On The Brink with Castle Island - Casey Rodarmor on Ordinals, Inscriptions, and Digital Property Rights (EP.399)

Episode Date: February 22, 2023

Bitcoin developer Casey Rodarmor joins us to talk ordinals, inscriptions, and digital collectibles on Bitcoin. In this episode: What is an ordinal? Why the name? The history of ordinals and inscriptions How does ordinal theory work exactly? How to participate in ordinals and do you need Ord The interaction between inscriptions and Bitcoin fees Should inscriptions be compressed? How should this happen? Why did Casey not use Counterparty? How Bitcoin is the most premium blockspace for NFTs Inscribed content property rights on Bitcoin versus Ethereum Will Bitcoin state just be pruned? Who will store inscription data? How bitcoin data availability might be improving over time How ordinals affect Bitcoin's security budget Why NFT traders are looking at Bitcoin for the first time Where the inscription 400kb limit comes from 'Ripping a 4 megger' Is Casey concerned about out of band transactions Do ordinals actually increase the possible efficiency of operating a full node? The primary bottlenecks in running Bitcoin Core and how inscriptions affect that Will ordinals catalyze more pruning in Bitcoin Core? Might ordinals make rollups more likely on Bitcoin? Do inscriptions dilute Bitcoin's primary purpose as a monetary system? Transporting other NFTs to Bitcoin with teleburns What Casey is most excited about with ordinals Further reading:  Ordinals.com Rodarmor.com Casey's twitter

Hello and welcome back to On the Brink. I'm Nick Carter. This is a highly anticipated episode with none other than Casey Rodermore, the inventor of ordinals. I'm sure you've all heard about ordinals at this point. It is a new standard to track individual Satoshi's, which of course can be used to inscribe arbitrary quantities of data into Bitcoin. Today we sit down to discuss the origin and purpose of ordinals and inscriptions, and we cover some essential questions, like will they bloat Bitcoin? What will the effect of inscriptions be on Bitcoin fees

on the security budget? What kind of a block space is Bitcoin? Could it be the most premium block space for NFTs? What are the property rights that inscribers get on Bitcoin versus other blockchains? Will the block space be pruned, or do we have strong available? availability guarantees over time. Who will store the data? Where the inscription 400 kilobite limit comes from?

All these questions and more in this episode with Casey. Let's dive right in. All right. Welcome back to On the Brink. This is a very special episode, very hot topic for you all. I'm here with Casey Rotamore. I actually didn't ask you what your affiliation is in our pre-recording conversation. So why don't you tell us actually? So yeah, I am actually still an independent Bitcoin developer, I'm fully a free agent. I sort of have some employees now that we pay all of them $32 an hour for historical reasons, but only like three of them that I have. So rapidly, you know, I'm a job creator now. So I guess job creator is now my title. I mean, I don't expect you'll be unaffiliated for long given what you have done, which is to kick the hornets nest, perhaps, or, you know,

one of the more important technical developments on Bitcoin in some time. Of course, ordinals. We're talking about ordinals. We only have 15 minutes. So let's dive in. What is an ordinal? Also, where did that name come from? Yeah. So there are two ways that we use numbers, ordinal numbers and cardinal numbers. And a cardinal number is used for amounts. Like I have 10 dogs or I have 35 cheeseburgers. And that's a cardinal number and an ordinal number is when you use it for a position in a list or some other set that has an order so that is the fifth dog or the 27th cheeseburger those are ordinal numbers and the reason that it uses the word ordinal is because uh everything in this protocol which is sort of this in in ordinal theory uh if you will uh relies on order it relies on the order in which blocks

for mind. It relies on the order of inputs and outputs of transactions. And so, ordinals seemed like an appropriate choice. And I really want to apologize to all the math cells out here because I believe we have ruined another math word. The first to fall was crypto. And now Ordinals is really sunk. I mean, they're going to have to really write a lot of papers to write that ship and make it mean the math thing again. I mean, the SEO battle is they're losing it. Yeah. I mean, I'm, ordinals.com is number two when you search for Google on Google with like the bare word out of control. Like our SEO is good. So this takes me back to my philosophy days

in college where we did a lot of number theory for in philosophy class. I don't know why. And it reminds me of especially talking about cardinalities, talking about the different size of infinity. Have you encountered any of this? Oh, sure. Yeah, the alif zero, alif nol. Yeah. What is it? Hilbert's Hotel diagonalization.

Anyway, that was a fun and really useless sudden. Maybe there's some use to being a math cell after all. Yeah. Maybe we found it. So this kind of was, it seemed like it was really thrust upon the Bitcoin community. But I do remember seeing you experimenting with it for a long time and then sort of quietly and then it blew up. Was it the fact that people figured out how to inscribe large quantities of data into Bitcoin that really catalyze the interest here? Well, so I created with help both ordinals and inscriptions.

