit about that and, that experience.

Okay, definitely. Yeah, I mean, so, you know, pre-pre, you know, like initially in the network itself, we kind of like had like a, like the first kind of like, you know, effective like training rules in the network, we kind of like had like a, you know, Livera, the way that's kind of like our, our style generally. So we, we, you know, we all kind of like agreed like initial, initial credit card, obviously you have a limit on there, just to kind of like signal, okay, we'll still like do our rental, we'll also kind of like, you know, cap off any sort of like, you know, security, issues as far as like, you know, like wumbo and things like that. I think like it's been maybe like, you know, three or four years now at this point Others they're doing kind of like no different type of trading, they're doing actually, you know, actual real place, re-retail size payments, they're gonna give cards and things like that as well too. Basically, one more thing where now, you challenge, no, no, no, Stephan, I guess kind of an odd thing, they want like larger channels, and it's usually good because now we're kind of like, you know, creating effectively like what's the backbone of the network itself, right? So, you know, so I would imagine like, you know, every single user is Business and things like that as well too. So you can, you can say, okay, well, initially we kind of like had the series of like, you know, paths and kind of like dirt roads and maybe even little like cops and roffers, now we're actually building kind of like these like larger like highways, which are kind of like, you know, will be kind of like the epicenter of transportation in the network and, you know, we're transporting Satoshi's, rather than like, you know, people or anything like that. So

Yeah, that's very cool. So as

to now where the channel part-participants, if they want to opt into that, they can now raise that limit and they can go, you know, one Bitcoin channel or even more. And, I guess as you were saying, this is really more like for big infrastructure, big nodes, you know, we're talking about like Bitrefill or Bitfinex or, Zap or people like that would want to have these big channels. Obviously, if you are a, you know, private individual or just an individual with a lot of bitcoins and you wanna run a big routing, routing node, well, nothing's stopping you now.

Exactly, exactly. Yeah, and I think it's also pretty cool that, you know, we can like reach this level of scale on the network, 'cause I think right now maybe it's like a little over a thousand BTC, like a net thousand BTC of like, you know, point one six channels, right? So like, you Good because now we're gonna, we can say, okay, well, now from here on, it's a little bit, maybe a little more bounded, kind of like now the network can, you know, grow at will basically, and we have some of the tools coming along, to kind of like let people, you know, know where to like better, you know, put their funds and kind of, you know, where they can actually be able to, things like that.

If we do want to do Wumbo channels and things, there are perhaps some additional security considerations that come with that because

Anything a little bit more about, you know, kind of like actually who they're, who they're accepting channels from, because like you don't necessarily want to accept like count anyone in the network itself, just because like, you know, if you accept like a bad channel, it can actually kind of like affect like the ability of like other nodes to route towards you, because maybe like, you know, that bad channel kind of maybe had two IPs or whatever else. So then, I think people should kind of like, you know, be a little bit more like, you know, wary as far as like And obviously, you know, just make sure you have like good authentication to your node, like actually, you know, lock down, you know, on the, the kind of network level to make sure you have like, you know, kind of like least principle of access to any individual on the node itself. obviously backups as well, you know. So, you know, one thing we released many years ago is basically this concept like SEP, which basically allows you to kind of, you know, close out the channel in the case that you maybe like lost, like, lost the data

point of view because otherwise, if you're in the node world, that's it, but now you can actually have kind of like a database fall over and actually have like, you know, more advanced architectures as far as like, you know, different nodes. That thing we work on as well as kind of like, you know, adding a lot more like support for like, you know, PSBT and kind of like, you know, different like pub key import stuff. Like right now with LND, you can actually like fund a channel from like your own wallet, so like let's say I can actually You know, only, only put keys on the thing itself, and it actually requires a kind of additional protocol to actually do any signing at all. So you can, you can, you can actually have like a lightning node that's online, actually running the payment itself, but actually no private keys on the node itself, and kind of delegates access to those keys elsewhere. So this kind of like lets you kind of separate security and have like different, you know, access control policies to make sure that things are moving, the way they should be.

I see. So, as I understand you there Keys that require, that are required to be hot, and so because, only some of those keys actually have to be hot, well, then now we can start doing things where it's a bit more of an advanced setup where people can, say, open a, a lightning channel from the hardware wallet and have it closed directly back to that hardware wallet, and in the meantime, they obviously some aspects of those keys have to be hot, because the channel, or because, you know, let's say LND is, you know, passing back and forward, Channel state updates, so obviously those keys have to be hot, but the other keys don't necessarily have to be.

exactly. Yeah. And like, we're looking at other different nodes as well, you can actually like, actually have LND, like, running kind of like a pub key only mode, like, let's say, I have like an Xpub, like, let's, let's say, let's say I have like two LNDs, like, one's kind of like actually like offline, the other one's like maybe online itself, I can Like not only like the privacy, but directly in kind of like have additional, additional like layer of security there too. So something we're definitely looking at, and you know, it's been something like kind of like a slow progress, like obviously to make sure we're doing it properly and kind of like also keeping up with the actual ecosystem as far as like the way the new standard is evolving. But, I think we're looking, looking pretty good here as far as like, you know, people, people being a lot more confident, kind of like having like, you know, more funds into their If you have like maybe like a few weeks to get the node, you wanna kind of like know when anything happens, you know, and do, do something pretty quickly 'cause otherwise you have to kinda like, you know, go in there and SSH to log in. I guess the other cool thing we've seen is kind of like a lot of people like making these new kind of like UI like node management, it kinda like lets, makes it a lot easier to kinda like actually like glance at the node and do other different operations and kinda like, you know, do, do things for a farm because obviously the

With the whole backups thing, as I understand, in the earlier days the channels didn't have that, and then so that was the new thing that, you came out with, with static channel backups, where, as I understand, it's like a static-- I think it's called static remote key, and then, so the idea is then it would then, if that channel were to close, it would then, the funds would come back and hit On, like, they'd kind of land back into your on-chain wallet, so that way now you can recover them. But the downside is you have to close that channel, that's the main downside there, right? Currently?

Exactly, yeah. And like, one, one of the things, like, you know, that wasn't as possible in the past, like, initially we kind of like had something where we actually kind of like had this like random nonce added to like the, the pub key of your wallet basically. So I could, so before, like, you know, in this SDB case You know, effectively like in your wallet, it's actually in the channel. So if they force close, you get your money back immediately. Otherwise, if you force close, you can actually do the whole CSV thing and then get through there. So it's, it's kind of, it's definitely like a step forward now because, you know, you trust the other party like much less, 'cause like now you're no longer relying on them to, like, give you this, give you this particular secret. But, you know, it is the case right now where if you wanna recover, We have a kind of like, you know, more replicated database system, and then beyond that, maybe you have other kind of like more on-the-fly replication going on, and, that's something we actually released recently in LND for, dot eleven. So we have something called, like, we kind of like have the initial, like, incremental, support for this, like, database replication, it's, it's needed, like, you know, it basically can like run, like, database kind of like clustered setting, so rather than like having just one copy of Things like that as well too. Now we're seeing all the nodes catch up to like, you know, allow people to have similar architectures running on Lightning, which definitely required a friend wouldn't do kind of like, you know, large enterprise types of deployments stuff like that.

Okay. So as I understand you there, well, comparing back to, let's say, earlier days of Lightning, I knew some people were trying to do things like R-Sync, which is like a way to try and copy it over, but that's like not very safe. And as I understand, what you're doing now is

Goes down, then you would still have, or at least you've, you've got a better chance of having the correct up-to-date state, and so that kind of how-- that's how you deal with kind of the quote-unquote toxic backup, problem.

