 Good morning everyone. Oops, that came suddenly. So who is here just because they were tired of queuing up for Moitilisch-Incorrect? So that's in the order of 10 at least. Nice. Okay, so let's see what we have here today. We have the lightning talk session on day 3. Could you please start my slides? So what we are going to do is we'll have 20 talks each five minutes. A very fast-paced session where everybody can pitch their own projects. And what I would like to explain to all the talkers today, to make things run smoothly, please sit in one of the front rows. Once your talk comes up, please quickly enter the stage. And yeah, go before the microphone. Please stay close to the microphone because otherwise we can't hear you. And don't turn around because then we can't hear you as well. So you can see your slides down on the monitor. I can see my slides right now. I don't know what's wrong. But we'll work it out somehow. It's early in the morning, so bear with us here. Okay, you can see your slides. All you have to do is stay calm, deliver your talk. You can advance the slides using this clicker, which remains here on the lectern. So just pick it up, deliver your talk, drop it again, and then get applause. And I used to have a profit joke there, but that doesn't really work without slides. So what else is to say? For all the audience, for people in the audience, ah, there they are, finally. Now I know what to say. So I had this, this, yeah, yeah, yeah, this profit, yeah. So how to listen to lightning talks for all of you in the audience? Simply be excellent to each other and watch the timekeeper, which is this nifty device here. Alex will briefly explain what this is all about. Yeah, I think you already know it. You have seen a lot of lightning talks. It just means if it's green, you are in the first four minutes of your talk. And in the last minute of your talk, in the last, someone mentioned, I say, 30 yesterday. But okay, in the last 60 seconds, it will start to turn yellow like this. So now you still can relax. You have now about 45 seconds left. And if it goes down to 30 seconds, it becomes red and redder and more and more and more red. And in the last five seconds, it begins to flash. And now it's your part. Five, four, three, two, one. Better than yesterday. Thank you. Okay, let's try it again, okay? Or were you satisfied? Do you want to try it again? I don't think we have time for that. So let's, let's just continue. All right. So it always works better in real life. There are translations available. So English talks will be translated into German and French. German talks will be translated into English. Please see the Wiki with the translations topic and also see streaming.c3lingo.org for information on how to listen to the translated streams or simply watch the translated stream video. All right. Let's begin. Let's do that. Let's start with the first talk. If someone needs water, it's behind the curtain. Thanks. So let's start with OpenH. Yes, hello. We're the development team of OpenH, which is a free re-implementation of Age of Empires that probably many of you know. We require the original assets and do this mainly because of unlimited possibilities and writing awesome mods for that thing. We base that on mainly C++ and Python 3. For the user interface, we chose QT. This year, we have three main advancements, which is a web interface for our continuous integration system. The Nian configuration framework was my master's thesis, and we have an event-based game engine concept for the simulation calculation. Okay. So let's start with the continuous integration system. Last year, we had a whole talk about it. This year, just a quick update. We have now actually included a web interface. It looks about like this. On the right, you can see the output, the real-time terminal output, and on the left, you can see the individual build steps with status and so on. Nian is our configuration system for the game behavior, which is a plain text format, a bit like YAML or JSON, which contains key value pairs that can update values through inheritance, and where modifications are done through patches. And this is optimized for mod creation because patches can also patch patches, so that means you can modify everything in the game with a single system. So in this case, the generic unit is specialized to be a villager, and the super villager is an even more specialized villager, so that when you evaluate the health points, it's 35 for the super villager. The modifications are done through patches. In this example, when you discover the loom technology, the villagers are stronger and then have 35 HP. But Nian can do much more because it allows patch transaction and history and has many more data types and allows mod combinations of overlays. The new game engine concept is tickless and event-driven. It will only recalculate events or the future if necessary. It will store the whole history and future of the game, and the rendering is actually only doing a snapshot of the engine and then drawing it to the screen. There are different types of interpolations. You can have a continuous line. You can have discrete steps for arbitrary types. You have maps and queues that can be accessed over time. Then, for example, the unit description in C++ is you have a discrete curve of hit points and a continuous curve of a position. Then you insert a value at a specific time into this curve and in the end you can access it at any value in between. For example, you can predict that the arrow will hit the villager in roughly five seconds. Then the hit points will go down to zero. So the villager is that and stops building the building he's working on. That's an event that's going to be triggered and issues a recalculation of this. The building finished event will be predicted into the future and if there's no villager working on it anymore it will be dropped. For example, an event will be the building construction finished. The building is getting damaged or a unit starts together. All the interaction in the game is described using these events and gets registered onto the buildings and onto the internal structures of the game like the building 42 triggers the event. Its construction has finished at a certain time when it's finished. Our next steps for the project are the integration of those two systems to implement actual gameplay features that most of you have been missing for a few years now. We also need a cool pathfinder based on flow fields and we also need a new renderer because the current one is really crappy, but we have a great guy already starting to work on that. So thank you for listening. You can meet us at the Stusenet assembly and we hope to talk to some of our community contributors that we have a lot of them. Thank you for your attention. Thank you. Next up is Atari and Commodore fuck ups. Hi, I've got a problem that I've got to have a seven minute talks which I have to squeeze into five minutes so I have to take a few shortcuts over there. My motivation is purely just to entertain you. Back in the 70s the TV was the only hardware that was capable of displaying something so if you had something like a game console you needed to have a modulator that modulates the video signal to the antenna signal because there was no video in connector that picked them. At the time the Atari 2600 was created, only five companies built such a device. Nolan Bushney as head of Atari made an exclusive contract with each one of those, so he was the only one who could get such a device. Back when Atari was sold to Warner communication there was then one guy in bookkeeping and I said what do we need five suppliers for? One is enough, fail. The video game crash of 1982. These two games are claimed to be responsible for it because they were so bad. I say these are responsible as well because they were so good. These games were programmed by the first third party game development company Activision which was created by, built by four guys from Atari who did the best games and asked for a different deal for their payment compared to what musicians had together from their record companies but they were declined very rudely. After that Activision was founded and there was a lawsuit against third party game development, Atari lost so new companies spread and like mushrooms and built games that were awfully bad. So you could think of it, summarize the crash like this one, you send your dad out to get a good game and he comes back with two cheap ones, fail. After the crash there was a Japanese company that approached Atari. They had done a video game system released back in Japan that went very well and they wanted to release it in the United States under the Atari brand but Atari declined saying something like we could do better on our own. So the Japanese decided to sell the console themselves. That console, big fail. Let's switch to Commodore. The 264 series was originally developed as a successor to the VIC-20 and marketing needed a successor to the C64 so they took the only thing they had and you don't sell a successor as an inferior system as a successor. This is what this fail looks like. Did they learn from it? No. They repeated the same mistake using the Amiga 600. This was originally developed as a low-end addition to the lineup. It's just basically a stripped-down version of the Amiga 500 and