So they're sort of both my design and creation. And to separate it out, I think it's good to do the little tech 101 bit of the show. So an ordinal is this trackable SAT, is this lens that you look through the Bitcoin blockchain with. It's kind of why I call it the ordinal theory. And if you believe in ordinal theory, then you see these individual trackable sats. You put on your ordinal theory monocle. And the sats, they pop out in high definition like Pokemon in the tall grass. but then you take your lens off and they're not there anymore.

So it is a sort of a convenient hallucination that I invite everybody to participate in because it doesn't change the base layer at all, base layer of fungibility, not affected. But if you want to do it, then you get these trackable stats and you can see where they start and in the block of their mind and how they transfer across transactions. And I actually, even though inscriptions came after ordnals, the inscriptions were actually the point, the whole time. I wanted to be able to create digital art objects on Bitcoin. And because of Bitcoin's UTXO model, it doesn't have the Ethereum account model, you need somewhere to put these digital art objects. And so first worked on ordnals. And then when Ordinals was finalized, then I created

inscriptions, which are a way of attaching a piece of data. It's a publishing tool. It's a way of publishing a document in the Bitcoin blockchain. And then, but upon publication, you assign that document to one of the ordnals. And then that ordinal, that specific sat, you know, becomes like a shiny Pokemon and not just a normal Pokemon. And then you can track it across transactions. And then, you know, by social convention, maybe it's a picture or some other art object or a piece of text or really anything. And then by social convention, the owner of that SAT is the owner of that content, of that inscription. And that's why they're called inscriptions, because in some way, you were inscribing

sats with this indelible additional piece of content. So is it fair to describe ordnals as an opt-in accounting overlay on top of Bitcoin? I don't know if I would use the word accounting, but I would say it's very, very far. similar to those other opt-in overlays, like first in, first out, when you are valuing bundles of shares or last in first out, or when you pick a specific share, you know, in reality, all of your, what's like a good normie stock? I don't know, Invidia. Okay.

So, yeah, in reality, you have 100 shares of Nvidia in your Ameritrade account, but they're all the same in reality. Yeah. But then you bought them at different times. And so is a sort of like accounting hallucination, you say, ah, this one was the first one. And that doesn't make them actually different. And you're only using it for this purpose, this accounting purpose. So it's like a, it is like an accounting system, like one of those for Bitcoin.

But instead of the purpose of it being to do your taxes, the purpose is to trade digital art objects and collect them. Because when I think about how Bitcoin works, like quantities of Satoshi's, there is no like literal, actual protocol mediated way to track them from one transaction to the next, right? So this isn't a sort of an additional tracking scheme that you're layering on top of that. Yep. And it's so, you know, normally we have UTXOs and they have like, you know, a thousand sats or 2000 sats. but there is no like which sat or sat number one or sat number two they're in this like undifferentiated sotoshi goo superposition and so the way you do it is you order all the satoshes yeah within the block and then you order all the blocks well we have to order the blocks already yep and so now you can define specific satoshes and follow them as they process through

the blockchain yeah and then at once you're looking at at the blockchain through the lens of ordinal theory, each input is essentially a stack of Satoshi's, like little coins. And instead of destroying them all, instead of destroying UTXOs and creating new ones, conceptually, you're sliding these coins across from the inputs and outputs. And that's exactly how the transfer algorithm works. It's essentially just this coin sliding. So the first input, you have, you have, you have, you have, you have, you have, you have, you have, you have, you have sets and you have slots. And you have a stack of slots, and then you have a stack of empty slots. And then you just go like this.

And that's how you move the sat, the sats into the new slots. So I'm going to reveal my ignorance here. How does one run ordinals? What is the ordinal software? Is there ordinal software? The answer at this point. is with great difficulty. So it's pretty rough.

There are DGens in the Ordecor, the Ordinals Discord. They are dying. They are fomowing so hard. And the reason it's hard, because they want to get running, they want to inscribe, the reason it's hard is because you have to run a full node, like a straight up non-pruned Bitcoin Core full node with TX index set to be true. And then that acts as your sort of your trusted partner to fetch you data from the Bitcoin network. And that you connect to it with Ord, which is the software that I and now more kind of we wrote,

which fetches blocks from it. And so as a result, it's a bit command line heavy at the moment. And this has actually been a strategic choice on my part. I am writing a light client or some other way of doing this without having a full node is a very, very, very distant priority. I would very much like it if there was this reference implementation that was the best implementation with the best developers and the best features. Maybe not like Bitcoin Core, maybe not the fastest moving implementation, but like really the best implementation. So it's been strategic that you have to have a full node to use. because I want people to run full notes. I want people to understand that if it's not your node,

