 Okay, while you're finding your seats, I would like to introduce you to somebody. The person on the screen, his name is Kawan, and he is one of my Brazilian God children. He's not really part of the talk, but I just thought I would show him to you, because he's usually the backdrop of my computer. My talk today is called 50 Years of X, a computer odyssey. It's a combination of my time in UNIX and computers in general. But I wanted to pick out why some of the things that were developed in UNIX and Linux systems over the years are really important from a computer science standpoint, and from a standpoint that a lot of people may not think about. I have done a lot of things in my life, one of which was to help to protect the Linux trademark, and the Linux trademark is the trademark of Lina's tour vaults in several countries around the world, and you can get it by going to the Linux Mark Institute for free for any legitimate purpose. Now I am really old. I was born in 1950, which makes me 68 plus years old, and this year I have been programming for a half century. I've been using UNIX since 1980, and I met Lina's in 1994 and saw Linux for the first time. We'll get more into that later. And I've been doing a lot of work in Latin America, particularly Brazil, and you'll see more about that too. I've had a wide variety of different jobs with different companies, some of them very large and some of them very small. But one of the things I'm pretty proud of is I'm pragmatic. A lot of people say that Linux is a religion, maybe so, but I am an atheist. And I'm going to be packing 50 years of history into this 50 minute talk, which is I'm going to have to go kind of fast, I may not cover everything, and maybe some people will have different ideas and different viewpoints, and that's fine, but you can talk about them in your talk. So this is a lot of anniversaries this year, and I may not have covered them all, so if I left your group out or something, I apologize, you can send me hate mail later on. But basically, just looking at 20 years of the Linux Terminal Server project, I have beers with those guys every once in a while, 20 years of the Linux Professional Institute, where I'm the chairman of the board, we're very proud of that. We have 150,000 certified people in 180 different countries. 25 years of the Linux version one of the kernel, in 1994 version one of the kernel came out, and if you're doing operating system design, version one typically means something that's good enough for general use. And we'll see how that flowed a little bit later. Bay Wolf supercomputers were first developed in 1994. People around the world were fascinated, and I remember that I was hauled out of class to look at this happening. Woodstock happened, big newly old, because I met him as a 21 year old university student, and you could count the disk space in megabytes, and when I say disk space, a disk of 180 megabytes operating systems, so we can lock in our customers. It was quite frankly more, let's put this feature in so we could make this miserably small. Now in those days, most software, you have to advertise your software, you have to write documentation for your software, you have to go forth and so on. Whether you had the source code or not, and most of the time, you did. The source code was, they determined what they wanted to hear. Deacus also ran this software, and there was no copyright and no patents allowed on it, so once you had it, you could make as many copies as you wanted to, so I go to the right. They were the users and the developers at the same time, they used their own operating system, how refreshing, and the first couple of editions of it were written in assembly language, and not even assembly language hosted on the same machine, because the PDP7 was too small for that, so you had a cross-assembler from one other computer into punching the tape for the PDP7, and they loaded that in and ran it, and of course like every program, it immediately halted, and then they used the switches to see what was going on, and after they got working well on the PDP7, that kind of ran out of steam, so somebody donated a PDP11 to them. The group that donated that PDP11 was the one group inside of Bell Labs that had unlimited money, money out to Wazoo, and could give anything they wanted to. What company, what department was that? Engineering, research, nah, it was a legal department. So lawyers do have a use, giving money to computer people to buy computers, so they got this PDP11 and they rewrote the entire operating system again in assembly language, and Dennis at that point said that's a lot of work, I'm going to write a better language to do that in, and so he wrote C. That's the type of guys these were, right? Oh, I just, you know, the language I read in the operating system, there it is C. And eventually, as they kept moving from computer to computer, it became more and more and more portable. And it was distributed in a source code at first. Ken would take a tape, put it underneath his arm, and go off to sabbaticals, at universities, and stuff like that, and of course he wanted to talk about operating