 Hello Bonjour Unfortunately, that's the last Portuguese you're here out of me today Unless I have to ask where the sanitarios are My talk today is 50 years of X Unix Linux and other things and why I feel that's important My name is John Dingau Hall Dingau is Mandovin for Mad Dog And I am the chief executive officer of a company called opt-in as of July of 2017 I'm also the board chairman of the Linux professional Institute, and I'm the president of Project Kawan and I'm the co-founder and advisor to a program called Kanina's Lucas, which we'll talk about later First of all, I am really old Next year I will have been in the computer industry for 50 years I've worked on a wide variety of different systems Computers that Physically would fill this room, but logically a way less powerful than even the simplest cell phone I've been coming to Latin America since 1996 and I have a unique Relationship with you I've been a wide variety of different things in my life I've taught operating system design and compiler theory, and I really love compiler theory And I've worked on very large systems and very small ones and Unlike some people in the free software space. I like to think of myself as pragmatic Which means I use software to get the job done That's what I do. I will give you a warning that I'm packing 50 years of history into this one-hour talk and My memory isn't what it used to be So for those people who are going to argue with me and say no this happened exactly this way or I disagree with you That's fine. You can send it to email There is dev null. This is my record. Oops. I don't need this. I don't need the Wi-Fi This is my record of my dedication to Brazil I Came here in 1996 and I saw the University of Sao Paulo 100,000 students 16,000 PhDs and I saw my very first Beowulf system there my very first physical super computer I saw at the University and I saw that they won many awards year after year in the super computer contests In 2002 I brought the Linux Professional Institute here to start certifications of Linux professionals and the first LPI certs in Brazil In 2005 I became the president project Kawa whose main goal was to create millions of jobs with a secondary goal of environmental use of computers as well as making them easier to use I Don't believe the computers will ever be easy to use Until they can get to the point where I can talk to them and they can talk back to me like I am talking to you In 2011 I started a project called Caninas Lucas to try and reduce the cost of single-board computers like the Raspberry Pi here in Brazil 2012 I became co-owner of the hop and roll beer club where we have 29 Craft beers on tap and I'm going to be spending this evening there It is in Curitiba In 2017 I married a Brazilian and I'm very happy with my partner And in 2018 I became the chief executive officer for super tie I'm going to take you back to the very beginning of computers This is before the transistors before integrated circuits transistors were there But they were physically huge. You could actually walk inside of them and we needed to have air conditioners that were measured in the tons of cooling power The memories were measured in kilobytes not megabytes or gigabytes or terabytes In fact, I remember at a point in my life Why I had to practice saying the word gigabyte and trying to realize how big it was and Now I think about how small it is The disks were measured in megabytes and We and typically back in those days you ran programs one at a time a Lot of computers even mainframes did not really have an operating system You linked the device drivers into your program and then you booted the program a little bit like Arduinos are done today and The operating systems if they were if there was an operating system was very specialized It had to be specialized the architecture of the machine because a lot of the architectures were different We were still experimenting with architectures back in those days and We were also specialized in the type of program or the type of problem we were trying to solve So you would have some systems that were oriented towards batch and winning jobs one at a time You have some systems are oriented towards time-sharing and some systems that were oriented towards real time There was kind of a myth going around in maybe the 1980s about how all these computer vendors were trying to lock you into their computer systems And this was why they invented all these different things in their operating systems. I Never remembered a single conversations that went like that I never remembered a single engineer who sat down and said let's see what we can do to lock our customers into our system The comments were always how can we make our system more efficient? How can we make it so that our customers can do the work better faster and easier? If we had tried if we said that we created a different operating system to lock our customers in Then we only would have needed one operating system on the PDP 11 And as it was there were over 11 different operating systems on the PDP 11 RT 11 for real-time are a sex for a combination of real-time and time-sharing Vistas for an educational system IAS for help for health You know there was Unix back in those days Why would we need so many if our sole goal was to lock our customers into systems? This is an urban legend. I believe I never remembered even one time having that Now I'm going to take you back to 1969. I was a young student at Drexel University in 1969 people walked on the moon for the first time and the whole planet Paid attention to that no matter what