Merken

Keynote: Towards a more effective, decentralized web

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
but I think a lot of very happy to be here and I'm in both of the budget committee since about 2
thousand and 1 I've given some probably talks so also Python Conference cycling Europe uroplakin mostly but
also some smaller ones and I'm still getting nervous in front of large audiences and I think the reason for that as somebody said this very intelligently is that imagine yourself in tight situation and you have lots of people staring at what does that mean that means they're probably going to kill you you know because naturally you don't have situation many people are doing this that's I think it's very natural to and that's kind of like the way for me to suit myself a bit OK that's how it is I try to get over it and show that many of you are doing the 1st talk FIL is even more strongly OK so apart from being heavily involved in Python part project partners many tools and things in the last 2 years I been going around in other communities uh mostly people from node estimate of the language Haskell and also meet small made up of people we don't focus on 1 language but actually on concepts and so I'd like to tell you some stories of what I learned from that time under the general topic of a more effective and decentralized web and for that I'm going to take you back to the
past in order to talk about the future so all this is some real rocket
science that happened almost 50 years ago that's the I think Apollo 11 that landed on the moon and was obviously rocket science because you have to to steer the rockets and it's still the speed record for humans traveling is still aligned with this property is it's something like 35 thousand kilometers per hour so and has been beaten yet and in fact I argue also later in the talk that at that time in the sixties and seventies actually the advances and the rocket science advances in kind of slowed down leading from 16 85 on every 15 years the number of scientific papers double you may relate this to was law like doubling of transistor was but it's the amount of scientific papers every 15 years since 16 85 double and this level beginning in the seventies so we don't have this increase anymore and I'm going to get back to why this is so at the end of the talk so he was doing this rocket science um was was programming the rockets to land and to fly that was Margaret and she was the lead
programmer and she's standing next to the source code I guess written in assembler and she let the whole programming and thought for
steering the rockets by the way print out the source code behind the source code of fire for
Linux you get something like 50 meters them another interesting thing is in the sixties more women than men were actually program OK any of you answer why that changed In the that you they marketed for many yes actually there if you look into feminist research that there's a general observation that the more professionalization and the more money gets into some kind of area and you all know that a lot of money actually went into programming the more mean dominant that area so basically the programming at that time was considered a minute just the thing to do and you know it's typing like on a on a typewriter and
so on so it cannot be that important and they didn't realize actually you know the meaning of what this program in what's sold at that time and this was like 1 of the leading rocket scientist another thing happened but leading
up to some some the convention in 1974 and I'm going to talk a bit about networking here and in fact 1939
telephones had pulse dialing so when you actually have a phone number of several digits and you you dial the 1st digit then this would not go to the relay and switch the cable you know that the next number would go to the next relay and switch
cable when the next digit would go to the next relay and switch cable so basically dialing your phone number means switching cable across the continent to the southern end point another telephone that was all operated by by the and like in the US it was also a period of 1 company and they had this big network and they were total control of all the relays and the hardware and they spend a lot of time to make room to make these relays and the whole network very reliable sold in 1974
the rocket science invention and change things and it was completely
despised by most like 98 % of the research this and it doesn't make any sense we have research we're doing real research into following that works and this stuff we
actually have a modem and you use the telephone network in order to do an
overlay network which was firstly that and then the internet protocol cannot possibly be more effective because it's using the same phone network also did doing something on on top of could this possibly be more effective well what they did they assume that you're going to establish connections and between routers well some phones connect to and then they're going to send data and it's not going to switch cable across the continent and then send audio data would we be going to chunk of audio data into packets and each of the packet gets a telephone number but it's a higher level telephone number it's the appearance the Internet Protocol Kidder and by doing that you gain a very big
advantage that wasn't very obvious to most of the researchers at the time of the beginning of the seventies which is you don't have all of the these set up if you think about the the pulse dialing that takes quite some time to just set up the connection to the table basically across the country if you already are connected to and you send the data packet with the virtual telephone number then you you already have the line you already connected and you just sending at all and the router makes a decision on how to fall into that packet so there's no set up costs the lines are already there are connected by the model that they envisioned doing this was a very distributed
network lots of nodes all the income interconnected that came from the military you know that if some nodes actually don't go down you can just rolled around that was the idea and you can run around it because the notes that you are aware of the connections and they get a packet and some broadest down the can just run it somewhere else that was the idea well turned out this was a bit of a dream but I kind of like everything is connected and like everybody can talk to everyone and all the nodes are not enormous and all this kind of thing what actually happened fast forward to 2009 and that's a picture of the Internet traffic in the UK 2009 and the logic of the sphere the larger the volume the more traffic goes to this telephone and point because the IP network is still kind of using the same idea in the original telephone
network you still actually addressing some kind of point you're saying they connect me simply some data get me some data from this remote endpoints why did this happen why do we have not
like this evenly distributed network but we have this rather we have like lots of stuff on networks works for basically lots of people access 1 kind of the I think the main
