Wind: Off-Grid Services for Everyday People

Video thumbnail (Frame 0) Video thumbnail (Frame 2888) Video thumbnail (Frame 4936) Video thumbnail (Frame 6199) Video thumbnail (Frame 8580) Video thumbnail (Frame 10265) Video thumbnail (Frame 13241) Video thumbnail (Frame 15019) Video thumbnail (Frame 16302) Video thumbnail (Frame 17869) Video thumbnail (Frame 19042) Video thumbnail (Frame 22376) Video thumbnail (Frame 25253) Video thumbnail (Frame 28212) Video thumbnail (Frame 29549) Video thumbnail (Frame 30983) Video thumbnail (Frame 32821) Video thumbnail (Frame 36294) Video thumbnail (Frame 37486) Video thumbnail (Frame 39357) Video thumbnail (Frame 40710) Video thumbnail (Frame 42165) Video thumbnail (Frame 43880) Video thumbnail (Frame 45404) Video thumbnail (Frame 48504) Video thumbnail (Frame 56153) Video thumbnail (Frame 58429) Video thumbnail (Frame 59925) Video thumbnail (Frame 61864) Video thumbnail (Frame 64312) Video thumbnail (Frame 66923) Video thumbnail (Frame 68203) Video thumbnail (Frame 69584) Video thumbnail (Frame 72860) Video thumbnail (Frame 74979) Video thumbnail (Frame 76204) Video thumbnail (Frame 77607) Video thumbnail (Frame 90212)
Video in TIB AV-Portal: Wind: Off-Grid Services for Everyday People

Formal Metadata

Title
Wind: Off-Grid Services for Everyday People
Subtitle
Integrating nearby and offline connectivity with the Internet
Title of Series
Author
License
CC Attribution 4.0 International:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Identifiers
Publisher
Release Date
2018
Language
English

Content Metadata

Subject Area
Abstract
The internet has become essential services, and offline methods of sharing data are rapidly disappearing. Other possible networks are often better suited when connectivity is not available or affordable. Radios, sensors, and computing are available in the cheapest of smartphones and routers. Wind is integrating nearby/offline data exchange with the internet services that we all rely on.
Keywords Resilience
Web service Software Artificial neural network Internetworking Projective plane Computer hardware Musical ensemble Connected space
Mobile app Software developer Projective plane Semiconductor memory Information privacy Mereology Information privacy Usability Revision control Process (computing) Software Right angle Freeware Modem
Expected value Focus (optics) Process (computing) Software Iteration Semiconductor memory Error message Information privacy
Mobile app Mobile Web Information privacy Hand fan Digital photography Term (mathematics) Core dump Software design pattern Communications protocol Local ring Traffic reporting Building Android (robot) Motion capture Mereology Information privacy Product (business) Explosion Software Computer hardware Self-organization Condition number Right angle Quicksort Videoconferencing Freeware Library (computing)
Software Network topology Internetworking Multiplication sign Software developer Software Projective plane Core dump Computer network Quicksort Mereology Freeware
Email Dialect Computer file Artificial neural network Real number Bit Mereology Computational intelligence System call Connected space Internetworking Freeware Local ring Physical system Modem
Point (geometry) Noise (electronics) Mobile app Game controller E-book Digital media Artificial neural network Source code Interactive television Software Internetworking Hard disk drive Quicksort Game theory
Mobile app Artificial neural network Multiplication sign Hard disk drive .NET Framework
Web service Personal digital assistant Internetworking Hard disk drive Data center Heat transfer
Area Polygon mesh Mapping Artificial neural network Multiplication sign Artificial neural network Vermaschtes Netz Similarity (geometry) Bit Computational intelligence Web service Pointer (computer programming) Internetworking Software Very-high-bit-rate digital subscriber line Tower Internetworking Personal digital assistant Telecommunication Core dump Figurate number Routing Local ring
Satellite Dependent and independent variables Mobile app Satellite Digital media Computer file File format Direction (geometry) Artificial neural network Bit Computational intelligence Connected space Content (media) Internet forum Mixed reality Synchronization Internetworking Universe (mathematics) Touch typing Videoconferencing Cuboid Physical system Thumbnail
Subject indexing Facebook Web service Software Internetworking 1 (number) Endliche Modelltheorie
Root Information Internetworking Multiplication sign Computational intelligence
Proof theory Trail Game controller Centralizer and normalizer Artificial neural network Nuclear space Multiplication sign Mereology Computational intelligence
Statistics Game controller Perfect group Mapping Information Multiplication sign Computational intelligence Electric power transmission Number Power (physics) Shooting method Estimator Software Internetworking Physical system Row (database)
Web service Internetworking Mapping Internetworking
Facebook Group action Internetworking Block (periodic table) Projective plane Mereology Computer programming
Building Internetworking Internetworking Mereology Connected space
NP-hard Mobile app Web service Message passing Polygon mesh Internetworking Artificial neural network Bit Endliche Modelltheorie Local ring 10 (number) Connected space
Focus (optics) Dependent and independent variables Software Multiplication sign Horizon Archaeological field survey Mereology Perspective (visual) Number
Wiki Process (computing) Archaeological field survey Schmelze <Betrieb> Bit Information privacy Product (business)
Point (geometry) Android (robot) Mobile app Functional (mathematics) Server (computing) Link (knot theory) Shape (magazine) Client (computing) Regular graph Mereology Centralizer and normalizer Internetworking Synchronization Videoconferencing Directed set Endliche Modelltheorie Metropolitan area network Installation art Default (computer science) Email Artificial neural network Server (computing) Projective plane Shared memory Data storage device Connected space Process (computing) Software Software repository Logic Repository (publishing) Hard disk drive Normal (geometry) Right angle Procedural programming Freeware Local ring
Multiplication sign Artificial neural network Data storage device Line (geometry) Connected space Message passing Software Internetworking Web service Cuboid Aerodynamics Quicksort Library (computing) Library (computing)
Area Trail Software Mapping Artificial neural network Personal digital assistant Business model Connectivity (graph theory) Musical ensemble Computer Physical system Connected space
Mobile app Surface Shared memory Business model Asynchronous Transfer Mode Physical system
