Episode 872 transcript

FLOSS-872

Jonathan: Hey folks, this week we're talking with Tris Willaker about open source and the law, including but not limited to topics about the GPL and legal cases, what AI means for lawyering, and the big court case over who the real Satoshi Nakamoto is. You don't wanna miss it, so stay tuned. This is Floss Weekly, episode 872, recorded Tuesday, June the 23rd.

I'm not Satoshi It's time for Floss Weekly. That's the show about free, libre, and open source software. I'm your host, Jonathan Bennett, and today, oh, I've been looking forward to this one for a while. Today, we're gonna be talking with Tris Wilcher, and Tris is a friend of mine, a new friend of mine that I met at the Ubuntu Summit.

And so you have to imagine that b- betw- at, they would do a couple of sessions, and then they would give us 15 or 30 minutes. And the, the tag on that was, "Go network. Go meet people." And I am an introvert at heart, so I have to work really hard during those times to go and meet people, but I was doing my best.

So I start walking around and talking to people, and I walked into this conversation, and it's like there was one person there I was wanting to talk to, and that we all introduced ourselves, and Tris was the other person. I didn't know him yet. And Tris said, "Yes I'm a lawyer." And I went, "Wait a second.

You're a lawyer, and you're at the Ubuntu Conference of your own volition? We need to talk. I need to know more about this." And so I got to talking to Tris about some of the things that he's done, some of the cases that he's been on that he is allowed to talk about. Sometimes in the legal world you're not allowed to talk about things, but there's some of these that happen in the open, and so you get to talk about them.

And then just his other open source stuff. Tris was a real pleasure to get to talk to, and I said during that conversation, "I need to have you on Floss Weekly," and he agreed. So without any further ado, Tris, welcome to the show.

Tris: Thank you so much for having me. It's a pleasure to be here. It's a sweltering day here in London-

In the UK. We've been issued with a cataclysm warning of how hot it's gonna get, so you'll have to forgive me if I gasp.

Jonathan: It was warm. It was warm when I was there s- surprisingly there, we Americans, we think of, sunny Britain. It's so far north, it's 70 degrees Fahrenheit all the time and just beautiful.

It w- it was kinda warm while I was there.

Tris: Yeah, it's a high fever s- temperature right now, so you'll have to forgive me if I gasp.

Jonathan: Absolutely. All right. So Tris let's start with I think your background would be interesting. So how did you end up in this sort of interesting position where you're both a lawyer and a software engineer?

Would you claim the title software engineer? At least canny. At least you understand what's going on.

Tris: I do release my own open source software, so yeah, I think so. I've got no formal engineering background, so it might be a steep claim. But it's not like I'm vibe coding anything Sure

So I'll accept it for the purposes of today.

Jonathan: Sure.

Tris: But yeah, my day job what I do day to day, I am a lawyer. Originally my background was in engineering. That's what I studied at university.

It was not software engineering, but a m- mixed discipline. And then I converted to law and IP and tech law specifically, and that's where I've worked at the interface of law and tech for my entire career now.

Jonathan: What, what led to that decision to, to switch from engineering to law, to legal?

Tris: In fact, there's a picture just behind me on the wall which I drew when I was maybe four years old, and it says, "When I grow up, I wanna be a lawyer." And I fought against that for some time. But then I I couldn't resist the pull of it anymore, and I fell back into that.

Jonathan: That's great. That is very cool. It is a, it is an interesting field, isn't it? The sort of... Because you've got over a thou- especially in England, a thousand years of history in the law, and then it's something that is at the same time so present and it affects every...

Whether you realize it or not, it's got effects on every part of life, and- Yeah ... people that have devoted their careers to thinking about that. I definitely see the appeal.

Tris: Yeah, you're absolutely right. A- and I think we're quite lucky as a generation to have liv- have lived through both the dawn of the internet and the dawn of AI- Yeah

Whi- which have- Yes ... bookended my life to date. I don't think a lot of people get to live through two revolutions in one lifetime, right? So- ... being a lawyer in that context is very exciting because although our statutes our actual laws are old, our computer copyright law is from the '80s the law is all about keeping up.

It's all about developing and reacting. And that's where I work. I work in disputes in courts in, in crisis situations, and we're always trying to think ahead, adapt to the changes. So yeah, I, I totally agree. Very interesting field and a wonderful partner for a tech interest.

Jonathan: Yeah. V- not very many years ago I thought rather regularly that the internet is the next big revolution and it's equivalent to the printing press.

And in my naivety I thought, "It'll be another thousand years before we see another revolution like that." Oh, how I was wrong. But we're accelerating.

Tris: We're accelerating.

Jonathan: Although I think probably, years from now looking back, people will probably see the internet and AI as basically being the same event.

That- Oh, yeah ... being in the midst of it, they're very different, but history will probably see them as basically the same thing.

Tris: One precipitated the other, for sure.

Jonathan: Yeah, absolutely. All right. So you you, you do specifically copyright and IP, intellectual property law, that sort of thing?

Tris: That's my specialty. And especially where that touches software and tech. It can be in the kind of case that we got chatting about when we were when we were- ... having our introduction at a printer conference or it can be somebody running off with source code and it's proliferated, it's been pushed out to I don't know, Git repos all over the place a- and controlling that, understanding the architecture.

That sort of thing. And when there are new issues coming up, you get to learn a lot. I'm always learning from intelligent people who've come up with new concepts from businesses who've started new ways of doing things. A- and it's just a real joy to be in that world and, a- and help it progress if I can.

Jonathan: I- is AI causing some ad- additional I don't wanna say headaches, but complexity in

Tris: the legal world? Yeah. It's certainly causing a lot of complexity and headaches for for lawyers who don't understand it. I'm at Bird & Bird, a firm in London and EMEA

w- where- Al- almost all we do is tech related in some way or another. And we act for a lot of AI companies, and we act in a lot of AI cases. There's been one big case which I can't talk too much about- Sure ... but is out there for and did catch some headlines. It was a case between Stability AI, w- who is well known for Stable Diffusion, and and Getty.