reason is economies of scale and that means that companies realized and will was 1 of the finest if I am the following endpoint everybody talks to but I need to handle all the note that some effort and they need to be very reliable and stuff like this but I get a lot of overview about who is accessing what and that turned out to be in terms of of tracking and and interjecting advertisement to be very profitable the most profitable because the model scale up and more people are falling you the cost for that does not increase In the same way that your profits go up by knowing more about people and there's 1 guy actually last year who put it very and clearly has been
doing that represents a lot of and he says that
this effect just the crop describe it means that there is a complexity task and complexity context that means that the more knowledge you have to handle you have to have load balancers all kinds of things and that the further out you go on to more nodes actually forming you and you have
additional costs but the text the complexity test tackle tax is regressive like going from 1 user to 100 is 1 thing going to 10 thousand uh it's a bit more cost but less than 100 then 1 2 million and that's kind of like seeing per person gets less and less and then 500 million
is it's even even less that's why it's progressively you pay less and less for getting more and more users but you know it's kind of like a constant income from from advertising and stuff like this and that's why the economies of scale mean the bigger you get the network effect and
everything that the more money you can make and there's a there's a tipping point basically at which point they become very profitable so fundamental and going to get back to him later structures all more detail the history of
networking there need jackhammers but who is was for 5 years he was 1 of the 1st Facebook some researchers he could in 1 of the 1st 100 in the grid 2011 and he was
looking around and get into that look at Silicon Valley the best people actually the best i t minds that focus and all that on how to make people click more you and it I think it's something that and is understandable a lot of money actually was to be made from working there very good working conditions and so on but I do believe that it's not really rocket science and it's rocket science to
a degree but it's very much focused on 1 particular purpose what about getting us into space flying cars whatever OK
so what we ended up with know this from what I call a million to 1 architecture so many different billion to 1 architecture so be spending a lot of effort actually on how to make like in the web how to make very scalable websites
because of this term regressors complexity task context and lots of start ups actually trying to tap into this and to become the mediator of communications the thing is people want to connect to each other like the future is distributed in the sense that we want to be able to talk to whatever the 1 they want to be able to talk to and to listen to see videos on YouTube and get connecting on Twitter and peer-to-peer basically socially on a social level we actually interacting among us but there there's there's mediator thing that actually I monetizes these interactions but it's still on the whole network effects actually driven by by people who want to interact in a very distributed and self determined manner and I think to a degree that the I think that is very
big and the Python community currently big Data pions on that getting enough good jobs is also focused on you know computing on large amounts of data collected from humans interactions and
so I think that that's just description and you can also probably see that I'm a bit too critical of this development but it's it's also description this is what happened so
far now let's go into space again
and 11 years even must also does the test power and many other efforts he wants to get immunity to the mask base so he wants to
have the master station on there and I ask you do you think that using TCP IP and HTTP the good old 40 your 20 year
old age to 20 years 21 and I P is like 40 something 41 and thing that these old networks are going to work on but can you call Gmail from others as about that patients data with a lot of
patience yes I think it will revived the idea of male and so OK so it's clear that want to have to want to think about networking
and we think about rocket science again and going to my then the particles that we have not going to work environments might say well but you know we can just kind of special solution from what we just do something else and we just have a good old e-mail servers and but of Chancellor people actually charge inside the mosque colony we
can just talk directly to each other and forget about http and DNS and this l and all the fun of self just don't do it OK but certainly
uh somewhat pragmatic solution but I argue that we have minus already in the ball we have not already
in many places in the world where connectivity is that the more people connect that you don't have the band with things disconnected go rule and so on soul and the idea getting more and more devices everywhere
like in in 1981 we had 302 computers and the Internet now we have billions and is still using the good old telephone particle a bit of a higher level than the pulse filing in 1939 but is still using this protocol except it's not actually true anymore when you connect devices the at Apple
devices all kinds of Internet of Things and so on there's all kinds of protocols that ask people spoken actually locally so you have synchronization mechanisms between your
mobile and your laptop and they don't go we to do the you know some memory remote and try to synchronize their although
sometimes you have to do that if you were meeting someone else and you want to transfer a file you actually figure out what's it seems to be more practical to go all the way up to you california's somewhere and exchange final there is that being able to send this wrong but you also know that especially apple I think is doing more and more in the area of this get appear and finding each other on local networks because it just much more effective you don't want to have this uploads they're very bad basically usually reflect the download links but very so upload links so you want to transfer some big files just very annoying to actually go through a small upload link so it's very ineffective so but this synchronization mechanisms they
are most of the time I mean a lot of the time they are proprietary and there's no which for that so there is a group I think it may be best in that
garden that says well when we went from 300 to billions of devices but maybe things changed in all our devices actually got more and more powerful but points like more and more CPU and they have lots of local connectivity and maybe this model of falling around the world all the time is not really how we want to
organize data so that's something that's going on in the mobile world a lot where Python is not a big part of that happening there are a lot that people actually say OK I want to have a local application that works and with offline and I never some checking or something but at some point I'm going to get a connection and then
centralized so it's not like this client-server no constantly falling somewhere it's um it's really about thinking offline
1st and thinking about synchronization the if you actually think about the number of very