Point (geometry) Web service Fluid Software Internetworking Range (statistics) Business model Cuboid Mereology Communications protocol Connected space Library (computing)
Connected space Internetworking Order (biology) Artificial neural network Data storage device Maxima and minima Cuboid Directed set Bit Endliche Modelltheorie Host Identity Protocol Modem
Web service File format Electronic mailing list Generic programming Communications protocol Communications protocol Fingerprint Address space
Dataflow Domain name Mobile app Computer file Local area network Multiplication sign Mereology IP address Host Identity Protocol Number Mathematics Peer-to-peer Internetworking Core dump Cuboid Information Artificial neural network Android (robot) Sequence Measurement Connected space Internetworking Software Cartesian closed category Library (computing) Address space
Domain name Building Artificial neural network Artificial neural network Bit Product (business) Expected value Software Synchronization Internet service provider Videoconferencing Self-organization Physical system Form (programming)
Time zone Personal digital assistant Multiplication sign Vermaschtes Netz Musical ensemble Line (geometry) Number
Laptop Point (geometry) Domain name Mobile app Computer file Multiplication sign .NET Framework Vermaschtes Netz Student's t-test Client (computing) Mereology Perspective (visual) IP address Number Sign (mathematics) Internet forum Internetworking Matrix (mathematics) Business model File system Cuboid Compilation album Mobile Web Polygon mesh Scaling (geometry) Matching (graph theory) Key (cryptography) Artificial neural network Projective plane Instant Messaging Bit Limit (category theory) Subject indexing Message passing Word Software Personal digital assistant Software repository Telecommunication Internet service provider Self-organization Website Musical ensemble Freeware Local ring Library (computing)
Cartesian closed category Musical ensemble
[Music] welcome to this session and to this talk hon scoffs Tina is going to give you a talk on wind off-grid services for everyday people and we'll be a network related talk and I'm very happy to welcome him on stage give him a warm and uh of applause [Applause] [Music] [Applause]
[Music] thank you so I will be talking about our project wind which is you said about really our work over the last seven ish years about integrating offline connectivity with kind of nearby networking into the Internet and making that something that you can use transparently I'll give a little overview some kind of randomly chosen or not so randomly chosen historical precedents that I think inform this whole discussion talk about some of the work that a lot of the work that we've done other examples that are presented on been presented at Congress and other places and also it hopefully inspire everyone here works on software of any kind maybe even hardware to also think about nearby and offline networks and how to integrate it into what what you're making so this work is kind of under the umbrella of the Guardian
project we are a free software project focused on privacy in mobile we make that means when we do software
development we started with the idea that we want to make apps where privacy comes first we don't want to compromise
at all on privacy as in the whole process include so that means you know we haven't become rich and famous because we've opted for grants rather than VC money and things like this one
central idea for us is that that that the design the usability of software does not is not in conflict with privacy but really and actually an essential part of delivering software that is actually secure in private if people don't understand what they're doing that it much more likely to make a mistake and and if you design things in a way that it's eat they after the user has to choose between the dangerous button and the safe button right next to each other then that can cause a lot of problems so
central to our design process is really thinking about user experience so this is you instead of saying you know the user didn't understand it we need to write a manual better we need to teach them better we say no we need to make it so that the software matches the people's expectations we need to present
it in a way that the expectations are clear and this means a lot of talking to people a lot of iterations and a lot of trial and error so Guardian practicum specifically focuses the users we've focus on our activists and citizens citizens and journalists and human
rights defenders people anything from like election monitors and etc etc we are often called upon to help inform
software that's used in businesses or business people I do travel used to use our software and frontline reporters and
one but really either so the core
mission is we really want to make privacy and and these kind of things a something for everybody and we've kind of found this niche where we can work in partnership with all sorts of organizations who want who will give us money to focus on these people like human rights activists and that we develop free software and we try to push it out there in ways that it can be reused both in terms of free software libraries so that you can incorporate into your into your apps but also in terms of design research design patterns and things like this so these are just a bunch some of the not all but some of the partnerships these are you know a lot of these organizations represent say the target user that and and so whenever we're designing software we're always working in conjunction with directly with people who are have strong privacy concerns that we're trying to address another you know accessibility and we also always try to find you know there's
already a free for software project out there doing an essential part of what we're doing we would much rather try to support another free software community than invented ourselves so this means we have to do a lot more talking as well but so that's part of coming to and that like this is to say okay you know where where are the parts that we all can work together and make something better yeah so I should also mention that you know I kind of wear many hats in in this is the I pay my bills usually through officially Guardian project but I but the way we work is that we we often do a lot of the work just because we think it's important I'm also a Debian developer I'm also a core team of f-droid and other things like this things that don't necessarily pay the bills but because we have some funding here means we can also have time to focus on on things that integrate into the whole thing and are also really important so networks so everyone's
familiar with the Internet and the Internet is now on our phones in our pockets all the time and I think more and more people are forgetting that there used to be lots of lots of all sorts of kinds of networks