systems, because he loved operating systems, and he just wanted to use his favorite operating system, and one of those places he went was the University of California, Berkeley, the home, the bastion of leftism. And you know, and this allowed, the operating system being in source code allowed other people to contribute to it too, other universities, even companies and governments and stuff like that could contribute because they got the source code. But Berkeley made this into a real project, BSD Unix. However, it was never really free. All those people who have got the idea that Unix was free to universities, that was just crap. You still had to, you still had to get the license, but for universities like Carnegie Mellon, Columbia, Berkeley, research universities, the price was $350, you know, for site-wide source license. And oh, by the way, you're not allowed to let your students see the source license, of course, you can see the source, and of course, everybody just blew that off. However, if you were a small, little two-year technical college, like Hartford State Tech, the price was $160,000 per CPU, and oh, by the way, you had to send them the serial number of your CPU. Now how many of you out there did a serial number of your laptop? Yeah, that's what I thought. Slackers. And then you also had to find the person inside of AT&T who would actually take your telephone number and your $160,000. And if that computer went down and was broken, you had to call them up again and say, oh, I'm moving it off of the broken computer to the one that's working, and here's the serial number of that. So naturally, this was great, you know? I mean, large companies could afford to do this, but not real mortal human beings. Now the interesting thing about UNIX was that Ken and Dennis were not really that interested in efficiency of the computer. They were more interested in the efficiency of humans. They'd have the same operating system move from computer to computer with the same commands and the same programming interfaces and stuff, and just use the compiler to take care of the architecture. And this was an interesting switch, because up until this time, you had to do significant training to train your person on the next computer. And they were also doing this just for fun. They weren't paid to do this. And a matter of fact, Doug McElroy, who was the director in charge of them, who hired them into Bell Labs, often had to run interference with the powers that be to allow them to continue their project. Doug McElroy, by the way, is the person who invented pipes and filters. And so if you like the UNIX command line, please say thank you, God, to Doug tonight, even though I don't believe in God. Okay, now other types of UNIX contributions, the shell. You know, an individual shell that was built outside of the kernel, so you could replace them with other shells that you wanted to. Pipes and filters from Doug McElroy. And he wrote some of the first original UNIX commands, just so he could show Ken and Dennis what he was talking about. Puttable network file system, NFS. I still remember the year of the network file systems, where we just had so many file systems coming in. People say, this is the one, and then we got NFS. And the X-Windows system, which was done by MIT as part of a project Athena. And, you know, to create these diskless workstations all running off of one server. And many, many more UNIX contributions, even though they weren't all out of Bell Labs or even University of California, Berkeley. But then the beginning of the end. Sun Microsystems, a little company, starting up. They got their computer board from Stanford as a project, and they wanted to have an operating system to put on top of it. Well, there was CPM, but that was a joke and would not have used all the power of their system. They actually went to Digital Equipment Corporation and asked if they could license VMS. Fortunately, they didn't. And because that would have been a problem too. A lot of that was written in bliss, which is very dependent on the VACs. And then finally, they went to University of California, Berkeley, and said, could we use Berkeley UNIX? And they got a license from AT&T to do that, and they started to do their port. But they wanted to do a binary only license, not because they wanted to lock their customers in, but because AT&T said, if you only distribute in binary, then your customers do not have to pay the $160,000 per CPU, and they don't have to tell us a serial number. Yes. And they came out with two licenses. One was a two-user license for a one-user system. Now, why you need that? Well, of course you need that because how else is UUCP going to log in and transfer files if only one person at a time can log in to the system? So you need the two. Or how can root log in when the system is jammed and you already have a user logged in, right? So you need two. That was a big argument. And then there was the unlimited user license. And some love this because they could take their work station, put an unlimited user license on it for like $350, and then they could have 10, 15, 20 people log in over the ethernet