what this makes, this really bad is that the true successor to the Amiga 500, the Amiga 1200, was released only half a year later. So this is what, how do, a way to piss off your community. And now let's get up to the most fun part. Commodore has developed a machine for business case use. You can think of it as one of the first laptops. For business use, because you had a modem built in back in the mid-80s. This is what the machine looks like. And this is what officially happened and I'm going to tell you the unofficial story. The unofficial story is that the CEO responsible for the project got his, the first prototype presented on the golf course, together with the CEO of Tandy. And the guy from Tandy just said, we did some market research. There's no such thing as a market for this kind of a machine. There's no way to sell this one. So half a year later, Tandy sold that machine. It looks pretty much at the same target audience. So another fail as well. So my conclusion is, Atari and Commodore were killed by guys in suits who think that they know business, but don't know a shit about designing, designing hardware and or software. Of course, this is just my opinion and my apologies to the translators for speaking so fast. Thank you. Thank you. Now you still have some time left over. Never mind. We will just continue with the next talk, which is attribution generator. Licensed notices for pictures from Wikimedia Commons and Wikipedia. Hello. So there's actually a treasure trove of images on the internet that you can use. So when you write a blog post or when you want to publish something, there's probably some pretty picture that you can freely reuse. And it is available on Wikimedia Commons and on Wikipedia, but it's actually not that easy to properly reuse that. And I saw that in the talk that was just before that. Sometimes you see stuff like source internet or this is a free image. I took it from Wikipedia. The Creative Commons license explicitly specify that you have to name the creator of the image and that you have to name the version and the name of the license. And then you are golden. Then you can take from this great treasure trove of images to picture everything. So to make things easier for people who are not legal persons, we build a tool. And we want to simplify the reuse of images from Wikipedia and Wikimedia Commons and give correct license and attribution. Because in this world where you give away stuff for free, there's actually a digital currency and it's not Bitcoin. It is respect. And we want to make it easy to give the correct amount respect to these creators who give away stuff for free. We have an English version and a German version that really released that last year. And quite a few people use this very simple web-based tool to create a correct attribution line for images that they find online. It works pretty easy. You have a URL attribution minus generator.org. There's a typo in here. Or you can use the amazing German word. And you just paste in the URL of the Wikipedia page where you found the image or you paste in the URL of the Wikimedia Commons page where you found the image. And then you go through a few questions and it asks you, do you want to use it online or do you want to print it in a book or do you want to use it together with other works or do you want to crop and modify the image? And then after that you will get an attribution line in HTML for copy and pasting. What have you? It makes things very easy for those of us who don't speak legalese. Just use this tool. It is a lifesaver. And it will keep you out of all the nasty things like cease and desist letters, abmanung, and it makes the world a better place because you will give credit where credit is due. We would like you to get involved. We think it is a very simple tool with a very simple use case. But maybe we can improve it. If you have ideas how to improve it, just tell us. Best thing would be to write me an email. And also if you speak an interesting language that is not German, English, Spanish, or Portuguese, contact us. And we would like to translate the interface into your interesting language. That's it. That's all I have. Thank you. Thank you. Next up is database data layouts. Hi, I'm Stefan and I'm a database researcher. And I like to say that designing databases is as hard or as exciting as constructing rockets. But I don't know how to construct rockets except for Lego rockets. But let me introduce you to database design. And there are many design decisions. And I want to talk about a small design decision. And that is a data layout, which means how you map your data to the memory. And especially this talk is about data layouts for relational databases. And here's an example of a table which you store on relational databases and consists of tuples. So you have different rows. And all these tuples have attributes, for example, first name, last name, and country for persons. And as you can see here, table data is two-dimensional. So we have rows and attributes. And the issue is that memory in computers is usually linear, so one-dimensional. And then you have to decide how to map two-dimensional data to one-dimensional address space. And basically, you have two design decisions. So this is simplified. And what you can do is you can store your data in a row store, which means you store data of a single tuple together, which is shown on the left side. And the other possibility is that you choose a columnar store where you store all the data of a single attribute together. And then you may ask how to reconstruct tuples. But this is quite easy. For example, to get the third last name, you have to know the starting point of the vector for last names. And then you jump to offset three and find winter as the third last name. And then you might ask, why does it matter? And the answer is it's not important how much data you access. Well, it is important. But what is also important is the access pattern. So if you do a sequential scan in memory, it's way faster than random accesses. And to show you this, I will present you two microbenchmarks. So one is an insert, so we want to add data to our database. And the other is a filter operation. So think about an SQL statement. We do a select filter from our person's table and filter by first name. And the interesting question is why it's important to do microbenchmarks. So think about design decision. And maybe you have a good idea. And it turns out to do experiments and show that the idea is good. It's way harder than the idea itself. And then when you have the experiments, it's way harder to productise it and make this experiment robust. And so how does it look like for insert? So for a raw store, we can copy all our data in one place. For the column store, we have to append the data to the different attribute vectors. So jump to different address locations. And as a result, so here, simplified measurement, we can see that throughput of memory is way higher for raw stores than for column store, especially when you have a large number of columns. So the small effect that column store is better here than the raw store for small number of columns. The reason for this is that we use mem copy. And mem copy has a small overhead when copying a small data chunk. So then it's better to do a variable assignment. But for large data copies, it turns out it's quite good. So this was a raw operation. So on the end, a raw database is better. And now look at the other way around. So operation where we have access pattern with benefits or access attributes. You can see this here. And this time for raw store, we have to jump in memory. So if you want to access all first names for the raw store, we have to access the first tuple, then jump to the next tuple. And there we have these memory jumps. And for the column store, we can basically just do and scan. And we did again a benchmark with different selectivities. And this time it turns out that really the column store has a higher memory bandwidth. And you can see again that also the number of columns has an important role. And you can all explain these with how caches and cache hierarchies work in modern CPUs. And basically what I want to show here is this is just comparing your data. And normally in the database, you have to store intermediate results. And then you don't just not only do the comparison, but also have to write the position lists or the intermediate result. And then it turns out to be a bit different. Okay, thanks. I hope you're excited and want to develop your database by your own. So have fun with it. Thank you. Next up is remarkable. Yes, remarkable. Hi, I'm Axel. And I decoded a binary format. So I'm talking today about the remarkable tablet as you see by my beautiful drawing on it. That's a tablet with ink display where you can draw on. And without advertising it, I was actually looking for the parts that are interesting, which are the software parts, because that's the part I understand. So the thing is the remarkable runs on our arm in Linux. We have full root access to it. And that makes it very interesting to tamper with it and to explore it. Short disclaimer. You can see it in the screen. What does it do? If you draw on a tablet, it basically records your complete thoughts. And that's something that you might want to control. So the thing is you can export the data