out there and that I think it's so I wanted to highlight a couple that were influential to me and also just to point out that you know there's there's things we can we can learn from them because some of these ideas people think well that's crazy to build a system like this that you know it's too hard well and you can look back and say well actually this was working pretty well not so long ago who hears heard a fight own it some okay
fight own is probably my yes my real Vai first exposure to anything like a network what it is quite simply was back in the days when you know connection to a network was a modem and dialing into something there was a thing called the BBS it was basically a one stop I mean it was one computer that lots of people could call in to and you could exchange you know you could chat you could find files things like this so a bunch of people figured out well you know if we want to exchange data to across the country that's a very expensive phone call but I can call make a local call for free or cheaper and then that will person that BBS can make another local call and that BBS can make another local call and it can and then we can have a global network that's actually affordable and that's this fidonet spanned a large part of the world I mean it went from us to Bulgaria I think a little bit into South America and just you know organized by people who wanted to be able to send little emails and things like this then for me one that's
near and dear in my heart anyone ever used beam beaming apps and and Palm Pilots alright so this is something I think is quite important this is like it has the idea of offline nearby networks built in from the beginning anything in as far as I remember it like they you know if you you had an app you could beam an app to someone else you'd be like hey here's this app you want this app and you point the devices and made a little noise you could also be in contacts and things like this and it actually worked pretty well and it was just based on like the same little infrared sensors as a room TV remote control and and I think it worked because of the technology was there but this more importantly I think is that they've got it smooth enough and this kind of interaction if you're already talking about like Oh give me your card or pull out your thing and pull out your Palm Pilot and point them at each other was already kind of plugging into an existing human behavior so it wasn't telling people to do some totally new weird thing one thing that's
getting a lot of media attention right now is a packet to Seminole the weekly package in Cuba which is really pretty amazing network of couriers moving hard drives around Cuba what it is is someone either goes to a country that has high-speed Internet or they have through some source they get it on and Cuba so frightening for people don't know most of Cuba doesn't really have internet access and the internet access that available is extremely expensive so people have figured out this thing well they just buy a hard drive and they fill it up with ebooks movies all sorts of things apps games and then someone
brings it into Cuba they have a whole network of this network that's of people who talk to each other and meet up and they physically exchange these hard drives so someone says I got the new weekly packet you go you have your weekly packets tour you know someone that gets who gets it from someone you go make your copy then people say okay I want to get apps they go to the weekly packets store they pay a little money you get some time on the hard drive copy what you need and this covers Cuba quite
well and there is no actual there's no wireless networking there's no there's no wired networking it is people moving hard drives around and and talking to each other and and agreeing and I mean so this was has often been kind of derided as a sneaker net just something near and dear to us and as something as a last resort but it
turns out that sneakernet is still the fastest way transmit data so Amazon has a service where if you have a lot of data and you want to moved Amazon they drive a truck they do I have a truck to
your to your data center and they plug it in and you load it up and this is an exabyte of data and I mean I don't know all the exact details but something like an internet for one example case it would have taken 26 years to do it over the internet and it took them six months with this thing so it's just really a truck full of wired hard drives and then
I feel like so this is more I feel like meshes is fading a little bit so this is something that was there was quite a bit of activity maybe ten years ago of people building DIY networks that can build a figure out how to route data through them but I'm kind of on their own and that's the core idea of a mesh network and so this specific map is the gwee fee network in catalonia you can see like the that is Barcelona where the mouse pointer thing is but so it covers quite a large area I think it's ten thousand nodes and to this it's provided actually relatively high speed networks in a place where the the local telecom companies haven't really been interested for whatever reason in doing it and there's there's actually a kind of it's kind of a hard thing so the you know this is built by and set up by mana whole lot of volunteers and that takes a lot of people's time and technical ability and what often happens in the in cases like this is that then they've done the hard work of actually proving that people want the Internet and maybe we'll even give some money to it to support it and then the companies often come in and be like oh look that's a market waiting for us and then and these things go away there's another recently I just learned about it yesterday I mean the world the outcome but there's a telecom in eastern Nicaragua called say sell a very similar idea kind of DIY community-based effort because the the local telecoms were not interested in eastern Nicaragua at all they set it up got it working and then they telecom saw oh wait people will actually pay for telecom service there and then they of course then set up a giant lots of towers and put a sell out of business and of course the internet
itself started as an offline network that was you know 1970 that was the internet that was all of it and it this was you know built node by node starting you know when they first turned it on it was two or three computers talking to each other and bit by bit it was built out into what we have now and on top of that you know so that that was a US