to use their work station. And other companies started to follow this. Now, digital had actually been supplying support to UNIX systems for a number of years, but it was hardware support. If your hardware breaks, we'll fix it. If you need to have a device driver written, we'll help you do that, but we won't do it. So we were providing the hardware support to universities and other places, but we were not distributing our own version of UNIX at that time. Sun beat us to it. Most of the companies that did this use BSD, not System 5. Why? Because System 5 was basically a swapping system at that point. It was not using demand page virtual memory, which the BSD people were. This was very important. System 5 only shipped two compilers, a C compiler and a Fortran 77 compiler. BSD shipped three compilers. They added Pascal to that. But the most important thing to a lot of people was that System 5 only supported UUCP and Berkeley supported TCPIP. And that was a real killer. The biggest problem that BSD had was that the University of California, Berkeley didn't have a mega-million dollar marketing budget to say, buy BSD versus AT&T having a multi-billion dollar marketing budget that said System 5, the right choice. And sure enough, every month, actually every week, for a number of years, there would be a two-page centerfold buy System 5, the right choice. Now at this point, UNIX systems were coming out in binary for reasons I just told you, right? It's the license fee. You can still get a source code license, but if you wanted the source code for one of these binary versions, you had to dig up the $160,000 and the serial number. And then you had to go back to the company to major UNIX and say, we want the source code for it and that was even more money. So most people just went with the binary license and the binary copy. This irritated one person, Richard Stallman. Notice how many asterisks are behind his name. And he decided he was gonna create this project called the GNU Project which stands for GNU is not UNIX. And you all know how that turned out. It was gonna be a complete operating system with a source code available. He wrote a license. He started the Free Software Foundation to help keep it funded. He had a cute GNU and things like that. And he started off by writing Emacs. And a lot of people say, you could have stopped with Emacs. It does everything an operating system does. It measures its own memory, it has its own processes. It does everything. You dive into Emacs, you swim around in Emacs. When Alan Cox took Linux, an early version of Linux, and ported it to the two megabyte palm pilot with no memory management, he got it to work, except he could invoke VIM, but Emacs could not be invoked because it was simply too big. But then Richard went on to do other things. Compireless, waste, utilities. Notice he's developing all this code which can run across other operating systems. So people who are using that developers' stuff could have the same code on all these different operating systems. It was like Unix, except it could even run on top of VMS or MBS or a bunch of other time-sharing systems. And this was very smart because he had started writing a kernel as a very first thing. What was he gonna run on it? Right, who could use it? And by the time it got to the point that he had applications on it, he'd have to rewrite the whole kernel again because it'd be hopelessly obsolete. So he started with things that everybody could use and he attracted more and more and more people to the project. And it also allowed a bunch of small companies to start up. Companies that would take this GNU software and other free software from other groups and companies and stuff and put them into CD packages, things like that with little books. Primetime software was one of them. Jim Joyce out in San Francisco had the Unix Bookstore where you could go in and he had candles lit and he served you white wine. It was great. And of course, O'Reilly story it up and had a series of books about Unix systems and he had all sorts. And also Cygnus was another company that started up and said we're gonna provide support to the compilers in the GNU suite. And they started selling support. I would love to have gone to work for Cygnus because always the board of directors met in a hot tub, it would be great. DEC, where I was working, as I said before, was supporting Unix with DEC hardware, the PDP-11 and the Vax. I joined in 1983. I was the 16th engineer into the group. And we brought out around 4.1 base BSD system in 1984, two years after Sun. Ultrix-11 was for the PDP-11 and Ultrix-32 was for the Vax. Now, as Kyle mentioned previously, the Unix wars started up because as these vendors came along, they were competing in the Unix space. And that market was made up a lot of scientific and technical people at first. Universities, Los Alamos labs and other national laboratories, educational people. Then there was the commercial side. Now, back in those days, the scientific side represented about 16% of all computing. Commercial, which was banks and companies and things, that was 84%. And digital concentrate is VMS traffic