you have on it. And you can import data there, for example, by PDF and PNG. But internally, it's stored in a binary format. And we really want to understand that, because that's what we control at the end. So unfortunately, that's not well implemented yet. And we'll maybe update it, but anyway, we want to know what's going on there. Also, you want to have more features when you draw on something. So for example, you want to have text recognition. You want to see the animations while you draw it. You want to have more brushes, whatever. So that's a short motivation for that, to understand how something stores it, so why Dog X works on the scrap. But actually, it's just fun to see what a binary format looks like. And I never did this before. So the idea was just decoded. So in the remark, there's some file for each notebook. So this is a page to write on it. And that's called lines file. And when you just open an empty file, it looks like this. So there's not much in there. Just a header and then some zeros. That's cool. But now we want to understand how drawings get added into these files. So what I did is I took a pen and took the smallest dot that I can do on the device. It really looks like one pixel. And when I did this, the file changed to this. So we see already there's a block of data, something in a hex editor that comes in there, and there seems to be a number one, which could be the number of times I hit the surface. So what can you do next? Make it a little bit more complex. So just try to add one more dot to a page. If we do that again, we see there's an other block coming in there and the one changed to two. So what did we find out already? There's a fixed header. Cool. There seems to be a counter. Often I touch the surface. Cool. And that's very important if you want to continue decoding it. There seems to be no compression involved, because otherwise the first part would have changed if I put in the second thought. So now let's do this a little bit more systematic. But I wrote a couple of lines of C++ code and built up a little matrix. For what I did is I seek to a certain position in the file. And when I opened this, I did this tables here. And every column is a different interpretation of the next byte or the next bytes. And I added offsets to the line. So just in case I don't know what something is, I just seeked on and on and on. How does this look in reality? So let's go to after the header and have a look in the file so we can see interpretations of what could be the next data type. Four seems reasonable, right? The float doesn't make sense. That means the next three bytes are crap. Makes sense because there's only crap in there. Next one could be one. Okay, the other ones are zeroes. Makes sense. And go on and on and on. The last one could be interesting. So even the ins are not reasonable here, but the float looks interesting. Could be a pixel position because it's inside the resolution of the device. If you continue like that, what you will get is something like this. So I learned in, in, in, in, in, float, in, float, float, float. Cool. So if we do this for one point, then we can start from the inner to the outer and try to understand what we got there. So just croop this a little bit. I know that the device has pages. Inside the pages it has layers because it's visible in the GUI. Inside the layers you can have lines. And lines are croop to dots because I can draw lines, right? So when I just start then to, to modify the dots that I put in the device, you can just systematically explore. And what you find out then is that actually these three floats are coordinates. The pressure changes if I put more or less the pressure on that. We have a, a rotation of the pen. And all of that is then grouped again into a line. A line only has one brush. And putting this on and on and then adding new pages, you just see what changes in the file. And with that you can meeting, get meaning from the pure data types to what's actually in there. Of course the next point has to be to document all the magic numbers and the number radars in there. So I did that. And the interesting aspect that's also in there is actually the ranges of the data you can get in there. So for example the device records the angle that you have to the normal of the surface and in both directions which is interesting because it doesn't support brushes which are not symmetric actually a point symmetric. But it could. So that's interesting. That's what we learn. So obviously last step is implement in your renderer. That's the original and I just played with colors and all of this is now open source and you can find it on lines are beautiful and get up. Thanks. Thank you. Next up is weapons patents. Hello. My name is Felix. I'm working in the intellectual property department and privately I came across the subject of weapons patents. I found out that there's very few poor awareness of the risks of weapons patents. And so I want to show to highlight some risks associated with weapons patents. I will relate weapons patents to crime. This is what constitutes one risk associated with weapon patents. And so let's go. I found some correlations. First I want to correlate the time when weapons patents got even more dangerous in the end of the last century then they went online beforehand they were paperwork difficult to access or behind paywalls. But in October 1998 the European patent office decided to publish them online without any restriction. The American patent office followed six months later in April 1999. If you correlate this with a German terrorist organization organization Red Army Faction which caused much trouble and who killed many people in the last century you found that this organization could not take profit from weapons patents because it was too late. But if you talk about Okada you will find that this organization came to public awareness in 1993 with the first attack on the World Trade Center. At that time there was no possible means to access knowledge about weapons patents by the means of the internet. But more than two years after all these patents went online they did a second attempt which was successful as we know. This is only correlation no proof of whatsoever. Here I show one example of weapons patents. Particularly of this patent first of all it's published by the US Army. It is told that this is is recommanded for unconventional warfare activities with means which means guerrilla fight and terrorism and if you have a look on this document it's quite descriptive. It gives a clear indication how to build such a weapon which material which quantities and all this. I would have liked I would have liked to find any evidence that the crimes of the destruction of the World Trade Center is related to weapons patents. I didn't find anything like this. There's no evidence but what I found when I studied the official paperwork is that the National Institute for Standard and Technology which investigated in these crimes that did not even think about the possibility that terrorists that suicide attackers didn't attack from the inside. This is what I marked in the red is that they assumed that professional people went into the buildings may have gone into the buildings planted some explosives left the building and went away before the final attacks took place. The possibility of suicide attackers inside the buildings was not taken into account. My conclusion I think it's a good idea to change the official storytelling which tells terrorism is a very very big comfort to mankind and we should sacrifice our sacrifice our yes thanks a lot. Thank you. Thank you. Next up a proposal on a novel heating method for coffee. Hi and welcome to a not so serious talk about novel coffee heating method which is an application for in-situ use of resources at the office with yeah responsible use of modern manufacturing technologies. So why do we want to heat coffee? Well I'm I work in science I code a lot and I tend to forget about my coffee. Coffee usually tends to get cold with time and yeah as probably everyone knows cold coffee isn't very good coffee. So we need a solution which utilizes the resources we have at the office and yeah that's first for that we need some sources of thermal energy. Well first I looked into humans because well there are a lot of humans at the office they produce as you can see in the chart around 100 watts at room temperature when seated but that's yeah that's energy that is at 36 degrees Celsius it's really hard to harvest because you don't want to put on a body suit or whatever to harvest the energy. So another one would be electricity but hey come on that's that's cheating because electricity is everywhere and I would need another an additional device. So the night before I saw Star Wars and then I remembered one quote I think it was like this the target area is only seven centimeters wide it's small thermal exhaust port right below the keyboard. The chef leads directly to the main processor. Hey my computer has that too. Okay now we found a source for thermal energy and how do we harvest that. Yeah as you can see in the picture that's a double-shelled mug which is which is hollow on the inside and hot air gets blown from below around the coffee and went it on the on the other side and now we need to get need to build this. So as I said I work in research we work