successful projects from the last 10 years many of them using synchronization and replication mechanisms that don't work according to this client server paradigm I guess many of you are using it for example and it is not like Subversion was before constantly you know going somewhere else to actually find out about stuff it was just that they all other parts of the fast enough and have enough storage you can just store the whole copy and do everything locally when do some commits just synchronize later offline for us and the torrent said well what's the point if it's
like 50 people of us now I downloading some video stream of of your skin just day or something what's the point of everyday body making this phone call back to California to get this right and we have the data here we just got talking to each other to actually do it and the torrent organized network distribution protocol that means when I access this remote resource I chunk it up into data I Compute hashes and I say OK I can provide these hashes and a register meant the next person comes along and says all I need all the data for these hashes and who has them northern stopped to talk to each other and we can basically 1 person downloads but the others just get it locally so that's a very much more efficient and that the particle and this already mean the Torah at 1 point had 50 % of Internet traffic and it is certainly a very well known the going to talk to much about the others but they're using similar techniques and do you know what
the central data structure that is behind which we all of these techniques what that is what sorry again Merkle trees exactly the Merkel directed acyclic graph on going to talk about this in a 2nd FIL Merkle
trees are the way we could trace our to have data objects and then instead of having a mutable name to I compute cryptographic hash so does the short recap on what a cryptographic hashes it's the computation that takes basically an arbitrary locked arbitrarily long data John it performs some computation and out you
get something like 32 or 64 . some number of bytes and these are not the unique identifier in the following sense you can always easily perform the computational from the data to the hatched but you cannot interview have the hatch construct the data so it's a one-way computation and that's kind of like the cryptographic mathematics behind that don't make this possible of course we happen be caches and stuff like this but this is a topic like having secure hashes that is uh a research topic like in many many places also sources l and what not so we do have really safe cryptographic foster pressures and you have the weekend change color data blocks and we can prune compute the hash and then we
can in the tree we can actually have the hash of 2 Hessian and then we have the root test but actually
combines all branches so when I when I modify 1 data item the past up to the root changes like if I modify the data block 0 0 4 then the F. has changes the be hash changes and there would have changes but the rest of the tree stays the same like the duty and it turned out that this data structures but also is what it is using and react if using like would are
replicated database and also to a degree from torrent and many others using and so the thing is this is all in implementation there is no when article for that and just a little long piece that little personal
comment here not was not the same name as the german chancellor and I'd like to take the opportunity that I totally disagree with the politics of assessment FIL so another development when you think about the other tree it's it's immutable data structure as soon as you change something you get a new rule test just like engage and you have a new data structure basically and that would reference that have identifies a complete freeze the and this concept of mutability has has been going on for many many years I mean functional languages even back in the eighties and nineties they use the concept of a mutable data structures but they have
become because of scalability and replication and Synchronization and other tasks they have been this this idea of a using mutable data structures has become very popular so you can see in virtually every language that gained popularity of was invented in the last 10 years the individually as king it's like a you you rather switched from having immutable situation to an end to the situation and then you do it ability on top and industrial it's very very popular found out on Facebook to the library called mutable J. there's also a very good think well tested Python libraries like pi our system which provides you an immutable dictionary and immutable this concept and so on and actually when you add an item to an immutable dictionary you get a new dictionary object but it's internally stored efficiently so it's not like a full copies of the whole thing it just takes the tree and the part of the tree that actually is needed and um the effect if you have a mutable data structures is not only that it becomes easier to reason about hold day country it also means that programming gets safer because when I pass and mutable data structures to some libraries I can be sure it cannot do any damage occurred just take my dictionary you also this idea of of mutability I think is very beneficial apart from replication and stuff it's very beneficial also for arranging an application and arranging the eyes and getting some certainty and you can do this today in Python and JavaScript other languages have
more 1st class support for the and you know what what when I
1 of the reasons I have fun and left with Python 2001 was this great idea of namespaces but I came from C + + and I worked with the introspective abilities of pi from an assault well that's great the module has a dictionary that contains a mapping from Maine to a class of objects or name to function objects then I can go to the
function object that has a dictionary that contains the name to tell you mappings the function tributes I can go to the class that has a dictionary that contains the names to functional objects the method so it was basically namespaces all the way down and I really like I thought this is correct and that can get a complete perspective on on on the whole the whole cold is actually related to each other and was certainly 1 of the reasons that may be very interested in in during a Python interpreter myself
so that was also 1 of the motivations behind contributing and co-funding no what always bothered for like the last 15 14 years is I don't know if you made the
statement or computers for a mature the namespaces out hunting great
idea that's more of spammers will last the small world models you know what what could this be and I really like I think every year couple of times a thought about personal what what what more can we do with it and what is what's the next thing you and I think um this mutability and mutability if a clue I think thinking about kind of mutable data structures and immutable namespaces can can provide some some some advanced basically and what namespaces mean and how long stable they are you know I can I have a reference to a namespace and that contains recursively data structures and this reference can signal me