And we acted for Stability, a- and we were victorious in that case. But it was a fascinating lesson in how your mental model of the law doesn't necessarily match up to reality and finding that out in real time. That was, I was lucky enough to be part of that team.

It was run by my colleague Toby Bond who does a great deal of cutting edge AI work.

Jonathan: I've seen some really interesting things with AI. I think I've mentioned this to you in another conversation, but there's a website. It's satire, but at the same time the point they raise is very interesting.

It's Malus, M-A-L-U-S.S-H. And this website it's basically claiming to be clean room as a service. We will take an open source project, this is exactly what they claim to do. We'll take an open source project and we'll turn the AI loose on it, and we'll have one AI bot create a documentation for it.

And then we'll take another AI bot and have it look at the documentation and create all new source code, and then you can put whatever license on it you want to. And they're basically making the claim that the, the modern rise of AI is going to... And I've heard people say this, "It's gonna kill open source."

I'm... I think it's i- imminent demise is maybe a little overstated there. Have you had to look at things like this?

Tris: Yeah. Yeah, absolutely. In fact, the Malus site we did chat about it. And the next, the very next day actually- ... a friend of mine sent it to me saying, "Oh, my God, have you seen this?

It's crazy." And the reaction from the people who didn't take it as satire- ... was really alarmist. A- and I think what that tells you is that it is hitting a nerve and it is satire done as well as it can be, right? Yeah, absolutely. It does take a while for the concepts to land even once you process the joke.

And I think that's for good reason because there's a philosophy about code and there's a philosophy about copyright. And where they interact it can be very difficult to draw the line. Now I don't think that the satirical recipe for a clean room is a complete recipe. There's a fair philosophy there.

A- and I think probably done the idea of understanding a process and then clean room coding a process- Could work. But I'm not convinced that an AI pipeline is a foolproof method.

Jonathan: That's probably fair. Yeah, I think I think there are some open questions just in general around what an AI does.

Is it sufficiently transformative to break the copyright from the training data? And, there's some cases where you can get an AI to spit out, character for character some of its training data, and in that case, obviously it's not broken that, it's not transformative at all. And then w- the other really interesting thing is the output of AI is, at least in some jurisdictions, not considered copyrightable at all, which I don't know, that has some implications for businesses using it.

One of, one of my trips recently, I sat beside a comic book artist, surprisingly. Really surprised me to have this conversation. I asked him "Are you guys using AI in your stuff?" He goes, "No, not at all, because we can't copyright it." Oh, that's interesting.

Tris: That is interesting. I think if you're using code as a tool, then you probably don't care whether you can copyright it or not, if it's just an internal tool.

But if you're if you're writing code as a product or as a service- ... then that is a pretty big issue. And it extends to all creative industries, right? Yeah. I don't think we should lose sight of the fact that software engineering is a creative industry.

Jonathan: Absolutely. Speaking about people being creative in software engineering, There is one case in particular that you have been on top of.

In fact, I think you were the technical lead there at Bird & Bird about this case. And that was when Craig Wright claimed to be Satoshi the mastermind of Bitcoin. And this was the story... the case that we got to talking about. And some of the things that, the stories you had from it were just so fascinating.

I, I wanted to I wanted to go into this a little bit. Can you give us some background? I think most everybody knows what Bitcoin is, but starting from that point, what, what happened? What was this case about?

Tris: Yeah, I'd be happy to set the scene. And before I do, I'll say that we're very lucky to be able to talk openly about this case.

This was a case about showing the truth to the world.

A- and everything was done, as you put it, in the open. We're also able to see the outcome of it from, in a very long and detailed judgment- ... resulting from the case. So with that in mind I think the best way to set the scene is probably to remind ourselves that open source is a community, and I feel every community has its myths and its folk heroes and even villains.

And I think if Satoshi Nakamoto i- is the mythical hero of the story for sharing his invention with the world back in 2008- ... releasing it under the MIT license, and then instead of becoming its BDFL stewarding it for a couple of years and handing over to the community before vanishing into the background then, That

Jonathan: vanish- that vanishing act has raised a lot of questions among the community.

So I pitched the idea while we were talking about this, like I think Satoshi wasn't a real person, I think it was an intelligence agency, and I've heard other people say that. And as part of your work on this case you said probably not."

Tris: I'm of the opinion that Satoshi was a single person.

I was exposed to along with others in my team- ... we were lucky enough to be exposed to potentially as much information about Satoshi communications with Satoshi a- as anyone ever has been- ... who wasn't his closest confidant. A- and in order to be able to honestly ask to see that f- for the purposes of let's be honest, justice- I gave a commitment to each person that sent something to me to say that I would never use it to try and dox Satoshi, and I've kept true to that. I've never actually tried to investigate who the real Satoshi is. So while my opinion is that Satoshi's a real person, not an agency I have n- I have deliberately never indulged i- in the temptation to, to scrub about.

Jonathan: Absolutely, and I wouldn't ask you to.

Tris: No, of course ... but there is

Jonathan: one person in particular that you can you can say with some confidence that is not Satoshi.

Tris: So if w- there's actually only one person I'm aware of in the entire world who is by law proven to be not Satoshi.

A- and that is, yeah, Craig Wright. So he's if anything- the villain of this story. He was certainly found by the court to be dishonest in many ways. But the core of that dishonesty was the methodical staking of his claim over the course of many years that he was Satoshi, that he was the creator of Bitcoin.

A- and true to this myth he built up his persona in a lot of different ways. And to my mind, he was quite clever in that he focused on the narrative- ... on, on selling the dream. And I think from what I can tell he definitely convinced some people. He, he convinced people who funded him.

He had a great deal of monetary backing. He started off by doing a kind of signing demonstration- ... to, to s- to journalists a- and even to one of the key Bitcoin maintainers where he dressed the whole i- idea of a,

of a PGP signing process up into this dramatic event. A- and it was very theatrical, and I'm pretty sure it was faked because a- anyone who has used PGP will, will know that you can just sign something and it works or- ... or not sign something a- and you've got no proof, right? But he began to convince people which was fine until he escalated.