Defense Department really funded effort but then lots of things started kind of bited on top of that so Usenet which is still alive and kicking today and is a fascinating idea of how to move so it's basically like forums and it's all organized by topics and inst but when it was created the internet was not really available and people wanted to say well we want to use this these digital media we don't have access to the Internet how can we do this so similar to something like fidonet this is computers calling each other syncing up data and and what you get basically is this so someone you you like a university will have a complete copy of you using it and then you can connect to them whichever is the closest to you and say well I'm only interested in this much of it and sync with them and then and then it syncing includes you're putting your responses back in that university takes your response and then forwards it it to every other connection and and it spans the world in these asynchronous kind of bit-by-bit connections going in every direction and a very a more recent one
which to me is such a clever hack both politically and technically it it's called Touche in its target it's basically about getting around internet blocking in Iran has anyone heard of Touche no it's not so it turns out that so the Iran often blocks aspect aspects of the internet satellite TV is very popular in Iran but the satellites are all operated by I believe UAE United Arab Emirates and they don't those countries don't get along so that is that they wreck the people who put two shaped together recognized oh this is an opportunity for us because ultimately a satellite is video is just bits and it turns out it's a standard format it's MPEG format bits so there's no-one checking that your bits actually our nice pretty playable video so to say is a system where they have a cable channel on a satellite run by the UA that's targeted in it's available in Iran on one of the on the main satellite that Iranians used to get television and if you subscribe to it and you basically a lot of these little satellite boxes you can download to a USB thumb drive so you can just download just reamed added to it and then they have a little app that extracts just like you know unzipping a file but it's taking data out of mp4 so this to me the whole thing is just such a crazy hack but that works that I found it very inspirational so so
then why do we now really want to think about incorporating offline and nearby networks like you know there's all these things before that was the only way or there was there was a direct need and now it feels like well it's so easy for much so much the world to just get stuff on the internet companies like like
Google have really pushed this idea that local they can provide always connected services and you know when they when you're always connected then they can run these huge servers and have these insane search indexes and all this stuff that you can't do in it where no one's been able to do in a decentralized way so you know that they so invested in this model of always connected and your net everywhere they're flying Bloom's Facebook as well you know that they have
their the idea of putting drones up to beam Internet always connected the internet around the world to me this I I'm I've always been very fascinated in the history of computers and the Internet and it reminds me of a one I think often overlooked but very
important story in the in the femme in the roots of the internet and in computing and that is a lot of this stuff like the internet was hat was being built around when the vietnam war was being fought and while we usually think of like a lot you know oh the internet was a bunch of hippies may Internet access or information available for free there that there is that what that was happening at the same time the people who actually were funding this the people in charge of the defense
defense defense Department funding had a very different idea their idea was about well if we have a nuclear war proof network that spans all everything then we can have the most efficient centralized command control possible we can build the network where every oh we have people taking care of everything and then when it'll be so efficient that the president can actually decide who which person to shoot weld is able to flip on a monitor and the airplane will be flying in the air and the president gets to pull the trigger this is this is this is really this is not a made-up story this is really part of what they were thinking about and one of the things that they did it so in the in that time so about 1970 the largest computer in the world was in Thailand because they had built a network all up through Vietnam and Laos and Cambodia where the Ho Chi Minh Trail was this is where the Vietnamese brought weapons from the north down to the south it's a fight so they there were the we had this idea we'll just build a network over the whole thing we'll wire up sensors and video cameras and audio sensors and then we'll know where everything everything that's happening on this on this trail
and then when we want to you know blow up some truck or something there there you go but the general can sit there and be like that one and we'll have perfect knowledge and that and so pi?? jamaat's money largest computer in the world they built the system and it was reporting all these statistics about all the things that they have killed and blown up here they have a nice map of all the
places that it had that were targeted via this system and then after a while someone tallied all the numbers that they had of all of the trucks and airplanes and everything that they and people they had killed with this and they realized that it was much larger than any of the estimates from the department's who just were trying to estimate the size of the military and then in after the war you can now see this in the in museums in Vietnam the Vietnamese that figured out oh here's a microphone and when it hears a truck someone shoots at it so they just played recordings of trucks so this this was really they had centralized command and control they built this amazing infrastructure internet up on the whole you know and this rural thing in a war and they were victims of their own information bubble so more directly what why do we need offline in nearby networks well what we have is is centralized very and getting seems to me even more so and it in some places in many places the world it's very fragile and pretty much any place in the world it can be fragile and when it surprising times I worked in lower Manhattan it during September 11th there was no internet no phones no cellular phones for weeks there was no power there were power lines this big in the street uh-huh so this this can happen you know
