 Ich nehme an, es ist hell hinter mir. Dann werde ich mich nicht umdrehen. Wenn Sie uns helfen könnten, wenn Sie eine halbe Stunde oder zwei helfen können, dann reden Sie mit mir, es gibt Kamerastoffe zu machen und das ist es. Gut, aber jetzt zurück zu Bernd. Du hast das Internet gebrochen und wir fixen es. So Bernd, fixen Sie das Internet. Und das ist fantastisch. Ein Applaus und dann die Stages für Sie. Danke. Okay, ich habe über drei Themen gesprochen. Wenn Sie auf der IPTE haben und wir die letzten paar Jahre auf der Net-Session fixen, habe ich bereits über die Höhe des Levels gesprochen. Und jetzt sind die Net-to-House Höhe des Levels sehr gut gemacht. Also bin ich auf der Höhe des Levels konzentriert. Das ist der Teil, wo Sie es nutzen können. Und diese Präsentation ist verwendet in der User-Interface. Ich möchte die Menschen als Browserreplassung nutzen. Denn unsere Browser sind zu groß und haben zu viele Attack-Vegatoren. Wir müssen einen neuen Weg finden, um Dinge zu verwendern. Aber erst lassen wir uns einen Überblick über die Motivation geben. Wenn jemand die Internet gebrochen hat, ist das nicht nur die NSA. Ich denke, wir haben es alle gebrochen, um zu viele komplexe Dinge zu geben. Und zu viele Protokolls. Und zwei große Browser. Und alles ist alt und über 30 Jahre alt. Wenn etwas über 30 Jahre alt ist, kann man es nicht mehr glauben. Vier Jahre ago Snowden pointed out how big the Internet is broken. And the politics responded to that by adding new threats to the Internet. So we now have a censorship, which approaches the way China wants to censor the Internet. Well, they do it, but it quite not works as well as they want. So now our politicians also want to censor the Internet and they have fake news and hate speech. So fake news means everything except Fox News, as we know. And hate speech is probably everything except applause for Donald Trump. We have crypto wars rebranded instead of increasing the numbers. It's now reasonable encryption. We have a number of dragnet surveillance laws, which legalize what they have been done before. We have the kill the link decision from the LGA Hamburg. We have people like Jim Sensenbrenner saying, you are not forced to use the Internet. And I don't know if it's good or bad, because 2017 was quite exciting for crypto. Crypto now means bitcoins. It no longer means encryption of your private messages. Forget about that. Crypto now means bitcoins and you can make a lot of money with it. This is from last year with the competition. The tour project faced some kind of Stasi like Tsar Setsung. And maybe due to their link to WikiLeaks, which faced the same sort of Tsar Setsung six years ago. And we want to have some solutions. Net2o starts to become usable. Gnounet starts to become usable with secure share. It's going to look a bit better than it looked before. So, for those who haven't been here last years, an overview over the layers of Net2o. So layer one is physical layer. That layer is not part of a real protocol. It's just the wire you have. And for the existing Internet, you use UDP as layer one. So you use the existing Internet with UDP as very lightweight as layer one of all those protocols which stack on top of it. And when we finally replace the Internet, we can remove UDP and do it directly on the wire. Next layer is then label switching, path switching. So you can join a number of short labels and switch packets through a path. The source has to route the packets, not on the fly. That is quite good for high bandwidth, for scaling. And when you have big tables on the routing tables on the Internet and when you have insecure means to announce routing tables like border BGP, then you want something more secure, faster and you want something different. The way the Internet did with routing and every hop was quite good for the time being, but you can do better. Next layer is we need to encrypt everything. There's no encryption. You need to have modern encryption standards, not the old one. You don't want to carry a big library of encryption you never use, because it's already broken. Next layer is the, let's say, TCP layer on the Internet. So you need some flow control, some congestion control, some retransmits if packets get dropped and that is here is timing driven instead of window driven. So even when you have packet losses, it's not going to slow down. And then I need a layer where I pack structured data and where I can send commands. And I do that with a stack oriented tokenized command language based on force, but not Turing complete. If you know Bitcoin smart contracts, they use about the same thing. Developed at about the same time, completely independent from me. So this, and if you look at the Java virtual machine, you have a Turing complete machine that is also using stack based language. Then you have the layer with files and messages as distributed data and distributed hash table. And on top of that you can put your application. So last year I gave a talk about onion routing as to provide anonymity, but onion routing has a big problem that is timing correlation. If you can sample enough package, you can de anonymize everybody who is using Tor. So that is for low profile targets, it's not feasible. If you're just using Tor to browse your Internet and go on Facebook without being attached to an IP address, you are still revealing a lot of your data to Facebook, but the de anonymization is not really feasible. So that kind of silly Tor usage is okay. But if you are a target, Tor does not help you. If you like Silk Road, if you are a journalist, which does things Donald Trump or the Chinese or Putin doesn't like, Tor does not help you, they can track you down. So one topology that is better is a mixnet, but a mixnet has a big disadvantage. Mixnets collect messages until the buffer is full enough and then they spread out the messages in random order. So no correlation is possible or not very easy. But that takes time