Exactly. And then, then, then, then another cool thing beyond that as well, you, you can actually kind of like use the database system to itself to like, you know, have like failover between the different LND nodes. Like, let's say I have like three different LND nodes, like one of them is actually kind of like the, the current, like, you know, primary like leader basically, like, you know, if that one goes down, it's gonna fail over to any of the other ones, which is really cool because you, you can then kind of like implement like I'm definitely excited about this. It'll be a chance, it's a lot more like, you know, it's a lot more meaningful to kind of like larger scale deployments and like what people really need to like, you know, be using large amounts of money, also just feel a lot more confident because now you can actually have, you know, all the state be pa-be-be backed up and backed up in a way that's actually very secure or like a little more safer because like you're not doing a manual, like the RISC thing, which, like

you

said, can get out of You're doing your own little benchmarking and testing on that as well, how, what are you seeing on the network?

yeah, definitely. I mean, so I think it's also kind of like people like reporting to us generally that they're having a lot more success. I think one thing definitely, the most asked kind of like, you know, people like getting a lot more used to using the MVP itself, kind of like allows you, lets you actually like, you know, split, split the payments into different, smaller amounts, and this is really good because now I think now we're like seeing a lot more success rate in people, you know, going beyond that, maybe to like the five, like the fifty to a hundred dollar, like kind of, you know, realm, and that's also kind of like, you know, pre-wumbo. I think one thing I'm curious about as well is like, you know, will people actually close out their kind of like, you know, pre-wumbo channels, like make larger channels, or will they just like add them on, add them on basically? So I think, I think we're still yet Like people are getting like, you know, better actually operating their nodes in the first place. Like, you know, everyone kind of has fancy like node being, like, "Oh yeah, you can, you can click the button and print money," you know, everyone wants that, like, you know, that dream of plugging a Raspberry Pi to give you, like, you know, twenty dollars a month. But obviously that's not the, the best time with the kids, 'cause you actually do a lot more actually managing your own channels. So I think people have got a lot better at, at To the auto education too. So it's kind of like a gradual thing, and I'm, I'm really, you know, excited to kind of see the progress we have going

forward. As I understand you there, it's essentially in the earlier days, and perhaps even now, there are, there are, there were, and there are a lot of zombie channels and kind of just, oh, that too. you know, just a lot of, nodes that got set up and then didn't sort of get maintained. And then so every time someone tries to route a payment through those, The tooling around how to manage your lightning node, such that you're doing things like balancing the channels better or that you've just kind of, you have, I guess your node has a better view of the network, so now it knows how to route those payments better. Exactly,

exactly. Yeah, and like, you know, it's, it's another thing as well, like we actually improved a lot in LND, kind of like the pathfinding aspect, because like, you know, before maybe like, you know, like very early days, like, you know, twenty seventeen, LND would basically forget all of its memory after maybe like fifteen seconds, so it was kind of like, like a goldfish, it'd basically like, you know, so even, even if it had a channel like work really well in the past, it would basically never try it again because it basically like And then also learn basically fail over successes, basically every time as you make more payments, you can again actually progressively like learn more of the current network map itself. This is also even possible to actually like kind of like, you know, combine like the type of like memory or like, you know, mindshare of like other different algorithms into like one kind of like central more or database basically to like help you, like, help you do a lot better pathfinding. That's not really done yet, but I think generally like people found a lot more success in that because now like LLMs basically

learn Have a dashboard to see what the current state of your lightning channels are. So for example, RTL Ride the Lightning, there's, ThunderHub, as you mentioned, LDMon, which is by Lightning Labs, and you also mentioned, I think there's, well, I've seen also Lit as well, Lightning, is it, Lightning Interface? yeah, it's like, yeah,

it's like a li-lightning terminal, and like, you know, you can kind of like smash into Lit. Yeah,

yeah. You're always good with naming, hey?

Yeah, yeah

So how, how important would you say, like doing all the rebalancing and, kind of the active management of a channel, of your channels, as opposed to, let's say, the more, let's call it laissez-faire, just kind of opening a lot of channels? Would you just say that's very Bitcoin, kind of capital inefficient to just have a ton of channels and not spend the time active re-managing and so on? Or can you give us a sense of that?

Yeah, totally, totally. I mean, it's kind of, kind of thing where it's like, you know, you need, you need to like, like actually like place channels in a place where they're kind of like most actually demanded, like they're most wanted to be used. Otherwise, I can like, you know, maybe put like 10% of the network, but if I put those, those channels in basically like, you know, cords that maybe aren't really well demanded, they're just gonna kind of, kind of like just sit there, right? So Like large amount that, that, and isn't actually getting trapped at all, I can like, you know, really have a large opportunity cost because I can, actually, obviously, I have my capital elsewhere, and would be a lot more effective, right? So I think people are now realizing as well, like, you know, they're kind of like, you know, watching like social media, they're seeing, okay, there's a big launch over there, they're kind of like on these like IRC groups and like Telegram or whatever else, kind of like really, like, you know, I want to like make sure that, okay, well, I'm actually getting a return, or I'm actually even like writing payments, because some people just want to, you know, feel that they're helping the network itself, and I think that's a really cool thing because it's kind of like a very, like distributed and open access thing where, like, you know, unlike, unlike the internet, basically anyone can come on, on, come to the network and put on channels and actually even start to write payments themselves, which is really cool.

And also while we're talking about managing your channels,

Default fees, but now we're starting to see a little bit more around manual management of those fees. So for example, I've seen, LN Big, who's a well-known, routing, well, he runs multiple routing nodes along the network, yeah,

twenty or something, yeah, yeah, yeah.

Yeah, and they've, they have like big channels and lots of, you know, liquidity and so on. And I've seen, he's been saying, "I'll charge the fees depending on how well balanced that channel is. So if it's not balanced at all Just comment a little bit around, what your view is in terms of Lightning Network channel fees.

Yeah, totally. I think it's definitely good that people are actually like now starting to like turn with the fees, because otherwise, like, you know, like we, we kind of like start things at the bottom, like obviously that's not gonna, like, that's not gonna, like, you know, persist forever, because like people actually at one point need to kind of like take into account, kind of like the cost of capital, also kind of like operational costs themselves too, because otherwise To like inform people like actually doing like client side routing, 'cause like for example, like you know, I, I maybe like wouldn't take that high fee route because like, you know, it was unbalanced, so that kind of like gives, you know, kind of clients like as far as pathfinding routes, additional information that makes you like, that's why they're doing. One thing we've seen as well, like, you know, one thing that like we run something called like Lightning Loop, which is you let you kind of like, you know, move funds in and out of the network Make sure like blue actually works as a kind of like a system itself. We're seeing the kind of like an automated channel placement to basically make sure that you're earning money. I think someone, one of them tweeted like maybe they're earning like, like thirty dollars, you know, in a month basically because like they're kind of like they're helping us like actually service the volume. I think, I think it's interesting because it's kind of like a multi-sided like marketplace or even kind of like, you know, like this like, a little more complex like actual, arena You know, advance as far as, you know, advertising things. I think that's great because otherwise, like, we definitely want them to be, to be self-sustaining. 'Cause otherwise, like, you know, long-term people aren't really earning like money out for writing, they're not really gonna do it themselves. But I think it's a thing where you definitely have to be like very careful as far as your capital placement, like management also coming from page schedules. Otherwise, you may be, you'll kind of like over-allocate your capital and like, you can actually put

Basically, people have to think a little bit about where, which direction they are placing their channels so that they are more Bitcoin capital efficient. And so theoretically, you want to open a channel in the direction that you believe the flow is going towards. So if you know somebody is setting up a new store and it's gonna be a really popular store, well, you wanna open channels in that way so that you can collect routing fees. and as you mentioned, there's a little bit more going into the thought around that. So would you say, like, routing fees, right? Now, they, they need to rise basically.