it's not your JPEGs. And also weird emerging property of all that is that it's the early wealth, so to speak, that it's accruing in the system goes to people that have a knowledge of running Bitcoin full nodes, which is a funny property. Is it possible to engage in the ordinal ecosystem without having Ord. Could you just DIY it with a Bitcoin full node? Yeah, you could. And in fact, I've made ordinal theory transactions, that is transactions that respect Ordinal theory to transfer Sats in specific ways, using raw Bitcoin core commands. So Ordinals is like a fractal of Bitcoin. When you zoom out, it's Bitcoin. And then you zoom into the little details and they're also Bitcoin. And what I mean by that is, you know,

that ordnals are very orthogonal to the rest of the functionality of Bitcoin, which as a result means that it interoperates extremely well with the full suite of Bitcoin tools. So, for example, you could take a, I guess I'm shilling NVK now, you could take a SATs card, which is like this Bitcoin card, and generate an address, get the address from this thing, and then send a rare and exotic sat, indeed an inscribed sat to this card. So, okay, but this is another wrinkle we have to talk about, is the sets within Bitcoin, within ordinal Bitcoin are not fungible with each other. Some of them are special.

We have a new convention, which is early ones are good. And are there other categories of interesting sets like round number blocks? Yeah. The first sat in a round number block. I actually have a taxonomy. of sats as such. Honestly, when I first became aware of ordnals, I think by reading your tweets, I thought that was the purpose, was not to inscribe, you know, arbitrary data on Bitcoin, but was

like an interesting, funny thing whereby we could identify interesting sats, or in particular early sats. I think that a lot of people had the same misapprehension, because I, when I started noodling on this in early 2022, I very explicitly wanted to create digital art objects on Bitcoin. And I knew that I needed to figure out how to do that within Bitcoin's UTX-O set model. And so I started noodling until I came up with ordinals in its kind of final form a month or so later, at which point I started being like, I started noodling on how to do the inscriptions. and the inscriptions part is really just like, well, where do I put the data?

And I figured out how to put the data. And then inscriptions is what everybody wants. And so they see this order. They're like Casey's talking about rare. And I've been like blathering about this to every single person I've met the last like 10 months. Like absolute maniac. And so people see Casey deranged talking about these like rare sats. And they think that's the whole.

that's like they think that's what I was trying to do. And that inscriptions are like some fun additional thing that I discovered later. But no, actually I had inscriptions, what became inscriptions in mind for the whole time, essentially. So the telos of ordinals is inscriptions. We can canonically say that. The telos is like the true nature of a thing? The purpose. I was being fancy.

No, no, no, no, no. No, no. for no reason. No, no, your volubility. Your volubility is one thing that makes you one of my favorite podcasters. So you should just like crank that up in this interview.

In sort of like Aristotlian philosophy, like the telos of a chair is to be sat on, basically. So the telos of Ordinals is to inscribe, I guess, JPEGs into Bitcoin or other content. I hope they do, well, JPEGs are good. Right now they should try WebP instead of JPEGs because it's small. That's actually, that's going to be one of my questions. You're running ahead of us.

Yeah. It's like, should we compress? I mean, I've been laughing at the fact that these images are being uploaded to Bitcoin uncompressed. That is astonishing. I think that it was just an artifact of where we are in terms of block space and fees. That can't possibly last, right? Yeah, well, so a lot of it is the artifact. to just wear ord the tool, which you can find at GitHub.com slash Casey slash ORD.

That tool lets you upload a file, and then it also implements ordnals.com. So you have this one binary ordnals that can both be, it's a wallet, you can inscribe, it has some other tools, and then if you run it locally, you'll get the full copy of ordnals.com. Like, ordnals.com is not special. You could have like ordinals. Nicholas Cartier.com and everybody would see your ordinals index, which I think is like an important point that sometimes people miss, that there's no like weird non-decentralized part.

But yeah, so it's really just that we haven't done it yet. Every format needs very, very careful consideration and support before we add it to be displayed on org so that we don't have to break inscriptions later, or that some inscription content types work in some places, but not in others. So really, it's just like slowly figuring out what the best formats are, guiding users to those formats, improving the tools so that they, for example, strip out access data before uploading.

There's been a lot of ordinal's improvers who have heard about ordnals and are kind of here to fix it. They're very, I love the ordnals improvers, but I hope to channel their energy, which can be frustrating sometimes because they come up and they're like, hey, why don't we do like extension blocks for Bitcoin? And I'm like, what the fuck? Are we having a conversation about drive chain now? Like let's just like do the low hanging fruit and compress the HTML before we put it in the blockchain. In a sense, the market will also solve for that, not to, you know, believe in the church of the free market or whatever. But bro, I'm, I'm in the pews.