like a complete snapshot like a get comments on all of the contained some namespaces so it's boldly stable there's no nobody can change anything about this when I have this reference um and I I think it's it's worthwhile to try to do some rocket science
trying to find out what how this could be beneficial I mean maybe it could be beneficial for example for suppose that we have many applications in Python that we use mutable data structures I think it would be develop very well possible that this opens up possibilities for removing the
guilt because when you know that some code is working on her mutable data structure and you are in a local functions working with local variables you know that they can do nothing that can modify base that you working on it's a vague idea it's not like you thought about thing and I'm happy to to discuss
that that's possible but I think even if that doesn't work out working with mutable data structures immediacy namespaces is beneficial for writing larger applications and for reasoning about a program so finally I'd like to introduce you to the interplanetary files
and I am I got to know this to undergo talk to really develop Germany and the and he gave some of the thoughts that that I'm presently he is actually motivated by his story telling them and I P
of S aims to bring all the goodness that we have been proven implementation of synchronization mechanisms and languages to the you hypermedia protocol that substituents http
so let's go to the the difference
between the 2 http as and that that location that follows with the government and then the power is the phone number it's I'm basically going to some destination I want to have a secure communication led to L and all kinds of stuff basically I'm forming some computer somewhere I asked it to authenticate itself and then I do um and access to the past and see what is there and that's inherently mutable operation everything in between that can you take like the stomach cancer new file the location can change to a different IP address this L certificates can change and so on so it's inherently a very it's it's 1 on 1 hand it's a it's a phone conversation on the other hand it's it's a very readable thing now contrast this with the basic new idea that idea as sense and some of the GPU we can introduce a new protocol called idea of S here and we use the content hatch which is pointing to the to the root of a model tree and then in that mandatory just like in get directory objects and file objects and so
on but once you have a content you don't need to follow anywhere I can ask anybody of you that any money at all of you have the data that's behind that hatch I don't need to make a phone call and verified the correct sender and stuff like this I can just ask world so instead of addressing the location I'm addressing data what course the question is where do I get this
content has from I said it's like 32 or 64 bytes or something like this and we already have problems remembering phone number and so what I give as needs to do this
it needs to introduce mutability on top all of the utilities and I give as if again just like I P was over the phone network can overlay network so they can still continue to use the internet protocol and and we can even continue to use for the time being and will give something better the DNS system so you have an accurate estimate installed you can actually use the baffling and the Colts be . things that's actually that actually triggers a DNS lookup to get to an idea as link in this case that I'm not directly of as link but the so called best that as assuming the cost and I have an identity the public
key and when I published mutable names I signed so the name to immutable some object mapping is signed by my public key and that's what is less
of hassle so the 2nd link you see is actually still immutable still immutable links and that and eventually and thought link the idea as turns into a the into an mutable link because now I have a concrete content Hessian and using the idea of as cooperative
and that is now I can ask around for the data and and if you install the binary currently
where I give as I can this and implemented you can use the idea of this protocol already by using your local host during that does
all of the connections non just
like I just like the inherent the from the idea of s architecture is kind of an odd last model uh I didn't get around to actually minimize painting and certainly not as good as the 1 on Monday but the thing is with idea as links new hypermedia protocol the web and continues to function like HTML CSS almost great stuff continue to function but instead of a set of like phone numbers at the peak so all this stuff you can actually use on that's the idea we can use ideas as idea and S and suddenly we can't use them all the goodness from being able to talk to each other the naming is currently done by cells certified and naming system this public key idea can I'm going to make a session tomorrow like and show me play the the more details as a as a bit um what like that described like in in 1 minute or so and then the central data structure is the mottled duck at some point you get the content hash and then you have the same data structure that get users that also means you get convergence web it's I think and and some on some level actually what this office is the same thing that you get from working
for the source code and having this on a network drive without any kind of versioning and then using nothing to control you know way get much more precise references who changed what what was the previous version and so on and then the the exchange of the
actual data blocks the leaf and the model tree that is actually done by the torrent what the torrent like mechanism so again I don't need to make a phone call so I can just ask Ron does anybody have this big around here and what I of as currently implements this something do this classic data structure called the distributed hash table all what accounts which for example should also worked in this room so it's like 2 different ways how you can find possible peers who give you something and I
think what because this is also used by Apple sold but you get like the global scale to the network just like the file sharing that like 10 years ago and you also get
local the connectivity to the peers around you and you don't need to be worried because if you ask for a data block you have the hash nobody can take you no compared to actually be today if I ask for the contents of the Google search and I just asked some notes I could get anything right I don't really know what's the what's the if answer is actually authoritative the routing and the this is test table and I mentioned is going to be explained by Nicholas finally you sum over and in the
afternoon that's climate going to think it what 3 or because the program was in Python and some analysts what this so we have some this kind of hydrogen and for networking the for the time
being just use the internet protocol because well it's 41 years old at the form proposed but still it's their everybody has it and have like all kinds of stuff I'd like to also introduce you to another fault from an
enduring he is he's a practitioner in them building this with the networks and he saw the problem when you do peer-to-peer network like the 1 that going to describe in the afternoon that's 1 problem is that what are your initial peers who do you start to talk