yeah, I mean, I, I think we, we see them kind of rising success, successfully. I think one thing we're seeing as well, like people are actually kind of, you know, using like the actual, the actual utilization of their, their channel as far as like actually if it's writing, to kind of like, you know, give signals if to basically, they should put more capital or not, right? So if you're gonna have a channel that gets maybe like, I opened the channel, it gets depleted almost immediately, okay

And that's probably the case where it's like, is it gonna be like, be a factor somewhere for kind of like on-chain fees levels themselves, right? So like, you know, if chain fees are very, very high, maybe like running fees come up somewhat, but I would see that maybe like there'll still be kind of an equilibrium where you'd, you'd expect still like, you know, certain, certain payment types to always be cheaper on Lightning itself. So I think as, as like, you know, chain fees rise a little more, like depending what's going on there, I think

Maybe we'll have kind of like distinct networks or maybe like, you know, not networks, maybe there's just kind of like subnets or different topologies as far as the different payment sizes. Like maybe, maybe like there'll be like a group of the red nodes that really like, maybe like small payments, and maybe they do market it once toshii, like a hundred toshii something like that, but maybe there's other red nodes, maybe like a little bit larger, and they, they only prefer like large payments. One of the things we'll like, you know, in the network, And sort of navigating, it should be able to kind of navigate that, like, you know, network to basically like find exactly what, like, sub quarter or like sub graph you should be using to send the payments most, most efficiently.

Wow. Okay. and so a couple things there. So it's like over time, as we see a lot more people join Bitcoin, we're gonna see more congestion in terms of the block space market for on-chain fees, and then that will kind of have this interesting dynamic in terms of what is the impact in terms of lightning routing fees, because lightning To calculate or at least account for the on-chain fees of opening and closing channels. Now, there's some level of opportunism there, right? Like a lightning node operator can wait until on-chain fees are low to try and open channels, but there'll be some cases where maybe they need to quickly close a channel even if the fees are high, and that's where they're just gonna have to basically cop the fee and pay it, but then that will all factor into, the light-- the routing fees, and One other thing I'm curious there, it's kind of, there's gonna be this funny dynamic over the next, well, let's say, if we really zoom out and we think over the next, you know, five or ten years, there's gonna be this funny dynamic where a lot of people w- might not feel the need to try to use Lightning until on-chain fees get high, just like how exchanges maybe they didn't have an, some of them didn't have an incentive to get SegWit until the fees went higher, right? So it's kind of this funny game of like, once

but then over time, you know, smart engineers such as yourself will find ways to basically pack in more for less, and then that will kind of bring it back down. So there's kind of like this sort

of, yeah, yeah, this thing where it's like, right? Yeah, so basically, you know, people kind of like utilize the, the resources like the medium a little, a little more like intensely, then we can make it a little more efficient. They have, like, you know, demand goes up, make it more efficient, it's kind of like this, like, you know, ongoing battle of that, in that kind of regard. I mean, so much of what you're saying, like, to my knowledge, we can see like Nice, NiceHash, one of the, one of the reasons they added, like, you know, Adopt Lightning because maybe the customers are basically eating out too much on JPs. So there's a really cool dynamic where it's kind of okay, well, you know, people will move to like whatever service they feel like and then we service them, you know, as far as the fee level and we're all just kind of like have this like different, dynamics as far as the system adjusting and like, you know, making getting more efficient and also managing how you're kind of like, you know, adjusting with all that load all the time at the same time

Fantastic. And so in order to make that work as well, so even in that example where NiceHash want to do Lightning payouts, I guess the interesting thing there is most of, well, most of the flow is going in one way, right? So it would be going from the mining pool to the miner. And so then we have to start thinking about, well, what is the, like, if they've got channels open or, you know, or they're just gonna be paying out to miners, then most of the time they're gonna be pushing money away from themselves, so they're gonna have to Swapping in and out in order to make that viable, and perhaps that's where something like Loop comes in. So can you tell us a little bit about Loop and, how that would help in that kind of scenario?

Sure, sure, yeah. I mean, like basically what you're saying is like, you know, in the ideal case, it's gonna be like fifty-fifty. That's kind of what our, like, ideal, like, you know, three-way collaboration is basically like you're sending, receiving the exact same amount that, that you're, you have coming in, so On their end, basically like, you know, people send money to them, and they wanna actually buy some hash rate. Like, I just think like, one thing basically Loop does, like, you basically, Loop's kind of like, like, you know, it's, it's, it's almost like like a release valve or kind of like an ability to actually, like, you know, like, remove water or like money, like, 'cause I think of like, you know, it's kind of like a series of, like, you know, money issues basically, to remove funds Funds into their channel, basically, and what they do, they kind of like send us, like it's kind of like a special, like an HLC contract on chain. We basically send them, an HLC, you know, over to their end on their node, and then once they pull the HLC, they basically get those funds now. They able to send more, and then we, we get those, funds ourselves. It's also kind of like, it's also like, you know, very transparentized and trustless too. It's like one way people use it Send out those funds back, back on chain. The cool thing about this, because like, you know, like they can actually use it to also kind of like minimize the amount they actually have in the line at any given time, because like, let's say every single day, every single day they can actually like send the money out, you know, from the chain into like on-chain, on-chain address, like they have that be cold. So Loop is basically kind of like, it's, you know, some people call it like a trustless like on-chain bridge, it's also kind of like

a Loop is a way that you can bridge on and off chain, and part of the benefit is that you can keep those channels open for longer without having to do what's called channel churn, right? Like closing and reopening a channel.

Exactly, exactly.

And so then, I guess just summarizing it then, if you are the kind of person who does a lot more payment than receiving money, then you care about having Loop in, right? So that means you would make a Bitcoin on-chain payment to this Loop- Server and then receive Lightning money, and now you can pay again. so I guess that, that would be like, I'm a retail individual and I wanna do some spending with Lightning, and I'm, and I'm got my balance down to zero or whatever, in that, or it, as in my local balance is going down low and I need more because I wanna spend more, well then I can pay a little bit on chain to receive more Lightning balance, and then, as I understand you, then the other way around is Loop out, and that would be obviously Let's say I am a merchant, I'm receiving lots of money, and then all my channels are now, starting to be full on my side, i.e. my local balance is becoming very high and my remote balance is getting very low. So this is kind of a way to, I guess, flush it out and get more capacity, right? Exactly.

Like one of the cool things about Loon is that we're actually able to do kind of like a combination of basically both like on-chain and off-chain batching. So basically, on-chain batching and off-chain batching, whenever we're actually doing Because like, you know, we're basically wanting to spend less money on transaction fees and it's good for them too, because then they actually spend less money, you know, kind of like, you know, doing any swoops themselves. Like one of the things like basically like the combination like loop and mpp is like really, really cool, because like, for example, like, let's say like, you know, I wanted to basically use something that wasn't loop, I would maybe have to like close, like, let's say I had

like, you know, new channel, I wanna Like sending out to other people, maybe I'm like actually like selling, you know, Bitcoin Lightning, like, you know, they give me dollars, I give them, you know, I can like, like, on the, on the network itself. One thing we can do, they can actually like do a loop, loop transaction, and then rather than like, you know, just send the money back to themselves, they can actually like, you know, create like n different destinations and basically send those like n different payments to other, other individuals and then send the remainder back to themselves. So be This is what Fire

system has, that was really cool. Yep. So as I understand you there, you're talking essentially about a massive efficiency gain. So in that example, I might have, okay, so if we're talking in the loop in example, I might have, or maybe the loop out example is better, but anyway, loop in, Yeah. I guess so that what that means is I might have lots of channels and I can refill all of them with one on-chain fee. exactly, exactly. And then in the loop out case, it's I'm a merchant and let's say I've got like twenty channels incoming to me and I've exhausted the incoming capacity and I wanna kind of flush it all back out in all of those channels with-- And so if we're kind of comboing loop with MPP, multi-part payments, then you can flush it all back out in one transaction, and so you can get a lot Chain efficiency for that, right?