I'm in the pews with you in that church. Love it. I mean, inevitably. if people upload a ton of uncompressed data, then fees will skyrocket and then the motive will be to compress it as happened on Ethereum, for instance. Yep. Where there's a lot of creative ways to have NFTs that are sort of on-chain native, but are small in a variety of ways. I was kind of pleased and surprised to learn that it is much cheaper in terms of fees to create an inscription, a fully on-chain inscription than it is to create an ERC-721 contract with on-chain data, much, much more expensive. I think at least five times, maybe 10 times.

So why did you not use something like counterparty? I mean, obviously NFTs have existed. Was it because you wanted to be able to insert much larger quantities of data? No, no, no, no, no. That was, I actually don't remember exactly how Counterparty worked. I thought it might have used Op Return or something. Yeah. There are issues that I have with Counterparty, and I don't want to like start beef

because I'm realizing that this is a really beefy space that I now found myself into, so I want to be nice to the Counterparty people. But Counterparty has a token called XCP, and even though the data is carried by Bitcoin, that puts it on the other side of this line that I don't, like to cross personally. The token line? The token line. Yeah. Yeah. The Bitcoin token line. You know what I mean? Okay. That's a good answer. Also, I mean, it's just a, I mean, the data itself for like the

rare pepays or whatever, that's not actually on Bitcoin, right? That's like somewhere else. Yeah. And I think that, um, doing that having data off chain is really fraught. Um, I think it's easier to describe if I just, if I describe a workflow that or like a thing that a user would do. So a user goes to OpenC and they see a picture that they like and then they buy it. And the seller sells it to them for $100 and they now are in possession of it. And the seller, let's say the seller, this was the only picture that they made. And the ERC 721 token that this user has now bought now points to a

a computer on AWS. And they are now dependent on like the goodwill or benevolence of the person who sold it to them to continue running that computer. And this, when in reality, the primary, the person with the primary incentive, possibly only incentive to keep that data reachable is them, the new owner of the token. And so by not informing users of this,

I think you're doing them a grave disservice, that they have bought something that they want to hold and enjoy for a long time, but it could suddenly break for reasons that they don't understand. And if they are not very, very technical for reasons that might be fully out of their control. I am troubled by similar things. I think about NFTs as like having varying degrees of assurance, basically, or like property rights. another way to think about it. And an NFT that refers to some external website is just you have no assurance whatsoever. Yeah.

It's actually even worse for ERC 721 tokens because your property rights in relation to where the content is are these sort of subsidiary property rights to your ownership of the token in the ERC 721 contract. which if it's mutable or it has an upgrade key or it has a bug yeah uh just takes gets taken can get taken away from you so it's like you like let's say you're in the fully happy place in the best let's say your er c sub 21 is stored in the safest way and it's the content and it's all on chain so you have your house and you own it that's what that's like that's your property right but then um it is there is a secure

vulnerability, right? And then that's just akin to like your H.O.A. Like a like a like a wildfire, right? Or like your HOA like trying to kick you out of the house of your house and succeeding. And so these, you have these like property rights and subsidiary property rights and subsidiary property rights. And the the subsidiary rights are worthless without these parent property rights. And so it's like, where's the content in your ERC 721 token is like the low level of property rights? And then is the contract secure? It's like the parent property right. And then you're subject to uncertainty

via the Ethereum development process. And yeah, and all these other technical and social concerns with Ethereum. So even if your data is all on chain on Ethereum, you're not safe. That doesn't really change anything unless you take care of all these other things. So that's actually one of the main things

wanted to address today is I believe there's an item in the Ethereum roadmap called the purge, which I think refers to purging state. I call it protocol level amnesia. That's good. So I mean, honestly, there might be some merit to that, you know, going stateless or whatever. So this is my main question, though. For Bitcoin, can users have assurances that there will be some class of sort of archival node operators in the through the indefinite future that are you know someone is storing this data basically is that you know how do you think about that basically ultimately users if you want something to be stored you must take personal responsibility for storing it you know they could the bitcoin network could cease to run or we could see

some sort of like pruning of like very very very old witnesses which is where the inscriptions are but um the assurances here i think are very good and they're actually getting better that that this old data will make will continue to be about available and one of the reasons that it's getting better is that uh i am told that uh ord and ordnals has led to the largest spike of bitcoin full node runners in a long time. And so these people who are interested in running Ord to trade inscriptions and own them themselves are probably pretty motivated to run an archival node,