now if if you think about this offline 1st situation where everything is battery-powered people are constantly switching on and off and nothing is really reliable as back in the 19 39 phone network situations when you some have this situation then there kind of a trade off like if you assume that everything like if you disconnect for 2 days and then you reconnect and you know all that addresses everything changed it will take a long time to get any kind of reliable connectivity to appears again and he said that some of its that's why that's the reason why in practice many of you networks they introduce just like skype stable nodes they're always there that operate but suddenly of course you have some kind of and these nodes are owned by people and they could actually and disrupt the network and his question that he puts up in his and I would say famous blog post I want to believe because he wants to believe in this completely fertilization assessed in practice that we work completely so he says rocket science questions can we construct a blind idiot got out of the so we do have some data structures out there
and we do have kind of like stable modes but can you make it so that they don't get to see very much if to do if the pyramid without the i you don't really it's line to kind of like applying make things but didn't really you don't really know where the nodes going on and 1 of the things of causes
and to and encryption and stuff like this but then there's the estimate and other questions so it's kind of an open debate on on how to do this so lastly I'd like to and reside very famous
inventor of and and researcher I think instead of blaming people who has also been doing of facebook or whatever we have also been doing great stuff and provided great services open source stuff they brought many of us into the position to have to unload money and so attacking them and saying you must change of something like this is probably not a good idea I think it's an I agree with Mr. Philippe that probably the better way to get a new system is to just build and then let it be better I mean this idea idea as has the chance to be just a much more effective particle and exchanging data you can implement get on top of ideas as you can implement Bitcoin on top of ideas the that doing the right so you can implement all kinds of stuff on this kind of new particles it's also going to help to get some
new um um actual protocols not just with implementation of actual particles into the web so we don't depend on all the proprietary effective some particles lastly is that would go I would get back to 2 discussion
off like we had this rocket science situation Margaret Hamilton programming the the the rockets to the moon and the science paper thing and that actually a lot of science stopped going exponentially but we certainly have a lot of innovations but are still happening and have been happening since the seventies but in a way that makes a very good historical argument with lots of data and facts has in fact the whole innovation has kind of leveled off it has been like doing something like real rocket science it's not really happening anymore universities are
very much restricted by which companies they get money for that a lot of administrative overhead he talks about bureaucracy and so on so he says the last 50 years haven't actually brought the advances that people thought in the sixties what happened you know compared to what we had between 2010 and 2 thousand uh sorry 1910 in 1960 or 1970 the advantages like like advantage like electrification got like electrification many places that wasn't there before right or we got like people flying the highest speed they few but then we stopped any kind of kind of like ambitious space program and many other examples kind of contradict the idea that we are living in a time where everything just exponentially grows all the time it grows and there is certain direction like you have new million to 1 architectures like new technology that actually helps there but in terms of getting an overall picture and and like making real technical advances that don't relate to monetarisation but just make everything more effective not so much is happening so I think the situation that we had in the seventies so that very few people invented the IP protocol and brought a lot of change might be actually happening again with people like I give as maybe it's not I give us but some something along the lines combining these ideas turning this into a about and also using this for all kinds of languages and non proprietary or open-source way I think is hopefully the future what if it if it
and you have to be that means that
any questions this but only
convinced you that cannot be the case what
by all the things you talk and adjusting quite understand the role of
the blind 88 God could expand again falling yes but the
blind got the idea is that if you have a peer-to-peer network like the 1 that regresses going to explain in the afternoon then you have basically if some lots of peers and you um segment disappears into sections and then everybody talks to the neighbors of all kinds of stuff but these nodes that are everywhere millions of nodes and you just have contact with few notes so you want to address something like you want to start something in the distributed dictionary but also what idea as as doing then you need to connect these maybe 20 nodes that you know about but knowledge into an online because the battery is gone on the Wi-Fi on something and then you turn online again that many of these no it's not just the so you try to contact them and know you don't get any connectivity and you try to and then maybe the 19th node actually know something but that node is only responsible for a very small subset of the whole network and then you need to rediscover which nodes are currently online and stuff like that so blindingly got idea says that how can we do stable nodes that are operating all the time so we have some kind of give low latency can just directly as they did you know that going to be there and they can quickly give us some information about the shapes of the whole situation but how can we do this with out making too much knowledge to these stable right because I mean Google and and and they are actually internally of cost completely just with the networks like if you look up the IP address of google . com you get 20 addresses and of course will has in there network complete coordination and they distribute all the information across the globe you know so you actually have this information you have this architecture of having central nodes that are stable and that tap into the actual this with the network that was operating but they are actually not blind you know because you actually connect to them and they provide the application so if we then do something like In the event was something like having end-to-end encryption and the nodes in the middle they don't really know that much anymore but it remains unclear because you have to get out of information still who who talks to whom and so basically the construction of something better that doesn't make so much knowledge in the middle and immediate and that's the idea of the Blind idiot got I guess that that Pope Francis which I that the 1st slide wouldn't approval this