Exactly. It's, it's one of the, like, it's really cool thing, kind of like, like, I guess, one of the things like, I guess it's really cool kind of something like Bitcoin, kind of like, you always try to like drive that like maximum efficiency basically, and like, okay, we realize, okay, well, you know, we can kind of like add like this additional dimension, so it's really cool effect, like, you know, just kind of like, like running into basically obstacles and problems, and like, you can

Or, you know, it's basically like pack more transactions onto a single byte. That's kind of like what's all about, what's Lightning, you know, all about, like Loop, you know, MPP, things like that as well. Basically, like, you know, having most efficient, like, utilization of the chain as possible. And then from there on, we kind of like, you know, have different effects as we make more efficient, we have to keep on demand, things like that too. But I think it's one of like really

cool,

tech to the The best way to, to use it?

Oh, yeah, definitely depending on like, you know, their competency or like, exactly like what they're doing. Obviously, terminals like, you, really easy way to like, terminal, like, that's the cool thing about it, it's, it's kind of like, it's like one buy and the package is like L and D, loop, you know, some other tools I'm working on as well, and they kind of like the UI and one, so if you, you definitely need kind of like a, like a visual, you're kind I want to, you know, I want to be able to, like, receive, you know, ten e-ten BTC at any given time, you know, from like three to nine PM. We can then have like separate, like, a contract basically do this in the background, basically make sure the user or like the merchant, the user always, always able to, you know, have whatever threshold they need either to be able to receive or send more money. So it's a really, really cool thing as far as

kind of, you know, I mean, it's one thing we definitely work on

You know, transfer the preferences to the system, and the system is able to carry those out effectively, actually do it. So it's a little more set and forget. It's like, you know, so this kind of thing where it's like we're trying to get set and forget thing, we're not quite there. It's definitely got like, like, like an uphill battle, but we're getting there, we're getting more progressive, and people getting more advanced. So I'm like, we're really looking forward to like when it's a little

bit more, you know, just like

straightforward, Discount if they're, if they're willing to like wait a little longer, because like, like for example, like let's say we have like one little bit of customer, like, you know, I, I, I'd be able to basically service them like at that particular point. But like, let's say we basically, okay, well, you can wait thirty minutes, we basically like bundle, bundle you with another individual, it's a good for us because then we're able to actually like do, you know, on-chain batching to like make sure those two individuals are basically on the same I think it's from, from, excuse me, from like, you know, like five basis points to, around like, you know, like forty or so, depending on like what's going on the chain, because like, we also wanna kind of, you know, make sure that we can incentivize like people, like, you know, doing loops, for example, on areas where it's gonna cost, cost like less change fees, right? So maybe it's the case where maybe, like, you know, we'll give you a quote, maybe the quote Like you, those are the things that keep it to account what's going on the chain and then kind of like these like optional on-chain services. So one thing we do, we kind of like look at the chain, we look at what's going on, we also basically then go try to promote batching on our end because it's better for the user and also better for us, and it's better for the user too because like there's also like less chain interaction in use and also kind of intermingle with other, other people interacting to fast and optimize properly as

well. Yeah, so essentially it varies on Efficient on both sides if you can, on your side, batch it across multiple customers and do it all in one, as opposed to, for each individual customer. But obviously, it depends on, that person's time preference as well. Exactly,

yeah. That, that's one of the most important things. Like whenever you're talking about fees or you know, talking about any sort of like on-chain activity, it's all about your time preference basically. And like the thing is, like, you know, if, if users are willing to kind of like have a longer time preference, You know, how long you're willing to wait? Do you need it now? If you need it now, or maybe you have to pay, pay a little more. If not, we can kind of, you know, have you along with like twenty other people and like do, give it a little cheaper, obviously like for that in the future.

Yeah. And the other, I guess, the other example could also be if you are just starting a new store and you fund it with some bitcoins, but you need incoming capacity, and so you might open some channels and then do a loop out to

Internally, it's like, you know, so let's, like, I, I open channel that basically do a loop out right from those coins, I can then take those coins and loop out, use that to basically open another channel and then loop that out, and I can keep continuing that effectively. So I'm kind of like, you know, stringing this capital around in like circles effectively, but at the same time, like, the money is still like moving around the network, but basically like gives you kind of like the ability, maybe like, you know, open channel one BTC and eventually get

With the use of, you know, Lightning Labs Loop, how does that compare with, say, other services? I guess there are some that might use what, what is more like the, the trusted swap in or trusted swap out. How, how, how do, how does your one compare?

Oh, yeah, yeah, it was a good question, 'cause obviously you can kind of like do it where, like, you know, I give you the Bitcoin, you give me like a channel or whatever else, and like so, it's something that we definitely try, try to like shy away from as Like, you don't have to move the user funds out on, on any regard, right? So I think the difference there is like, you know, it just matter like if users are comfortable, maybe people are a little more comfortable with kind of like, you know, doing like smaller amounts with, with trusted swaps, where you just always try to make it, you know, not risky as possible, just make sure we kind of like minimize like any type of risk on, on either end of our things. And I guess the difference as well too, from those that, you know, I think they're Which basically kind of has different cycles, like you know, as we're, as we're kind of scaling up, people are recognizing, okay, well, well, like, you know, that's really good now because they're actually doing a bunch of routing, they're actually more capital to us, we kind of like utilize the capital and we kind of like a cycle, like, you know, this kind of growth in the system, itself. And, but yeah.

Okay, cool. and while we're speaking about channels, I've seen some mention of channel accept

And any like lightning node like you, you even the online node basically just to accept every single time that you have incoming, which is the kind of like a, like a simple thing to do, 'cause you can say, okay, well, you know, like they're actually putting money towards me, I don't really have any cost, and maybe like every, everything's fine. But I think like, you know, you may have a cost in the future, they actually like, they can't like interact with you, so what you actually need like, you know, put those funds on chain It can be, okay, well, you don't have enough capacity. It can be, you know, like you're not connected to other individuals. Basically, people are kind of like now, like, be a little more se-selective about their, their, their channels. So you, you, you kind of seem like, well, you seem maybe the same like a bunch of capital Bitcoin itself, you basically want to select kind of like the best capital, so like, you know, add to your portfolio because I think the portfolio is like a little bit more healthier, that basically means you can route more, route more, more effectively.

And so essentially it means as a node operator, you can be more selective about who opens channels with you. You might say, "I want a minimum balance in that channel," or "I want you to be running a certain type of node." does it also allow, conditions in terms of things like Well, I don't know, maybe this is getting more into like BOS scoring, but does it also allow things like uptime and things like that?

good question. I mean, it's like, you know, the API itself is like very flexible, like, you know, I could like for, for say, like, you know, like, let's say I have a new activity, like, create some, some other thing, or I can even like a budget score. So one of the things I think eventually people will maybe usually basically like, you know, have a lot more com- Anything. So I can say, okay, well, I only accept nodes that have like a even number of the last, you know, hex digit of their, of their, their pub key, right? It can be that, it can be that arbitrary. Or it can be, okay, well, I only accept nodes that are at least two years old, that have one, one BTC in capital, that are connected to this and this and this, that have a score of this, that like, you know, have this particular centrality. So we

can, we can get really, really advanced, I This individual have like a node for two years, this could be, all the time is like one, one and a half years old, they're gonna kind of like show that they're kind of like really committed to the system itself, they can like maybe use this to, use that to extrapolate as far as the quality of the node in the future.

Someone might be concerned and saying, "Well, hang on, does this mean this is how Lightning Network gets permission? Because maybe it's hard for you to run, you know, your node because now other people might be really, selective in how

Theoretical, and I'm kind of exaggerating a little bit, but how are you thinking about that kind of, that idea of kind of balancing off the ability for anyone to be able to, you know, use lightning, but also having, giving people the ability to be more selective about who they accept channels with?