And what he did was he moved from shelling a story fr- from the sort of reputational side of it to actively attacking others. A- and I don't mean verbally attacking or physically attacking. He, he did do certainly verbal attacking. But the really problematic ones were when he started to interfere with people's lives, with their rights by suing them personally.

And he was suing people, right? Not just companies, but people, actual cryptographers, journalists, a- and open source developers who doubted him. He would take them to court to, to prove that they were wrong to do so or that they had libeled him. And with all of his financial backing it came to a head because he began to win, right?

He began to- ... get judgments against people, that they had defamed him by the way they had called into doubt his claims.

Jonathan: Yeah. I have a ... I wanna interrupt and ask something. He had financial backing.

Tris: Yeah.

Jonathan: Why would Satoshi need financial backing? Doesn't the man, whoever he is, or woman for that matter, I don't know own like 1,000 of the first minted bitcoins?

Isn't Satoshi like unbelievably wealthy?

Tris: Yeah. Why indeed? A- and there are many holes in the story. I think if, f- out of all of them, I feel like that is the most forgivable hole in the story, because you can imagine that i- if Satoshi who set up... Let's remember, Satoshi set up Bitcoin in its early days- where it was an unstable system and more vulnerable to a 51% attack. He did so by effectively sinking mining energy into it. A- and if he started to move those coins, which were effectively, as I see it, ballast in the system-

...

Tris: Y- you can see that might be considered destabilizing. So perhaps of all of the of all of the questions, that is the most reasonable.

A- and true to life those coins haven't moved, right?

Jonathan: I remember at the time hearing critics of his say, "If the man wanted to prove he's Satoshi here's the real easy way to do it. Spend one of the first bitcoins, and then nobody would have anything to say about it."

Tris: Just spend a sat.

Jonathan: Yeah. Spend-

Tris: A- and of course, there was this there was this principled response to that whi- which was enough. And this was really important, right? Because what Craig Wright was doing was selling the narrative. It was enough to sell the narrative- ... to say, "I, as a matter of principle, won't prove it cryptographically, because I want to prove it with different evidence."

Okay he failed to do that but that was his story.

Jonathan: That's a weird that's a weird place to stand on, but okay.

Tris: Yeah. It... But if you've got no keys, then that's what you're left with.

Jonathan: True,

Tris: true. But you highlight a really good point, a- and the point you highlight I think implicitly is this: you don't need to prove you're Satoshi to benefit from this story.

What you need to do is you need to prove that you're enough of a risk That people need to listen to you. So a developer may not believe you. Somebody who's got the technical savvy to understand how cryptography works may never believe you. But their boss may. The boardrooms of the companies which are handling these exchanges may.

And it doesn't really matter whether you believe him or not because if he sues you, the judge may. And I see that as almost a form of social engineering, right? So social engineering works by taking power from a person at its core, right? By deceit. And I think if you can spin a yarn and get a judge to believe it, then you've effectively got the judge to award you that power just as in the same way an employee might give you a password into a corporate network.

Jonathan: Yeah.

Tris: It's an inter- it just resounds much more widely, right?

Jonathan: Yeah. I suppose we see that, And of course this is my opinion on these, but we see that with things like junk software patents. And here in the US there was a problem for a while. One company was sending out letters to individual businesses over a patent that covered, I think it was scan-to-email.

And so they were sending out- Oh, really? Yeah. It was terrible. They were sending out basically w- letters demanding a $10,000 patent license purchase for an- any company that they thought was using scan-to-email because you're infringing on our patent." And somebody finally-

Tris: And let's say you receive one of those, right?

What do you do as an open source developer? Y- you can choose to fight it on principle or you can choose to try and settle it or you can choose to pull your product. If you're a small business you might try and settle it. If you're a huge business you might fight it on principle. But if you're an open source developer you're gonna pull your product 'cause it is not worth it.

The open source bargain is that you contribute and it doesn't come back on you, right? And whether that's more or less enforceable in different countries I'm not sure, but if you are faced with something that so dramatically changes that paradigm- ... then y- you're gonna just back out of maintaining.

Pass it on to someone else, pull the software, archive it, whatever you do you're just gonna say, "No thanks. It's not worth it." And that's what was happening with Bitcoin. Some of the main contributors to Bitcoin were saying, "This guy is a litigation risk against us personally. It is too much. We are no longer going to be contributing to this system as much as we love it because we can't risk our livelihoods."

So it's destabilizing and there's a chilling effect.

Jonathan: Yeah. W- was there, and you may not be able to speculate on this, but like what do you think Wright's endgame was? Obviously he was trying to amass money and power, but like he he ran the risk of killing Bitcoin, and I ... it's just such a, it's such a weird, it's a weird thing for somebody to do to- Yeah ... to claim to be this and to, threaten all these people. I don- I don't understand it.

Tris: I'm not sure there was an endgame. I'll never know, right? Yeah. And you're right to call it speculation but sitting here in my speculation chair I, I'm not sure that it ever would have ended.

That there were many irons in the fire. I don't think you need to look far beyond power and money to get to the end of the story. But maybe it would have continued for a long time. A- and thankfully we'll never know, because I hope it has now ended if not in the way he predicted.

Jonathan: So y- we talked about this idea of an open source contributor might just pack up shop and go home, and that, that was happening for some of the Bitcoin contributors. What what changed? Was there some big group that said, "On principle, we're going to stand up to this and stop it"?

Tris: Yeah, and it is exactly that impact. So the Crypto Open Patent Alliance is an alliance of, B- blockchain ecosystem companies. N- not just in Bitcoin but Bitcoin is obviously a foundational technology. A- and these are by and large a group of competitors. Th- they're competitors in their industries in many ways- but they come together to form this alliance to share patents, which is why it's called the Open Patent Alliance- ... or although we always abbreviate it to COPA to share the patents among COPA members- ... a- and pr- promote stability in a system because this is a system built on open source- and these members are all about keeping the system going allowing everyone to benefit from it, a- and making it work i- in a way which it won't unless there is stabilizing a sort of stabilizing economy behind it. So these guys sat up and took notice because when you when you are starting to discourage people from contributing their intellectual effort to the foundational open source system, you're destabilizing the whole mountain that's built on top of it.