country you're in you were glued to the television set to watch this miracle happen Also in 1969 the ARPANET was started the predecessor to the internet In 1969 Linus Torvalds was born. Yes, Linus will be 50 years old this year It makes me feel incredibly old Because in 1969 I wrote my first program I Took a correspondence course in how to program the IBM 1130 in Fortran That's spelled with all capital letters Not Fortran to Fortran for Fortran 77 Fortran 94 high-performance Fortran just Fortran Because with that and cobalt Why would you need any other language in 1969? I shaved for the last time but more importantly to the people in this room in 1969 in Maryhill, New Jersey at Bell Laboratories a man named Kenneth Thompson decided he wanted to play space wars on a computer and He decided along with his cohort and friend Dennis Ritchie To write this operating system which became known as UNIX Now in those days Most of the software which was distributed was distributed in source code form The reason for that was that there wasn't enough of any one type of computer to really make it worthwhile to create a binary distribution of anything I Was working for a very large computer insurance company in 1973 and I remember purchasing I remember purchasing a compiler from a company that was selling it The compiler was very efficient and generated code that ran very quickly and The compiler only cost us 130,000 US dollars back in 1973 To put this in perspective that same year my parents bought a three-bedroom house With living room dining room family room full basement on a quarter acre of land for $32,000 So this compiler costs three houses But it was worth it because it produced code that ran 30% faster And when you bought a 2.5 million dollar computer Having code to ran a third faster was like getting back one-third of your computer And so it was justified We signed a contract for that for that compiler We had to negotiate with the company for days weeks To sign the contract for it and after we signed the contract and paid the money This engineer showed up with a magnetic tape And they spent a week at our site getting that bag being the compiler on that magnetic tape to work on our system Because the first thing you had to do was assemble a small version of the compiler And then use that compiler to build another one. This sounds I'm sure familiar to many of you But after the engineer spent the week here, they left and got back home But they did not take the tape with them The tape stayed with us Because we were going to have the source code in case that company disappeared And we needed to fix the problems There were no software copyrights or patents on software back in those days That did not happen until about 1986 When ROMs started to come out for games and they wanted to protect that So before that time, you could not protect your code with copyright or patents At least in the United States There were few professional engineers In fact, I remember a professor at my university telling me John, you will never be able to earn a living as a professional engineer I'm only 67 years old now. I'm still waiting to see if he was right But back in those days you Were a physicist and you wrote programs to solve your problems You were a chemist and you wrote programs to solve your problems You wrote the programs to solve your problem And then when you got finished with it, you would say to yourself. Well, what am I going to do with this? Now you're probably aware that selling software is a very hard job You have to set up a support team, you have to write documentation You have to be able to accept bugs You have to advertise your code and so forth and so on And all of these things take away time for what your real job is As a physicist or chemist or whatever So you might contribute your source code to a library Such as Deacus or Share or Mindstorm or any one of the libraries Or you could take your code and put it up on a bulletin board That people could dial up at 1200 bits per second or whatever speed they had Or you could contribute your code to a user group Either a physicist user group or whatever But in any case you contributed your code And sometime later you might go to a conference And somebody would walk up to you and say, oh, you're the person that wrote that code Oh, it was such a great code. I have some ideas about it Do you mind if I give you some pieces of code to improve it? Of course not. If you think about it, you just doubled your staff If a third person worked up, you tripled your staff You have three people working on that piece of code now And then maybe somebody would say, well, I don't know how to code, but let me buy you a beer Well, let me buy you dinner Or don't you want to come and work for my company And these are some of the reasons why we write free software today We need it for our own use We're looking for help in making our ideas viable And yes, it could lead to a job or other things Now, as I said, back in those days there were many different operating systems Because the computers were so weak and miserable and crappy and small That they really couldn't handle a general purpose of what we think of today As a general purpose operating system So Ken Thompson and Dennis Ritchie started with this little cast-off computer Called a PDP-7 at Bell Laboratories It was sitting in the hallway, nobody wanted it anymore Because the memory was really small It couldn't even host its own assembler And so they went to another computer system and used a cross-assembler To punch the paper tape that would then be loaded onto the