both to store their own data and to make sure that ecosystem data, to protect the value of the ecosystem, because even if your piece were to be lost because it's no longer stored somewhere, that would reduce the value of another piece. There'd be like thud, you know, among inscriptions. So I think actually those, so yeah, I think there will be a lot of people running archival

full nodes. A lot of them will be inscribers and inscriber appreciators, inscription appreciators. And so I think that those assurances are good and getting better. I guess the other way, I don't know if you subscribe to the school of thought, but to the extent ordinals drive up transaction fees and inscriptions which I think is already happening

then there's just more security budget to go around I know some people don't like to think about it like that I think that's a really perfect way to put it totally reasonable in purchasing power terms real terms relative to a basket of apples

and oranges or whatever Bitcoin is more secure and the miners are being paid more So that is also potentially, you know, accretive and it maybe is kind of a more nebulous way to Bitcoin's long-term sustainability. So there's that angle. So one of my hypotheses around why people, including NFT, people that have never touched Bitcoin and are now, you know, running full notes for the first time, one of my hypotheses for why they like ordinals is because Bitcoin is perceived as being kind of a very pristine block space as in really strong assurance. If you, you know, set your homestead down in Bitcoin somewhere, no one's going to take it from you. That data is going to be there forever, even relative to Ethereum, which is sort of the dominant NFT domicile. Is that your view as well? Is that your explanation for why like the NFT community

is sort of coming over to Bitcoin via ordnals? Yeah. I don't know if it's, I think there's a lot of different factors. But I think that that, like the idea of that Bitcoin is like the, like the American West after the Homesteading Act, maybe not. Maybe things were a little wild there. But that it's a place with very good property rights that you can homestead and be free from molestation by the government or by attackers from whatever reason. Yeah, I think that's like absolutely a huge part of it.

Another related thing is that NFTs are, I hate NFTs, digital art objects, digital collectibles are all about scarcity. And I think that people understand both that Bitcoin is scarce in block space and in currency. And that, and also the design of inscriptions makes inscriptions themselves scarce because you must, get it into a block and there's this limit to the size of each block. So people also, it's the safety, it's the security. And it's also the scarcity, the brutal scarcity of, you know, these things, of the currency, of the block space. And it's like a proof of work in a way.

It's just hard to mint an inscription, especially as fees increase. Yeah. Like a, what's the maximum size of an inscription under standard rules? Is it 400 kilobytes or something? That's right. Yeah. Where does that threshold come from? It's a one-line check in Bitcoin Core.

So in Bitcoin Core, it gets a transaction that it's thinking about a new transaction that it might want to accept into its mempool and relay. And then it does a series of checks on it. It goes like, okay, is this transaction well formed? Are the coins that it's trying to spend unspent, et cetera, et cetera, et cetera. And there are some consensus rules, which are like, things like, does the transaction not create coins out of thin air? But then, and every node has to have the same consensus rules, otherwise they fall out of consensus.

But then there's also policy rules. And policy rules are those rules which every individual node runner can set differently. Sort of like how you run your node, the particular details of how you run your node. You know what I mean? And so this is a policy. policy check, not a consensus check. And before we upgraded to all these fancy new signature types like Taproot and Segwit, I think that there was a problem where if you, a problem of a bug where if you,

I don't think I'd call it a bug, I'd say design flaw maybe in this old signature types, that if you had a very big transaction that had a lot of these old signature type inputs and you had to like hash it very many, many types, for all these different signatures, you would rapidly make this transaction very expensive to process. And this difficulty in processing scales with this old style transaction, scales with the size of the transaction. So at 400 kilobytes, it's manageable, but at a meg, it's not great. At four megs, it's not great, like the full size of a post-seguit block.

So it's just this check that limits you to 400,000 bytes is just this one line check in this part of Bitcoin core that checks the code. I don't know where it is in the code base. If you're a real Bitcoin core gangster, you refer to them by like, oh, it's in policy. cp or whatever, but I don't know where that is. So and then, yeah, it's this check and it says, oh, it's above this. I'm, you know, if it gets mine, cool, but I'm not going to accept it into my mempool. So, okay, that's, I was very curious as to where that threshold came from. But of course, if you know a miner, and this has been the source of controversy as well,

you can email them the transaction data and they can mine up to four megabyte transaction if they want, which we did see the other. We call that in the, in the ordnals community, ripping a four megar. have there how many of these have there been i just knew about the one with the taproot wizard i'm i'm pretty out of touch on the subject i think i hooked like my only knowledge of ripping four megars is this this group chat where i encouraged an artist dgen friend of mine and a technical dgen of mine to start friend of mine to start exploring ripping four megars and the title of the gc is called four meg rippers and they're going

insane in there and sometimes I check in and I'm like Jesus Christ I don't even want to know. This is my level of visibility onto everything like going into a GC and being like, all right, they're whaling out in there and then checking ordnals.com and seeing the latest inscriptions and being like, all right, they're whiling out out there. And then being on Twitter and like people are like panicking because somebody burned a half milly monkey. Like I'm, I don't know what's going on. Like I'm just trying to stay focused on like adding features and fixing bugs. So many things are just like straight up news to me. Yeah.