And we've all seen that XKCD cartoon. We know what happens when that domino collapses. So the th- these guys stepped in, and they came and they spoke to us, and we put together a team at Bird & Bird, and together COPA and Bird & Bird we started this case to try and prove that Craig Wright's story was false- a- and to try and show the world that was the case. So it was actually led by my colleague Phil Sherrell, who heads up our our London headquarters a- and there were several of us. I will say that I will say that the guys on the other side had more than twice our number. So we often felt we often felt like we were underdogs in this fight.

Jonathan: And Bird & Bird is not a small firm. That's probably not something you guys are used to, being the underdogs in a fight.

Tris: Yeah. That's right. We are often pretty lean in how we handle things. So we like to put together a team where y- you know the ideal the dream team scenario, where everyone's got their place, and is an expert in their field, and we all integrate.

Jonathan: Yep.

Tris: And that worked really well in this case. It was an absolute delight. So I ended up leading on the technical side and the- ... forensic investigation, some of the technical factual investigations too, which we might come onto.

My colleague Ning who was magic with the more open world factual i- investigations.

And then before I joined the team Graham Smith who you may have heard of, he's been an internet lawyer for decades, since the dawn of the internet. A- and he actually created- And by, and

Jonathan: by internet lawyer, you don't mean someone on the internet that thinks he's a lawyer. You mean an actual lawyer- I mean-

that does internet law.

Tris: The law of the internet. He's seen it all. Yes. ... and he's still very active. And he'd actually put together this amazing database, I think a SQL database- ... drawing all the threads of the factual stories together and cross-referencing to see where they didn't add up before the case.

I- it was a really amazing team, and we were able to keep it quite lean by, by segregating everyone's specialties.

Jonathan: Yeah that's super cool. What, what did the actual legal structure of the case look like? So like you can't just, from what I understand, you can't just sue someone because you think they said something that wasn't true.

What was the actual- Yeah ... how did s- how did standing work? And and obviously for those of us in the US it's gonna be a little bit different because this is British law. It was in a, an English court. But how did that part of it work?

Tris: It's a fascinating legal question, not one that most people usually touch on, so good one.

A- and if I can try not to get too technical, you're absolutely right. You can't just sue someone because you don't like what they say. That's not how the law works. And the question of standing was very important. But in the end it came out all right because Craig Wright had embarked on a campaign of claiming copyright in the Bitcoin whitepaper itself.

A- and as anyone in Bitcoin knows, it's a sort of mark of being part of the club that you host that whitepaper on your site-

Jonathan: Ah ...

Tris: to show that you're part of that ecosystem. So anyone who was hosting it, he was writing to them and complaining that they were infringing his copyright. So we checked with him, we wrote to him and we said "COPA's hosting the whitepaper.

Are you saying they infringe your copyright?" And he said, "Yeah, I'm gonna sue you," at which point we, we were able to take it to court to prove- ... that he didn't own that copyright. A- and who owns the copyright on the whitepaper? Satoshi. A- and it boiled down to the same question. In fact it was a bitty, a bit of a tangled ball of wool.

There were many threads altogether. He was suing open source developers and also the companies who happen to be COPA members. A- and it did coalesce into one huge case w- with one central factual issue. Wa- was,

Jonathan: Was that a challenge to get the court to bring it all into one case rather than have to fight him 50 different times in 50 different cases?

Tris: Yeah. I can't say exactly how the court would've approached it, but if I was the judge seeing this, I would've said, "That is a tangled mess. How can I shake it out?" A- and maybe that's what the judge did. It's certainly the effect of how the judge handled it. So we did coalesce the cases into one trial.

Certain aspects were stayed because they were gonna fall away if the main case resolved the way we wanted it to. And it all came to a head after a few years in one big six-week trial which was an amazing trial for a few reasons. One is that it was held in the biggest courtroom in the building-

With everyone piling in. And the seats at the back of the public gallery totally full. Two, because it was actually live-streamed, and i- in a, I, in a term which I've never heard of before or since, there were more than 1,000 people following the live stream daily. Which is pretty thick stuff- in a court of law if you're not a lawyer. Hats off to those guys and girls who dive in every day.

Jonathan: Yeah, absolutely. Going into this, was there a part of you that thought maybe Craig Wright is Satoshi? Maybe he's

Tris: Wright? Yeah. A- absolutely. In fact a- as a matter of principle.

I'm a lawyer, so I don't have any skin in this game. I came to this totally cold. I don't have any Bitcoin. I don't know about the technology, and I'm learning from everybody around me. And I don't know the difference between Satoshi and not Satoshi at that point. You come to it with objectivity.

You come to it open-minded. And you see the difficulties, but you can also see how they might be surmounted. And it's only as you begin to go through the evidence, the mountains of evidence that he dumped on us, a- and you start to see that everything that supports his claim turned out to be a forgery- that you start saying, "Yeah I'm beginning to make up my mind now." So it was a process. It was absolutely a process.

Jonathan: How would you have handled it? So obviously you personally came to the conclusion, and I know this from talking to you, you personally came to the conclusion that Wright's not Satoshi and so that made this easier.

How would you have handled it being on the side you were of the litigation if you had looked into the evidence and gone, "I think it's him. I think this is him"?

Tris: Y- then you would have put the case as best you could, but you would never tell something to the court that was untrue, right? I- it's a bit like asking a criminal lawyer, how do you defend someone who you think is guilty?

Yeah. You just have to be very careful. You give them the defense to which they're entitled- ... and and you do what is right and just, and you leave it for the judge to do the right thing. So luckily we weren't in that position. We didn't have to introspect too much. It was all outlooking because it became pretty clear, at least within our four walls- it became pretty clear to me who were the good guys pretty early on.

Jonathan: Do you remember the moment or the piece of evidence that you looked at and went, "That's not him"?

