 So this is what I'm going to talk about, and yeah, let's see. So I think the challenge that we face in today's world is to try and build an economically sustainable free software business that works in symbiosis with its awesome and exciting community, right? This is the challenge, right? Because there are many different actors with many different motivations and interests. And I'm afraid I'm not going to come up with a perfect answer, one size fits all. I'm sorry, but we can at least explore the problem. Perhaps there's a perspective there that you haven't heard before. But first I want to talk about this heresy 101, this economically sustainable thing. This idea that someone needs to pay something in return for something scarce. This is not uniformly the case. Many people pay for things that aren't scarce because they just love them. But what you'll notice is that even charities try and make something scarce. The membership gives you the right to do something. You can get into castles cheaper if you join the national trust, or whatever. So somebody really needs to pay for something in return for something scarce. But why is it a heresy? Well, it's quite possible. You can imagine a plausible world where everything is done by volunteers. It is not a completely implausible world. And we have the most amazing volunteers in our project. So it's just extraordinary to talk to many of you and hear your passion for doing good things for open source. And that's great. That humbles me. I was once a student and I had once lots of time. And I volunteered too in many ways as a passionate person. But time isn't money, but there is some kind of equivalence there. If we want time from passionate people, it helps to have money to allow them to work and so on. But still, lots of floss has been written by volunteers. An open SSL is a great example. I'm not one that I want to mock in any way. Although, you know, there have been some pretty serious vulnerabilities there. And there are some really interesting problems as they try and regulate open source software, particularly key critical infrastructure, security software that's written by volunteers. And there's exciting problems of liability that my friend Simon Phipps and many others are wrestling with even now in Brussels and elsewhere. Lots of businesses want to buy time somehow, get time to fund their programming addiction, to create better software and to grow and to do more. So it seems crass to talk about money or time, where we get it for, to some people. And I understand that. But I think the economics and the flow of money or time is very important to us as a project. I will use this word volunteer. And I want to recognize that volunteers, as some of them, are paid. And so in Apache, you could be a staff member in one context, but you're treated as a volunteer in another. Like if we take, for example, not my staff, I mean my staff do this too, but TDF staff often contributing not just to LibreOffice, but other projects and doing good things everywhere and going far beyond what they're paid to do. So there's just loads of goodness and value there. And so if you use this meaning of volunteer, of course, all software can be created by volunteers. It's just a passionate people working hard and so on. So there are lots of different ways. And I've got about 15 of them. I'm going to rush through about five that we found that work for funding open source development. So they all rely on a scarcity, I would argue, or many of them. And there's a scarcity of skills. So consultancy is this. I need someone quickly to solve my problem. I can't do it myself. I have to buy someone. And of course, there you are. You represent your customer's interest in a project. That's great. You deliver something. You have a clear spec. The customer is happy. The project is happy. All good. You walk away with clean hands. A week later, QA finds 10 bucks and the users start screaming at you that there's a problem. Right? This is one of the downsides of consultancy. It's extremely hard to grow large consultancy businesses that are safe. It's called a non-scalable business. It's extremely risky if you do fixed price projects because even builders building houses cannot somehow estimate the cost of building the house correctly. And building houses is not a new technology. It is well understood. There is a deep pool of experience around this. But the guy who wins the tender to build your house is going to struggle, which is why you should get three tenders and pick the middle one. Just a tip. Anyway. But this is how our companies go bankrupt in consultancy. And the scarcity of skills may not last. Who knows? It's very, very hard to know. So there's one model. Another model is the scarcity of binaries. This is also known as an enterprise distribution. So Red Hat Enterprise Linux was, I think, one of the first people to pioneer this. It was very controversial in 2003. I remember it well. And, you know, SUSE does a similar thing. It is not easy to get binaries free or it wasn't when I was there. And there's this editorial composition of compilation of packages typically produced by other people and then a pay wall to get the older stable enterprise content. And the latest and greatest is free, of course, and open and so on. And, you know, you can subscribe to get old news. And the good news about buying a product is that there's a presumption that you buy it again. When you buy consultancy, there's a presumption that you solve the problem and then you don't do it again. When you buy a product, the assumption is that you renew. You buy the same product every year. It's in the budget already. So it's very much easier to scale these businesses. It's called annually recurring revenue, ARR, which is what everyone wants. Of course, there are some real problems around cloning, CentOS, just taking the work and publishing it. Red Hat is now hiding its source packages to make that less easy. There are claims about cuteness