question here um I like actually real-world idea missing it makes a lot of sense for all their children as the example to electron sharing data bit videos and so and but I wanna actually because we have also a lot of them so you presented as something that would replace basically it's tedious and the trend particles but we have a lot of things coming up for its API requesting like things going life having very short being very short latency a lot of very small calls and so and I guess it's not really adapted to that so you see like 2 systems and in parallel uses meditation possibilities yes a
point I mean how do you actually talk to entities and they have some storage and you have you you you successfully API because you modify resources and stuff basically I don't think that something like I P S I give as is going to replace everything know you knowledge using for having human readable names that has just currently and I think it's going to coexist for a long while the like I am using this this um presentation is actually image with you can see that but it's a it's soft from my
local IP as you and um and and is allowed to be here so it's it's all work in progress right I mean people are working
on this they're thinking about the problems and and they're trying to make a stable implementation and they're using current and HTTP for example 1 of the things is can you do a Firefox plugin that allows to on the 1 hand that allow you to use I P of S in Europe but if you don't have that you have some kind of 20 filled that actually redirects to some available notes because the idea of s network is also currently relate into the HTTP again as network under data . I give as start by all 0 so you can actually just access the central site and they provide you you into the whole situation I think it's going to be for a long while it's going to be a hybrid situation and the question about API access well the thing is if you consider the model that all endpoints are so powerful and like in 10 years our phones and laptops and so on will be very very powerful I think somebody concluded that in 50 or 60 years on the left phones will be able to do what the has a status and us today right and the more intelligent these devices get the more you don't want to actually delegates stuff somewhere else and just change data into get into architectures where the endpoints just talk to each other you know and range from state change there is secure manner because you have like margin and state changes and it's at the top actually and what time frame this will I mean you must says we have 11 years last left and and we already have 41 years of IP is so you know it's it's really you need to get rolling they are trying to do this from
yeah but from the top it seems that i the Office's for distributing public data but currently you know I if I put something private on the on the Internet and probably Facebook or Google or they can see it but I'm may be more concerned that my competitor over my x doesn't see it out there any solutions to this problem has a good question I have to admit that
I don't actually mean my in my think about it I would actually say the the way how would do this now is a true idea as I published some contact points and then I would probably currently use HTTP to perform perform some kind of authentication so that old I'm actually talking to the correct end-point and provide my authentication credentials to that point and then that basically use it on top of know so I can still just get my private data using you as far as well as we currently have and I think it's true that which is a um I give as and IP and this is basically 1st of all it's something for the public and data are that much of that I mean if something is going on in terms of volume range and to and triptych and provides communication I think you can just use it if you want to actually of my mobile device wants to talk to the mobile device of my partner somewhere in Germany I really want to talk to the device because that's actually where she is attending you know and I'm just so on some level OK that we actually have this phone call it's actually a formed what really want to talk to the device so in that sense and just using in that case I would just use IP so I guess I P or something like that it's not going to go way like for everything so have another overlay network that provides benefits natural elements I
have the microphone arrays I don't think so thank you for the for the talk about my name is about the i it's really great concept is interplanetary fire systems but there is applauded understand missing because it seems like the principle of this in the community and is chunking like you have to partition the data into smaller chunks completed cryptographic hash and then in order to provide a reduced the amount of information to be exchanged between the nodes you obviously look for the chunks that are already in the local network which seems like looking up the cached data that this like a locally so is there any research done because you can go to very extreme like you use each and the data in such a level that each gene teacher number 1 has its own past they instead of sending number 1 you would be sending the is there like a sweet spot all research done how long you would go uh maybe it's what what where should they find what amount of data should be chunking into you have this I think I mean that the timing of
data if you have like a large data blocks and you put this with a PUT request and to the local host give as given the chance of for you use some kind of chunk size that people have been found just like a TCP IP the default windows and stuff like this you just have some default chunk sizes and then used so you don't get the higher level structure with the you don't get to think about how that is done by the you know what archived content defined function OK so you know I'm not knowing everything about what's going on and Thomas most of someone the things that come this is all very much in flux basically hold all you have to know which kind of numbers makes sense and you need to have real networks to actually see what the effects are on how you do your chunking and your BitTorrent exchange and stuff like that um I can use it for a while elements in what recently from me but that that's that's certainly research topic some embezzlement and I have the general idea of as on field we can ask such questions so I'm sorry but too much in the media um
could mean we said the last question but darpa has gone out of research on delay tolerant networks and I have you know what that is an error the sea which is called on the vertical which is a 50 50 that's a lot of research and there are even some false they need to do that we see on that kind of stuff out how did you see all these coming to 11 these because these bundle protocols the later and that was focused more on not having an end to end point from 1 point of the network to the other like for example satellites or for word so the lights went down and you do not have connection between all the 70 so the time for for example and more life on so that way you will use the net work on it and that was the CCD all these protocols to that I think that's really relates to how do you do with stable and just his hash table DHT is what all of the sharing protocols and some other protocols have been using and I think there's been a lot of research and all that I met last month amid immediacy is a researcher in the someone this book we