with metal 3D printing so I got it printed in aluminum. Yeah yeah it's yeah you need to do some reworks on that so a mark on the mill is also was a first for me and now the very important part we need to verify this solution. Okay I set up an Arduino with three temperature sensors which measures the temperature of my mark of reference mark and the ambient temperature. This data is sent to a computer in in terms of one seconds because thermal processes are really slow and that should be sufficient. A script visualizes that in real time and also stores the data in an hdf5 container because yeah scientific data and I can process it further afterwards. Yeah as I said I need I needed a similar shaped conventional mark for that for reference and then filled both mark both marks with the exact same amount of coffee using highly scientific methods and for because we need this to be realistic we also added milk 24 millimeters of this to be precise. Highly scientific methods look like that and the milk was added yeah just with this range and then I wired it up set it up at the computer and started the testing so to generate sufficient thermal energy I ramped up some some simulations and got some results. Before I show you the charts I know now that coffee with 60 degrees Celsius is too hot for me definitely you you burn your tongue between 50 and 60 degrees Celsius is super nice warm coffee I like that one down to 40 is still acceptable because yeah it's still drinkable but 30 feet three degrees Celsius definitely too cold yeah and also drinking coffee with a wire stuck in the mug is really really weird because it's everywhere and in your way and you don't want to do that and it's especially weird if that mug is made from warm aluminum. Yeah so that is the chart you can see the red line is my heated mug which performed yeah as expected it it only sank to an higher temperature than room temperature but that wasn't enough because you can see the regular mug which is in green that that after 40 minutes the coffee in the regular mug started to get colder than in my heated mug yeah that's not really good so my conclusion to that the proposed device did perform did not really show the expected performance because the aluminum outer shell acts as a heat sink transferring the heat from the coffee to the outside so I need a new device with less thermal conductivity and while maintaining in the shell. Thank you next up is Interleger. All right hello my name is Evan Schwartz and I'm going to be talking to you about an open payment protocol called Interleger so today payments are broken this is roughly what the payment space looks like it's super fragmented and there's tons of small regional networks that are all separate from one another each of these works well within themselves like Sepford in Germany ideal in the Netherlands Venmo in the US each one of them works well but it's impossible to pay from one to the other as a result you have just a few really major global brands that basically dominate global payments if you look at Visa's slogan everywhere you want to be they're not saying we're the best we're the cheapest we're the fastest they're saying we're the biggest today payments is all about network effects and reach as a result no one can really compete if I started a new payment network today and said oh come use my network your first question would be who can I pay with it it's all about network effects the real fundamental problem is that all of the payment networks are disconnected whether they're banks blockchains mobile money networks etc all of them are disconnected and in many ways this resembles what information networks looked like before the internet so what we need is internet working for payment networks we need a system that makes it so you can pay from any network to any other network so then it wouldn't matter if you and I are on the same payment network this is what interledger is interledger is a protocol for relaying packets of money across many different payment networks just like the internet protocol does into interledger provides three main things which I'm just going to go through very briefly here if you're interested in more details come to the session later so first there's a ledger agnostic address so this is a way to say where you would like to be paid no matter what payment network you're on the second is an IP inspired packet format that's super super simple and just communicates to the intermediaries how basically the details of the payment the third thing is multi-hop security because dropped packets when you're talking about money are a little bit more serious than dropped packets when you're talking about IP packets as I said more details if you come to the session later some key facts about the interledger project it's an open open project we work on it mostly under the ages of a w3c community group there's 275 members that have a lot of different backgrounds ranging from banks central banks payment companies blockchain companies etc because interledger is meant to be a super generic payment method that can be used for lots of different applications so to talk a little bit about implications of this this is a graph of what internet working did for data basically costs plummet and volume explodes because as you add internet working you add interoperability it makes it possible it makes it not matter which network you're on increases competition and the costs come down as the cost come down they get so low that you can suddenly start sending information for much for many more use cases than were possible before imagine doing some of the things we do over the internet over traditional mail just impossible so this is basically what we expect to happen with payments with interoperability so to leave you with three kind of pieces of food for thought I'll pose three questions to you first what could you build with free and instant global payments imagine that you could pay or get paid by anyone in the world in any currency in any amount small or big that's question number one number two a lot of people have talked about the problem with the internet's business model today today it's all based on advertising there's only a few companies like spotify and such that can actually convince people to take out their credit card enter those details for security reasons user experience etc imagine if your browser could just pay for the services that you use in tiny little amounts and you could just control it with kind of a dashboard with sliders that's the kind of thing that this could enable third as I mentioned briefly interledger is high is heavily inspired by the internet itself and there's an interesting interesting issues that come up around what lessons should we draw from the internet what should we copy and what mistakes should we avoid very interested to hear people's perspectives on that so session today at 6 30 pm in seminar room 14 15 if you're interested you can also find us online interledger.org and on twitter thanks thank you next up bergikon hi all good morning so i want to present it our small conference that we are organizing in Serbia in Novi Sad and by the way i'm LNM so what is Balkon Balkon is meant to be a small hacker conference in southeast of Europe to build we try to build their community because a lot of people and students from that part of the region doesn't have so much money to travel there on such a conference as a ccc or some other places so we started to organize first Balkon in 2014 and next year will be our sixth conference so i want you all to invite to join us for the next Balkon the dates the dates for the next Balkon is 14 15 16 september it's uh in Novi Sad in Serbia so it's let's say one hour drive from belgrade and you hear you have here all the important deadlines but why Novi Sad and why Serbia because we are from Novi Sad we decided to make a conference there but we are not living in Serbia anymore but we wanted to give community there some feeling of the hacker conferences from the Europe so you can see the pictures from Novi Sad what can we promise to you can promise to you very good food cheap very good drinks a lot of parties and a lot of fun we also have a great speakers that are coming every year i don't want to tell now the names but on our website you can see all the lists of the speakers that are was last five years by with us and they are coming back because it's uh it's not a big conference last year we have around 450 people it's familiar feeling there you have a opportunity to speak with everybody so i think that's important and i want to all invite you to help us to build a community in Serbia because there is not so much hacker community we have two hacker spaces one in Novi Sad it's small one it's two years old let's say now and one in belgrade more than five or six years i'm not so sure so we are trying to build a community there but we need people from abroad to show the people from in Serbia how it's looked like so that's the reason why we are organizing the conference and we also have on the balcony a small hacker area so we also have assemblies and we also want to invite all the assemblies to come to join us to have fun these three days so if you have any questions how to reach the Serbia because it's maybe for someone it's not so easy there is a cheap flights to belgrade it's one hour drive