in the middle of New York City in a lot of the world is expensive it's getting you know easier and easier to censor it it is everyone knows it's being surveilled in so many ways and and of course it's lots of people are trying to push it to not be to not be neutral like have to prefer certain services of others so it's also not evenly
distributed which is a map of the world where based on percentage of people using the internet very unevenly distributed and
so part of why that is well it turns out that it's quite expensive in certain parts of the world and often in those parts of the world where people don't make a lot of money so really I mean affordability is a huge block to getting on the internet in a lot of places this is something that Facebook has figured out with their internet org program they basically figured out that you know if it's worth it to them to pay for the internet bills for people around the world to access Facebook but then of course they don't pay for the internet internet bills as a whole just accessing Facebook there's even I mean another
example of a project that we worked on so in the u.s. that there are large parts of the US where people are that do not have coverage for example farms a lot of large farming areas and so we worked with a group of representing farm workers in Georgia the we were helping them track the work that they did so they could get paid fairly and for the
most part they did not have access to the two mobile connections or the Internet this is in in Georgia not hardly the most remote remote parts of the US there's also lots lots of places
in the world so here's an example of Syria where there's actually the internet out you know the connection to that world outside of Syria is pretty tenuous there's Tartus basically has three connections and then there's one in Aleppo and that's it and in some countries they've designed them deliberately like this so Libya after the fall of Gaddafi they were able to kind of go in and look at how the was built it was built for to have one central building that could control every aspect of telephones in Internet
and another nice example were these in Hong Kong these protests that broke out and they where they occupy this causeway so it's kind of a very unexpected thing to be have tens of thousands of people you know in a highway and there was a lot they were trying and it was political demonstrations they were trying to get themselves hurt and they had put a lot of effort into trying to communicate and also keep abreast with what the police were doing and there was quite a bit of a tension but about I forgot the name there is a there's a service of a chat app that had a remote what mmm no this one was specifically about having a local connections fire chat yes thank you fire chat and it got a lot of publicity
here but it and that one it was quite interesting because it actually it worked as a regular chat app but if if you couldn't connect on the internet it would you could also try via Bluetooth just to send messages through the network so it's it's a nice model of saying okay well internet if you have it and ad hoc mesh if you don't so now these are a lot of nice ideas and hard
technical issues how can we do this that's part of the question that we've been working with and our most recent work has been related to a Mozilla competition around offline networks and we started we chose to focus on Latin America and to kind of be broad our horizons and we started out we had
we had done some work of the time we started out with some survey so you get some ideas of what what were the issues like what were the concerns about people so these are the respondents from these countries of that we talked to then we're just trying to get some general ideas about you know when you're using a phone you know what are what are the problems that you have and the cost of data was number two and batteries number number one so we had commented with a perspective very much thinking that we had heard I we probably had just learned a lot about our kind of how expensive it is to use phones and a lot of part of the world and that was our focus so then we were surprised to see oh actually yes that's a concern but battery time is is the number one so from from this also we
do say you know how much data do people use so 2 gigabytes or less is about half the people that responded to not not so much and then how much they pay so most
the people were paying more than 20 a month which you know knew is not so much money in a lot of places that's that's a lot of money that's a substantial portion of their income so about 10 percent of a lot of people's income so
and from this we so we did a lot of interviews and things like this and we also so we wanted to we find that having real stories is very informative to the process but there's a lot of privacy concerns so that actually saying oh can we interview and spread your your story about your concerns around the world so what we did actually was based on our surveys and interviews we put together fictional people and these are three of them and try to keep them close you know close to what we were hearing and so these are these I think there's eight or ten or so all available on on the okay thanks website it's linked on the the talk page and it just kind of tells a little bit of their story their their phone their income and things like this what apps they use and what they do with the phone and these were very for us very helpful in thinking about what how to prioritize things you you know battery we had we put more effort into battery life good serving then based on that we went back we've the longest-running piece of this
work is in F droid where we started in about 2012 with this what we then called swap app swapping and from there we've tried to think about so after it is in App Store it's all about getting a blob of data from somewhere to your device so you can use it you know whether it's an app a video of you book or whatever so in after ID we we really have this opportunity to think about like well what are all those 'fl ways that we can move blobs of data from here to there and how can we get them all into one user experience where you don't really have to be aware you don't have to say well I have to set up I have to you know oh the Internet's not working I have to set up the Bluetooth connection in a special way so that I will be able to use it so we wanted or mirroring or Baba so it's - it's and this is really taking shape more and more I mean on the left there you have some of the many ways that that data gets transmitted in the in the f-droid ecosystem so of course you have you have me up top is f-droid org it is there by default you open up the phone with Internet device is search for an app install it if all is the Internet's working fine it will just download it from there in each in in the after idea or