and people want real time experience. So what to do? The easiest solution is just to push out garbage with a constant bandwidth and if you get something in, you push out the contents instead of the garbage. That's what I'm going to do instead of having just onion routing and timing attacks, add enough garbage so that there's no timing correlation possible. And to get more bandwidth, I plan to have a multi-pass solution so you're not going to use just one switch circuit but a number of switch circuits which each give a guarantee of a low bandwidth. But together as aggregates you have enough bandwidth to have fun, to have real time fun. That is plan on the lower levels. All the other lower level part is finished. I just not finished with anonymity part. So now we go to the browser, the user interface part. The code name is Minos 2 because 20 years ago I wrote a user interface in the time where Delphi was very popular and I decided when they have Greek I use Greek mine is Minos and I completely rewrote that 20 years later using OpenGL as basis because you have OpenGL on any smarter device like smartphones, laptops, even on Windows there's still an OpenGL driver available. It's possible to use also Vulkan but I think it's overkill because it's not a game engine. It's just render your text. I use FreeTypeGL for the fonts. FreeTypeGL is a very small project and they are not working very hard on it so I forked it off and edit the things I needed. I use Soil 2 for images. It's also a very small project and the small project has a big advantage that the attack surface is not so big. I want videos to textures and unfortunately that is a lot more work so I need to use the big frameworks like OpenMax on Android and CheeseDreamer on Linux. When you are familiar with the coordinate system on Windows or on X you usually have this coordinate system with the origin on the top left and in OpenGL it's like Mars, it's on the bottom left. So you have to rethink a bit. Type setting is done like Latex. You see, I can even do the logo of Latex this kerning, this rising, lowering. So you are not going to have a poor type setting. Fortunately Latex is not really complicated. It's just powerful but it's a relatively simple program in today's standard. It's printed in one book so it's a few hundred pages of programming code normal things we nowadays see. You couldn't print them out like a Mozilla Browser. You would run out of paper before you print out all the source code. I want to have very few classes. I want to have a very lightweight, object-oriented system. I want to use combination, combinatoric stuff, not many different, slightly different things. So, those are the base classes. Editor, text, emojis, icons, image. You can have animations. So, essentially everything that is visible. And you have boxes, horizontal, vertical and stacked on top of each other. Latex doesn't have stacked on top of each other but the other two Latex has. So, it's based on the way you do Latex. And you can either render to texture or to display. And so you don't have too many things to consider. And by combining all these things you still have a lot of power to do your rendering, to do your markup and build frameworks to display things. Because it's not like HTML is for everybody. You know why HTML is so bloated because everybody wants to render something different. Somebody wants to render a map. The other person wants to have a small game and just pushing it through. Already structured text is not the right way. The layering is wrong. You put the structured text on top of the objects you want. So, I start with the document object model not with the structured text itself. For animation and things you want to minimize draw calls. So, I have drawing different things in layers. Each of them is one draw call. Only the image takes one draw call per images. So, I think we can go to the next point. In 2017, the big thing was crypto and crypto means some money on the internet now. It does not mean send your messages. So, I thought about why do they want this Bitcoin? Why do they want the blockchain? What's wrong with it? What can we do about it? What's needed to do? What's not working well with the system? Can we fix it? Is it broken? I think yes, it is broken because Bitcoin is just a proof of concept and it way outgrows what it could do. So, it's not broken in the sense of Bitcoin as proof of concept is wrong. It's wrong when you scale it up. That are my topics. It's not just technical. It's also ethical. It's society and it's about history because I think one of the shortcomings of Bitcoin is that the concept of money is a libertarian concept of money. It's not very scientific. It's not modern. Let's go about the history. Fortunately, when we talk about the history of money there are things first. We don't have to leave China and essentially all Songdynasties a thousand years ago. We can learn a lot from this history if you look closely because they made all the mistakes vested in the industrialization 800 years later. So, how did money start? Money started with the shells. The cori shell. Money is about 30,000 years ago as the first cori shell object found. Use a jewelry and use to exchange things because cori shells are somewhat rare. They are very difficult to imitate, to fake. Even today you would not be able to fake a cori shell. Even with a high level of sophisticated material you cannot fake a cori shell. It's nanotechnology. It's the living thing. It is really difficult to fake. So, that is an object with an inherent value just as jewelry, but it is inherent somehow. You can exchange it with other objects with inherent values. That is the oldest concept of money we have. The bad thing about cori shell is that it is gaining value the further away you get from the shore. Because it is a maritime snail and you have to collect it on the shore and when you are in the mountains a cori shell is pretty valuable. So, the people