with a bus from the belgrade airport to Novi Sad or if you have any other question relating to accommodation organizing conference feel free to contact us write us email we'll want to see you all in Serbia in Novi Sad next year thank you thank you next up is Capricorn hi so earlier this year ransomware spread out pretty much all over the globe according to the news luckily myself i wasn't infected but i made a program which tries to detect what changes are made in your file system and today i'll tell you something about that and i'll end with a call for activity so first off a short table of contents of what i'm going to present first a bit about myself who am i then what is capricorn how does it work and how can you join in which is the initial goal of my presentation so who am i my name is max kerstin i go under the nickname of libra and i'm in the third year it security bachelor in the Netherlands i'm interested in reverse engineering and analyzing malware i did android banking malware for my internship and i plan on writing my final thesis about desktop malware because i'm also fairly interested in that which brings me to my program capricorn it's an anti ransomware command line interface application i didn't feel the need to build a graphical user interface um though it's it's still possible uh it's built modular so it can be expanded rather easily um open source um you can find the at the end of the presentation a link to the github repository and um it's written in java uh supports both um linux distributions and microsoft windows um it uses local honeypots a bit more about that later on and one other rule i used is to not require administrative or root privileges so uh how does it work i deploy a honeypot uh folders with files in it uh on the system in multiple locations uh so all of the user home folders are used so that would be your uh desktop folder your documents your videos uh any of these and the root of your file system itself uh is also used so if you're using windows it's probably the c drive and if you're using linux that's just a slash on your on your device um so then i fill the uh honeypots uh with uh files with a random extension i um used my own hard drive as a super reliable source of what extensions are used most um and i got roughly 700 extensions there um and then i fill the um files with words uh based on the english language i um i used some sites uh and i found out what the uh allegedly most used words in the english language are to avoid uh detection since other programs which use uh similar techniques as i do um they often have words in them like this is a test file of do not remove anti ransomware um so ransomware could be able uh theoretically uh to check that and then uh not encrypt these files and since i want them to be encrypted first um that is my uh my goal so uh what do i do if there's a change in these files uh i simply shut down the machine uh usually um uh crypto uh software doesn't have a uh persistence mechanism because files are encrypted anyway so they want you to pay so the current state of ransomware um so it's first executed obviously it scans the file system um it targets uh a set of um file extensions so pdf docx uh i don't know uh whatever they want and um then it starts encrypting um uh downwards from the from the location it started usually uh but it could be user folder the root of the hard drive and then eventually you've got the message well please pay us or you're never going to see your files again um and um so far this uh this works uh at least for the criminals sadly but the program also prevents it uh i tested it with uh uh quite some versions um also the world famous wanna cry um would have been stopped at least in my test machines um so what i might foresee in possible changes in ransomware is that they also include the header of the file so that's been something i've been working on lately to uh have the correct header for the correct file extension because otherwise it could be uh that the plain ascii text in the file is um seen as a honeypot and is uh skipped and if all the honeypot files are skipped the program uh doesn't work anymore so that brings me to my last slide is how do i participate um after the lighting talks are over for you can contact me outside in the hallway you can email me or on get up thank you okay then let's continue with the next talk and the last talk before the break riot the friendly operating system for the iot hello my name is martin lenders i'm uh researcher at freio universitet berlin and uh also a main contributor to the riot operating system um so most of you might want to know why do we need an operating system for the internet of things at all and why riot so if you look at the internet of things today we see that most manufacturers just use their own solutions mostly proprietary sometimes standardized but almost always over some central cloud architecture and also using some proprietary software and basically no one knows what they're doing and our vision for the iot is that basically everything is built around free software and free uh and well standardized protocols i put some riot logos in this lightsaber you can also can use other open source operating systems of course so that we basically if we want to connect our heater with our light switch um but why do we need a dedicated iot os anyway why not just use linux well if you look at uh the devices linux usually supports we see that there is a big difference between the devices riot supports and the devices uh for example linux supports and so our motto is basically if your device cannot run linux use riot right um but what does riot bring to me um well we firstly we are lgpl licensed software so we are basically free software we also support a variety of well-known and well-tested tools for our development process so we don't require you to install some eclipse environment or use some self-written IDE to develop for riot and we also support a variety of platforms like the makers favorites aduino nuclear as the astinucleo or the udder food feather but also educational platforms like the caliope mini and its british counterpart bbc micro bit and uh we also you might have uh seen it on hacker day that some of one of our contributors uh hacked the ikea light bulb triad free for riot so you can write run riot on that now and um here for the ones who are a little bit more technically interested um riot basically has a real-time capable kernel architecture uh a highly configurable modular design from the ground up it has a partly poeple projects compliant api uh fully featured ipv6 network stack and also javascript support via the samsung javascript interpreter and also some other third party software which you can also include via a bsd port like package management um so if i could spark your interest you can look at our website riot os.org look a little bit through our api documentation at api dot riot minus os dot org and uh look at our uh github tutorials if you want to uh test yourself out in the environment uh also we have several mailing lists our most active one is the devil mailing lists uh where we discuss most of our topics but we also answer questions at the users mailing list and if you find any vulnerabilities in riot uh you can confide with us at the security mailing list security at riot os dot org thank you thank you okay so now we're gonna have a short break of 15 minutes uh we will see you all back at 12 45 please give a big hand for all the talkers okay the break is about to be over so please switch to the slides again and then we can start with open source do it yourself laptop okay hello uh i'm lucas hartman and this is a quick intro to reform uh my open source do it yourself laptop why a computer should be a simple tool um the users should have all the power over it you should understand you should be able to completely understand and repair it and there shouldn't be any corporate or government intrusion that you don't want there are some similar projects like the novina that's also based around imx6 there's the pyra the new 900 phone so this is my contribution to the overall open source hardware movement the status there is there are two prototypes one is based around a development board and is quite a bulky box this is done and works uh the second prototype is currently in the works i have it here in this suitcase so i can show it later uh after the talk it's based around a custom board and it's much thinner here are some first sketches how it all started um by industrial designer anadantas we iterated on a lot of different styles and ideas how it could look like and for the first version settled on a very simple uh blocky design this is the very very first version uh made from uh laser cut acrylic uh 3d printed keycaps just to get a feeling uh of the proportions and so on um we then started to print our own keycaps in the form labs form 2 resin printer this is how they look freshly baked um this is the completed first prototype uh insides you can see there's a lot of cables and components floating so that's why it's so big um and me using derby and linux on it and it can show its own photo uh here are some more renders uh showing how it could also look like um uh there's general specs it is based around the nxp imx6 so c it has four arm cortex a nine cores at one gigahertz four gigabyte of