can declare well I'm also available here in this mirror you know if you cannot reach after ID or org you can try the repo on this server so that's available in there but the client knows to say ok I can't connect to this one it's been so what's the next in the list and we'll try that and it'll keep on trying and then users if you want to set if you have your own server you want to set up your own mirror this closer to you you can set up your own mirror and add it anyone and share it to anyone and that same logic applies you don't the user doesn't have to know that the main one failed or even you there's no notification or anything it just keeps going until it can't find anything if it can't connect at all then you're notified so that still is a centralized server that is F dreaded org is a central community of people any the everything that F Droid makes is free software and we try to make it easy for the people to use it so for example there is a Guardian project app repository or repo anyway you can just kind of subscribe to it you can send the link via email click it and after it says do you want to add this and then that also can have its own mirrors and you can have as many of these as you want each with their own mirrors each with their own setups and then we have the nearby functionality so every phone has radios that can talk to each other there's Wi-Fi radios there's Bluetooth radios and these are set up to exchange data so that you know after it has the nearby tab to help you walk you through the process of making a connection so you can exchange apps for any APIs or video or things like this that is installed on your phone you can make it available and connect to someone else's device they see it in the regular app store experience click install and also then there's the problem of you know you need to do this procedure you need after a talking-to f-droid so if you want to send so yeah I have an app I want to send you I don't have EFT right ok well I can just send you f-droid straight so that's a relatively straightforward thing let's see on the left of just saying okay well I'll send you after it first and then the rest will be much easier to sync up and then of course it needs to also play nicely with other app stores so most the world of Android people are used to only having Google Play but large parts of the world many app stores is normal China for example India other places and for this to be a whole ecosystem they all have to play nice so if you know if you're getting absent Play and then you also install f-droid then f-droid will do what it can to to work nicely it will you can of course then if you install an app from Google Play you can of course share it on using any of these things and then we recently saw a nice kind of
a surprising example that we someone came to our chat room and just said oh by the way I've set up a store in Cuba and he so because we we put the effort into the whole tools to make sure that it just if it can talk on a network it'll work regardless if it's whether it's the Internet or not this man in Cuba of decided he was going to open a mobile phone store his and he's taught good way to promote his mobile phone store would be to have a local Wi-Fi access point that you could connect and get free apps and so that he was on the out edge of town he said well if you come here you can connect for free to my network you'll get free apps and then oh if you need a cable or alpaca thing I think he does that then my store is right here and so this is it kind of interesting this is an interesting model to me because it's for the most part it's offline it is one Wi-Fi access point with a hard drive plugged into it more or less and it works and then every so often he gets he syncs with the internet so it's a kind of like once a week internet connection ID along those
lines there's these little things like all sorts of little boxes that have Wi-Fi in storage in them now and this one's called library boxes software for it that it makes that idea of this kind of occasionally syncing to the Internet and having a local store very very simple and very cheap yeah so yeah I
also want to point out like we're not the only people doing this work and and really what we hope to see is a lot more like a lot more people focusing on this kind of thing and doing it in a way where we can in turn interoperate so briar is messaging actives talk from last year here at Congress from Torsten Grotti which has it's a very similar idea it uses the internet when it's available but it also can do local connections for sending messages there
are this we work a lot with Tibetan activists in Dharamsala in northern India where a lot of Tibetans are base they are quite remote didn't have networks they've set up their own whole systems of both a network and computer systems that cover their area then I mean what we're seeing now you
know even the Silicon Valley doesn't have perfect connectivity so you can get fine maps on things like on Google Maps and we're seeing you know Apple and and
Google now are also pushing this nearby idea is it's I guess hitting the mainstream and specifically so you know they have huge engineering resources which is a wonderful thing to see on that but they always come this far in the case of Apple but they often you know make it work really nicely but then they stick in arbitrary restrictions like you know iTunes is not allowed to do nearby because they don't want you sharing music or and and Google of course track adds an extra track Inc that's their business model and from from Apple there's a very important
lesson to always when you're designing systems think about spam and abuse this is a this is apparently become a thing in New York where people are doing airdrop Apple airdrop is they're nearby surface dick pics to anyone who's on the subway who happens to have left their their airdrop in I guess there's a like kind of listen all mode so yeah you open your phone and it shows you the picture right there right less than ideal and
there's also two apps called share it and zapier which have many millions of users which do need this nearby sharing quite well but that's what they look like to me that doesn't look like they're really focused on the nearby so much I mean when they started out it was very simple and it was all and it worked nicely and then it's the same you know they're startups it's the same business model now they're trying to track and monetize you with like endless the wipie things so you know we need to do it another way this the if we really want to have something like oh that's neutral and nearby and working offline the it's not gonna it's not gonna come from Apple and Google and it's not going to come from the VC startup fund we see funded startup and for to to highlight really