PDP-7 And they would run it and they would see what was wrong And then they would go back and change the assembly language back on the other computer And you say, okay, well, that's cross-assembling, cross-copiling And all of us have done that from time to time Except the method of carrying the data back and forth was punched paper tape And the machine that read that punched paper tape could read it at approximately five characters per second So if your operating system was any size at all It took a long time to punch out the paper tape and then read it on the other system And just to come up with some type of an error So they decided that the little operating system was going to have a very small kernel With a minimum of functionality inside of the kernel And that did several things It made it easier to get it right And eventually when Dennis Ritchie wrote the C language They translated it into C so that it was more maintainable And also more portable across different hardware And these were the types of things that Unix started to do Because remember, operating systems up until that point were specifically tailored to a piece of hardware to make it work faster But in Ken's and Dennis's mind, it was more important to have the operating system portable And to have the people portable That you could have the same operating system with the same interfaces across all of these different pieces of hardware And of course they distributed it in source code Ken would go out with this little tape underneath his arm as he went on sabbatical He loved teaching, he loved working with students And he would try and demonstrate operating system design using his favorite operating system, Unix And this allowed universities and companies and governments to contribute to this Now unfortunately Unix was never free It was never free of intellectual property, it was never free of cost In order to work on a Unix system, if you weren't a research university You had to pay $160,000 per CPU Now remember the house, right? This is five houses you are paying just to get the right to put the software on one computer If you wanted to put it on another computer, it was another $160,000 The third computer was another $160,000 This is the telephone company, they really didn't understand volume discounts But the worst part of it was that you had to tell them the serial number of the computer that you were putting it on And to find the person inside the telephone company that actually would accept that serial number And check it off and say okay, you can buy the software, you can pay me the $160,000 It was very difficult If your computer system broke and you wanted to transfer it to another one You had to call that person up first and say yes, I want to transfer it, here's the serial number of the next computer So this caused kind of a problem in the Unix industry But in 1982 Sun Microsystems had this very nice little board that was developed at Stanford University And they wanted to bring this out as a piece of hardware and workstation And they looked around for an operating system, they actually came to Digital Equipment Corporation and they wanted to get our VMS operating system But unfortunately two things, Digital did not want to sell it to them, that was a big mistake And secondly, a lot of our operating system was written in assembly language And therefore it was not as portable as it would have been It would have been very difficult to make the VMS system work on this workstation So they looked around some more and one of the people starting the company was a person named Bill Joy Who had worked on the BSD Unix project and he said let's use BSD Unix And so they went to AT&T and negotiated for a binary only license That was considerably cheaper, it was about $350 for an unlimited number of people And even less if you're only going to put two people on the system That's strange, two people, why two people? Why not just one person? Well, you could be logged in and what happens if somebody wants to send you a message as UUCP? You need a second login for that and that's why it was two people Now other companies started to follow this because they didn't want to be left behind And most of these companies were BSD oriented, not System 5 Why? Because System 5 was still a swapping style of virtual memory You had to have your entire program someplace in the memory And you could swap out your program and swap another one in But it was not demand page virtual memory System 5 at the time only included two compilers, the C compiler and Fortran 77 But almost, most importantly, its method of networking was called UUCP, Unix to Unix copy So if I wanted to send a file from one place to another, that file was first copied to another system, a hub And then from there it was copied to another system and kept hopping across BSD on the other hand had demand page virtual memory, which was very important It included three compilers, C, Fortran 77 and Pascal Pascal was a very big language in those days and a lot of universities were using it But again, almost most importantly, it had TCPIP And that made a lot of difference to a lot of vendors because it allowed you to set up reasonable networking Now at digital we were trying to decide which one to support and from a technical standpoint, BSD was the way to go There was one other thing that might happen that might have steered us in a different direction And that was the fact that the University of California had no marketing budget for Berkeley Unix But AT&T definitely pushed System 5 constantly for five years with a two page