in the mountains came up with an idea they mined metal. They made imitations of cori shells with bronze. So, we had the first metal coins. Also in Greek we had an independent development metal coins. So, metal coins is where you mine where you have the inherent value because the metal is rare and there is a lot of effort to take out of the mountain. The problem with this rareness is if your economy grows fast you run out of money. You have a lot of valuable things to exchange but nobody has spare money. And the Chinese came up the banks came up with promissory notes so it was not the state the banks who issued paper saying we give you a thousand coins if you hand in that paper. And the merchants liked to use the paper and the banks liked to use the paper because they could over commit their coins. You don't need to give everybody coins only those who actually wanted coins. And next step was that the state discovered oh, we can do that ourselves instead of minting expensive coins we print cheap paper we steal on it have a controlled procedure and punish everybody who makes fake paper and yeah, what they discovered in the process is we don't actually need the coins. The paper is good enough. That was the last one in the Mongol Empire in China they dropped binding the currency on silver they just issued paper that mild inflation of about 2-5% per year because that improves the economy people spend more when there is an inflation when the inflation is not big the function of money still is good so over 200 years money became a factor of 100 inflation but it's not so bad year to year is not bad it's just people who want coins don't like it and deflation is much more problem than mild inflation big inflation of course is also problem money works by the constant what you want for money is that you have a constant factor to exchange it with other valuables mostly constant slight changes are okay but big changes like the bitcoin spikes up and down that's not okay so yeah, what's the problem with bitcoin the lower problem is the value is very instable and the bubble it looks like a very textbook bubble development so it's now in the phase just before collapsing when it follows the textbook we don't know, we never know if it follows the textbook exactly but it looks like a textbook bubble but from a technical point of view the proof of work is way more dangerous than this bubble development because proof of work is dangerous to the environment you have to I think we have now 30 terawatt hours per year for just mining bitcoins this is an incredible amount of energy we need for a relatively small amount of transactions that are signed with these expensive hashes and it's not even very secure because the attack vector bitcoin looks at is what happens if someone fakes the last block that's not the most profitable the most profitable is fakes the entire chain because then the mining here was very cheap and the mining here before the bitcoin exchanges started was even cheaper so when you fake the entire chain you have about 16 million coins now you need to invest more work than was invested in the real chain but most of the work in this current chain has been invested in the last three months it's not before mining was not so profitable so faking the entire bitcoin chain and getting all the bitcoins as reward is a way better attack than just faking the last block so it's not very secure then we have the problem with deflation deflation works like this inflation is money is cancer deflation is in fact nobody is going to spend bitcoins to buy something everybody is using their bitcoins to speculate they do day trading maybe and then they need transaction but to buy things to sell things bitcoin is useless so typical sign that even spam sending is now you get spam join the bitcoin wave or the israelium wave and you don't no longer get Viagra spam so this scheme is very profitable yesterday somebody from the bitcoin group said they want to do the exchanges on chain so this decentralization that is a good development because if you think only works for exchange for speculation all you need is the exchange and when it's decentralized cryptocurrency the exchange should be on the blockchain that is not in some bank like Mount Gox which went bankrupt if these exchanges will go the same way when bitcoin collapses again so there are things to do and that's about ethics so the bitcoin people are pretty religious often and they are like these banksters so they say have you a moment to talk about our lord and save us a free market and I think they have 3 seconds and they will go day trading again so it's maybe not so bad as moments or Jehovah Witnesses but there is a religion about free market so everything and the bitcoin universe already has a verse just after few years has a verse distribution of wealth and the liberal economy managed in the last 30 or 40 years and inequality means that society is going into self-dom if you aren't a bitcoin millionaire you have to work for the bitcoin millionaires and that happened in almost any society where money was deflationary and the economy grow fast so money is attracted by those people who already had money and they make more and more we had the robberbaran economy in the wild west and in the Song Dynasty China we had about the same problem in north China they solved it by a revolution and killing all the gentry and in south China they solved it by an organized change of how the state worked to a more socialist model but something needs to be done to be fair and equal so this deflationary system is not fair it's not equal it makes people rich and most of the altcoins follow this model because when you issue an altcoin you want to become rich quickly so you make a white paper you collect bitcoins you give out your coins your now worthless coins and hope that you find bigger idiots buy them and sell them to even bigger idiots and that's how these Ponzi schemes work