the way we're thinking about it like what we want to put up first and foremost is is that ultimately we want people connecting to people and thinking about how people act rather than encouraging them to act in a way that's that works with your business model and also putting this into the design of how
the protocols work so this is some early thinking about from paul baran who's involved a researcher involved in the design of the internet so it's interesting 1962 he's thinking about you know centralized networks decentralized networks distributed networks but these all assumed that those points don't move but people move and now that you know we all have phones we have to consider that that movement is always going to be part of this so that means at some points you will have a great connection to the internet so you can use it and maybe it's that centralized hub and spoke other parts it might be some more decentralized and other points you know it may be just connected connectivity to one library box that has specific services that you may or may not need there's a whole whole range of things that need to be considered if to make this whole kind of fluid experience actually work to break it down I mean
you would we have these kind of concepts as our and of course first is what we call it chime and that's the idea you have - it
has to be discoverable if someone sticks a little Wi-Fi box there if your device doesn't know how to find it then it becomes you know it's quite a bit of work to actually use that thing and you need something to store data in order to have an effective model where I say you have the once-a-week internet you also need a way for people to send their stuff out to gather and share it out and and so one thing I think is that the
most from this most recent work that is the most generic Andry shareable is to actually say okay what is the format of data for announcing yourself and a services you're offering this is our first draft of such a thing the wind chime announce protocol and so this needs to and now it's you know I am available on this Wi-Fi node so the SSID bssid I am available at this address I am available at you know whether it's
Bluetooth or not and this is the kind of service I am providing and it has to be in a standardized way so that it can happen automatically you don't want to present a whole list of text where you know many many many things to the user that's not gonna ever be useful then we
also have this available in a library for Android called Yonder which just tries to make these core pieces really easy to do in an app so that's the discovery piece the sending files and the receiving files part so
then the question is you know how can you get started doing this in software that you're working on and some of it is just a matter of trying as we've learned where you know most of the time you
think well you know if I have a network connection even if it's just a local network with you know small nodes it should just work right if I'm on a measure at work it should just work cuz it's all the internet well it turns out it's pretty normal to check for specific things that only exist on the internet and then that ends up being the only thing preventing your software from working in a nearby or offline way so that's the first thing just like well set up a network of oh you're get a library box one node see if you can talk to it with your app you doesn't work without a domain names of we're used to you know CCC dot de as but if you just have an IP address is the sequence of numbers will your app work because that actually the domain name makes it a lot harder to set up individual lightweight nodes and then you know in the flow of using your app is there data that can that can be sent and cached and you know in a way that's not say leaking private information things like that and just starting to think about that is that is the first step of saying and with a lot of little tweaks you a lot of software can work then on on nearby and offline networks without big fancy changes of syncing with Bluetooth and things like this so to wrap it up I just want to say that really what we wanted we know that this is possible so we want to build
this network in a way that reflects our values instead of always turning us into a product and that and and you know I that means building a network that they're theirs will do something useful if the big provider is not available making it easy that means building systems that you don't have to ask in advance to set them up just you know by the 30 euro tp-link turn it on and it should work in some form you know there's lots of things like official things like domain names which help but it shouldn't be essential and then thinking a lot about you know well how can we keep this affordable I mean affordable affordable accessible and without any kind of arbitrary things like well you can sync anything you want just not iTunes songs so affordability then you know it might feel like well you know it's less important here but the system is made up of affordable little bits that gives you a lot more flexibility in how you actually you build it so it really applies kind of everywhere and it close you just to
leave a little something that inspires us is these are these starlings in in England that every night gather and just fly around in these swarms of these self-organizing swarms which are you know many different individuals some leading some following the leaders change it's very highly recommend watching the videos are amazing of these of these birds and it's if we can make networks with that kind of fluidity of organization then that would be quite an amazing thing thank you [Applause]
[Music] thank ya that's what's up for your talk
and we have still some time for questions if you have to leave then do so quietly and we have two microphones one over band one in the middle and of course the signal angel just line up and I will call you to ask your questions don't forget to be quiet while walking out maybe a little show of hands if somebody wants to ask a question but cannot reach it because there are people saying in front of the mic yeah yeah okay just walk up to number two that's sitting right over there and go ahead yes you talked about the battle zone a case where they have some kind of a mesh network already as far as I understood