Oh, yeah, super good question. I was, I think it's, I think it's kind of like going like different tiers, that maybe like you have like, you know, T1, T-T2, T3, so much like, like, like, like basically Nodes, kind of like, you know, connecting, you know, interacting with everyone else, maybe you have like different paths. I think, I think it's gonna like be kind of like a gradual scale up, where like, you know, you, you probably like can't go from like zero to like, you know, Wumbo mega node basically, right? You can kind of need to work your way up. Maybe, maybe initially you kind of like connect to other nodes that maybe are like at your similar like level of like a square or else, that kind of progressively like you work your way up to Maybe the console can launch by kind of like, you know, massively, like, you know, opening like, you know, a wider number of channels than the network itself. And one way you can kind of like, you know, guard against that pretty effectively is like, you can say, "Okay, well, I only accept channels from a node that maybe is, like, you know, like, of at least a few months old and has a number of channels, kind of like, reduce the ability for someone to, like, basically, like, you know, rapidly, like, gain like It's Bosworth, you know, type like, they're not great at beginning, you know, they gotta work their way up, gotta get, get, get some credibility, but kind of like, you know, the different like conference thing, then they kind of like, you know, work their way up to get, to be better known.

So if you wanna be a lightning hero, you gotta slowly start, you gotta start one step at a time and work your way up.

It's like, yeah, you, you know, you gotta be like on Twitter, gotta be on Slack, gotta kind Like maybe four hundred or so radium, radiums, which we kind of like monitor, we've been there from the start, we kind of like consider those kind of like the like top tier of the network right now, which maybe that's like, maybe that's around, that's around like four or five percent of the network, which we consider like really, really good. I think we, we definitely want like you know, grow that over time. It's something that we watch, it's kind of like what we call like the strong set, of the network itself that we kind of Building roads so we can basically add channels instantly, right? And there's very like low cost compared to like, you know, construction, like material, whatever else. So it's kind of a thing where you can kind of like, you know, very like rapidly adapt the network in terms of like where the actual demand is most needed, as far as like where people are routing to, and also kind of like, you know, what the variance or barrier to entry is, which I think will always remain low if you have a computer and some Bitcoin, you can definitely do the network and actually start routing payments or

Speaking to that as well, oh, and to that idea of good lightning node operators, can you tell us a little bit about updates in terms of BOSS scoring?

def, def, yeah. So something that, like, you know, it's something that we haven't officially announced, but like I think we've kind of talked a little bit more. It actually started like a while back, initially we were kind of like working on, working on like the mobile application, which we don't, we don't really do as much anymore. Basically, the question was, okay, like, Like, you know, look at the graph itself. The thing is, like, you know, just because you have a lot, a lot of channels in the graph, maybe, like, you know, you, you look, you look connected, doesn't mean you're actually able to actually, you know, effectively route channels in real time. So basically, Bonscore is about, like, you're actually adding a little bit more kind of, you know, real time, you know, information into kind of this currency itself. So what it's doing is kind of like, you know, looking at You see, you're connected to who and who and who, like, you know, if you can actually wrap your head around effectively, that's what BOSCORE is actually trying to do, tell like, you know, actually like, you know, get like, what are like the real no's, like, you know, ignore the pretenders basically who maybe aren't like quite there yet, itself. Something that, you know, we're definitely like working on, kind of like developing a lot more. We actually kind of like had this like recent analyst where we kind of like, you know A lot of people would like to see what's going on there. And I, I obviously think like, yeah, ideally, like this should be like the entire network, ideally, but it's obviously, it's, it's a thing where like people got, they got like some of the knowledge maybe from the tools, a little bit contrary in the future, in the future, but we definitely want to like make that a lot more easy to like, you know, rise up and actually, like, you know, be able to like evaluate your node. I actually think, I actually think that, you know

Over time, also maybe you can recommend other individuals, maybe like new business or whatever else, okay, connect to these guys. These guys are actually pretty good, maybe they actually like, you know, better than someone.

So, it's like over time, the BOS scoring might help People who wanna use autopilot, so they, if they want to be more like a casual node operator, they can kind of rely a bit more on the block scoring, I guess, rather than them manually doing things. Is that partly how it might work?

Yeah, yeah. I mean, to be, basically like, you know, you can basically view like, kind of like the block scoring, like, you can view like the no, no ratings of it to kind of like, you know, looking at like the real time activity, because like, you know, someone can like look like they're a Looking at the network itself, I think it's kind of like more of a supplementary thing, is you look at kind of like what the actual activity, what's going on is. I think it's also like very, very dynamic. You may see, well, maybe like the current like network topology doesn't, doesn't really change too much, like you know, network is changing, you may see required change, and that could be taking a long time, also can be relatively expensive. While the bots could basically change, you know, from the hour or like you know, or even the day, whatever, So like looking purely at the graph, we've made-- because I feel like a lot of people maybe typically look at, okay, well, you know, that node has a lot of channels, it's gonna be really good, but that's not entirely true. You can have a thousand channels, but I can basically just like have my node in the, you know, in like my closet and I never touch it basically, and I'm not really gonna be, you know, that effective as running on that.

one other thing I saw from the documentation was, this concept of HTLC interception. So can you tell us a little bit about that and what might it be useful?

Oh, yeah, it's super, super cool. So that's something that, you know, we, we talked about a while back. Like we always kind of like have these like concepts internally, then we maybe like make an issue with like, the, this company like agrees to kind of like have like a pretty, you probably remember what they made, made the only thing they actually made is basically the idea is that, HLC in certain, let's get like about like allowing individual basically like to spend the like forwarding flow of the network or sorry for HLC, then decide basically where to put it from there on. Like there are a few, you know, different use-cases. Like one, probably one biggest use-case I think was kind of for mobile phones, because that like, let's say, you know, one thing obviously right now in order for me to send payment, I basically have to have my phone on the entire time, it has to be like fully safe process. What Bob will, I mean, sorry Bitcoin, I can send Bitcoin right. So basically making it a little more easy. That's kind of like just like one use case, right? So basically the ability to like, you know, kind of like, you know, preferentially like forward or like halt the, forwarding of different HLCS up until like some other condition is met. So like if, if you heard of like holdable invoices, basically holdable invoices are for kind of like, you know, receiving payments, HLCS in, in a certain way basically for people like forwarding payments. That's like one different

Okay, so if I understand that, it's like, as an example, someone like Breeze, for example, the, provider of a lightning services and, they run a, a wallet. So as an example, the user might be out and about and they might wanna take a payment, but their phone is off at that time. And so then what happens is Breeze might be taking that incoming payment and using this HTLC interception, then it will then notify that person, and that could just be like a notification on their phone, like a- Push notification to then turn on their phone, turn o-open their Brave wallet so that they can now receive the forwarded payment. And as I understand you, the, the technology or the concept being used underlying is the HODL invoice. And so that's this, idea that- Oh, well,

it's not exactly an invoice. It's, it's more that like, you know, basically like, so like what, what works? Anytime you get a payment forwarded, like it basically is okay, well, should I forward it like automatically? Should I cancel it

or should I settle You basically can use kind of like, like a middleware like layer basically where you can, you can add like a custom forwarding logic into LND itself. For example, like, you know, let's say someone like wanted to have like a, like brand new like routing protocol, right now we basically use like onion routing or circuit routing, they could basically like add additional information into the payload and then intercept the HLC and then put it to basically some other routing network, which maybe doesn't exist right now or maybe it's like, you know, some, some academic thing or whatever else, and then like let And put some additional logic in there to decide what, what actually works versus still having a intermediary for it in

the end. I see, yeah, yeah. and so, well, I guess one criticism I've heard of that idea when I was chatting with, some other lightning developers is they might believe, or they see it like that's locking up liquidity along the network. So what do you think about that? Is that just locking it up at the end point, or do you see that like, well, that's additional HGLCs across the network because that payment hasn't been released through