it works quite well until it crashes so now about the actual technology beef what do you want to prove when you have a blockchain what is the challenge the challenge is you don't want double spending you have digital money is just a number numbers can be copied without problem we want the number to be unique uncopiable just in one blockchain and bitcoin came up with this proof of work it sort of works and it burns a lot of basic minor energy it is very costly and it is marginally secure and what is also doing is centralizing the process again the idea was decentralized but now we have five minor pools all in China and they do a centralized proof of work they distribute the actual hash computation to decentralized parts but the selection whose transaction goes in is done by five pools and if they don't want your transaction to go in it does not go in so Ethereum such as proof of stake and yeah that works because we have said that it's called banking system the banks have a stake we trust the banks because their own money is at risk if they go away with our money but that also has problems because money corrupts and since 2008 the banks don't trust each others even though they have this proof of stake so proof of stake is probably not much better than what we have now with the banking system it's not a replacement what I suggest is proof of well behaving we must see that the attack for double spending is not just trade your own blockchain it's also you must make this blockchain public so you need a man in the middle attack so people already know where to get the most recent block and they trust the person who gives the full node who has the most recent block only this full node can misbehave and also when we look at this game series when you just sign one block you can decide do I cheat or do I play honest each time you can decide a new in the bitcoin universe the signature is completely anonymous it's just a hash nobody knows who did it so it's just one decision each time you can cheat now you can play fair next and nobody is knowing you cheated this time we know from game theory if you have a repeated prisoners dilemma so you are known when you cheat now nobody wants to play with you and this repeated prisoners dilemma is what makes people trustworthy even if they are gangsters even if they don't want to be honest if you look at the mafia this ometta and that works by repeated prisoners dilemma you don't cheat on the godfather you don't cheat, it's not possible you will be dead and the punishment is much bigger because you want to repeat your actions you don't want to make it one time cheat and then be honest member again so sign the blocks know the signers use the PKI system have crypto parties know the persons who do the signing you don't need to know the persons who spend the money that can be anonymous but trust is much better if it's not anonymous so last question is how do we scale the blockchain in the bitcoin blockchain you know it has about a few transactions per second when we look at the competition micro payments on the internet like widget pay they peaked off last chinese new year with one half million transactions per second and that is for one tenth of the human population so if everybody wants to use this online micro payment system and send friends, small gifts a few cents each and you will want to have five million transactions per second and it's not going to work when it's on one computer when it's one blockchain no matter how much bandwidth you have how much signing power you have a centralized solution is not going to work well so we want it decentralized we want it not just replicated ledgers, we want partitioned ledgers and we still want to be able to check every transaction not like the lightning network where we do all the transactions in the bitcoin network offchain and maybe we have to orbit all the corrupt transactions that didn't arrive and then the blockchain would be needed and would quickly overrun the limit of transactions onchain is not enough to orbit a lot of broken offchain transactions so what I suggest is to use the standard practice of bookkeeping you have double entry bookkeeping and that allows you to cross check all your bookings even if you have many books so if you have a distributed partitioned ledger all of your transactions going from one ledger to the other must sum up to zero no matter what you do the accountant has no money in the system so all the entries come from outside everything you add has a balance, has a negative you have a debit and a credit entry in the system put money into the system you say we promise you we give it back to you that is the credit and this money now is in the system is the debit so you can quickly check that all the bookings are okay and you can partition the ledger by coinpupkey and you can use an n-dimensional ledger space to have only small parts of the ledger space talk to each other in each cycle and then you switch to the other dimension for the next cycle that's what I'm doing I'm now developing the concept that's what I want to add as application part before I thought using no tala is good because it moves all the hard part over to the banks but one problem is the banks don't trust each other so in capitalist America bank drops you you have the problem with Lehman Brothers with the complete trust of the inter-banking transactions eroded in 2008 so that's why the bank want blockchains they don't trust each other and the last thing I want to talk about is how to mine coins so normally a blockchain is just for assets you can do normal real world assets in a blockchain you can do euros, dollars you can do real estate in a blockchain you just must make sure in this blockchain because if you hard fork a real estate blockchain the real estate would be in this bitcoin and bitcoin gold and bitcoin cash 3 times which one is the right bitcoin we don't know so you must when you do real things in the