ddr3 ram which is unusual for this for an arm system it has a vivante gpu and uh safe uh lithium iron phosphate battery and it's completely open there are no binary blobs in the system the gpu is completely open source has completely open source drivers um it's currently based on the tiny rex som but we'll move to the open rex som so it can also be replicated um yeah it has a bunch of i o for usb pc i express sata gigabyte ethernet lvds and htmi at the same time and the sgtl 5000 sound chip and there's a few um gpi os and spi i i square c for your own experiments um all the files are um are open uh as step files and you can print your own enclosures and parts the board was designed by me and uh keycard which is also open source all the components are easy to source and the input devices the keyboard and trackball are just usb devices and this is how it looks like in keycard is my first four layer board um uh here's a render and this is how it looks like in real life uh i was able with a lot of patches to bring this board up uh and runs some emulators on it here you see it almost fully populated usb hub didn't really work um and i was able to test a lot of software in the last weeks um debian linux runs fine with the mainline kernel uh firefox blender emacs io quake 3 um and player liver office all of these work quite well um and um i'm now working on the board revision number two which fixes a bunch of problems that the current board has and the case revision number two um and we'll start taking pre order soon for a complete DIY kit here is just a preview of the thinner uh enclosure thank you um you can contact me here mntm n.com slash reform or talk to me here later i have all the parts with me so thanks thank you next up is shit might hit the fan driven code good morning good morning everyone i want to talk about shit might hit the fan driven code uh because this year i had the feeling and i think i'm not the only one that shit is very close to the fan or as uh another talk here framed it very very nicely we might not go the direct route to utopia there might be um some detour and so i was writing some code mainly android apps to deal with my fears and um to basically make some digital umbrella uh because main thing is i don't trust uh the elected leaders we have currently and i just want to have a backup so um we will go up the pyramid of needs uh and i think it's not wifi at the bottom it's actually food and shelter um so i was checking all this like how to make fire and stuff and there were some apps already on the android app store uh but not a non aftroid which is really a problem for me um and i didn't like all these apps so i was writing one it's um based on the us army field manual but completely demilitarized so if they speak about soldiers it's humans and everything else um it's based on a wiki um this is basically the content um and i added some uh for example power because often people ask yeah there's no power when uh shit was hitting the fan but there are solutions to that you find that in that section basically um so there are ways around it how to generate your power let's edit content so please that's one reason why i'm speaking here also add your content it's a github wiki and uh that gets compiled into the app and it's also not um it's called offline like um sorry manual to help overcome situations without internet offline that doesn't mean like offline it works offline so the content is compiled into the app um the next thing i want to speak about is ipfs droid because i think like that's one step ahead of the pyramid now now we survive we have shelter we have food we have everything now we really want digital content and one really nice way to get digital content is um ipfs content addressable and uh that's basically the um android app to access that um it can resolve via the go ipfs note that's compiled into the app or also via via the gateway uh to not have such a big binary blob in there so please look up ipfs uh and ipfs droid another option might be swarm from yeah next thing i want to talk about is uh wallet um that's like the cryptocurrency thing is not really well received here but it's not all about money there are other things you can do with that and also like banks might explode after shit hit the fan um and you still want may want some means to um exchange stuff and don't go back to a raw barter and uh also none of the apps for ethereum were suitable for me before so i was starting the wallet project i started as an organization um it's also a lot of coddling code um if you don't like javascript and you like static typing languages um in the ethereum and ipfs space often is uh javascript is dominant but most of that stuff is written in codlin so now we are at the top of the pyramid basically we have some time uh all needs are solved um one game what you can nicely play is go like the game of go was also really interesting um alpha go now with the ai um you can also find that at the hive an installation which plays the alpha go um games with um that app basically as a continuous playing all the apps uh you find them on google play for your convenience um but you can also compile them yourself and in the middle from the convenience also always all the apps are available on aftroid uh your choice um the next thing i want to talk about is not um um android app that's um i saw that the cannery watch um stopped basically because it's too much work currently to monitor canneries and i really like um like ethereum not for the money like stuff but you can do like a smart contract please look up erc 801 um that's an erc to make um canneries so it should solve all these problems that are outlined in this blog post um yeah so follow up uh you find all that stuff on my contact pgp key and stuff on leaky dot de um contact me also like at this conference here just speak to me i might look scary but usually i'm quite friendly um have a nice day and a great conference thanks for listening thank you five lightning talks in one very nice so let's continue with uh technological sovereignty volume two hello so i am maxigas and i will actually talk about a book not software and of course it has digital versions as well but uh we printed out a bunch of them and they are available in the anarchist assembly in world two next to komona and um well this is a book written actually i should show you this um written by people who live in the um echo echo industry a post capitalist colony in katalunya that is called colorful it's a kind of housing cooperative and um productive cooperative complex that we are uh buying from the owner over 10 years and um technological sovereignty itself started with this idea from margarita padilla and he was saying that uh there are consumption cooperatives that buy the uh their food from local farmers together and why not people do the same with the technology so we found uh via campesina peasant movement that is an international political movement of peasant and they had this nice definition of food sovereignty and what we did was to adopt this idea um to technology so i will read this quotation from via campesina and try to imagine what happens if you replace food with sovereignty so food sovereignty is the right of people to healthy and culturally appropriate food produced through ecologically sound and sustainable methods and the right to define their own food and agriculture systems it puts those who produce distribute and consume food at the heart of food systems and policies rather than the demands of the market and corporations so for us to adopt this idea to technology was a way to go beyond the limitations of free software because free software was a very uh inspiring and powerful movement uh uh at least in the 1990s and by now it has been become part and parcel of capitalism so we were looking for a more holistic um approach um to alternative technological trajectories um the first uh three years ago we published the first volume and this is the second one that came out now um it was supported by the ritmo foundation from france and it was translated to italian by people from hack labo in uh bologna um we collaborated with a lot of other groups for example with coati who are a um translation collective so the first volume uh three years ago we put out in french, spanish, catalan and uh now in the anarchist assembly you can actually find the new italian translation of it um the second volume that we published now um will come out in uh it's already out in english and spanish and available in the village um there is already a digital version in french and catalan and italian will come up so one of the reasons we came here is not just to distribute the book but also to ask people if they want to help with german translations or translations to any other random language that you desire and speak um we published the book um also there is an original version so if you want you can get the digital uh book and you can have the original french articles in french and if somebody wrote in spanish you can have the spanish version so all together we are releasing in six different formats in print in html in pdf in markdown in mobi and um in uh something else e-pub so it's basically um six times six books that we made from the same material and they are all available in a jit repository so we use um actually i will show you this