and compromise and it's legal, but is it right? And, you know, discussions and fighting and so on. But this at least is a model that has worked for distributions that have these features and editorial compilation of lots of software. It is starting to not work because this editorial compilation is not what people want. They want small docker images with nothing in them. They don't want lots of stuff that works together. They just want one thing. So let's see what happens to the economics of Linux distributions in a container world. Another approach is the proprietary periphery, which is the free software way of saying open core, you know, major on the bit you don't like. And so this way you basically have a proprietary product, but you open wash it by having an open piece and it's a very simple value prop. If you don't pay, you don't get the proprietary piece. Problem is to do this, there are nasty compromises around the community and your competition will focus just on the proprietary piece and not the whole thing. And yeah, depends how you balance that periphery. This is not a model I like. We've tried it. I've been in companies that have done this. It's not been great. Another way is to deport your privacy. So ad-supported fat PC apps, you know, if you're not paying for the product, you are the product. Turns out, if the product is you, we can use open source and we can drive this. And today I'm thrilled to see that KHGML has totally dominated the market and is winning a Linux desktop technology breakout. Like we want to be, you know, 90 plus percent, it's hard to find a proprietary web browser. In the same way you can't really find a proprietary compiler in today's world or a new proprietary language. This is awesome, except for the fact that it's all paid for by deborging your privacy through search, right? And sadly, LibreOffice doesn't have a search or, you know, I like this not to scale there. It doesn't seem like selling your document privacy would be a good business model that people would, but maybe there's some innovation we can do there. There are sort of abundance models like this donation to donation funded development, which is a great way of doing things. You know, if we build it, they'll come and they'll, you know, we need a big brand and a real big project to make that scale. But this gives you free money. It comes through the door every day and that's brilliant. There are some downers there. Some people think they have to pay to download. Some people think they're buying a product. So maybe they get a liability. There's discussions about who you employ because the whole community creates this thing. And then some people have the privilege of working for that. And there's relatively low conversion rates. I'll show you some of the numbers later. It's, okay, maybe it's 15% of the dollar yields versus just selling convenience. So then there's convenience sales, which is like what TDF does in the app store, making it easier to download and maintain and secure an update. And there are lots of open source projects that do this. And here I will whiz through a whole load of other models. You know, enterprise binaries, selling exceptions, scarcity of bandwidth, selling CDs, scarcity of confidence, scare wear. This actually exists in the open source world. There are people out there terrifying people about compliance, complimentary services, set up in docs and know how. There are some people that sell software without the, you know, like it's all open source, but you just need this massive lump of proprietary data to make it work, which is basically AI in a box, right? So, you know, beware of the open source, but you know, the data is under some weirdo license. One of the things I want to point out here is that even the people whose business model I radically disagree with are often really good people that want to create the minimum scarcity necessary to drive their product development and to make an impact in the world and improve free software. But I'd like to point out that some degree of compromise is inevitable here. So we can mix and match those things. I'm slightly concerned that even the apparently massive projects with way more people than you can imagine possible, 2,000 committers a year, you know, have problems. You know, they're under strain. The kernel people are complaining that they don't have enough, you know, resource for maintaining Linux kernel. And, you know, that's pretty concerning. Bob Young says, you know, you make money by branding. I think I've been through this many times. I'm sure many of us are asked, you know, how do you make money selling free stuff? And that's a good question. I think, you know, it's a really good question. So I'd just like to look at one of these scarcities. So this is a monetizing the artificial scarcity of convenience. Or put another way, selling Libre Office in the Mac App Store. So the convenience here is that you click once, you pay some money, and then it's secured, updated automatically. You know, all of it's done for you. And of course, you can get the same software, actually with even better software, with the Java database built in, et cetera, if you go to TDF and just download it, right? So you really have a choice. And it tells you that. So, you know, we're not trying to fool anyone. We are selling convenience. And we sell it for a price. And we make it scarce. And this is all built on a work that Calabra has done over, well, quite a long period. And here's how it looks. So in 2015, we started doing it with this dialogue. And the dialogue would encourage you first to donate to the Document Foundation. And second, to consider buying the Enterprise Z equivalent, right? To support Project and so on. We've got about 10 donations a month. And we've got about 50 purchases of the upsell. Now, interestingly, we still get, like, some of these numbers are very hard to gauge. The value of the brand is very difficult to ascertain because we never tried to