advertisement in Computer World Magazine, System 5, the right choice But even with all of that people tended to use the BSD system base There was one person that was not very happy with this move to binary distributions though You probably can guess who that was, Mr. Stolman at MIT He didn't like that, he wanted to see the source code for the operating system He needed to write device drivers for things and he couldn't see how to do it And so in 1983 he introduced the GNU project for GNU is not Unix And he wanted to build a complete operating system and have all the source code available In 1985 he created a free software foundation to help to fund this and to give some structure to it Now an interesting thing about that was he didn't start with the kernel of the operating system You know you might think oh well why don't you start with the basic first start with the kernel but think about it If he had started with the kernel and he got finished there would be no applications to run on top of it And you know so nobody would buy it and by the time the applications came his kernel would be obsolete And so he said he started with something that almost everybody at the day used a text editor called Emacs Some people say that he could have stopped with Emacs because it's like a whole operating system I personally am a Vim person, don't boo But then he went on to create a compiler suite Now that's another odd thing, why would he create a compiler suite? Back in those days each vendor would have their own C compiler as an example Deck had one, you know HP had one and yes there were standards in the language But the compilers were not perfect And so in the middle of your code you might have to put some if-defs to get around the fact that Digitals C compiler did something this way And IBM C compiler did it something different And the syntax and the semantics, even if the syntax was correct The semantics of the compiler might be different And by using the GNU suite that ran across all these different operating systems You could eliminate a lot of your if-defs Eliminate a lot of your branches in testing Which meant real savings in money And so companies like Boeing that were required by the government To use operating systems and architectures from different companies They used the GNU compiler suite And they paid a small company called Cygnus to provide support for that Eventually Cygnus of course was purchased by Red Hat And still those people still exist today After the compiler suite he went with utilities Bash Shell, other utilities And bit by bit he created enough to replace almost every necessary version Of the UNIX operating system And this also allowed even more people to be portable across this Because from vendor to vendor there might be some small differences In the utilities of things that would upset people, upset shell scripts And in the last days of sun A lot of times people would order Solaris And immediately throw away all of the sun utilities And to get the GNU suite and put it on top This top-down approach also allowed small companies to start up Providing support of different types to free software And they packaged up to software people like Roanuck Creek, CD-ROM and others And made it easier to use for people who weren't used to compiling things People provided support companies for the software and made money that way I worked for Digital Equipment Corporation from 1983 to 1999 And we supported UNIX at first on some of our machines for the telephone companies and universities We would write device drivers for them We would help them use their UNIX systems with code that came from either AT&T or Berkeley But you had to have a source code license for it And so you either had to be a research university Or you had to pay the money to AT&T So when Sun Microsystems brought out their binary-only version We determined to bring out our binary-only version And we released that in 1984 This started to create the UNIX wars All the different vendors trying to push their UNIX system And gain part of the market share Most of this was in the scientific and technical space You know, businesses, chemists, things like that Or in software development Where all the tiny little tools in UNIX Could allow you to develop your software And deploy it across multiple systems To build it on all these systems Because the GNU suite was there to act as a compiler Back in those days, the GNU suite did not generate code that was very efficient Using the GNU compilers might generate code That would only be 70% as efficient as a commercial compiler But as computers became faster and faster That extra 30% performance was not necessarily useful or necessary And people made the trade-off of the ease of maintaining their code Versus the speed that you would get If you needed an answer faster, you just bought a faster computer That's what the answer was As I said, this was mostly in the technical space The commercial space, business and so forth and so on Still was mostly in mainframes batches And about 84% of the market And that's still true today About 84% is oriented towards commercial work And 16% is still in the scientific world About that time, an organization called UNIX System Labs came up AT&T felt they were losing control of the UNIX market Actually, they lost control a long time before that But they joined with Sun Microsystems to form UNIX System Labs And Sun Microsystems went from SunOS Which was a highly tuned Berkeley-based system To supporting System 5 with Solaris Which in the early days we called it Slow-Laris Because they were going from