one so we have this toolchain based on the jit book library mainly that we use to translate markdown that makes it very easy easy for translators and collaborators to work on the book um just to give you a taste of what is in the book um the first volume uh looked like this and how is it yeah um sorry so okay in the um in the second book you find five four three you find a lot of different topics from uh thank you thank you all right um next up is formal verification for the win hello i'm wildhacks and my talk is formal verification for the win so it's well known that the number of bugs generally increases these lines of code and we have a whole slew of code around us um so embedded devices are what we have in our smartphones or cars or elevators or railways or wherever and they have code in them and there's a lot of them uh and even in 2004 cars had 70 computers in them in different networks and by now you can be sure that you drive a car uh that does not work mechanically as in uh when you press um um when you press on the gas uh there is a command center computer which operates the engine it's not mechanical and there is code in between they actually boast about it um so uh there is um stuff from the mundane like HDDs uh to the internet backbone which is critical uh and there are actually standards around it uh the common on the right is how often a bug is supposed to happen at maximum uh and it's really hard to test that a bug happens only once uh every 12 000 years when you're a manufacturer but when you release something in the wild with millions of devices consumers find bugs another issue with the testing processes you test for known issues and you generally don't know issues which are interesting uh so there are cases like uh a Toyota one where the car did not stop when you press the pedal just did not because they messed up concurrency and people did die uh almost confirmed death and a lot of injuries and they tried their best to keep it out of the court so they settled a whole lot of cases uh also look it up with Michael Bond Philip Kupman there is so interesting uh slides of inexperienced witnesses uh and they audited the code uh another legendary case is Serac 25 basically it's an X-ray machine which sometimes worked in kill cancer mode when it should have worked in scan mode and also killed people and also because they messed up concurrency basically um also little cases like uh sometimes Boeing Dreamliners uh AC system stops working because implementation issue sometimes petroid missiles used to miss their targets because implementation issues uh and also Chrysler messed up uh they did not put a firewall between the internet connected youtube player and the car engine because they used a can uh can bus which is not uh indicated by default uh so form verification is about formalizing this relationship uh and it's recommended to use uh and there are two kinds of it basically system verification limitation verification system verification is about your concept your architecture your protocol uh and implementation is about your code adhering to the protocol and it's basically an extension of types you are used to and i'm going to skip that uh you can track stuff with your types and there is also that the concept that uh in really high assurance and really high assurance stuff like space rockets uh you realize that your hardware and compiler are also uh potentially uh not adhering to the spec so you check them too i'm not going to talk about that form verification has been used to great effect uh in a lot of cases for instance amazon has used it to fix some bugs in their web services and uh also verify that some optimizations could be made uh and it's static analysis is used in planes in the space station in basically every phone every smartphone out there uh and also in your cpu caches it's verified that they behave correctly also some interesting military applications uh what i want to say is um things are getting harder uh and basically if you don't want the people to die you have to have really bad assurance process which makes iterations slow to a crawl and if you want to get to the space faster we have to use form verification basics is higher in form verification is in your engineers right now uh check out these materials they're cool thank you thank you next up is small modifications i would like to share with you some things that i have found mostly while walking around cities near the pacific ocean um i'm kind of curious how you might relate to these from your perspective um so here is uh the entrance of a driveway in Seoul Korea and there's a small slope on this ramp and so it looks like someone had cut up tires and fixed it to the concrete with screws so it doesn't slip and um about 50 years ago a guy called bernard rudovsky said something like this what we need now are not new technologies but new ways of living and i guess that's a question that i often wonder when i look at these things made by non-specialists using very primitive materials somehow they do something they change something in our experience of being alive so here is a photo taken by my friend chris berthelsen while he was living in tokyo japan and someone had added this tire to the entrance of the driveway which makes a difference um i will just read out another quote um by a guy called george kao gunham he said that life is improvisation meaning acting as circumstances permit at one point a blade of stone is sufficient as a surgery tool and stainless stainless steel was not sufficient yet we live like castaways so even though here are two in two different cities when the picture on the left is from mexico city the one on the right is from soul korea again people working in relationship to cars they make these modifications to um offer some kind of protection in the first example and this was found near the same car park where old posters has been reappropriated for storage and i feel like the same kind of caring could also be extended to other creatures so here's a bird feeder found in tokyo and often also when we think about boundaries like fences and borders it could be a site of separation but also a site of exchange between people and offering something to your neighbors and here is a soccer field made on a permanent in penong pen kambodia and the blue tubes are very ubiquitous in that city for water and here the same tiled ground has been kind of appropriated for anchoring this tent and this is something that i did with a group of my friends in new zealand before i came to living humbuck last year where we modified the existing furniture in the middle of the city so it could become more of an open home for strangers to meet and here we discovered fruits on the edges of a car park um i just read another quote by an anthropologist whose name i have forgotten he said to craft not beautiful and convincing artifacts but evocative and open-ended materials for further experimentation to creatively set the scene for distorted here and now and if you have any idea what this could be you're welcome to come and find me in the tea house thank you now a party for quality of life is needed okay so welcome for this talk you might suspect it uh because there's a the word party in the title it's something like a insta party or crypto party as especially as we are in a hacker conference but uh i have to disappoint you it's about a political party but i tried to make this a little bit uh entertaining so i want to explain my point using references to popular popular culture namely the famous series game of thrones who of you does know this series game of thrones okay so i can keep the summary of the plot very short from my point of view it's kind of a mixture between a soap opera and the fantasy story and the soap opera elements are that there are a lot of couples having discussions and making much love to each other then of course there's the iron throne which is the symbol of the power for the whole continent and then the uh there's this game who is to occupy this throne and this game is played of course by discussions and by discussions including swords and riding knights and even fire-spitting dragons and from my point of view the the important thing of this series is that this whole game who will finally occupy this throne is more or less pointless because in the north there is the horrible knight king and he has an army of hundreds of thousands zombie warriors uh trying to to um yeah kill all people and extinct all life on earth so the whole series is about setting the wrong priorities and this is the relation to the reality which i see because the priorities of actual politics are keywords like economic growth jobs competitiveness and recent time also national identity but the real challenges from my point of view are the collapse of the global ecosystem which is a crucial condition for the existence of humanity and also the looming collapse of democracy or even civilization i just want to give some buzzwords like climate change biodiversity multi-track resistant bacteria or in the second part the resilience against autocracy the ability for a rational discourse is declining and raw prevention if we ever had something like this is also getting weaker so one solution approach would be to change the optimization criterion but there's the problem that the attention for inconvenient truth