sell without also selling LibreOffice, I think, or LibreOffice. So, but it's not a very, very perfectly pure test. But this is quite interesting. We got some like 12,000 downloads a month. And out of that, we got around four Euro cents per download. Which is quite surprising, as a thing, because at the time we were earning about 800,000 Euros a month from around 20 million downloads, which is about four Euro cents per download, which is quite interesting. Now, after a time, we realized this was not necessarily the best thing to do. And so we, and it wasn't doing well for TDF, although we donated 15% of our income to TDF. It's actually a net loss, if you consider the equivalents there for them. So, well, we thought we'd try something else. So what about charging money for convenience for LibreOffice vanilla in the app store? And so at zero dollars, we got 12,000 a month. At three dollars, we got under, you know, 650, 700,000. And at 16 dollars, we got, well, many fewer, right? So the market demand curve, as the price goes up, you would expect the demand to go down. And I'll show you the real-world data, which, you know, you have all these nice smooth curves in the economics literature. Yes. But when you look at the next graph, all you'll see is the yields that we get from actually charging for something that's non-zero and not asking for donations. It's around six times higher for 12,000 visitors. Okay? And of those 12,000, maybe only 1,000 bought something, hopefully, a large proportion of the other 11,000 went to TDF and donated four Euro cents each. That would be the hope, right? So hopefully, we can't measure that. So here is the market demand curve. There are several things there. I would, the best numbers are this donate yields and units. So the yellow line and the green line. And so the yield is essentially the money you get out of it, if that makes sense. So if you increase the price very significantly, many fewer people buy it. So if we're here at $30, which we tried, and this is the threshold where people start complaining instantly about the price. So we didn't go for $100 to see how the graph goes out there. But it's interesting that it's going up here. Like you get more yields, like still plenty of people buy it. It's surprisingly flat, the curve here at the bottom. So you see the green line here is five people a month buy it, five people a day buy it for $30, right? And that gives you quite a lot of money here, right? Does that make sense? Are you following me? But the surprising thing is that whilst as you increase the price, the demand goes down. There are these quite chunky things here. And this is quite a lot of data. This is not just an artifact. And so you see that the yield goes up here. You know, like there is something funny about human beings that they like things priced at certain amounts. Okay? So people have a mental model that a chocolate bar should cost this much. And if you make it cost a little bit more, you sell many fewer chocolate bars. And the yields is way lower, right? And this is partly why, I mean, you can read lots of outrage about shrink inflation and making chocolate bars smaller and, you know, slicing bits off them and so on. But there is a reason for this. It's so that it fits people's mental category of what something costs. So it's interesting. And you can see that there's an optimal price here that's just somewhat under $10 for maximizing yield. And you can see other places where it's, you know, there are spikes. Cool. Well, that's particularly irritating. Does anyone have a laser pointer? Or maybe I should move the mouse over here. I'm sorry. Here we go. Better, better, better. So you see here, so here is five people a day buying something. And if they buy it at $30, then we get more money than if we sell it at $25 and we still get four or five people a day. And we just get a lot less money, right? Which is not that intuitive that raising the price increases the yields. But it's also not particularly intuitive that around here, around $8 or so, there's a local maxima. So if you look at the green line, which is probably the best line here, like actually you can sell more units if you price it here, which is just really not intuitive. As I say, there's quite a lot of data there. And I think it just fits people's mental perception of what something should cost or what is cheap or what is a bargain and better somehow. And so yeah, so you get just a much higher yield at certain points, which is, yeah, it's not your traditional market demand curve. So no, I just thought that might be interesting to people. We live in a non-linear space with real human beings that behave unusually. Now for something totally different. So one of the interesting questions in economics is why the West conquered the rest or why the industrial revolution started in Britain and moved rapidly to Holland and then more slowly to Europe and then the rest of the world? This kind of question. I wasn't there, I don't know. And neither do these people, but they write books about it endlessly. And there are lots of explanations for it. And the primary explanations are these, institutions, strong institutions, inclusive ones, culture and geography. And it's of course clear there is no one explanation for any complex phenomenon like that. And it's not trivial to map property rights or say geography to what we do as a project. But nevertheless, it's interesting nonetheless. One of the big explanations for why some cultures seem to grow and invest and succeed and thrive, if indeed you think those are good things, is about innovation and encouraging people to innovate and put stuff back in. One of the main problems of institutions is that they can become extractive and exist to serve themselves only and to pull out or serve some portion of the population. And the obvious case here is monarchy. Like the king decides everything that happens or the queen. And so it doesn't matter if you innovate in some way that's