a highly tuned Berkeley system Which was very efficient On to this thing that was basically sample code from AT&T That had never really been a commercial product And their engineers had to spend two years Bringing it even close to the performance that they had with SunOS Digital, HP, IBM and others There were seven companies at all Formed OSF in 1988 to kind of combat this They were afraid of being locked out of the UNIX market by Sun and AT&T Defining what the standards of UNIX was going to be So they formed an organization called the Open Software Foundation And they set up a set of standard APIs and standards And a test suite You didn't have to use their sample implementation As long as you made sure that your APIs and your operating system Could support the test suite And from that, many of their companies were able Not have to change their operating system much at all But still were able to call it OSF-1 Or later on, UNIX Digital actually used the sample implementation that came from Carnegie Mellon Based on Carnegie Mellon's mock And we came out with a product which we called DECK OSF-1 Or later on DECK UNIX Now in 1991, there was a small corporation set up by some of the people from Berkeley Mike Carroll's, Kirk Bikinsek And a bunch of others called BSDI And they brought out a little operating system that they sold both the binaries and the sources For only about a thousand US dollars They were rapidly sued by AT&T Because of course they hadn't paid the $160,000 for the license And this lawsuit went on for a number of years until 1994 The lawsuit, as it dragged through, allowed the BSD people to actually rewrite most of the code Eliminating all of the AT&T intellectual property Until there was only about 17 files that had anything that looked like AT&T source code in it And when the judge finally said, OK, these 17 files are the bad ones They quickly rewrote those two Producing a distribution called BSD-LITE Which became the basis of free BSD, open BSD, and others By 1991, most of the GNU operating system was in place And a young university student in Helsinki, Finland decided that the operating system that came with his new computer Was really a piece of crap And he wanted to have a UNIX system But he couldn't afford it And so he decided to write a new kernel And of course you know that story It was Linus and he started to write Linux Now why did this happen? Why was it 1991? Why at that time was this possible? First of all, you had powerful but cheap computers The 286 was only capable of supporting a swapping operating system Remember, in the early days of UNIX, swapping was bad, demand page was good The 386 was the first Intel processor that could support demand page virtual memory When Linus got his brand new 386 for Christmas He recognized that it could support a reasonable operating system Now to be fair to the operating system that came with his 386 That company had to support all the 286s, 186s, 8086s that came before And those were all swapping systems Just because a few people with a lot of money could buy a 386 They weren't going to change their entire operating system just for that group of people And so Microsoft continued to put out DOS and Windows And Linus said, I want something better The other thing that happened was that over time the 386s that people purchased for a lot of money Became older machines and you went on to buy a 486 or something newer Now you had the 386 that was a secondary machine Something that maybe you could use to do a project or like write an operating system And so there was a lot of these machines starting to appear And people say, yes, I can use this to write this operating system and contribute to this cause Another thing was that faster internet was coming to the home It wasn't just dial up anymore People could actually get DSL and other types of broadband internet into their house It wasn't just, they didn't have to do the work at the university They didn't have to do the work at their company after hours They could do it in their home And there was a lot of information online and examples of operating systems of different types Minix was out there, you could see the source code for that The BSD code was out there, you could see the source code for that The World Wide Web was maturing in that time It was actually first started in 1989, so by 1991 It was easier to get this access You didn't have to do things like take two binary files and put them together And decompress them and things like that that you had to do with really decent porn And so in 1994, which was 10 years after George Orwell's 1984 We had system vendors that were seeding the desktop to Microsoft They were giving up and saying, we're not going to have Unix on the desktop We're not going to have VMS on the desktop We're going to have NT or we're going to have Windows on the desktop And maybe NT as a server And a lot of these vendors are beginning to panic because they saw that Microsoft was going to eat their lunch Even Tim O'Reilly had started to publish, he had been publishing books on Unix Unix Command, how to use Unix, Unix Systems Administration And all of a sudden he stopped publishing those And started publishing books on Windows and Windows NT Now I don't blame Tim for doing that, he had a company, he had people working for him And, you know, he needed to keep them working But on the other hand, he had given up And then along came Linux And in May of 1994, version 1.0 of the Linux was published And that was a signal for people to say, I can start using this as a real operating system to do real work Some