So are you concerned about out of band payments now due to this potentially attractive four megabyte transaction type? Are you worried that miners are now being compensated in some other way? Because there was a lot of concern about this with your unlocks are first at it. Yeah. Yeah. Um, yeah. Yeah.

So, uh, not really. I think that it is undesirable, but I think that it is a very, very short-term problem that fixes itself because inscriptions don't use any old signature types. And because the soft 400,000 kilobite limit was there due to certain old signature types, a four megabyte, a four megar, as we call them, that does not have any of these old signature types and the inputs. is actually safe.

And so we could loosen the check in Bitcoin core to say, okay, well, if it has all the new signature types, then four megars are all good. Or more clearly, you can create a transaction up to the block size limit. Or if it has some old signature types, then we say, okay, the old 400,000 limit applies. And if miners start getting compensated out of band and maybe some miners get better at getting business this way, like sort of getting people to come to them to rip form egers, and they're getting disproportionately, they're disproportionately profitable unless their share of hash rate is increasing, then it's a one-line PR to be like, hey, let's loosen this for inscriptions. And that problem just goes away. Everybody can rip a form-egger direct. into the mempool and just pay a fee and the miner gets to select if they want that to include that to the detriment of all the other transactions that they might include in that block. Are you familiar with Eric Wall's thread on how ordinal slash inscriptions increase the efficiency of operating a full node?

Mm-hmm. So it's a little counterintuitive. I think I'll try and recreate it in my own words. So because witness data isn't computed by full node operators, and this increases the ratio of witness data. It's just kind of ignored, I'd say. This is generally ignored. Yeah.

This increases the quantity of that relative to transactional data, I guess, in some cases. This means that full note, disk I.O., which is the primary bottleneck in running a full node, that is actually lessened as inscription data makes its way. way into the chain? Is that like an accurate summary of that concept? Yeah. The first thing that I'll say is that everybody focuses on bandwidth. Well, actually, everybody focuses on download size. Like in literal bytes. Yeah. Like the actual file size. Yeah. The first thing they hear is that like, oh, your computer can't run Bitcoin core. You need a 500 gigabyte hard disk. And that's like the primary resource that they know about. But there's many other resources involved in running a Bitcoin

core node. There's there's ongoing bandwidth. There's bandwidth for blocks. There's bandwidth for all those transactions that you want to see, but that don't make it into a block. There's, what is it, I.O., like you said, which is like Disca. And also signature verification, which is very slow. So are those not the primary bottlenecks, rather I.O. and signature verification under most just like normal circumstances.

Discussing what is the primary bottleneck is kind of, Well, normal circumstances. Normal circumstances, I think a lot of the time it's actually down that speed. Okay. But the thing is that every time you have like eight bottlenecks and you're always getting throttled by the one that you're the closest to. And then when you fix that one bottleneck, all of a sudden you're bottlenecked by something else. So under different circumstances and like hardware and network connectivity, you'll be bottlenecked by different things. So you could have like a monster, monster connection to the internet, massive, fast SSDs, an incredibly slow CPU.

And your bottle neck is signature verification, which would be weird, but it happens. But yeah, no, I think I think download size and disk space, because you also have to, if you have a slow home connection, then those are sort of the marquee items, the things that like jump out to people. Is also the second order version of the analysis, do you agree with this? Basically, the idea that maybe ordinals will be a forcing function to catalyze more aggressive pruning methodologies in running Bitcoin Core. And as such, will not directly, but indirectly, increase the efficiency of running it because basically pruning will become more prevalent in like the node software. Yeah, I think that's extremely likely. There's a really easy first win here, which is you have a node that has assumed valid turned on. You have this assumed valid height or assume valid block and below that block.