in that versus Unix was more for educational scientific. And a lot of other companies did the same. The databases back in those days were using raw partitions on Unix systems because there was no synchronous write and to do files so that you could never tell when the data was actually out on the hard drive. So databases would do their own file systems and their own file management, their own buffer management and things like that. They were a lot like Emacs. And the Unix wars heated up with two major groups coming together. Sun, who had originally based Sun OS on Berkeley, BSD and AT&T with system five decided to join together and in fact lock out all the other vendors of Unix. To do that dirty deed, Sun had to convert their customer base from Sun OS Berkeley based to Solaris, which was system five based. That took them about two and a half years. And the first releases of Solaris was so inefficient and so bad compared to Sun OS that everybody called it slow Eris. And it was the only port where you did the port. It was all pain and no gain. Deck, HB and IBM really weird bed partners came together to form the Open Software Foundation in 1988 and created a standard set of APIs as Kyle mentioned. A lot of them were based on POSIX but there was a lot of upper level APIs also based on other things that became the Unix distribution. And they created a test suite to test that to see if your system was really compliant with the APIs and they had a sample implementation of code that they brought from all over called OSF one. And you had the ability to either use that code or change your own operating system to be compliant. That was really up to you. A lot of companies like Hewlett Packard and IBM changed their operating system they had already. Deck decided to take the sample code and engineer it to be more efficient and to base it on that. But if you pass that test suite you could be called Unix. And there was a lot of people even VMS got to the point where they could pass the test suite and they could legitimately call their system Unix. Now in 1991 Unix Systems Labs sued BSDI. BSDI was a little company that was using a variant of BSDI, selling both the source code and the object for $95, oh sorry, $195, almost $1,000. Now of course back in that day $1,000 was a lot of money but it was a lot less than $160,000 plus all the other charges you had. And USL did not like that so they started to sue this little company and the lawsuit dragged on for years with the University of California Berkeley standing in the background by giving them encouragement. At the same time the people at Berkeley were changing the software to scrub out every single line of AT&T supplied source that they could. And so eventually in the end there were only 17 files that the judge said oh yes that looks like system five code and at Berkeley said really? They're gone. And that created BSDI which is the basis of a lot of the open source BSDs that we have today. But unfortunately this was 1994 that this was happening and in 1991 there was this young man in Helsinki, Finland who had gotten a brand new 386 processor for Christmas and the operating system that he came with it, he was not very happy with that. And one of the major reasons why he was not happy with that was that that operating system was missing something that the processor could support. Now what was it missing? It was missing demand page virtual memory. The same thing to make all the UNIX vendors typically go to Berkeley instead of system five. Why was it missing demand page virtual memory? It's because the company that wrote that piece of shit out that operating system was still supporting its customers that had 286s and 186s and because that code still ran on the 386 although slowly and poorly they were not gonna support that until a lot more people had those 386s. In the meantime our hero in Helsinki, Finland said hey I could write a kernel for the operating system just for fun and you know the history, he sent out the thing they almost called him freaks but he said it was Linux and stuff like that. I'm gonna have to speed up my thought. But why could the Linux kernel happen at that time? It could happen at that time because there was powerful cheap computers coming out. Earlier you had to spend a million and a half dollars to get a computer, now you only had to spend a couple of thousand and the 386s were kind of coming to their end the 486s were coming in. So he said okay I'm gonna do it, I have this computer at home I can do it. The faster internet to the house dial up was not the internet norm anymore. You were starting to get DSL, you were starting to get cables, you could have reasonable internet to the house and people wanted to work at this at the house because they didn't want to work at it at work. There was much more information online about how operating systems actually worked and code from different places that you could pull in and the World Wide Web was maturing and wasn't just useful for porn anymore. So in 1994 which is 10 years after George Orwell's 1984 system vendors and seeded the desktop to Microsoft. They were doing these servers, we're