would use basically these questions like how can you how can you how can you make this very effective in and the default case of things disconnecting and connecting them no and that has been invested a lot of research on actual practical work in the last 10 15 years but some of the main change that idea as brings it actually ties all of these ideas together it's kind of like crucial ideas all around to get local and um tries to get the whole stack basically lined up so everything works together a very stable way so that's exactly which I tried and and that's something that is probably totally workable right now but it's it's a very active project it's not like the whole world is working on it so just like in the city so it always depends on people actually doing things and regarding Python actually um there was a goal implementation there is an upcoming not yes implementation and there is a Python sketch implementation nothing basically so would probably be interesting to start working on getting some of the things also looking in Python because then you can start using it emerges from produce well but currently I would just use the demons written goal and interface with that no from Python just use the network to keep up with the with the highest I just about directly into into Python applications this is the last question is not
and you don't see a problem with version the web sometimes when is good if you can say see helping about it's the electrons conditions FIL happens if want some to see my old buddy carried notes and just use occasion has an inviting is mapped them and
foster if you if you for example if you have your self-certifying namespace and to publish and new names to content hashmap in that like in did you would have something like a comet object that points to the last common but you don't have to you can just make a brand new object that doesn't contain the history to the old 1 so it's about so that is not and then falls still your choice when you create the data point been missing the new thanks
from the somewhere the lighter 1
2 facts so there you lose you in a position to present them to the momentum you lose my and thank you thank you thank you so
that you
can do
Schätzwert
Roboter
Suite <Programmpaket>
Bit
Formale Sprache
Red Hat
Fokalpunkt
Computeranimation
Eins
W3C-Standard
Fundamentalsatz der Algebra
Knotenmenge
Mereologie
Vorlesung/Konferenz
Projektive Ebene
Baum <Mathematik>
Fundamentalsatz der Algebra
Datensatz
Kategorie <Mathematik>
Zahlenbereich
Ordnung <Mathematik>
Gesetz <Physik>
Computeranimation
Übergang
Software
Programmiergerät
Whiteboard
Flächeninhalt
Assembler
Hochdruck
Meter
Luenberger-Beobachter
Vorlesung/Konferenz
Quellcode
Optimierung
Arithmetisches Mittel
Bit
Puls <Technik>
Puls <Technik>
Digitalisierer
Ablöseblase
Zahlenbereich
Optimierung
Dialekt
Computeranimation
Hardware
Datennetz
Puls <Technik>
Digitalisierer
Mathematisierung
Gamecontroller
Zahlenbereich
Frequenz
Einfach zusammenhängender Raum
Protokoll <Datenverarbeitungssystem>
Datennetz
Datennetz
Zahlenbereich
Router
Ordnung <Mathematik>
Overlay-Netz
Computeranimation
Übergang
Internetworking
Modem
Einfach zusammenhängender Raum
Internetworking
Bit
Punkt
Datennetz
Virtualisierung
Zahlenbereich
Mathematische Logik
Dialekt
Internetworking
Entscheidungstheorie
Informationsmodellierung
Knotenmenge
Kugel
Puls <Technik>
Datennetz
Router
Spezifisches Volumen
Gerade
Tabelle <Informatik>
Internetworking
Zentrische Streckung
Informationsmodellierung
Weg <Topologie>
Punkt
Datennetz
Term
Soundverarbeitung
Softwaretest
Internetworking
Bit
Regulärer Ausdruck
Kontextbezogenes System
Komplex <Algebra>
Lastteilung
Task
Metropolitan area network
Knotenmenge
Diskrete-Elemente-Methode
Firewall
Vorlesung/Konferenz
Soundverarbeitung
Konstante
Zentrische Streckung
Facebook
Punkt
Datennetz
Firewall
Regulärer Ausdruck
Vorlesung/Konferenz
Datenstruktur
Computeranimation
Gammafunktion
Benutzerbeteiligung
Web Site
Minimalgrad
Konditionszahl
Güte der Anpassung
Vorlesung/Konferenz
Computerarchitektur
Zeiger <Informatik>
Baum <Mathematik>
Raum-Zeit
Computeranimation
Soundverarbeitung
Managementinformationssystem
Telekommunikation
Datennetz
Interaktives Fernsehen
Kontextbezogenes System
Komplex <Algebra>
Term
Übergang
Videokonferenz
Task
Minimalgrad
Prozess <Informatik>
Ultraviolett-Photoelektronenspektroskopie
YouTube
Softwaretest
Bit
Datennetz
Systemaufruf
Raum-Zeit
Computeranimation
Verdeckungsrechnung
Deskriptive Statistik
Metropolitan area network
Arbeitsplatzcomputer
Booten
Softwareentwickler
Baum <Mathematik>
Leistung <Physik>
Internetworking
Datennetz
Direkte numerische Simulation
Güte der Anpassung
Server
Booten
Partikelsystem
Programmierumgebung
E-Mail
Einfach zusammenhängender Raum
Internetworking
Kraftfahrzeugmechatroniker
Bit
Protokoll <Datenverarbeitungssystem>
Güte der Anpassung
Schlussregel
Ausnahmebehandlung
Internet der Dinge
Computerunterstütztes Verfahren
Elektronische Publikation
Synchronisierung
Übergang
Internetworking
Puls <Technik>
Gruppe <Mathematik>
Partikelsystem
Kraftfahrzeugmechatroniker
Flächeninhalt
Notebook-Computer
Lokales Netz
Binder <Informatik>
Elektronische Publikation
Synchronisierung
Einfach zusammenhängender Raum
Binärdaten
Internetworking
Punkt
Gruppenkeim
Stellenring
PASS <Programm>
Kartesische Koordinaten
Zentraleinheit
Metropolitan area network
Informationsmodellierung
Zustand
Mereologie
Datenerfassung
Personal Area Network
Kraftfahrzeugmechatroniker
Punkt
Apache Cassandra
Zahlenbereich
PASS <Programm>
Synchronisierung
Computeranimation
Magnettrommelspeicher
Client
Datenreplikation
Mereologie
Server
Programmierparadigma
Projektive Ebene
Speicher <Informatik>
Baum <Mathematik>
Zentralisator
Distributionstheorie
Punkt
Protokoll <Datenverarbeitungssystem>
Datennetz
Graph
Apache Cassandra
Systemaufruf
Computeranimation
Internetworking
Netzwerktopologie
Streaming <Kommunikationstechnik>
Rechter Winkel
Hash-Algorithmus
Vorlesung/Konferenz
Partikelsystem
Datenstruktur
Hash-Algorithmus
Mathematik
Mathematisierung
Zahlenbereich
Computerunterstütztes Verfahren
p-Block
Quellcode
Computeranimation
Objekt <Kategorie>
Netzwerktopologie
Metropolitan area network
Druckverlauf
Caching
Hash-Algorithmus
Identifizierbarkeit
Booten
Kantenfärbung
Softwaretest
Hash-Algorithmus
Datenhaltung
Schaltnetz
Mathematisierung
Verzweigendes Programm
Implementierung
p-Block
Ausgleichsrechnung
Computeranimation
Vorhersagbarkeit
Netzwerktopologie
Metropolitan area network
Minimalgrad
Hash-Algorithmus
Booten
Wurzel <Mathematik>
p-Block
Datenstruktur
Haar-Integral
Soundverarbeitung
Softwaretest
Facebook
Hash-Algorithmus
Algebraisch abgeschlossener Körper
Formale Sprache
Gefrieren
Schlussregel
Kartesische Koordinaten
Physikalisches System
Synchronisierung
Computeranimation
Data Dictionary
Objekt <Kategorie>
Task
Netzwerktopologie
Skalierbarkeit
Mereologie
Datenreplikation
Programmbibliothek
Funktionale Programmierung
Optimierung
Datenstruktur
Softwareentwickler
Objekt <Kategorie>
Mapping <Computergraphik>
Managementinformationssystem
Lineares Funktional
Interpretierer
Namensraum
Hash-Algorithmus