outside of communities like this one here is nearly zero and we have prevailing mental infrastructures which favor consume and cheapest cure and uh everything like this so uh the changing the optimization criterion isn't easy but nevertheless our approach is to establish quality of life as a political political concept and therefore uh we think they intend to found a new party which has or which is centered about this concept uh might be a way to go because currently our society tries to optimize um economic growth and the number of jobs but these can only be the means for reaching quality of life but not be the the goal itself so quality of life and our understanding is just a mass compatible interpretation of sustainability or more precisely um establishing a lifestyle which will work on a global scale and on a long time scale and obviously our current lifestyle does not fulfill any of these two conditions and uh one very important key aspect is to deconstruct the dogmatic growth orientation so um yeah if you want to have more information you can ask me or just visit our website which is uh partyoflifequality.de so plq.de thank you very much thank you next up is borg backup i will quickly i have to set this up in the browser uh bad opsec everybody sees my bookmarks so now let's go it should work with the clicker i tried it yeah okay okay okay uh welcome i want to tell some words about borg backup uh it came into existence because we made a fork of the attic backup software and uh back then uh attic was uh not very well known but some people already discovered it and were quite happy with it there come this there comes this citation uh some greek guy said i found the holy grail of backups so it looks like he was quite happy with it and um how it came to life is uh attic was a five-year-old project it had a quite nice design and nice code but the problem was a bit the development was going on rather slowly and still some bugs left and the problem was you could make a lot of pull requests but they uh did not get reviewed and not accepted so it was basically going slower or even be stalled uh and after some months i think of waiting we forked it and uh that was two and a half years ago and now it's quite uh fast going and it's more or less a community project and pull requests get merged and it's quite a lot of activity uh the feature set it's easy and fast to use it's a command line tool uh it does content defined chunking uh more about this later and it does deduplication and not only based on a whole file but on chunks of files it does compression uh in january i think i will release the next version and it it will even include a z standard it does encryption it authenticates the encryption and it has a file system key value store and you can access it locally or by ssh it's bsd license documentation is quite nice good platform support linux bsd uh solaris kind of stuff um site when even it's python a little bit of siphon and see to uh speed up the critical parts and we have continuous integration system in quite a nice test coverage you can mount uh backup repositories so you can just look inside and copy files with your files manager uh somewhere it's about a deduplication because compression and encryption you maybe can imagine how that is done uh the deduplication is interesting because it does this junk based deduplication and for example if you have a virtual machine image and you just use rsync with this hardlink trick for deduplication it will often copy the whole virtual machine image to your backup disk because the whole file has changed but boric deduplicates the chunks and so you will notice that most of the chunks of the file did not change and it will only uh copy the changed ones um you can also work on logical volume snapshots you can deduplicate a directory even if it was renamed because it does not care for the name it only cares for the pieces of content that you have uh you will have maybe some inner deduplication because you have some duplicates in your file system but it's not a main point in any case you will have a historical duplication because every backup you make will be mostly similar to the backup you made the day before so there's a huge deduplication there and you can also have deduplication between different machines if they happen to run the same operating system or if you have the same data on them uh how is it made uh it reads the file and then it applies a rolling hash algorithm over a small window and it cuts the file into pieces and the nice thing that the piece cutting is not at fixed offsets it's happened it happens uh by content so if the content is moving the cutting place is also moving and then it stores the stuff into key value store and that's basically it uh currently we have uh the 1.0 release out since a while one uh 1.1 i mean 1.0 is the old one and yeah 1.2 are some crypto and parallelization enhancements this is the home page you can also grab me here thank you true thanks so next up is uh unstillable for langen but i think it's an english talk it just has a german title i'm here to talk about insatiable desire now this isn't a technical talk so there won't be any dirty pictures sorry this is the dirtiest that we'll get i think many of us are actually a little uncomfortable talking about desire including myself so it doesn't come up a lot and what i love about our congress is talking about the future and creating um solutions to our problems using technology and i would like to contribute to an element that i feel is a little underrepresented in that um so if we look at ourselves we have this narrator uh in the front of our head that comes up with very reasonable explanations for our decisions but our decisions in many cases are influenced by our desires uh not only um sexual desire but other things food money power love friendship so good things all i'm saying is that our lizard brain is in the driver's seat a lot of the time and that has implications for our society so a large part of our society is built around catering to our desires and selling our stuff and technology is coming up with better ways of serving our desires every day and you might say so that sounds pretty good what could possibly go wrong uh well if you take any statement um you need food to live and take that statement to its logical extreme then it becomes absurd so if you need food to live why don't you eat yourself to death and we are eating ourselves to death because food has become so good that it's addictive we can't stop eating even if it kills us and more is usually better it's just that our lizard brain doesn't know when to stop uh and the same is true for other things so if i have one billion dollars i won't be happy until i have two billion dollars or take drugs ever since legalization of cannabis started um industry has taken over for obvious reasons and there's been an explosion in the variety and potency of thc products which sounds good but we have no idea what that is doing to our brains um and i think it's easy to blame capitalism and consumerism for our problems but what if we have that the wrong way around what if they are actually symptoms of human desire and they are bounded just like we are because we are engineered that way and society is built around that well the implication is if you take anything to its logical extreme it becomes absurd and our world is built around fulfilling human desire um which is impossible because human desire is infinite um so there might be some bad outcomes from that say climate change and thinking about the future depending on who you ask in the future everything will be free we will live in the collaborative commons or in the future we will try to become gods or super intelligent machines will take over and murder everybody um but i feel like we're missing something i think that human desire will actually play a large role in shaping that progress and determining our future and i see this going one of three ways potentially so one way is we can just continue as is and enjoy the ride and it will be fun for us not so much for our children but anyways uh or we can find a way to limit or change human desire which could be good but could also be very scary or we can put our faith and technology to find a way to satisfy human desire without the external consequences um for instance build a virtual world that's just as limitless as our desire and then we can live there and um be happy without destroying our actual planet in the process i don't know if there if those are good solutions i think that maybe there are some better ideas um what i'm asking you is next time you discuss some technical solutions think about insatiable desire thank you thank you next up is dhcp anonymity profiles and implementations are you there um does anyone want to give this talk we've got the slides huh yeah i just uh okay he's apparently talking about the hcp servers and lots of computers that are connected to them and even the printer also you really do want to give the talk yeah we're windows okay yeah i'm sorry it didn't work out maybe he i didn't get an email or anything he didn't uh he just didn't show up yeah i think that's it so the usual suspects employed iris yeah well that's all that uh that's uh you know a tragic end to our session i would say so uh please give a big hand for all the talkers who showed up