Tris: So we got his disclosure, I remember, and I did... The first thing I did was I just spun it all up on my Linux box- ... which I actually had to switch to j- just to get through the volumes of evidence that were being poured on us-

and to be able to pull tools to analyze it. And I spun it up, and I just I just put some filters through it for the earliest documents that mentioned Satoshi. And I remember very clear, very clearly pulling up the first, I think it was a Microsoft Word document- ... opening up the internals of the document and seeing right then and there within a few moments that the internals of the document had all been sanitized away to appear from maybe 2007 But they had all of the editing artifacts, u- unfortunately for Craig, still compressed into the document itself.

And it's all in the forensic reports. You can see the references to the Financial Times from 2016 and The Economist, and you're like, "Oh, yeah. That was prescient." Now, Satoshi was prescient, I think, but not that prescient.

Jonathan: So when you found evidence that things were forged, that was pretty much the moment?

Tris: That was an eye-opening moment. Very exciting. But actually the evidence came in many threads. It wasn't just forensic analysis.

On that technical side, though, what really struck me was the extent to which everything was based on open source tooling. So we were there, and we were defending an open source industry against somebody who was suing open source developers o- on behalf of a patent openness client.

So we're already into that field. But then we're receiving forgeries created with with LaTeX, with C++ with OpenOffice. A- and we're using Linux, and we're using open source tooling to be able to analyze them. And that was my moment when I ... when the whole open source bargain and community really clicked for me because you could just take these things, apply them, a- and get the result you needed.

Maybe tweak them if you needed to 'cause of your specific use case. A- and I had to learn, I had to learn Python and I had to learn Bash scripting to get through all of this. So that was my learning curve. That's- And it was a delight. Yeah,

Jonathan: that's great. One of, one of my ... Oh, I can't re- I can't think of his name, but one of my absolute heroes, there is a, there's a judge in the US that did the big Oracle Google case.

Tris: Oh, yeah.

Jonathan: ... Oh, what?

I don't remember. S- Steven No. Oh, goodness, I can't remember his name I-

Tris: is the name Alstrup or Alsop?

Jonathan: That, that sounds right. Judge Alstrup. Yeah, anyway, he he presided over this big case about Android between, Google and Oracle, and at the end of it he's... One of the things that he said is, "I learned how to program to be able to understand what was going on in this case."

And, he's... And then says the- there was, like, 11 s- lines of source code in question at the end, and he looks at it as part of his ruling, and he says, "There's only one way that you could write this. There's no other... you could ask a, a first-year engineering student to write it, and he would give you basically the exact same thing.

So that's fair use." And I just... I love the fact that he learned to program to be able to try the case, and one of my, One, one of those stories about the law that I really enjoy, somebody

Tris: doing it

Jonathan: I think

Tris: this is because we're in the information age and lawyers all around the world need to know that if you see things through the lens of the information age, if you start seeing your cases that way- then you become a better lawyer because you can use software tools to become a better lawyer. You can assist people who are using software more effectively. Yeah. And I th- I hope that everyone starts to do that. I once had a case a long time ago now- ... where somebody brought me in and they said, "This is a real problem.

I know you've got... I know you've studied engineering. Maybe you can make a head, heads or tails of this because our client says that they never copied anything, but their code block is identical." And I looked at it and I was like, "This is just linear regression." "There is only one way that you can write this."

It's your same point. I, and I do think that comes up again and again, actually.

Jonathan: You actually... You see something similar not exactly the same thing, but something similar in music copyright. ... It's like the, "They copied this tune," and there's a, there's actually a project out there where a lawyer that understands music put it together and "Okay, six notes in a row.

There's 12 notes. There's only, and I forget exactly how many, but it's there's only a few thousand ways that you can put this together. We're just going to write all of those tunes and copyright them all, and then put it in the public domain, and then we'll just be done with this idea of, "He stole my melody."

It's six notes in a row. It's... You can't copyright that.

Tris: I like that. I like that. Did you hear about how Ed Sheeran was in court and he brought his guitar with him, and he played through the various ways that you could put some chords together? It was a very effective- ... piece of personal advocacy.

Jonathan: Yes. Yes, absolutely. Every l- how does that go? Every pop song for the last 30 years is the same four chords, something like that.

Tris: Yeah.

Jonathan: It's very much the same thing. It's the same idea in music theory as we would have in computer theory. "There's only one way to do this. Every song is basically the same chord progression."

Tris: Yeah ...

Jonathan: that's

Tris: similar. It sounds good.

Jonathan: Yeah. So one of the, one of the other neat stories that you had, and I'm probably jumping partway into this, but B- Bjorn Straustup, y- he was a, ... Y- you made personal communication with him. Yeah. And something about the story here was, like, you needed an expert witness, and you couldn't get one, and the only way- So it, it is the case- What's the

What ... Remind me the story.

Tris: So we ... W- Background is we'd gone through these 100 main documents that Wright said was ... his crown jewels, and we'd basically proven that every one of them was either unimportant or fake. A- and we'd served maybe 2,000 pages of expert evidence.

There was a conspicuous absence of code in any of his documents. So he, as far as I can piece it together had read our manual of how to forge and how to be caught forging. A- and must have seen the whole seen the whole where there was no code, a- and tried to fill it, because the thing about source code is it's plain text.

There's no metadata that goes with it. A- and that came very late in the case. We were already gearing up to trial. It was an extremely late disclosure, a- and I believe that it was late because he'd just finished forging it. In fact we were able to prove exactly the dates when it was all forged.

But the the problem was that in order to prove that certain functions that were called on in that code- couldn't be correct to 2008 or 2007.

It was necessary to un- understand things like what is the standard namespace? W- when were these things published? And it's very difficult to prove that without an expert who will educate the judge about what these things mean. So w- what do you do? You either go through a big procedural question of does everyone get evidence on both side?

There wasn't time for that. Or you ask the man himself. So we decided we'll contact we'll contact the C++ developers. We contacted the person who wrote the library in question. And we contacted Professor Stroostrup and they very kindly responded, because open source is a community.