You don't check the witnesses. You don't even, you download them, but you don't even look at them. And then if you're a pruned node, you don't need to serve witnesses to anybody after you download them. not an archival node. So if you're a pruned, non-archival node with a soon valid turned on, the current thing that Bitcoin Core does, which is a little weird, is it downloads witnesses, but then doesn't use them for anything. Which is a kind of a vestigial, unnecessary thing to do. I wouldn't say it's vestigial. It's like things only happen on open source projects when somebody has the time and inclination to like do them. And so like I think I saw Peter Willa comment,

Peter Willa comment when he realized like, oh, we should really just be, he was like, oh, yeah, we should do that. Has somebody done that? Like, that's kind of the open source way, you know? Anyway, so, yeah, there's an easy change to Bitcoin Core, which would turn a pruned node with assume valid turned on to if there are a lot of inscriptions and thus there's a lot of witness data. when they had a stretch of blocks that was below their assumed valid height that were very, very heavy with inscriptions, they wouldn't need to download witnesses at all because they're not going to look at them and they're not going to serve them. And so that might become a kind of appealing security tradeoff. You're doing this assumed valid thing, which is a whole other kettle of fish on what that is.

And if that's a valid security tradeoff, I think it is. opinions differ. But you, yeah, it becomes a appealing security tradeoff. And let's say we have like two years of crazy inscription volume, right? We have two years of 75% of all blocks are full of inscriptions. 75 blocks are witness data. 75% of blocks are witness data. Then once, you know, the Bitcoin core developers release a new release where all of those blocks are below the assume valid height, all that witness data is no longer downloaded. or checked by pruned, assumed valid nodes. So they get this whole, this like,

I forget if I said one year or two years of free sync or faster sync, where they just can rip through these transactions, ignore all the JPEGs, and they sync faster to the present day. It's very counterintuitive, but I buy the reasoning. Do you also think that this could be a forcing function to other scalability approaches? Like a lot of people are contemplating how roll-ups could be built on Bitcoin.

One of my weird theories is that this will actually be one of the things that induces more excitement around that and potentially a soft fork, which would be required, because Bitcoin looks like it'll be congested for the foreseeable now. Yeah. Yeah, it's congested now. I personally actually, honestly, we have believed that we have entered a new fee regime that will be the new normal going forth. Inscriptions are very popular. They really filled an unmet market need that many Bitcoin developers did not take seriously.

And they don't understand the demand for digital collectibles. And so, yeah, I think fees will continue to get high. And that will catalyze everybody who needs low fees. Like it's, it's, um, transactions compete for blockchain. based on the basis of economic density, like a large transaction that doesn't pay much, has like a low economic density and a small transaction that pays a lot has a very, very high economic density. And inscriptions, an inscription that somebody doesn't really care about is a big transaction with a low fee and it has low economic density. And so the way you can get more block space, the way you can get block space is simply by making

transactions that have a higher economic density. density. And so protocols that do this like lightning, which batches many, many, many off-chain transactions into an on-chain netting out have extremely high economic density. They represent, you know, you're at a sandwich shop. They take lightning. They open a channel on day one, and then two weeks later they close it. That channel open and close represents the hopes and dreams of like all of the sandwiches that they sold in the intervening two weeks, possibly many sandwiches. So like and then other things like roll-ups, every roll-up design I've seen has been like I'm not sure it's ready yet. But yeah, they they for sure are a very economically dense way of making Bitcoin transactions.

And also for example, other things that are kind of counterintuitive like Fedomint. which is the sort of Chaumian eBank on Bitcoin that interfaces with the world on behalf of its customers by Lightning. That lets many, many users use the Lightning Channel, which makes it even more in a non-fully custodial trust model. So that makes it those transactions, like a Chamon, a Fedomint instance, channel open and channel close could be incredibly economically dense. It could represent transacting on behalf of a community for six months. What do you make of the critique that inscriptions dilute Bitcoin's message? I mean, this is sort of, I've seen a lot of bad critiques of ordinals and inscriptions. Like, oh, you're, you know, somehow breaking the social contract or breaking the rules or you're, this is a DDoS or you're making blocks too big.

Like, you know, in a way we didn't anticipate. We can leave all the bad ones aside from now. Yeah. What do you make of the critique that's just, you know, quite simply, okay, Bitcoin has a singular purpose. And by making Bitcoin about something else, we are diluting that, diluting the message and making it harder for Bitcoin to compete against maybe other, you know, generalized platforms like Ethereum or whatever. My answer to that critique is that Bitcoin is not for any particular thing. It was created by people who had certain use cases in mind, but it has become a public good. And in that way, it has transcended the wishes and intentions of its creators.

And so if you think of Bitcoin as a thing that is for something and for another, I would, I suggest you reevaluate what you think Bitcoin is. It is not a, I don't know, a very special purpose tool, like a ball peen hammer or something, it's like a lathe that can build anything. And these people are just upset that we've been making like, we've been making round tungsten cubes on the lathe for a while, round tungsten objects on the lathe for a while. And now people like square tungsten cubes and the round tungsten mackses are like upset and confused because they thought that making a good thing that the lay it was good for.