distributions, like Debian, were announced in 1993, the year before But the real distribution came out in 1994 Soft landing systems, Yggdrasil, those are systems that were announced and shipped in 1993 But they didn't survive Slackware did, and Debian did Now over the next few years, GNU Linux, and I put in the GNU there to make Richard happier I know he'll never be happy, just happier But it really should be, after all, GNU, MIT, BSD, and a whole bunch of other things that go into the operating system Linux, it still didn't have any applications It had the shell utilities, had a bunch of interesting stuff there But no real applications on top that people would recognize, no office package or anything like that However, it was a lot cheaper than getting Spark and Solaris for things like shell accounts for ISPs Or to set up email and web servers Or to set up a firewall Or to set up a DNS server And remember, you now have all these old 386 machines that you want to repurpose You don't want to just throw them out, you don't want to use them as a doorstop You're repurposing these old machines to do work while the new machines you're buying are doing your real work In 1994, we had a problem Supercomputer companies were going out of business, ECL and Cray Because they would spend millions of dollars to create a supercomputer to solve problems And then only five people would buy them And three of those people were universities that never really paid any money So, two people, Dr. Donald Becker and Dr. Tom Sterling, created a concept of the Beowulf Using commodity off the self programs or systems And put this little operating system called Linux on it And solve real world problems Like weather forecasting and hydrodynamics and other things And that created another market for Linux In 1998, databases like Informix and Oracle started to port to Linux And for somebody who had worked at digital And had gone through this process of bringing out a new operating system and architecture several times I recognized a pattern that was forming And this pattern was the pattern to becoming accepted as an operating system in business In the year 2000, embedded systems started to embrace Linux Up until that point, you had proprietary embedded systems like Wind River and stuff like that But then, something magical happened These little embedded systems were expected to talk to each other And more importantly, they were expected to talk to the internet And when you do that, you have to have something known as an IP stack A TCP IP stack, and they are not easy to write And you need to make your system secure And you need to make your system efficient And this was going to take a lot of money and a lot of effort And all of a sudden, they realized that there was this operating system that did all of that And guess what? It was free, and it was open And they said, we're going to use Linux And almost overnight, Linux became the most used operating system in new designs for embedded systems Now, I know that you're all sitting out there and saying, this sounds wonderful And what we really want is the year of the Linux desktop Or the Linux desktop And that is a hard thing to do Because for a long time, there was only really one standard on the desktop It had been ceded to Microsoft You had millions of people out there They did not really understand what they were doing when they ran the computer They had this tab, they had this notebook filled with little things they had written down Push this button, this thing appears Do this, that thing happens And if anything was different from that set, they were lost My mother and father were these people And they would not learn anything else Because what they were learning was button pushes and things like that However, something else happened The worldwide web that was started back in 1989 Now has matured to the fact that it was now browsers setting up a new user interface And cell phones were setting up yet a different user interface And people had to learn how the scroll bar worked Not just put your mouse here and drag it down And they had to understand what radio buttons do And they had to learn not one interface but many So that helped out a lot, but not enough Because the number one reason given as to why people didn't use Linux on their desktop was lack of games I have a slight problem with that I think of computers as something that I use to solve real problems and games are not one of them But I'm not the average user, I guess And then people would say, well I only need five applications to be able to use Linux But when you start asking around, the first three applications are always the same Office package, mail, browser, but then you get into the more different ones And finding that last and fifth application that everybody wants to use is often very hard Another threat at the same time was Apple's upswing to turning computers from being data processing things into being consumer products The iPad, it has to look sleek, it has to be great, it has to be easy to use Apple systems are not easy to use But they are consistent and they are consistent across all of their products One of the things that Apple did in the very first days of Apple was to create a style guide That said this is how the operating system will look and this is how the things on it will work If you remember Apple brought out the Lisa computer, it was a graphical computer And they said to all their third party people, just port your code to it, that's fine And it was a disaster But then they brought out the Macintosh And they said, when you port