disruptive. And almost all innovation is disruptive. The examples they give and do read some of these books, Why Nations Fail is what I read most recently. Not totally convinced, but it's interesting. So for example, you invent an amazing new spinning machine. It speeds up the process of spinning cotton. It's absolutely wonderful. It works. You can show it to people. But in order to make money on this, you need to get a monopoly or a patent or whatever. And you go to the queen and you show her this thing. It's amazing. It's awesome. It's brilliant. We can do this. And she says, well, but my mate owns the spinning thing that we do already and it employs lots of people. And they support me politically. And so the number of times that many of these innovations have been independently invented and then squashed because they offend some existing order or do something is legion and tragic. And so the economists tell us that what you need a strong property rights and a fair and inclusive system where everyone has a stake, that they can be confident that if they innovate that they can then benefit from it is absolutely vital, as well, of course, as keeping balance. And there are lots, lots of things are there. Yeah. Anyway, do you read the book? I think you'll find it really, really, really interesting. I think another part, and I think the inclusiveness, so inclusive political institutions, then it's thought lead to inclusive economic ones that then generate innovation and wealth and so on. And there's lots of other examples of semi-market economies. So in China, for example, this man realized there was lots of building going on. And the trajectory of it was going to be huge. And he set up, he thought people are going to need cement, lots and lots of cheap cement. So he invested a fortune, set up his own cement company, built up a big business, making really good cheap cement and started to take over the market. But unfortunately, there are a whole lot of state-owned companies that employed lots of people that made inferior cement that were politically connected. Who crushed his business, nationalized it, took it all away. And this doesn't stimulate other people to innovate, let's say. It's not necessarily a great way to go. I think another thing that's important is culture. One of the controversial, let's say, but perhaps important things is we have in our statutes is meritocracy. And meritocracy, I'd just like you to reflect, not on the dignity and depravity that's equal across all people. And we're all wonderful, wonderfully created in the image of God, I would argue. But when you come to build a culture, who you empower and who you listen to and how you wait decision-making is really important. And how you build that culture. And I'm sure it's obvious to many of us instinctively we feel that the person who is hyperproductive and constructive, and these are hypothetical extremes. I don't know anyone that's all of those or anyone that's all of those, right? But this does happen. That you want to encourage and empower more of the people on the left, perhaps, than the right. And that the qualification or the way that we balance input should be based on some metric, such as the volume of time spent or contribution or constructive. You can think of many different ways of calibrating this. And of course the membership committee has an important role accepting members. But also we have a role as a culture to try and listen to people who are expert in their areas. I just leave that as a thought. I think becoming a shouty-ocracy where the loudest shouter is the one that is listened to is really bad. Of course becoming an English-ocracy is also bad where we listen to the people who can most speak, easily speak, that's true. Or a developer-ocracy where it doesn't matter but you have to be a developer. This is bad. We need inclusive institutions that empower those who do the work and where we listen more to the people who are contributing constructively. And I think that's a really important part of our culture that we need to defend. I think it totally has talked at great length about makers versus takers. So I'm going to skip over this. But only note that there's a lot of tragedy in the world at the moment around this. The Terraform movement recently includes a great long helpful list of all of the companies that have recently moved their software to be non-open source. And that's really sad. There's a bit of a taker epidemic in the internet companies we see today. And of course there's lots of diverse community challenges. Working with people of different opinions is exciting and we have to bear with each other and so on. So there you are. A heterogeneous wonder through lots of topics. Again with no time to talk but I would love to speak outside. I'm still around to late this evening. We should talk about these things. We should have space to discuss all of these things. But our industry is really changing. Enabling people who don't contribute to benefit from our project at the expense of those who do contribute is in my view not tenable. We need some degree of scarcity. I don't know what you think it is but there needs to be something. Regulation is coming to eat us alive. I haven't really heard much about the CRA, the PLD, the AI Act which is coming at the conference. But these are big topics that will have the potential to radically change how we go about our business and develop software. We should have inclusive governance which is stable and doesn't change arbitrarily where the rules are clear to everyone and all of the people are included given my caveats about culture and meritocracy. And we should be really aware of breakage. We have quite fragile software ecosystems that build all of these things that try and minimize scarcity to maximize openness and freedom and goodness in terms of delivery for everyone. So I think we need to be slightly careful. And that is my conclusion. Thank you for your patience. I have rambled enough.