So I just think I just think it's a misapprehension about Bitcoin. What is the deal with burning NFTs on other chains and putting them on Bitcoin? You've been involved in this, right? You've been trying to facilitate this. Am I right in saying that? Oh, yeah, encourage. Yeah, facilitate and encourage. Okay.

Yeah. Well, let's just say, it's called teleburn. Teleburn. Yeah. And myself and Rob, Roberto Hamilton came up with it. His name's Rob, but I don't know if it's short for Robert or Roberto. So I'm going to go with Roberto. Me and Roberto Hamilton came up with it one day at Bitcoin Park. And in about six minutes, we sketched out of design. And then later that day, somebody had already burned a half million dollar board aid, teleburned, sorry. That's remarkable. That's remarkable. Things went pretty fast. Basically, they sacrificed it on Bitcoin. Let's use nicer language for what they're doing.

They didn't sacrifice it. They were in the sad ERC 721 place, and they went on a spaceship and not like those Russian apes that like never came back and rocketed to the orderverse, a much, much better place where there's freedom. And really, it's like a defection from Soviet Russia. like to America. That's like the real metaphor here. I love how philosophical, this is like the ape of Theseus. Is it the same ape? Right. The ape of theseus. Oh yeah. No, in this case, I don't think you can invoke the ape of theseus because it really, it goes away. Well, I don't know.

Yeah. Philosophy is hard. I think it's the same. It's the same ape because I believe ape holders have like IP rights over their apes. So it should be up to them to re-domicile the ape to a new home. Right, right, right, right. Yeah, full agreement. So yeah, so the way that Teleburn works is you make an inscription, and then it has a particular fixed inscription ID. And then you generate this Ethereum address from that inscription ID. And that address is generated in such a way that it has two very important properties. There is no private key. So there is no corresponding private keys. So it's like a burn address. And it corresponds, you can show that it refers to exactly one, exactly one inscription, one particular inscription. So it gives you both, it says, on the ERC 721, and it actually works for any asset, including ERC 20. On the Ethereum asset side, it says this thing is, the owner, the last owner is never going to move it again. And find me on Bitcoin at this forwarding address.

It's like those little cards that they make you fill out when you move and tell the post office about your new, your new, your new mail receiving location. So unfortunately, we have to wrap because I didn't budget enough time for this episode. I do have a lot more questions, actually. But I guess we'll just end with, there's a lot of exciting stuff happening here. I think this is catalyzed more excitement in the Bitcoin space than I've seen in a long time. and it's made even kind of former Bitcoiners turned whatever non-bitcoiners. I'm seeing a lot of them actually return to Bitcoin saying they've made their first Bitcoin

transaction in years and years on account of this, which is really cool and interesting. So clearly a lot to be excited about. What is it that excites you the most about Ordinals inscriptions? What are you looking forward to seeing here? I'm looking forward to the future additions to the protocol that we have planned. to ordinal theory and to the Ord inscription protocol. I plan to do provenance in particular in a very exciting way that I think is very elegant. I have designs for other features that I also think are very elegant like a global name space.

The sats actually all have names, all of them with the letters A through Z. If you go to ordnals.com and you type in some characters, you can see the block in which it was mined and you can see the sat. So it already has a namespace. I'd like to give a namespace to inscriptions. What do you think about, or names to inscriptions? What do you think about the thing that inscriptions have, that's like a name? I was thinking a run. Okay.

Do you like it? You're the creator, so you get to name it. No, no, no, I need help. Like the idea is, I don't like, normally this would be called a symbol or a ticker symbol. Oh, I see. But I don't like financial terminology. Like I call inscriptions digital artifacts instead of NFTs. So this is actually an honest question.

Like, yeah, what do you think, ruin instead of ticker symbol? So you go, it's a string of characters A through Z, all uppercase. And you can look it up in like that search box. And then you go to a particular inscription. And I was thinking Rune. That's the inscription's Rune. Well, you know, it's very unfair of you to put me on the spot on my own podcast and then make me be creative on my show in a whole time.

We can follow up via email. Yeah, I like her. I'll think of something and I'll email you as well. Okay. Okay, well, on that note, I am very excited to dive into this. It's been very fun and it's been a delightful conversation. So thank you. Thank you.

I'm a long time listener of the pod. Also give my best regards to your co-host. You guys are like the best. team because you're like kind of like more laconic and like kind of softer spoken and then he gets into like he's like talking about like you know crime and he's getting fired up it's a it's a great it's a great combo so it's a real pleasure and an honor to to be on yeah I like to think of myself as the calm and measured one but sometimes you know I don't know sometimes we play different roles this has been great awesome I wish that I'd booked some more time but

we will do an update. Yeah, we can do an update. We'll do an update in six months. Perfect. Okay. Thanks, Casey. All right. Talk to you.