And when people ask you about your software pe- people are generally in this community extremely helpful- ... at sharing knowledge, both technical and historical. And this happened again and again. So we had forgeries made in OpenOffice because that's the software Satoshi used.

So we were able to contact the then maintainers of OpenOffice- ... who were able to give us their build logs and prove that the build that he'd used to forge it didn't exist at that time. And nor was it possible to predict the build hash o- of the software that, that was output by that compilation process.

That was OpenOffice. The same happened in LaTeX because he then pivoted to, plain text LaTeX source. Which didn't end well for him either. But Professor Stroostrup a- and many others joined in to help us prove fact after fact. A- and these things were bricks in the wall that, that added up to a proof that each and every piece of evidence he relied on-

Was false.

Jonathan: I have a little note here that there, there was some interesting sleuthing done outside of the digital domain.

Tris: Yeah, that's right. So what do you do if all of your digital forgeries are getting picked apart? Hand write some, right? W- we had this amazing ... A- and this c- this particular document actually went through three court cases I believe without being pulled apart, and it was very simple.

It was a single sheet of A4 paper, and on it were written meeting minutes, handwritten, very terse, and they were on this form which was headed up minutes- ... prefilled table And with a single sheet of A4, handwritten, there's not much you can do- ... to prove that it's fake. So we thought about how do you go about this?

A- and without going into the privilege detail the upshot was that as a result of an amazing teamwork of people all around the world y- across three or four continents, we were able to track down not just when this pad was printed and by whom, but we were able to follow the trail of companies that had been acquired one after another, and back up that stream, contacted the lady in Shenzhen, China, who had operated the printing press-

which printed that particular pad of paper. Ah. And she'd kept her PDF proofs, and they still had all of their metadata intact, and we could show, As a group, we could show that this pad couldn't have existed before 2010, which was three or four years after he, he claimed it to be. It was the most magical moment when that landed.

I think it was like 11 o'clock one night. I didn't sleep at all that night. It was so exciting.

Jonathan: Yes. Oh, that's great. You do the little happy dance.

Tris: Oh, yes.

Jonathan: Is it... Because that's the sort of thing you don't know for sure until you get it in. This is a... i'm sure that was a long shot.

You g- My guys must have considered that a long shot.

Tris: It's totally a long shot ... maybe there's- But you're open-minded, right? You follow the truth because you'll either find out something which is true and is against you, in which case at least you found out the truth. Or you'll find out something which is true and for you. But if you're open-minded and follow the truth that is the right way to to approach these faction investigations.

Jonathan: Absolutely. The the trial itself, you said it lasted six weeks?

Tris: Yeah.

Jonathan: How did that

Tris: go? It was, it was pretty incredible. We started off with Craig Wright himself on the stand, as you say, for six days of cross-examination. And then that's a long time. That is a long time, yeah. Plenty of breaks. It wasn't inhumane. A- and he was- ... very good at telling his story. But I think it's fair to say that the people who went into that courtroom thinking that he was he was Satoshi, were able to convince themselves, "Wow, this guy s- has such a facility with answering questions, that he must be Satoshi."

A- and the people who were maybe more open-minded could see the holes beginning to emerge at that point. And we went through all of the documents and every facet of his story and he clearly prepared. Excepting one case which I will dwell on because it makes a great story. So let's say that you...

Just imagine that you are a person who's forged. W- whether you've done it with your own hands or you had help doing it, w- we'll never know but imagine you're that person a- and you've got an answer for everything, and then the one unexpected question comes, and you react on the fly, but you realize there's a hole in your story.

What do you do? You go home that night and forge another document. And that is what seemed to happen during this trial. So a few days later, after that first week w- we had a situation where the lawyers on the other side said, "We... There's something we need to bring to the judge's attention."

And they did and they did it in open court. They said, "These documents have just been disclosed to us." And they provided them, and of course we had to go through a whole another forensic investigation. By this time it was maybe our sixth. Because of the layered disclosures that we'd been having all the way around.

A- and it was it was... The upshot of it was a document that he'd forged during trial So he had to be recalled once to answer for certain additional problems that had arisen, and then again right at the end, right at the end of the case, he was on a third time under oath to answer for this forgery that was made during trial.

So it went right up to the wire. It was an incredible experience to be conducting a forensic a- examination back at the office while the trial was in progress. I hope I never find myself in that situation again.

Jonathan: So it was established in a court of law that Craig Wright lied on the stand under oath?

Tris: Yeah.

Jonathan: Why is he not in a British jail right now?

Tris: He, ... I don't know exactly where he is. I think he went to Thailand for a time. He was found to be in contempt of court and has not come back to the country since then. But I think the way the judge put it was like this.

"If Dr. Wright's evidence was true, he would be a uniquely unfortunate individual, the victim of a very large number of unfortunate coincidences- ... all of which went against him- ... and/or the victim of a number of conspiracies against him." And then the judge took a break and says, "The true position is far simpler.

It is, however, far from simple because Dr. Wright has lied so much over so many years that on certain points it can be difficult to pinpoint what actually happened." A- and he went on to find over the course of maybe 230 pages it's the length of a novella, Okay ... that Wright had lied and lied.

A- and it was a very cathartic result. W- what had really happened was that all of Craig Wright's eggs were in this one basket by the end. He just had to convince this one person, this one man, that he was more likely, 51% likely- ... to be Satoshi. The archetypal 51% attack on Bitcoin you might say.

And he picked the wrong mark because he was up before a judge who himself has an engineering degree and a long history of practicing in IP at the forefront of technology who really dived down into the detail, read every word- ... looked through, I'm certain looked through the source code line by line, really understood what was going on.

A- and at the end of that six-week trial, the judge had been, He stands he-- I was gonna say sitting. You, we say a judge sits in court but this judge stands, and he'd been standing and listening to all the evidence silently throughout. And at the end of the case, it was his turn to speak.

And the typical the typical conclusion of a case is that a judge will say, "Thank you for all of your submissions and evidence. I will consider it and deliver a written judgment." And that's what happened here. And then he paused, and he said, "But I have seen all of the evidence, and it is quite clear to me that I can make the following binding declarations."