gonna unit servers and proprietary servers, we're just gonna make the desktop Microsoft and Apple. Windows NT was eating into the server space and they were starting to become afraid of this that Microsoft was gonna own the entire operating system market. Or Wiley actually started publishing books on how to program Windows NT and Microsoft. But then as Linux came in, he kind of pushed those away and started bringing back his books on Unix and just calling them Linux. But I don't blame Tim because he had a business and people depended on him and he couldn't allow O'Reilly to collapse because those people would be out of work. And sometimes we have to remember that when we started talking about nasty things about companies because they employ people and there are people that count on them. They should all be doing Unix or Linux but that's a different subject. And in 1994 version 1.0 of the Linux kernel was released and so we went there. In 1994 in Deacus in New Orleans now there are two cities in the United States that are adult Disney lands. One of them is Las Vegas and the other is New Orleans. And I was going to New Orleans to Deacus and when I went there I found this nice young man with wire rim glasses and sandy brown hair and I found out he liked steamboats and I took him out on the Natchez and that is him and the left. And I talked to him about porting this project called Linux to the Alpha, a 64-bit system and a RISC processor so he could get the intel-isms out of his curdle. And he agreed to do that so I went back, I got him an Alpha system, I got him some engineers to help him from Deacus, small group in the semiconductor place and I started to learn the power of the free software community where people who had never seen each other face to face proceeded in this project. They went out and they actually bought their own Alpha systems which were not cheap to be able to help out with this and that was amazing. And at that time, GNU Linux was just beginning to explode. I might say GNU Linux, I'll see that, but it should really be called GNU, MIT, BSD, da, da, da, da, da, da, da, Linux. Right, that's the theory. And just like Kyle said, it was cheaper than the Solaris Spark, you know, equivalents, but it had everything you needed for ISPs. Remember ISPs, those are the people in between the telephone company and or the cable company and they gave you services, remember them? They gave ISPs everything they needed at a fraction of the cost and allowed people to repurpose their all 386 processors to do other things when they had their new 246 processor on their desk running Windows. But at the same time, this thing called Beowulf Supercomputers came out, we'll get to those in a moment. And in 1998, databases started to port because they started to realize we can run our database engines on this thing now and save our companies, our customers lots of money. Embedded systems in the year 2000, almost overnight, Linux became the most used operating system in embedded system design. Why? Because those companies were used to putting their embedded system operating systems in, but all of a sudden in this timeframe, all of their customers said, you know, I really like to be attached to the internet. What? Oh no, you can't take our little proprietary embedded system and expose it to the internet because it's wild out there. You need security and you need a whole bunch of, and you need a TCPIP stack and those aren't easy. And the customer said, yes, and we want them in these ARM chips. No, you would use Intel surely. No, maybe Motorola. Yeah. And so all the embedded system people said, oh my God, what are we gonna do? Wait a minute. This is operating system already runs on all these chips. It already has a TCPIP stack. It already has multitasking. It already has all this stuff and it's free to a certain point. And so they started using Linux. They didn't turn away from the proprietary systems. They still supported them, but they used Linux a lot. I promise you I talk about Baywell supercomputers. In 1994, supercomputer companies were dying. Supercomputers took millions of dollars to create and then millions more to support and manufacture and they typically only sold five of them. One to those agencies that we dare not say their name. And the other 40 universities that never paid for them. And they were just so expensive and then something like the Cray One. That's a picture of the Cray One there. University of Toronto received one of those. It actually generated enough heat to heat 20 homes in the middle of the winter just to keep it cool. And all these companies, Cray, ECL, CDC, they were dying but people need supercomputers. They really do. There's problems we need to solve like weather. How many of you remember mercury capsules going up? Yeah, you're the old ones. The mercury capsules, they put the astronauts up there they fill it full of fuel. The astronauts are sitting on top of this really huge bomb. They say, we're gonna set off halfway through the countdown. They say, weather's coming in. Well, you gotta stop the countdown. All the astronauts sitting up there waiting for the thing to go. Sorry