I forget the term, but worst case, like you know, like with HLC, like you, you always kind of have the worst case delay of like the actual, like, you know, like till TBD, so, so till TBD of HLC itself, right? So like, and again, someone should basically be ready to basically like, you know, always incur that worst case delay. So like, the, like, you know, something like HLC in general, basically holding the bridge, don't really make the worst case delay any worse, maybe it makes it a little bit like less For the time value of Bitcoin on ongoing basis, then I think the kind of like, you know, resolves a number of issues because like at that point it's like, okay, well, you know, you're basically leasing my coins out right now. Yeah, I'm getting compensation, doesn't really matter. So I think we, you know, we're definitely like missing some of like additional incentive, which I think, you know, would kind of like make things, things a lot kind of like more cohesive. But I think we're definitely getting there, and we have a bunch of other

ideas in that Payments and if somebody does a lot of this kind of HLC locking up, they're not actually paying because if the-- unless the payment goes through. And so it's kind of that challenge around, kind of locking up or jamming up other people's nodes or channels because it hasn't gone through yet. And what you're speaking to there is around like what kind of incentive could be made to allow or give some incentive for the fact that you are locking up somebody's HLCs on their channel because there's a limit on the number of HLCs per channel, right?

Exactly. Yeah, So it's basically trying, trying to like, you know, it's basically like spam prevention, right? And like, what, what can we use for spam prevention? We can basically use lightning itself. So basically like, you know, let's say like, different nodes, like, let's say they kind of like, like, enter into like some different, like, you know, critical stage where they say, okay, well, in order for you, for me to even forward a payment, I need like, you know, two percent of the payment upfront, right? And maybe You know, people need to be compensated, you know, for the usage of capital, otherwise, like incentives, like, you know, won't, it won't last, effectively, or we kind of like have some other, you know, kind of like, you know, things that can be exploited by, different network participants.

And I guess while we're on this topic of HLC forwarding and, you know, attacks on the Lightning Network, I'm also curious to get your thoughts on some of those recent papers, for example, there was the Flood and Loot,

As I understand and, so I spoke about this, so listeners, I spoke about this on the recent episode with Christian Decker as well, but I'm curious what your thoughts are on flood and loot and these kinds of attacks around, and how, you know, you think you're thinking of ways that Lightning Network may mitigate those kinds of things.

Totally, totally. Yeah, I mean, it's like, it's one, something that, like, you know, like I guess we want to lay the ground, like, it's not something that's like brand new to basically It's, it's like something I was actually noted in the original light-lightning article here, as far as like, you know, it's like people remember in the past, people had things like time to order, else people were trying to, you know, stop kind of like the, you know, attack of kind of like, people basically going to change it all at once. Like it's not something that's new, but I think when the cool thing with cold new paper, they kind of like, you know, actually start to like formalize some of the actual attack itself and also give Maybe like do so in like kind of like a wide spread manner, so we can like, like, we can have like different engagements basically, and try to get, get, do that all at once, so they can't, you know, add a thousand BTC all at once basically because people aren't gonna be doing that, but that helps out. The thing that helps out is like, you know, one other thing with, like, you know, kind of requires the ability for an individual to like not actually be able to get their HLC into the chain at any point. We have something called anchor channels

There are things like, you know, there's also kind of like a number of like actual limits on channels themselves. I can say, okay, well, I only accept a max of, like, you know, like, million Satoshi, like, ten thousand, like, that kind of thing. So I can actually kind of, you know, like, limit my exposure basically, make sure like someone actually can't, like, you know, go through and, you know, actually do this all at once. There's also kind of like additional things like prepay as well. So I The attacks aren't like a high enough, maybe like a kind of negligible HP, but we can basically increase the cost of certain, certain types of attacks like giving, giving people the, the ability to actually either detect or maybe even mitigate some of these attacks, and then we'll be like in a lot better place.

Yeah. So essentially, a range of mitigations around things that can be done to make that kind of attack more costly or just less technically feasible. And some of those, as you mentioned, things like having channel acceptor policy, the use of anchor outputs,

and HDLCs on a given, channel, right?

yep. And like, you know, there, there are also kind of like, you know, base layer changes possible. For example, one thing we talked about in the past is okay, like the ability to kind of like maybe delay time locks if we felt like the, the blocks were like, you know, basically over a certain high, high watermark, that sometimes like has other kind of like, you know, issues because like you can say, well, you can tell them then delay indefinitely, Basically, like, you know, make sure we can actually mitigate these things at multiple levels and also give people the ability to, like, detect if they're going on and also, like, you know, otherwise. So, but yeah, I mean, it's one of the things where I think, you know, kind of, kind of like by like a Pincho-Pincho principle, like, you know, if someone has like a thousand transactions, only two can get to the chain at any given point, so someone can't get it. So it's basically

like, you know, how

yeah, well, I guess for a channel jamming attack, that's probably another similar one where maybe you might restrict who connects with you or how would that, how would you sort of think about that?

yeah, I mean, so the one, so basically like one thing we talked about in the past, like, so, so let's start right now, like, what, what, that kind of required like an individual basically open channel to you and kind of like, you know, maybe like, you know, exhausting all that, crappy channel instantly, right? So it's On the HLC on the channel at the time, maybe you know, as it was settled, like you know, everything goes, goes properly, maybe I raise that to twenty, maybe to twenty-five as well. So I can kind of like, you know, limit exactly like, you know, how much you can utilize the channel, 'cause I, 'cause anytime you're routing towards me, I start to kind of, you know, put my capital somewhere else, basically, like, actually, you know, make the HLC look itself. So if I, if I kind of like, you know We can have a long range, we can have a long term type, or maybe over several months, but definitely like in a lot more, you know, time and like energy or almost basically that kind of like an act of need to like get up to the level what being a good node is. So we can kind of make it harder for someone to kind of, you know, come to the network and basically like, you know, monetize all the resources, because the kind of thing where it's like, okay, well,

if you're using my, my capital, you basically, I need About that, what is an anchor output, and what's kind of the implications around fees?

that's a really good question. I mean, so basically, so one thing we know, like, is that like, like when, whenever you get a corporate, corporate, corporate channel, you can actually kind of like actually negotiate fees ahead of time because you, you can then kind of like say, "I want five cents, I want ten cents, maybe we'll be in the middle of seven." But like one thing, one thing before end of channel, like, you Like three months ago or else, but right now they're very, very high, right? So right now, like, you know, if someone like launched something like this, you'd make, you wouldn't make actually bump up, bump, bump, bump up fees at all, maybe that means you would actually get into the chain. So when the anchor app was, it basically like adds like two new outputs on like every single, on the channel itself, basically like one for me and one for you basically, and it's kind of like a very low amount, basically like slightly bump up fees. And like progressively, I can basically use the fee to bump it up over time. Just because this, this allows you to, for example, like, you know, increase the fee rate on your, your HLCs and then also command transactions at all, and just make things a lot safer because now you can actually, like, you know, bump your fee, and also let you actually pay like lower fees over all, overall as well too. So actually, the ability to like actually control the fees whenever you need to do kind of like what is like a unilateral like forced transaction, which

before wasn But the benefit, the plus side is that we can then change the fees. And so I guess the question then is like, do you think on net it'll be better for the network because people can change their fees more easily, versus it, compared to the current model?