blockchain it must avoid hard forks but how do you use the coins the typical altcoin is mined by creating a lot of coins and before you start with the project or while you do the mining and I don't want to have expensive mining for signing the blockchain I don't want to use expensive rewards there so you have to you need a different process of mining the idea of mining is you have to do something difficult you have to do work but it should not be automated work because it just burns energy it should not pollutes environment so if you mine gold is not ethical mining mining gold is one of the dirtiest things you can do in the real world and mining bitcoins is about the most unethical thing you can do in the virtual world because you really burn fuel for value which is compared to the rest of the acumenomies about nothing and my idea is a number of free software projects other scientific projects live from donations now we give the donation to the project we don't get anything directly back we get the software back we get the knowledge back but we could just issue a voucher a tradable voucher return back that voucher and you could trade your donation if you decide I want to donate now because I think this project will become great next year and many people will donate next year I donate now next year to all the other people who also want to trade with these vouchers and donate to this project and that is of course self-interest of those people who want donations it's also in the self-interest of the people who give donations and it is in the interest of society because we get funding of free software we get funding of other interesting projects where we now have difficulties because people now donate the money is lost forever they cannot reclaim it by selling the voucher to somebody else they must make a tough decision this would make the decision maybe easier for people to donate and add an incentive to be speculative and donating to projects which are early and not donating only to the big projects which already work well and are well known when the project is well known the price of its donation is higher than when it is new I think that way we can use the hype about bitcoin and other coins instead of just wasting more and more energy and that's it I have two links now and you find a lot of more links on when you serve sauce so, yes, I am ready for questions thank you very much Dear Bernd, questions from the audience now I have a question about Minos if I got you right you made this presentation with Minos is that correct and how did you do it or what would I have to do if I wanted to make a website at the moment it's just the document object model so you have to fill in the document object model with entries and I can show you the source code but later you shouldn't do that in source code you should use something like markdown but I don't have a markdown pausa right now the source code of the pages itself first I define some macros like here this latex now that is a macro definition I want to render the latex logo I need some string, I create a text object, a small text object and I need to do rays and kerning but that is just filling that is all written in force and we will have a talk about what is force in the evening at 1830 all the natural is written in force and fortunately this helps me to write this macros quickly I don't need to write a new programming language that is how you how you then create a page so wrap up, you have a frame on the bottom with a color you select the color of the font, you set some baseline skip you set the font size as large, you create a text you arrange it center change the color of the font and of course when you have a markdown the markdown parser would just go over the markdown and create fill in the document object model with the text it found in the markdown but here everything is done manually everything is done with commands write in the source code fill the document object model this is manual work I am doing now because I don't have the higher levels, the markdown parsers I will add that later when everything in the document object model is ready I can add a parser on top of it I think markdown would be my favorite for this kind of application because everybody knows markdown people seem to like it unlike html was the like markdown so that would be when I have the markdown parser ready and you want to make a presentation write it in markdown what is the overrun the arbiter attack you said the lightning network is susceptible to overrun the arbiter attack not familiar with that look you have 5 transactions per second in the blockchain the lightning network works by not checking the actual transactions and if you think your transactions didn't come through you push the lightning transactions to the blockchain so let's say you have a queue of 5 hops from A to B on the lightning network and it didn't arrive at point B all the nodes in between to the blockchain you have need 5 transactions in the blockchain but let's think we do a million transactions per second in the lightning network that is entirely possible no problem and 10.000 of these 1 million block transactions don't arrive because some evil persons took the money and ran so now we have to push 10.000 lightning transactions into the blockchain to do the arbitring and the blockchain can only do 5 so the risk of getting caught when you run away with the money in the lightning network is very small and this overrun the arbiter actually is a real world phenomenon if you order something in China and you don't get what you want you could go to court in China they actually have a court system but it's completely overrun so the likelihood that this case goes to court is about 0 and that is why the Chinese merchant cheat you it's not because there is no law it's because the arbiter is overrun thank you thank you the proof of work we have now is already centralized in bitcoin we have just 5 