And he did. He declared right then and there, "Craig Wright is not Satoshi Nakamoto. He did not author the Bitcoin whitepaper. He did not author the Bitcoin software." A- and he delivered it, he brought it to an end right then and there with 1,000 people on the live stream, I think. A- and us all in the room.

And I still get I still get goosebumps thinking about it now.

Jonathan: That's great. That's a sort of once in a lifetime moment.

Tris: Quite right.

Jonathan: Yeah. That's awesome. Now, you you've done some other open source work that's in the sort of the legal sphere as well. So you do work at Bird & Bird in software.

Yeah. Did they give their stamp of approval of, you did this on the clock, but you're allowed to open it, release it as open source?

Tris: Oh, no my side project is my own and done off the clock. Okay. So I do make tooling internally- ... for n- neat jobs that need doing.

I like to make self-contained tools- ... which can be run client side. A- and with a bit of JavaScript and just understanding how libraries work- ... combining your knowledge of how the legal process works with that technical capability allows you to make tools- a- and just deliver solutions, often very quickly. A- and it's often like document management or document manipulation. So I have a l- li- little library of functions that I can use. That's all great a- and helps us and is very very fun to, to work with my colleagues to to solve those problems.

But one thing that I made for myself was a, o- originally Bash and then Python a- an implementation of a pipeline to get your documents from loose PDFs- ... to a bundle of documents that you can take to court. A- and what's what's great about this is most lawyers will give it to a trainee or a junior to do and they will deliver the documents- Okay

ask for it to be done, and then receive it back the next day and never have to get their hands dirty.

But that is the most soul-draining part of a young lawyer's job. It's awful. Ah. I- it is the bogeyman task that everyone tells stories about bundling late into the night.

So I, I realized this was actually really useful. And I thought I've spent the last three years benefiting from open source tools myself. I've got the bug. Why don't I open source it?" So I did. I re-implemented it again in, in JavaScript- ... so that it runs client side. And taking advantage of some of the quite cool WASM r- recompilations that are available now.

Oh, yeah. Especially for Move PDF. And with a bit of front-end magic, it allows people to create their court documents into a bundle which fulfills all of the court's requirements, and to do it in a- in about 30 seconds. And that saves time in two ways, right? So first way it saves time is if you're a lawyer and you know how to get it done, it's a quick process.

The second way it makes a sa- a time saving for people and I find this very wholesome, is that if you don't have a lawyer, you can't afford reputat- representation- ... then you'd have to learn what the court requires first before you could start even doing your document preparation.

And this pipeline does it for you. So all you do is throw your documents at it, get them in the right order, and it does all of the hyperlinking, cross-referencing, and all of that for you. So it is a time saver mainly for litigants in person. Also sometimes for lawyers, and I know some some major law firms in London definitely use it.

So that, that's it. It's my way of giving back to the community because I don't have a great deal of time for pro bono work at the moment. And it's built on open source, so why not release it open source? Yeah.

Jonathan: Absolutely. I imagine it also helps avoid those awkward conversations of I'm sorry, Your Honor, the table of contents says page five.

That's actually on page seven."

Tris: It's such a waste of time. And it's not just an admin problem, because it's an access to justice problem. Let's say you- ... haven't got a lawyer, and all you wanna do is go to this, possibly the most stressful place you'll ever find yourself. I can't think of many more.

Perhaps hospital. But you're going to this stressful place, and you need to make your argument against a talented adversary. And if you can't communicate to the judge with the documents in front of you, if you're always losing your focus because you have to add three to the page numbers, you're not gonna be able to communicate effectively.

A- and actually it should just help people not fall into that problem and I hope give them a better chance of conveying their case to the court. I can't improve their case for them, but hopefully it will help them do what they can do for themselves.

Jonathan: Yeah. Yeah. That's, that's- Something I've thought about quite a bit is one of the problems with the legal system around the world, I think, is the barrier to entry.

And a lot of times that's cost. But even just become- having your tooling together and your understanding together well enough to be able to present a case. We have this... The ideal is that anybody, no matter who you are, you can go to a court and you can present your case. And if you, if the, if right is on your side, you win.

And unfortunately the reality is that oftentimes it's whoever has the most money that they could pay for fancy lawyers is the one that gets to win. And,

Tris: Of course in the Copa case, if right's on your side, you're gonna lose. But the the barrier to entry problem is huge. Not just in terms of tooling, but in terms of c- conceptualizing.

A- and actually the legal system is very similar to, to to software engineering. What it's attempting to do i- is create a highly precise form of words to convey a concept. And that is in essence w- what a coding language does. It, it applies a particular syntax to that problem. The language we use is English.

But it's about conveying concepts with precision. A- and, and- ... if you squint at a contract next time you have to sign one and get your get your software engineer mindset on you'll notice that we have effectively imports function definitions in the main loop.

I w- we don't call them that. We call them party names. We call them defined terms- Yeah ... and clauses. But they operate mechanically in exactly the same way. So I feel like there's a lot of shared mindset there. Yeah. And perhaps then of all people the people who, who suffer the least barrier to entry are software engineers.

Jonathan: Yeah that's probably true if only because we are used to parsing through the difficult things and understanding how they work together. So one of the, one of the quotes that I like to use on the show and it's Simon Phipps, and usually he's talking about the importance of community when he says this.

But he makes the statement that "Software licenses don't compile." And his point there is, you don't know exactly what the software license is going to mean until it be- appears before a judge. And, the judge has comments and some of it doesn't survive, some of it gets understood differently than you expected.

Is that software license a contract or not? Some of those various questions that, that show up. It's interesting though to think about that in a way that is what this is an attempt to do. The law is a, an attempt to make s- these concepts almost compile. That's a rather intriguing thought.

I wonder if someone could formalize that at some point and make it so that they do actually compile into a perfect meaning.