guys, weather's going out, weather's still here. We missed our window, have to take you out. That's because they gathered all the data to predict what the weather was but it took 48 hours to compute it. And that meant they could tell perfectly what the weather was gonna be 24 hours ago. But by shortening that computation time by using more and more computational power eventually it got to the point where they could tell what the weather was gonna be right now. And then shortening it even more eventually they could tell perfectly what the weather was gonna be 24 hours in advance. This, the thing that helped them do this is something called fluid dynamics. And if you're like me, you hated fluid dynamics because it had math out the yin yang. And that was one of the reasons why I went to programming. And this, fluid dynamics isn't everything. It's in weather, it's in the tides, it's in glass. Glass is a liquid, you know. And all these things is done fluid dynamics. And they needed a super computer for this and two people, Dr. Thomas Sterling and Donald Becker created this concept of the Beowulf super computer parallelizing a lot of computations of bringing the answer together. And inadvertently, I influenced this a bit because I was visiting a friend in Hawaii and his roommate was a computer science student that said, here kid, here's a CD with red hat software, what is this? Oh, kind of interesting. That guy's name was Pac Goda. And eventually he went to the Los Alamos laboratory and that's the Beowulf system he built with that red hat CD I gave him, called Loki. And Loki he was using to see what would happen if meteors crashed into the earth. And first he was using New York City for that, then after 9-11 he started using Los Angeles. But now, today, all 500 of the fastest supercomputers in the world run Linux. It used to be that two of them ran Microsoft, I think because Microsoft paid them. And if you look at that, the Gordon Bella Award, that was particularly bad because Gordon was in charge of the Microsoft Supercomputing Program. In there, you can see Donald Becker's name, you can also see Patrick Goda's name as winning prizes for best price performance in creating supercomputers. Gordon was a good sport about it, he kind of laughed. Now, we keep talking about the years of the Unix desktop, or that should be Linux desktop typo, sorry. But at first it was only one standard of desktop and that was DOS and Windows. And I remember my father writing down on a piece of paper, when you see the word login, put your name, when you see the word password type, he's writing this down on a piece of paper. He, you know, without that notebook of written paper he couldn't use his Microsoft Windows system. But he could use it because every Microsoft Windows system was the same. And the thing that broke this was cell phones and browsers. Browsers broke the concept that you could write down every little thing you did on the computer because when the browser changed, then so did your notebook was useless. And cell phones did the same thing. So today people actually understand a little bit what they're doing when they use these devices. The other thing that really keeps killing the Linux desktop is lack of games. We're slowly fixing this. But it's always been amazing to me how many people say, I'd love to use Linux, but I need my game. And people say, well look at all the applications for Linux and sure there's like 250,000 different applications if you go to places like GitHub and Sourceforge does. However, people are only looking for five, the five applications they use. The first four are relatively easy to find because tens of thousands of people use them and they're on Linux. It's that fifth application, and it's typically a game, that's not there that they just refuse to give up their Windows system for. The Linux Terminal Server Project in 1999 was a project that allowed you to take cheap and old computers with relatively thin software when it's so they can be very old and very cheap and hook them up to a server and basically be able to use the software for the server. And so that's 20 years old now. Now another threat that's coming to the Linux community is Apple's customers. And Apple, I have to give them credit, they took computers and turned them in from a business engineering scientific thing into a consumer product. And people started buying computers because they looked so pretty. And they took FreeBSD and hid it underneath of their graphics. And they were headed towards domination of the phone market, but Android foiled their plans. Why? Because Apple was using the same strategy that they used for years on the desktop and the laptop. Apple only has one partner, Apple. If you're not Apple, you don't make any money from Apple. Microsoft used the other philosophy. If you could even say the word DOS, you could put our software on your system. And so everybody else who made hardware advertised Microsoft, helped Microsoft and put the little Microsoft in size stickers on their computers. And that ended up with Microsoft earning 90% of the desktop, Apple earning 7% and a variety of other systems