and you're told and you told us that probably the telecom providers will come there and take take like the piece of cake there and built their own business model there but what are the limitations so why is it possible because I think if the infrastructure is already there it should be cheaper to maintain its then then to pay extra for the telco providers so this you talk about the Nicaragua case know for example the Barcelona case Oh the Barcelona case so well the problem is you know if you have a company if you have good competition and they can build the networks cheaper and maintain them and and you know do them in a way that's with net neutrality and things like this that's great but I think part of the motivation in Catalunya is that they don't have much network competition and nick nicaragua very much so there's to comfort as a Spanish company and a Mexican company and that's it they're huge and so that these companies don't are not responsive to what people want to do and often put arbitrary restrictions and things something like that does that answer your question or but I was kind of wondering why it's not competitive why do I like the own dependently but infrastructure is not competitive I think in the case like of Nicaragua the telecoms really they have lots of money they really don't want anyone to have a set a precedent of being effective competition and so they were like oh they're actually getting subscribers let's crush them I mean they're monopoly is a huge huge matheny it's you know that's that's part of it so I think it doesn't matter the cost really they don't want people getting the idea that they can start competition and thank you yeah okay microphone number one so you currently focus on mobile technology do you consider the desktop or the laptop out of scope because it's irrelevant or just because you want to make it a feasible project so for the general idea as a whole absolutely it's fully in scope and yeah we want to see that me personally and Guardian project as a kind of organization we are developing work is on mobile and that's why this was have such a mobile bit but we'd love to see desktop and things like this number two please thank you for your talk I have actually two questions one is as far as I'm the student right now the Bluetooth is the way to go for this decentralized networks at least what people are doing right now and what is the biggest thing that we have a comp that people have accomplished until now like what is the biggest decentralized network blue toothpaste maybe in this chat client or something and the second question is you mentioned a lot of technical problems that happen and so how does their routing actually work with this Bluetooth stuff do you have like gel based routing or do people everyone knows everyone or yep thank you so yeah there's a let me start with a simpler question so we it's it's not just Bluetooth II so it'd be blue - so you can do Wi-Fi just if you have a access point without the internet that works there's also the things like Wi-Fi direct which seems to be can work nicely but not always so really you have to kind of consider all of them if possible and not why because bluetooth is often the easiest connect but really slow where Wi-Fi is harder to connect with much faster than the largest I mean using that as the largest it's like terabytes of data the word I mentioned with the forums copying around that is generally copied around on the on the internet that's the largest one that I can think of and that the last question was oh wow ting so that's I think for a lot of - Liat really with the this wouldn't really be on the kind of transport layer where you're thinking about routing the the idea more is to say like well so if you want to do ad hoc routing then you would build a mesh network because mesh networks are all about Matt ad-hoc routing firechat does some I believe they do some ad hoc routing in a very small scale in Bluetooth and some of that I think is freely available if I remember and so really the idea is more that so like in the case of F try to think is easiest is that you make we there's a system of so we have a concept of a repo which is a collection of absent etc cetera files and it has a kind of globally unique ID it's the site of the key so that means when the client and kaya encounters some files you can easily identify it as oh it belongs I know this is f Droid or because it's the matching signing key and it can just say wherever it found as long as it can talk to these files I mean what I'm then it can exchange data so there's no routing there if it's on a mesh network and it gets an IP address on that medical network and it gets to the same files the index nose so it's about in the app you have to think about okay how what are the things we need to discover what are the bits we need to move around and how can we do that in a way that's reusable you know it's very easy to just connect to the internet and say oh yeah use this domain name assume it's always there and I think for that for so after is one example I think briar is probably very good example where they're integrating internet and Bluetooth and local Wi-Fi to do messaging okay just to check is there a question from the internet no questions from the internet so then number two was there before um one of the things that I noticed this what you're talking about apps and after all it and distributing apps and then connecting to local nodes what about ipfs that and and some of the other distributed networks have you I'm sure you looked into those and you were talking in the beginning about like integrating with other initiatives like matrix and so on was on the slides and talk about that and yeah um ipfs for people who don't know is interplanetary file system the idea of one file system there is accessible over the world I think ipfs is this great example for this kind of thing and we would love to integrate it it's just it's only so many hours of the day in the day question really we've been focusing or so right now I mean it's something we very happily if someone knows ipfs and wants to try and integrate it into something I would happily work with them we my personal focuses is on the lower hanging fruit I think or at least from my perspective based on what I know and that's like some of the mirroring stuff some of the like the library box local internet free box the Bluetooth so that's based on my or the pool of skills and interests of the people who've done the work that's part of what I hopefully am trying to you know what in conveying here is is that this needs to be decentralized and effort there's so many ways and and and it's like okay how can we find what ways to tie all these things together and make it so that you know messages can go from a Bluetooth mesh over the ipfs to then the centralized internet and back and work without people thinking about it okay unfortunately our time is now up so finds of course off here at the
Congress if you have more questions later at any place around here I guess so and yeah let's give a warm hand of applause for [Applause] [Music]
[Applause] [Music] [Applause] [Music] [Music]
Feedback