Tris: Ethereum would say that smart contracts are the way. But the i- if licenses are your compiled programs, then lawyers are your CPUs. So what you're really trying to do is you- as I think out loud i- is you're trying to have something that reproduces no matter what processor it runs on in the same way. A- and a well-established license with well-established clauses- ... will be met with the same legal response, with the same legal advice, no matter which lawyer you go to, as long as- Right

they know their job. I think that's the closest analog to compiling a license. I- if anything, the license is its own machine code i- in that paradigm.

Jonathan: Tries to be. Do you follow some of the cases, like for example right now there's a quite notable case in the United States where Vizio is being sued by one of...

Oh, I forget the exact name of the group, but one of the free software groups, basically saying, "The things in your TV are GPL. You need to release all of this source." And it's, it's notable because it's one of the first times that the GPL will go before a judge, and we get to see what the judge thinks of it.

Tris: So that's, I think, a Software Freedom Conservancy.

Jonathan: Yes. Yes.

Tris: SFC. SFC. I had a chat with with those guys at one of the conferences. A- and, ... it's a fascinating debate. I am certainly not qualified to, to comment on the legal aspects of it. Sure. But I, I think I think there are not only the f- like, the fundamental questions of what does this license mean, but it also is...

It's a big stand, isn't it, for the industry. We are going to enforce our licenses. We are not just gonna sit back a- and let you in, in the terms of SFC, abuse our license. And the outcome of that I potentially could change how people measure the risk balance.

I don't know. E- especially since it's it's in an important court, isn't it? I don't know if it's federal court or Californian court, but it's a court that people are gonna be waking up and listening to.

Jonathan: Hello Just going to look at what court that's actually in. I don't remember.

Yeah, that's what I was gonna- I think it's, I think it's a California court for now. But the way these things tend to work though, US law is based on a lot of British law and the idea of common law, what the courts have decided over the years tends to become binding. And what a California court decides will get referenced by, a Florida court looking at it, or a federal court looking at it un- until someone else comes along, a higher court, and says no, you got it wrong.

We're gonna change that." I, in a separate conversation I asked you something about whether you thought the the next version of the GPL was being worked on, and I'm reminded of that because depending upon how court cases like this turn out, that may be exactly how we get a GPL v3.1 or GPL v4.

And I think- I

Tris: actually have no insight into the question, but-

Jonathan: I

Tris: think- But I think you would... You're absolutely right that waiting for the outcome of a case like this would be a great thing to do if you were working on it.

Jonathan: Absolutely. It may very well be one of the things that spurs on that, that next revision.

Personally, I think AI is something that needs to be addressed in the next version of the GPL. Probably have a couple of different licenses, whether, you want people to be able to pull your work into an AI training des- set or not. Like how does that work with copyleft? How does AI training work with copyleft?

That's an unanswered question in some ways.

Tris: It's... In some ways I think it might be unanswerable.

Be- because it, in many ways the core principle o- of open source is reproducibility, right? So does that mean that if you're gonna be looking for an AI model to be open source, you need to be able to reproduce, what, the weights?

Or do you need to be able to reproduce the training from scratch? Where does that re- reproducibility line start and finish for something which is a totally different kind of software doing a totally different kind of task?

Jonathan: Yeah. The OSI actually took that particular question on, and they now have definitions for what it means for an open sou- or a model a, an AI model to be considered an open source model.

And they made some decisions about how exactly that would work, and people are not happy about it because it is, it's so new and it's so different from anything else that we've had to fight with before. So that, the, there are OSI definitions for what it means for a model's license to be, OSI approved.

But people are not, people were not happy about

Tris: it. It's a highly political question be- Sure ... because people have different opinions about what AI means- ... what it should be, and what openness means in that space. I thought it was a really good thing that they went ahead and published it.

Yep. Be- because a lot of the questions were being asked in the- abstract. And this gives someone a framework against which to test their their thoughts and assumptions, right? I feel like it probably created as many opinions as it put to

Jonathan: bed. Of course. Of course. I know, Tris, that you've got to hard out here in just a few minutes, so I think, and we've gone for over an hour. I think we can go ahead and wrap it up. It has been an absolute pleasure to have you on the show to talk about-

Tris: My pleasure to join you ...

Jonathan: things legal. We we'll have to do it again. Give it a few months, we'll have to have you back. It'll be a lot of fun. Love to.

Maybe once the the Vizio and SFC case wraps up, we can have you back to talk about that. That could be interesting. But- Let's

Tris: talk about it.

Jonathan: Yeah, we'll do that. All right. Thank you so much for being here. I appreciate it so very much. Again, it was a real pleasure. Oh, I gotta ask.

I gotta a- I will get emails if I don't ask. Okay. Favorite text editor and scripting language.

Tris: Oh I love to script in Bash. That's my favorite that's my favorite scripting language, for sure. I have been experimenting in the last few weeks with Helix. I'm more of a Neovim kind of guy. And when I just wanna get a quick note together I find Kate, because I use Kubuntu.

I find Kate to be extraordinarily fully featured. And very good for markdown with a preview.

Jonathan: Helix is the post-modern editor and of course in their FAQ it's, "Why post-modern?" It's a joke. If Neovim- ... is the modern Vim, then Helix is post-modern.

Tris: I have had trouble getting to grips with its reverse bindings- i'll be honest.

But it does have a handy pop-up.

Jonathan: Yeah. That's fun. All right, Tris. Again, thank you so much for being here. I appreciate it. It's been a blast.

Tris: My pleasure.

Thank you.

Jonathan: All right. We've got some fun stuff coming up in the future. We've got Andy Rick and Aaron Bassett from QNX next week.

We'll make sure and ask them why QNX is not open source. And then the week after that we've got Andrea Gallo of RISC-V. Looking forward to that one a lot as well. And then in a few weeks we're talking with Michael Meeks of Collabora. I believe I met him at the Ubuntu Summit. And then we have a returning guest, Francois Poulin of Smoked Meat, and that is yet another GitHub but related security tool.

You don't wanna miss that one as well. We appreciate everybody that is here, whether you watch or listen, get us live, or on the download, and we will be back next week on Floss Weekly.