earning the other 3%. I predict that in the cell phone industry, Apple will own 7% and Android will have more. And Android will have the rest. I've been told I'm out of time. I'm sorry. I've gotten through 31 of the 157 slides. I have had a call to keep going. Okay, I have been told I can keep going. Thank you very much. So Android will have the rest. And why? It's because the hardware market keeps developing new ideas for different designs of phone that a lot of people say, yes, I want that phone. I want it because it has some feature that I really like. And maybe the feature is price or it's distribution or it's channels or whatever. But Apple will be very happy with their 7% because it'll be very profitable because people pay outrageous prices for them. And they'll be a very wealthy company. But now we have the cloud. And just like Kyle was talking to us beforehand, all of the OS and utilities are hidden behind that cloud and you don't know really what's there. And you're gonna be leased this software forever. Now, isn't this great for certain companies named Microsoft? Because Microsoft, 34% of the desktop software in the United States is pirated. If you go to Brazil, 84% of the desktop software is pirated. If you go to China, it used to be 96% of the desktop software was pirated. It now came down to 84% because China's under pressure from the World Trade Organization to respect intellectual property laws. Vietnam is still 96%. Because if you live in Vietnam, you make approximately, or at least you did, make approximately $4 a day in which you could support your whole family. But you were then told to be in business, you have to have a computer, running software, and Microsoft Office would cost you $400. That's 400 days of not doing anything like eating and sleeping and buying clothes and stuff like that just so you could buy this shiny plastic disc. And of course, the Vietnamese knew that shiny plastic disc didn't cost $400 to make because they could go down to the corner store and buy it from their software pirate dealer for 25 cents. Quite frankly, if I was in their situation, I put a patch in my eye and go, are, are, are baiting. But with the cloud, I have to get a little bit of money from everybody every month. And there's the business model, right? And I don't have to distribute CDs. I don't have to do any of that stuff. The life is sweet. And I can give you the illusion of privacy and security and backups and stuff like that. And if you believe that you have privacy in the cloud, I have a bridge to sell you in New York City. We need to take control of back from these companies of the cloud. We really do. And there were some interesting projects out there that helped do that. So the summary of history up to this point is that UNIX created portable code and portable people and set the APIs for the future. GNU Linux relit the desire for control of your software. That maybe you don't have the expertise to fix the problem that's making your software or your hardware obsolete. But if you get enough people together, they will work together to fix that problem. And that's important. Flush and free culture is another form of capitalism. Don't be fooled that there's only so many economic models, communism, socialism, capitalism. And if somehow free software is socialism or communism, that's crazy talk. It's capitalism. And it's known as cooperativism, where the company is owned by the people who make the software or it's owned by the people who use the software. And they are in control of it. And that's what we need. I remember that my managers laughed at me because I was supposedly marketing digital UNIX. They laughed at me about Linux. But then all of those managers now work for IBM. But this talk would not be finished unless I talked a little bit about the future. We're going into a situation right now where we're starting to get a lot more users than we even know we have. I was sitting in a bar the other night. I was waiting for a political rally to start that was gonna be in the bar. This guy next to me starts talking about his Linux system. Really? Yeah, I love Linux. I'm using it and do all this stuff on it. Oh great, you ever hear about Linus Torvalds? No. Know anything about the Canoe Project? No, but I really love Linux. He's just using it like a tool. Like people use their cell phones. They don't know that Linux is underneath the van droid. They have no idea about it. It's just a tool. And maybe they shouldn't have to know about it, but maybe they should. And maybe they should understand the basis of the software underneath. Some companies say they love open source. This is relatively new for some of these companies. And I want to remind you that in the ancient Greek there was three types of love. There was Agape, the love of God for humans, the love of mother for their child. The all forgiving, all giving love. Then there was Phelous, brotherly love. We get the, the city, Philadelphia, philanthropy, philosophy, all those types of loves of human beings for each other. And then there's Eros, the dirtiest love, the most basic love, the sexual love, the love, the door, door. And I think that some companies are at the Eros level. Don't bend over in the shower. We need to wear condoms when we're around them because they will screw you over if you allow them. And we need to protect people from their love. We are close to world domination. 