your applications here, you have to meet this style guide Where you will not be a Macintosh application And people looked at it and they grumbled and stuff And the first Mac application came out and Apple looked at it and said, yes, you get a nice big gold star Everybody said, what's the big deal? But then the second Mac application came out And it looked like it was written by the same person and everybody went Now they understood And so Apple turned the computer from a computing process Into a consumer item Now in the latest round with their iPad, iPhone, i everything They were looking to really move the market again But one thing happened Android foiled them Because as you know, the only person that can do business with Apple is Apple You can only buy Apple disks, you can only buy Apple printers, you can only buy Apple monitors I'm sure you can get out of things, but they don't necessarily work with Apple You go to the Apple store, you have the Apple genius helping you But what if you want to make money and your name isn't Apple Then you use Android And the same statistics will end up in the future Where Apple will probably have 7% of a very profitable market Which will make their stockholders very happy And Android and other operating systems will have the other 93% Now we have the cloud Everybody loves the cloud Where we hide the operating system and the libraries and utilities away from the customer So that they don't know what's going on They don't know where they've stored their data We give them the illusion of privacy and security But we don't tell them about the NSA And what we need to do is give the cloud back to the people As I said, I'm the chief executive officer of Optin And we produce SuperTie open source peer-to-peer cloud software It allows you to take your resources that you have, your computer resources Combine them in a better format to be more efficient And to be able to buy and sell or barter your computer resources to other people But from my viewpoint, it's even more important that with this you can actually keep your data away from the NSA And from other people who want to spy on you And it's not just spying by the United States government It's spying by people that want to sell you things and want to know what your psyche is We have a hardware router We call it the blockchain router It's a broadband router, but it also provides NAS backed RAID It can be an IoT gateway Supporting Arduino shields and Raspberry Pi GPIO pins, which we have built into our system But most importantly, it allows you to be energy efficient in your blockchain mining So you can mine Ethereum tokens of different types, utilizing the router And it only draws 18 watts of power to do all of that That's about the same amount of power that your Linksys or your Daedalic routers use It is an open design The design is up on GitHub But it's going to be built in Brazil, utilizing Brazilian parts for the most part As part of a project called Caninas Lucas And we'll hear more about that in a few minutes Our company is led by three open source advocates Alex Carasulu, who has worked for the Apache Foundation for a very long time And has written a lot of the very difficult code in Apache Sally Kuderi, who is the director of marketing And she's been the marketing person for the Apache Foundation and other open source projects And myself as CEO Now, here are some shameless plugs of things that might affect Debian a lot We have here in Brazil, headed up at the University of São Paulo A project called Caninas Lucas About five years ago, when the Raspberry Pi came out I invited the Raspberry Pi people to come to Brazil to attend the campus party To show them how enthusiastic and how knowledgeable the Brazilian people were I begged them to produce the Raspberry Pi in Brazil Because as you know, the Raspberry Pi that cost $35 in the United States Is currently selling for about 200 hais here in Brazil Mostly due to taxes And that makes it too expensive For high school students or university students To use it the same way it's used in the United States or in Great Britain After two years of working with them They said, no, we're not going to do it They didn't tell me why So I went out to a Chinese vendor called La Maker And I got them to agree to license their design to us And then we would manufacture it in Brazil It has taken three years to get through all of the logistics, taxes Everything else But we now in the next month or two are going to be producing this in high volume We produce a better system than the Raspberry Pi With twice as much memory Two to three times faster in benchmarks USB 3.0 instead of USB 2.0 as a bus It works in a higher temperature, 70 degrees Celsius instead of 40 degrees Celsius And it is also better with electrostatic discharge And it will cost five dollars more than the $35 Raspberry Pi Now, Professor Zufo, who is the head of this project Went to the Brazilian government Who is working on the Internet of Things Plan for Brazil They're investing 10 billion, the equivalent of 10 billion US dollars in the program And they expect to get about 200 billion US dollars out of it They wanted the professor to generate the hardware platform For the entire Internet of Things program for all of Brazil And he has decided that he's going to generate three different parts One of them, he calls the Flea, it's a tiny little sensor computer And that little computer will have a lot of sensors on it He'll be able to run off of batteries for several years And it'll be very, very inexpensive The Labrador is the machine I just described It's