Yeah, yeah. I think, I think it's way better. I mean, like, I, so I guess one thing, like, it doesn't actually cost more for you to open a channel, it costs more for me to, to you maybe to close the channel in a certain case. But I think the one thing is that

You got that wrong. You basically have like a very, like, you know, bad outcomes, right? But in this case, I can actually, like, maybe, like, have the fee amount, fee on my closing back and just be, like, you know, what's right above, like, the relay fee because I think it's, I think it's gonna get demurrer, then I bump it over time. So I think, so I think overall people will pay less chain fees if they actually have like more control, you know, what's going on. I think also, this is

To actually, you know, get, have better chance of getting into the chain, like you, you maybe, like, almost kind of the thing was like, okay, well, you know, how much am I gonna pay to, like, you know, settle like a ten BTC, twelve C, whatever else going on, right? So it's kind of like a different, like, cost there, but I think it's definitely like a big step forward. I mean, one of the things that we, we kind of discovered, like, like during the process, they actually got like serious,

Obviously, the difficult thing is that we don't really know what miners are running, but if all miners just ran full RBF, everything would be on, would be a lot easier for us. We could just like kind of, you know, assume that if we have a higher fee, things get in, and it's just kind of a little bit like, just, easier, like, you know, parameter-wise.

Also, I saw some interesting discussion around upgrading channels. So the idea is that we could upgrade the chan- so I guess naively, the, the kind of We do this without actually closing and reopening, and actually the two channel partners can agree to upgrade to a new style of channel.

Exactly, yeah, yeah, because I mean, remember like, you know, we talked about, like, because like, I guess one thing is like, people don't realize, like, Lightning's kind of like a multi-channel type a few times in the past, we recently kind of like had the OG type, and we had a setup to start from a key, and we're basically at, about to like add anchor outputs, right? But things like, yeah, many channels We'll talk about the chain itself as well too. Basically, this is kind of like the realization that, okay, well, you know, the only thing that's actually said in the channel is basically the funding output, because like it doesn't actually do the multisig itself, we can't change that output. Anything else beyond that, it's kind of like a little bit more, you know, manually, but actually fully dynamic. So basically, this process of, you know, effectively like having like a zero downtime upgrade for your channels and also doing it in a way that's actually decentralized the entire network,

A bunch of good people, kind of like incorporating that into what I'm working on right now is a protocol drop. Basically, once this, once this happens, you may to then actually, you know, upgrade your channel on the fly. And so I think one thing, like, the number one thing I think is basically like safety. Now people can always upgrade to basically save a channel type and else do it in a way that doesn't have to, doesn't, doesn't actually incur on transaction. But I think the other cool thing that also kind of like, you know, allows for a lot

I see. and so hypothetically, then let's say we, we wanted to upgrade to p t l c, for example. So, maybe, maybe the hypothetical is like maybe in, you know, one or two years' time, we get Taproot, and then the Lightning Network, we wanna upgrade to p t l c. So would this be a mechanism for people to be able to change to p t l c channels without closing and reopening, or how would that work?

yes and no. I mean, so- So I mentioned like initially, basically it seems like kind of like the funding output itself is like you said, it's like, you know, we're doing something that people seem to kind of like need, they just don't, excuse me, like they just don't, don't like taproot features. So maybe initially we kind of like actually modify the actual funding output itself, but I guess it also depends, like, so for example, like, you know, so one of the things with taproots, they kind of like really like, you know, really like, there's kind So I think one thing it basically allow people to maybe even like have different channel types in the future. Like it's like, so assuming people already have like a, you know, kind of like a, a schnorr or, or whatever output in the chain or whatever else, they can, they can upgrade to all, all, all the, all the other things whenever they want to. The other cool thing is that, like, you know, it doesn't require the entire network doing it at any given time, right? So let, let's say maybe people have like PTLC,

maybe So obviously the nodes would have to talk to each other and as you said, they would do that feature negotiation to understand, oh, okay, I can speak PTLC, so you can speak PTLC to me. but would that mean, like, how would that- Work with routing, if y-y- does that mean your entire route must be all HTLC or all PTLC, or how would that work?

Yeah, yeah. So, yeah, so that's like how it would be, like, so it's one of the things where, like, you know, it's nice because, like, we don't have to like have like these like very, very like hard, like, you know, upgrades, like, for example, like, there's no like hard fork price kind of like a more like a gradual thing people upgrade over time. So You know, optional or required feature bit. So initially maybe it'll be an option, but maybe over time people say, okay, well, I only accept p t l c type payments because I, I'll do the other things. There's also kind of like a number of really cool, you know, things you can do based off of that, like different types of, you know, things like payment or whatever else. I think it's kind of a natural intensity going over. I think the main thing is like, you know, people can basically upgrade on, on their own, you know, leisure basically Could basically support both, maybe initially, maybe like one year at a time. I was like, "Okay, well, I'm only gonna do, I'm only gonna do P-- gonna do PCLC. I wanna see if all the writing orders, like the major writers, never have to go to PCLC only. That kind of reinforces the, that kind of, you know, adds momentum to the existing client, makes the client out also update some help me to, to actually, be able to wrap payments."

Cool. one other point I thought was interesting to just talk about is, when we talk about how many payments can go through Lightning, right? So, you know, it's people, people make the argument that, look, we can basically, we can move like millions of transactions per second on theoretically on a, on a Lightning network, and where, where if we compare with like, you know, credit cards, Visa, Mastercard, or whatever, people might say, whatever, fifty thousand transactions per second or something like that. But I think that's also ignoring a very important Right? Because with Lightning Network, and I think I've seen, other people make similar points, someone like Jack Malas or Rusty as well, that it's vastly different when we're routing on the Lightning Network because there's no fraud risk, there's no chargeback risk as well, there's no waiting for block confirmations.

Exactly.

Yeah. Yeah. So it's an important qualitative difference, I think, that, maybe people are missing when they do that comparison of, of Visa versus Lightning Network, let's say.

Exactly. Exactly. Yeah. I think the other big thing Maybe have BTC or whatever else, and basically start instantly making payments, you know, anywhere around the world, versus like, you know, being a minute basically like what's like actually like a gated, you know, system that can like censor other individuals and then maybe even like, you know, have like different choke points going on as well too. So it's one of the things like, it's also kind of like hard to pin it down because obviously like, you know, BTC is like one agency that can, like, you know, accurately, like, you know, see what's going on there, Seeing ourselves kind of like, you know, loop traffic go up as well too. You, you, you can basically view kind of like loop traffic as kind of like, like a, like a lower bound, an actual number, number in the traffic within the network itself. If someone like, maybe like, you know, does like a 10 BTC through the loop service itself or something like that, that, that means like, you know, they at least like had, you know, 10 BTC of activity going on elsewhere in the network itself. And so it's kind of the same thing where So I think it's one of the things where I feel like the actual capacity maybe like, you know, is like something like a, a lagging or indicator itself. You get like, you know, maybe that queue's going up a lot more, maybe then we're actually pumping, but, but, but we're gonna get the capacity, come like, you know, utilization, and then we actually attract more capital from there on. So

Yeah, really interesting comment. And so when it comes to things like dealing with censorship and obviously, Bitcoin, people we believe in the idea of permissionless money

With the current model, or just with Lightning Network, is, is that, the fact that people have a public key, or, you know, your node has to be kind of like the, the node, I guess, the node ID is public, is that potentially a vector that people might try to censor and say, "Oh, you know," let's say you're running a Lightning service, and let's say, you know, someone could come and say, "Oh," or like a government or a regulator might say, "Oh, I think this, this public, this node and that, you know, we want you to censor any payments that go to that particular node, as an example. Do you, do you, do you see that as like a, like a possibility or do you think that's really not, a very realistic, scenario?