500 are the fastest computers in the world. The most used operating system in complicated embedded systems. Over 60% of the service system is built and worked even in the cloud, okay? We are close. The desktop still eludes us a little bit but we keep winning further on that. Don't give up hope. But we need to get people to understand that there are business models around free software. One of the things that gives me the greatest pleasure in the world is having people come up to me at conferences like this and say, Mad Dog, in fact, conferences like shows, like Seabit, Mad Dog, I listened to you 10 years ago and now I am the chief technical officer of my company because I brought free software into my company and we saved so much money and we gave our customers better service. Mad Dog, I listened to you 20 years ago and I own my own company. I employ 60 people. Thank you for saying this. Mad Dog, I listened to you a number of years ago. I went off to start at a cloud company called Digital Ocean and I recently sold it for a lot of money. So we need to understand the business models. We need to publicize those business models. You can make money with free software because if people think they can't, then they won't pay any attention to it. It really is what drives the marketplace. And you need to talk about open source of free software every place you go. Where a pen went on you? And particularly a tux penguin. Oh, where a BSD demon? That's okay too, because our real enemy is closed source. That's the enemy. I don't hate BSD. I hope the BSD doesn't hate me. As a matter of fact, I was the person who at the Lin's Professional Institute who said yes, we should be doing BSD certifications. Oh, I approved it, I didn't. Somebody else suggested it, I'm gonna give them credit. But I am working on various projects. I am the chairman of the board for Lin's Professional Institute and we're facilitating open source professional membership programs. I write a blog for Lin's Magazine. You can read stuff there. I'm working on a project called Kanina's Locos which this month we're gonna start bringing inexpensive Singapore computers to Latin America at a very low cost, bypassing the high costs of their import taxes. I'm working on a project called Project Calwan named after my godson that you saw in the previous slide, or the background that helps university students afford university. In Latin America, most universities are free of tuition, but 40% of the students who qualify still can't go because their parents don't have enough money to pay for their apartment, their food, their transportation, their computers, their internet. Project Calwan shows them how to make money taking the information and knowledge they already have and helping small business people run their businesses better with free software. I am now the chairman emeritus of wit.com. We will be disclosing more about that later. And I am working on my retirement project, plus BS. Kanina's Locos is the hardware platform for Brazil's IoT initiative. It's made up of two computers, one of which is this little Singapore computer much like a Raspberry Pi, only better. And another one's called the Polga or the Flea. We named all of our computers after dogs. We did want to name it the Chihuahua because everybody hates Chihuahuas, so we named it the Flea. And we are writing device drivers and things like that as we put more information up on the net. Perhaps you'd like to join us. It's going to be used in IoT initiatives like agriculture, smart cities and health. And it centers on, but it's not limited to the Debian distribution. A shameless plug, there will be Debcov, that should be Debcov, I'm sorry, Debcov 2019 in Curitiba, Brazil, Debcamp, Debday and Debcov. And if you come there, you'll be able to meet my godson, you'll be able to drink beer in my group hub, Dumpart, Urna Rove, and you'll be able to program Kanina's Locos computers. This is my retirement project, plus BS, Mad Dog's Monastery and Marina of Math Music, Microcomputing, Microbrewery, Microwinery, Microdistillery, and Big Chop. And in 2025, I'm going to stop traveling around the world and talk you to wonderful groups of people like you. You'll have to come there and drink beer with me. And finally, it was wonderful to come to Faustem again after so many years. The first time I came here, I remembered that the hallways were crowded and everybody was drinking beer. But that was when there was only one building. And now there's eight buildings and the hallways are still crowded and you're still drinking beer. And I'm glad that that hasn't changed. It is nice to have people come up to me and tell me where we met and how Faust has changed their lives. And I want you to think about something. If you want to see the most important person in this community, when you get up tomorrow morning, you look in the mirror. Thank you very much.