a Raspberry Pi on steroids And then the SuperTie blockchain router has been donated to the platform So that they can use that as the eye end And he needs help in putting Debian, which is going to be the selected platform On all of these systems, on to this Internet of Things set of platforms And so he's asked me to come here and reach out to you And say if there's people that are interested in joining this And helping out with this here in Brazil Because all of these systems will be manufactured in Brazil By Brazilians And this is going to create a center of expertise here in Brazil That can be used in other electronics projects Instead of having well trained people go to the United States Go to China, go to Europe Because that's where the interesting projects are We're going to create them here in Brazil And then SuperTie has decided to use Debian as our main platform for distribution of our software So in summary Unix was important because for the first time Human beings were thought of being more expensive and more worthwhile than machine cycles That maintaining your investment in human beings And be able to use that same investment across all the pieces of hardware All the scaling points was more important than being a little bit more efficient in the use of them It's set in a lot of ways, the APIs for the future over a wide variety of systems Today, Linux runs on 60% of all the servers being shipped It is the most used operating system and embedded system design It is a core of Android It is used on all 500 of the fastest computers in the world There used to be two that ran Microsoft Because Microsoft paid them to run it But even they gave up And GNU Linux has relit the desire of people to have control over their software You know, we talk about software freedom But there's a lot of people out there that do not understand what the word freedom means Certainly Donald Trump doesn't seem to know that So I have stopped talking about software freedom And instead I talk about software slavery Because people understand slavery better than they understand freedom When you're a slave, you're told where to go, what to do, when to have children, who to marry You're not allowed to have property, it all belongs to the master When you're a software slave, you're told when to install the system, how many people can run on it, what systems they can run on And when you can't use your hardware anymore And the control is taken away from you and is put back in the vendor I like software freedom And free and open source software and hardware and free culture have set forth another version of capitalism Do not let people say that free software is communism, it is not communism It is cooperativism And in cooperativism, people work for common good There is many forms of capitalism, Donald Trump only thinks of one So here is another shameless plug, this is the last one really It's a series of things because next year, 2019, will be 25 years or quarter century of Linux as version 1.0 or before It is the anniversary of the creation of the Beowulf supercomputers And it is the anniversary of most major distributions starting up Next year will be 30 years at the World Wide Web And 30 years I've had the Unix license plate on my car that says live free or die Also in 2019 is one-half century of the people walking on the moon, Unix, the internet, Linnus Torvalds being born My programming and the last time I shaved But last to next year is also Debconf Curitiba So let's party Thank you very much, if you'd like to ask some questions I'd be happy to answer them There's a question in the back, that lady By the way, I'd like to say I was very happy to see the Debian women up here Because two of my favorite people of all time, Ada Lovelace, the first programmer And Rear Admiral Grace Murray Hopper, the first modern day programmer Yes ma'am I would like to ask a question about how can we get involved about getting Debian and the Labrador board So I'm sorry, what is the question? How can we get involved with the Labrador project that you mentioned, the project that we, that are going to be manufactured, the board that are going to manufacture here in Brazil That board is being manufactured by a branch of the University of São Paulo called LSI Tech And if you go to www.caninaslucos.org, caninaslucos of course is crazy canines All of our boards are named after different dogs I'm the mad dog and the professor is, I don't know what he calls himself But there you can see more about the project And if you can't find the information about how to get involved, then you can send me email at maddog.li.org And I will look for it What I would like to do is see maybe this Debian group here form a subgroup that would be interested in helping with this IoT platform And that might make it easier to set up a mailing list and stuff And figure out for yourselves the ways that you would like to help with the project We need people to write device drivers We need people to help us create distributions I've always, I've been happy with the Raspberry Pi people where they created the Raspbian distribution I want Debian to run on that You know, not call it Raspbian or anything like that I want it to be Debian for ARM architecture running on that Or any other architecture that happens to be fit into the Caninas Lucas project We're actually looking at the risk 5 architecture for the fleet Because of the heavy intellectual property costs of using ARM So you may have some experience then in working with the risk 5 architecture in the future Another question? Thank you very much for listening