I mean, I guess it's definitely a possibility, like, you know, regulators who we've seen will go to like, you know, pretty, you know, kind of like drastic measures. I guess one thing with that is, like, you know, obviously the grant of you would actually maybe like typically have the IP address attack, but you can also yet use Tor if you really wanted to as far as routing, so it's not necessarily like, you know, actually revealing your, your, your location. The other thing, like, the actual public keys themselves are actually ephemeral in Inter-interception thing, right? So let's say like, you know, I'm actually like node C, but I wanna basically have it be, be, being routed through node B itself. I can like route it through node B, catch the UCLC, tell, hey, node C, continue to forward in this direction basically, and kind of like, you know, maybe like, let me send some of that kind of like liquidity to the other nodes there too. So I think because like, you know, nodes can always kind of like, you know, have different identities, they can also kind We're always, I think no matter where I'm at, I think there'll always be kind of like the ability of the network to actually kind of like have, you know, direct, like, more easy access basically, even if maybe like certain chains or whatever else actually have like slightly more, more like, like layer of it, because you can't really control the network neutrality. There's no kind of like, you know, like, direct like mining or whatever else basically like centralization. I think people will always be able to kind of

like, you know, find their Or they doing what they do, what are we doing? 'Cause I think people definitely want to have it, such that be, things are a little bit private. I think obviously the big thing in general is like, okay, well, you know, like every single live transaction that happened in the past like three years was not recorded on the chain, and, you know, probably even hard, like we, we probably saved like gigabytes, you know, as far as like what's actually not gone on the chain. So I think that's like a big, big thing itself, because

like now Also, I guess just turning to the ecosystem, obviously you're, you're a deep technical expert working at the protocol level, but I'm sure you have a view or you have thoughts around what you would like to see in Lightning Network in terms of the application space. Is there anything that you would like to see there?

super good question. yeah, like I, I definitely want to see kind of like people doing a lot more kind of like conversations like mobile, as far as like one thing we did, we did initially, or it's like now, it's like, you know, one thing LND has capabilities, the base has capabilities to have like, do kind of like what's like a, like a custom MacRune, so basically MacRune is basically the ability to kind of like, you know, get someone like an authentication token, that token like allows maybe like, you know A thing, a game. Basically, a game like, you know, request access to my LED node, my coin. Basically, hey, I wanna send, you know, up to like one BTC, like, you know, maybe like, you know, two times, even if it rolls, I can basically give you like a very, like, private credential that basically gives you the ability to do that. So it's just something really cool on that, I think kind of like a more general overview and usage on that. And another cool thing I want to be, probably should mention out is something we have, you know, called Elsatic, basically ability to kind of like have, you know, combine basically HDB four or two and like, like, like together, like have a very cool kind of like authentication system, going on there. Basically, you can, you can ask like, maybe I can like, kind of like, act like rent, like, you know, this space over time A static kind of address, but basically we'll be able to like, kind of like, allow people to do like, maybe subscription payments, we can kind of, you know, put, push payments directly into the network itself. It's gonna come with something we have called AM, which I think is a really cool thing, 'cause then at that point, people can effectively have, like, what's more similar to kind of like what they think a Bitcoin address is, but actually all, we'll also be able to reuse for other individuals, and then you know, also implement things like maybe like a subscription Like maybe, maybe something a little more specific to their particular use case.

Yeah, I find the really, the really interesting, the one you mentioned around having some way that people who are used to the old paradigm of just saying, "Oh, here's my Bitcoin address, donate to me," but have that in Lightning, because I guess one of the things that, people who aren't as technical or aren't as, you know, and in fairness, it's still early days, but some of the, I guess, obviously, there's huge upsides to Lightning, Yeah, your node has to be online or persistent, and there's kind of, there's sort of these different requirements around how it works, and so people have to think in a different way, but if some of those things can be mitigated a little bit so that you can just put up a, some kind of an address, to receive money to, kind of like that HGLC interception, thing or the hotline voice ID.

Exactly. Yeah, yeah, yeah. Like, like, like, here's an example. So let's say I have like some subscription service One basically sort of like maybe like my ID, my account ID to them, sort of the station basically kind of like what the registered payment is itself. Then my phone basically waits about on demand, kind of like to basically push the payment over to the other individual. The individual basically catches the payment, checks my ID, so like, okay, well, this is like Roasbeef's actual identity, this is, his payment revolves, and credits me on the server itself. And that can be done kind of like in like one simple server basically, I scan something on my phone, click Like machines actually accept Lightning payments themselves. So what they can do, they can like recognize, okay, well, this is a payment hash that I created, that only I could create themselves. And then basically the user presented, the pre-image, the pre-image can only be, be available to them if they actually paid actual payment. So therefore, I'll give them this candy bar. Alright, so they have a bunch of cool things that is there as well. People like, you know, kind of have these really cool, kind of like maker type, hobby type thing that they

can also

It's like one chart, one, it's kind of like reusable QR codes everything.

Excellent. And so I guess just looking at the ecosystem, are there any examples where you think someone's doing really great user experience?

good question. Yeah, I mean, so I really like the Strike stuff. I like, I like how they kind of like, you know, have this like hybrid between the credit card and actual ID itself, where people maybe don't really even really know like what's going on in the background as far as what they have going in there. So this is people like doing some cool I, I'm really, really kind of like, kind of like dual anything like, kind of like involved, like actually like, you know, doing like, you know, more like native like web type, authentication payments, as far as what's going on there. well, I think Phoenix also do, do some cool things as far as, you know, really like pushing, pushing, pushing the envelope, sort of like UX. So basically have a system like similar to, something's redo, where they can, they can actually like have a user, you know, And I think we'll see a lot more of that going on as we start to add some additional complicated protocol which makes it support these more advanced use cases.

Yeah, and so over time, it, you know, well, in the earlier days, it was, you had to be really way more technical and there was a lot more stumbling blocks around this, but over time, it seems like things are becoming a lot easier and it's becoming a lot more of a slick user experience as people are figuring out ways to, make it a nice experience for the lightning user. So, I guess, Of, you know, let's say another one year's time, do you think, there'll be way more people on Lightning or similar to what we have now?

Oh, yeah, really good question. Yeah, I think there'll be more because I think we were kind of like figuring out, like, so, 'cause like, oh, yeah, I think you're right, and definitely, like, you know, it was kind of like really moving like a new paradigm. People are like, "Okay, well, you know, I have two balances, how do I communicate my A whole bunch of other tools basically allow people to kind of like, you know, acquire channels a lot more easily, so basically do a lot more, rebalancing on their end, basically make things a lot more stable. So I think we'll see people, people actually operating, on, on the endpoints of it itself. I think we'll see kind of like, you know, a lot more usage of like lightning in the web, as far as like, you know, people maybe like having like a Western type thing, maybe even like, you know, using different commissions, Not just, you know, like, not just like directly tipping me, but you kinda like, you know, have like a lot more, you know, like have the, the payments actually be, be like more core, core part of the game. So, yeah, I mean, I think the one thing is like, I've always been continually impressed by all the developers doing. So I'm still, I can like think of like some of them, but I kinda like, you know, giving them the tools and they just surprise, you know, as far as like all these things they're making. There's so many people doing really cool stuff.

Excellent. I think, yeah, we're seeing a lot of growth in terms of, the ecosystem and people being able to e-easily use Lightning. so, Lalu, I've really enjoyed chatting with you. if listeners would like to find you or Lightning Labs online, where can they find you?

Oh, yeah, you can find me on Twitter, twitter dot com slash roasbeef. That's, roast beef with no T. I'm also roast beef with just about everything else, been that way since like you know, linking from there too and check out our repos, and, yeah. And also Twitter, Twitter dot com slash lightning is also us, that's Lightning Labs on, Twitter. We, we have the snag, you know, the, the, the cute little handle, so that was nice. Well

done with that.

Yeah. Thank Elizabeth for that. She, she pulled off, you know, she pulled off some hookups.

Nice. Yeah. Well, look, I really enjoyed chatting with you. It was really, really educational for