Algebraisch abgeschlossener Körper
Klasse <Mathematik>
Modul
Computeranimation
Virtuelle Adresse
Stabilitätstheorie <Logik>
Informationsmodellierung
Befehl <Informatik>
Namensraum
Hash-Algorithmus
Computerunterstütztes Verfahren
Datenstruktur
Computeranimation
Lineares Funktional
Variable
Namensraum
Hash-Algorithmus
Vorlesung/Konferenz
Kartesische Koordinaten
Datenstruktur
Code
Computeranimation
Virtuelle Adresse
Kraftfahrzeugmechatroniker
Namensraum
Hash-Algorithmus
Protokoll <Datenverarbeitungssystem>
Güte der Anpassung
Formale Sprache
Implementierung
Kartesische Koordinaten
Permanente
Elektronische Publikation
Synchronisierung
Computeranimation
Dateisystem
Hypermedia
Reelle Zahl
Datenstruktur
Optimierung
Peer-to-Peer-Netz
Telekommunikation
Nichtlinearer Operator
Umsetzung <Informatik>
Subtraktion
Digitales Zertifikat
Protokoll <Datenverarbeitungssystem>
Zahlenbereich
Computer
Elektronische Publikation
Netzadresse
Computeranimation
Netzwerktopologie
Objekt <Kategorie>
Informationsmodellierung
Dateisystem
Wurzel <Mathematik>
URL
Kontrast <Statistik>
Verzeichnisdienst
Leistung <Physik>
Schätzwert
Datennetz
Protokoll <Datenverarbeitungssystem>
Softwarewerkzeug
Systemaufruf
Zahlenbereich
Physikalisches System
Binder <Informatik>
Overlay-Netz
Computeranimation
Internetworking
Verzeichnisdienst
Verschlingung
Nichtunterscheidbarkeit
Direkte numerische Simulation
Direkte numerische Simulation
URL
Inhalt <Mathematik>
Objekt <Kategorie>
Mapping <Computergraphik>
Public-Key-Kryptosystem
Verzeichnisdienst
Protokoll <Datenverarbeitungssystem>
Verschlingung
Direkte numerische Simulation
Inhalt <Mathematik>
Binder <Informatik>
Computeranimation
Einfach zusammenhängender Raum
Zentralisator
Bit
Punkt
Protokoll <Datenverarbeitungssystem>
Güte der Anpassung
Zellularer Automat
Zahlenbereich
Physikalisches System
Binder <Informatik>
Computeranimation
Übergang
Office-Paket
Informationsmodellierung
Benutzerbeteiligung
Menge
Hash-Algorithmus
Hypermedia
Direkte numerische Simulation
Computerarchitektur
Datenstruktur
Baum <Mathematik>
Kraftfahrzeugmechatroniker
Versionsverwaltung
Klassische Physik
Systemaufruf
Peer-to-Peer-Netz
Quellcode
p-Block
Computeranimation
Netzwerktopologie
Informationsmodellierung
Hash-Algorithmus
Direkte numerische Simulation
Datenstruktur
Datei-Server
Tabelle <Informatik>
Softwaretest
Einfach zusammenhängender Raum
Zentrische Streckung
Datennetz
Peer-to-Peer-Netz
p-Block
Elektronische Publikation
Computeranimation
Metropolitan area network
Hash-Algorithmus
Vorlesung/Konferenz
Inhalt <Mathematik>
Direkte numerische Simulation
Optimierung
Tabelle <Informatik>
Einfach zusammenhängender Raum
Internetworking
Stabilitätstheorie <Logik>
Protokoll <Datenverarbeitungssystem>
Datennetz
Web log
Adressraum
Peer-to-Peer-Netz
Feldgleichung
Computeranimation
Kreisbogen
Internetworking
Portscanner
Bildschirmmaske
Knotenmenge
Vorlesung/Konferenz
Direkte numerische Simulation
Datenstruktur
Schätzwert
Managementinformationssystem
ATM
Unterring
Facebook
Stabilitätstheorie <Logik>
Ortsoperator
Physikalischer Effekt
Open Source
Datenmodell
Mathematisierung
Implementierung
Physikalisches System
Extrempunkt
Ausgleichsrechnung
Portscanner
Knotenmenge
Chiffrierung
Web Services
Rechter Winkel
ATM
Partikelsystem
Gerade
Parametersystem
Benutzerbeteiligung
Protokoll <Datenverarbeitungssystem>
Reelle Zahl
Güte der Anpassung
Implementierung
Partikelsystem
Grundraum
Offene Menge
Protokoll <Datenverarbeitungssystem>
Open Source
Formale Sprache
Mathematisierung
Systemverwaltung
Term
Raum-Zeit
Computeranimation
Richtung
Computerarchitektur
Overhead <Kommunikationstechnik>
Optimierung
Gerade
Offene Menge
Adressraum
Kartesische Koordinaten
Netzadresse
Computeranimation
Knotenmenge
COM
Vorlesung/Konferenz
Einfach zusammenhängender Raum
Konstruktor <Informatik>
Shape <Informatik>
Vervollständigung <Mathematik>
Datennetz
Peer-to-Peer-Netz
Ereignishorizont
Teilmenge
Rechenschieber
Chiffrierung
Programmfehler
Rechter Winkel
Grundsätze ordnungsmäßiger Datenverarbeitung
Garbentheorie
Computerarchitektur
Information
Koordinaten
Baum <Mathematik>
Videospiel
Bit
Punkt
Twitter <Softwareplattform>
Systemaufruf
Partikelsystem
Physikalisches System
Kombinatorische Gruppentheorie
Speicher <Informatik>
Bildgebendes Verfahren
Videokonferenz
Randverteilung
Informationsmodellierung
Web Site
Rahmenproblem
Datennetz
Arithmetische Folge
Rechter Winkel
Notebook-Computer
Plug in
Computerarchitektur
Ausgleichsrechnung
Phasenumwandlung
Computeranimation
Offene Menge
Telekommunikation
Facebook
Datenmissbrauch
Punkt
Natürliche Zahl
Systemaufruf
Element <Mathematik>
Overlay-Netz
Term
Office-Paket
Internetworking
Übergang
Metropolitan area network
Spannweite <Stochastik>
Authentifikation
Spezifisches Volumen
Soundverarbeitung
Lineares Funktional
Datennetz
Zahlenbereich
Fluss <Mathematik>
p-Block
Element <Mathematik>
Physikalisches System
Computeranimation
Übergang
Knotenmenge
Datenfeld
Total <Mathematik>
Bildschirmfenster
Hypermedia
Hash-Algorithmus
Lokales Netz
Information
Ordnung <Mathematik>
Datenstruktur
Default
Logik höherer Stufe
Einfach zusammenhängender Raum
Videospiel
Satellitensystem
Stabilitätstheorie <Logik>
Punkt
Datennetz
Protokoll <Datenverarbeitungssystem>
Gemeinsamer Speicher
Mathematisierung
Delay Tolerant Network
Implementierung
Kartesische Koordinaten
Computeranimation
Hash-Algorithmus
Vorlesung/Konferenz
Projektive Ebene
Wort <Informatik>
Dämon <Informatik>
Default
Baum <Mathematik>
Faserbündel
Schnittstelle
Tabelle <Informatik>
Fehlermeldung
Objekt <Kategorie>
Telekommunikation
Namensraum
Benutzerbeteiligung
Punkt
Konditionszahl
Versionsverwaltung
Vorlesung/Konferenz
Baum <Mathematik>
Auswahlaxiom
Computeranimation
Software Engineering
Roboter
Offene Menge
Impuls
Ortsoperator
Red Hat
Baum <Mathematik>
Computeranimation

Metadaten

Formale Metadaten

Titel Keynote: Towards a more effective, decentralized web
Serientitel EuroPython 2015
Teil 70
Anzahl der Teile 173
Autor Krekel, Holger
Lizenz CC-Namensnennung - keine kommerzielle Nutzung - Weitergabe unter gleichen Bedingungen 3.0 Unported:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen und nicht-kommerziellen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen und das Werk bzw. diesen Inhalt auch in veränderter Form nur unter den Bedingungen dieser Lizenz weitergeben
DOI 10.5446/20126
Herausgeber EuroPython
Erscheinungsjahr 2015
Sprache Englisch
Produktionsort Bilbao, Euskadi, Spain

Technische Metadaten

Dauer 1:03:52

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract Holger Krekel - Keynote: Towards a more effective, decentralized web Many would like to see more decentralization but what does it mean, really? In this talk, I'll discuss the recent rise of immutable state concepts in languages and network protocols. And how the advent of hash-based data structures and replication strategies are shaking the client/server web service paradigm which rests on managing mutable state through http. By contrast, building on git, bittorrent and other content addressed data structures provides for a more secure, efficient decentralized communication topology. There are projects, thoughts and talk to create new web standards to bring such technologies to mass deployment and fuel a new wave of decentralization. What can Python bring to the table?
Schlagwörter EuroPython Conference
EP 2015
EuroPython 2015

Ähnliche Filme

Loading...