pools the actual work is decentralized but the choice which transaction goes into one of those pools is made by a central instance and work has economy of scale and if you say who is going to do the work with the proof of work system it is going to be those people who can do it the cheapest so that is why we have miners in Iceland, in China in Mongolia because they can do the work which is actual machine work it's not human work, cheapest and that is centralizing by itself so when you have proof of trust you don't give a big incentive to the people who sign the chain because signing is not expensive the operation is not expensive but what must be expensive for them is to gain and retain the trust so that means auditing that means you have to be known in person that means we check where you live if we can get you into jail if you cheat if you say I'm from Nigeria I want to join and we figure out there's no way to get you in jail in Nigeria but when you cheat us we will not accept your signature we will rank it down at least so we can use existing systems of keeping people honest and we can also build our own we can make sure we meet the people at crypto parties we know they are honest we know what person that is and we know it's not this thousand people is just a sibling attack and there's really only one guy behind them because we want to show this guy we want to see this thousand people, we want to see them and that makes sibling attacks difficult as PKI is not an easy problem is not a problem that can be solved with technology alone we know that from our crypto parties with PGP and we know that people don't quite like to come to these crypto parties and but when you do something which is a has a high responsibility we can also ask you to do the effort to prove you are honest and real and about science and ethics ethics, we are talking about economy economy is game theory and so that is mass but the result of game theory is ethics if you do it right so ethics is a scientific topic and you can use mass to do it any more questions from the audience thank you i have one more question isn't it that there is actually some value behind bitcoin because it's used on the black market so it has basically some unethical value one could say mostly and i mean there are these speculative transactions but you also have this in other markets i don't know like silk road or things and i mean in real world markets you also have trading bots which also do only speculative transactions and they make up a large quantity of the volume yes of course market means that you have speculation and anonymity of currency means you also have people who want tax evasion and illegal business that is clear but that is something in the real world currencies also facilitate you can use cash and it's just maybe more expensive but when the transaction fees of bitcoin rise higher than the transaction fees of sending a suitcase full of money to somebody else maybe the drug dealer will go back to cash and not use bitcoin we don't know it's just take the cheapest that's available and i'm not for full anonymity for completely hiding like monero or zcash where you have an expensive operation so you can even hide the amount of money transferred and hide the networks where does the money go from go to i think had this overkill for normal for honest transactions you just want to vanish in the mass like what i said before with the tor network there is anest user which is not a target is safe so i want to have a system where the honest user is safe because there are so many of them and the target is not safe so if you do silk route your onion routing can be broken and your bitcoin can be broken they did it with silk route so even bitcoin is not secure enough for this kind of silk route which is why i heard that most of the drug dealers now use monero and no longer bitcoin i don't want to be part of the race for the most unethical ethical business that others can do and they will suck away the bad guys from the simpler and less cost efficient systems when you do bad things you want to pay the cost for doing your bad things the good people don't want to pay the cost so make it simple make it secure enough for normal transactions where you would have to sift through a big heap of data with a lot of computation power to find everybody that is okay for me i don't want to be perfect with anonymity money exchange it's not ethical because just look at what kind of bad things people can do okay audience there's a more question this in total sounds a lot like not sensor resistant very controllable you control who has trust so it's easy to sensor you don't want to be used like real money you cannot do bad things with it so you decide what is a bad thing to do and it shouldn't be used for that the arbiter doesn't know what you do with it the arbiter only knows you make a transaction and due to the trust system if you as arbiter decide you are not going to let this in and the next arbiter decides i'm going to let this in you may lose your trust that is a danger so if you try to sensor as arbiter there's not just one arbiter there are several arbiter if you try to sensor you might lose trust because the other arbiter all takes this transaction you become a minority and if you misbehave you lose your trust you get thrown out of the arbiter system so you wouldn't do that you know how much money it is but you don't know is this buying a car or is this buying drugs but you just said that you don't want to support people buying drugs on the black market and you're fine if the anonymity is not that big i'm fine if you can figure out who gets all this money and when the police already knows he's selling drugs they just want to have a proof that he's doing it then they can figure out who is this guy like cell crowd the police knew there is cell crowd it exists we just have to find the guy who aborberates him Bern, thank you very much applause for Bern