 Okay. Good morning. Good morning everyone. I feel like you're all so far away in this other room. Good morning everybody. Hello there. Do you see okay? Can you see the screen and everything alright? Yep. Okay. I feel comfortable. I mean we're here together for two days. We met a lot of you last night but I think probably after you, we didn't meet. Just short introductions for ourselves and we'll... I don't think we need to do the warm up and stuff anyway. Okay. I'm Mikhail, Mayor, and this is Mr. Scalero and we are your humble facilitators and teachers for the next two days and you are our teachers too because I'm sure there's going to be lots of interesting questions and things that are coming up. I don't know. I feel like just jumping right in unless there's anything we should talk about first. No, I think that's it. I think we should go ahead and get started. Okay. However, I would like to make this as interactive as possible. Yeah. So please, if you have questions that come to you that are relevant to the topic being discussed, please don't hesitate to jump in and inquire if anything is unclear or could use further detail. We'd like to hear from you. So don't be silent. We might say that it'll be covered in like, you know, the next session or whatever, but... Hopefully. But anytime you have a question, just feel free to interrupt. No problem at all. So please go ahead. Okay. Well, cool. Welcome to Cardinama Workshop. We're here for two days and we've broken down the workshop into a number of sessions which are grouped approximately into four half days. We start off with nothing. Right now we have nothing but ourselves, our computers, our existing knowledge. But from this point into the end, the idea is that we will take all the steps necessary to build a location-based application. Starting off with actually creating geographic data and then processing and working with geographic data, visualizing it, that is making maps that you can look at, and then finally building the apps with a goal being a location-based app that actually could be pretty flexible for using for other purposes. There's a number of resources which might help you to follow along. In fact, all the slides are on GitHub. So there'll be some links. You'll be able to jump ahead. You'll be able to look at different resources. Can you read it in the back? Is the text too small? It's at github.com.com. It's basically a markdown file of the slides for this talk. There's a live loop CV, ISO, which you may have, which has all the software we're going to use over the next couple days on it. Real quick, actually, on the subject of the live CV. Is this a guide here? So I'm thinking we probably don't want everybody to run off and download that ISO over the already bogged down Wi-Fi. So is it possible to pass around USB sticks? All ready to be done. Okay, great. Wow, you guys are so thoughtful. We don't need that. Yeah, we're not going to be using it very well. Yeah, we'll need it for editing, but actually, that's actually just a... When we edit, I don't think everyone is going to be editing at the same time. Maybe on team-up. Team-up. We'll team-up for that. So yeah. So all right, well, that gives you guys time to copy it around. And you can have virtual machines or whatever to do that. Yeah, cool, cool. Okay, there's one thing I'm hoping into this one. Okay, it's happening right now, being copied into Glass Drive on this very computer. So there's the slides, the ISOs coming around, the workshop schedule you've probably seen. So I think we are running a little bit of wait at this moment, but we'll catch up. And the mailing list. Those are kind of the main resources for the conference, for the workshop. So start off with creating GeoData. We're going to be going through how you create Geographic data in OpenStream. And OpenStream, I've heard about it before. Pretty much. That's great. Well, it's like Wikipedia for Maps. It's a global collaborative project that was started six or seven years ago, which has at its goal creating a global map of the entire world, contributed by volunteers and free to use by anyone. The project started because mapping data wasn't available. This was even pre-Google Maps. And a guy named Steve Coast had the realization that with a device like this one, which we'll be playing around with in a little bit, these are the GPS devices, just with having this, which is a consumer-grade GPS device, you can actually make maps for yourself. And if you convince all of your friends to also make maps, then you'll make the job a lot easier. And before you know it, you'll have a map in the entire world. And that's a crazy idea that somehow is starting to really work out. That's the Mellonliner. I mean, anyone can contribute. There's no moderation like Wikipedia, which has a little bit of moderation. Anyone can contribute at any time. And all the data is available for access through an API and through data downloads under Creative Commons licensing currently. Creative Commons licensing allows for reuse of the data for both commercial and non-commercial use under the license that we have, only with the stipulation that you share back any changes that you make to the data. So that if you get some local stream map data and you notice that the name of the road is incorrectly spelled and you correct it, you have to contribute that back into the Commons, not even directly back to the street map. You just have to make it available. Creative Commons licensing has worked for a while, but there's a lot of issues when you deal with geographic databases, which some lawyers get excited about. So we're moving to a new license called the Open Database License in like next month, actually, in one month exactly. If you look at the legal wording in Creative Commons licensing, particularly around things like derivation and attribution, it's not exactly clear when you come to maps when that applies. So for instance, if you have two photographs that are both created, one's Creative Commons license and you have another photograph and you like paste it on top, you match them up or whatever, that new piece is a derived work and should also be shared under Creative Commons licensing. But in mapping, a really common thing that you'll do is you'll have a base map and then you put a bunch of points on top of it. They may come from two different databases. Under Creative Commons licensing, it's unclear whether that combined work is then shared like. We think it shouldn't be, in fact, and you should be able to layer on other data sets with other licensing without triggering share like. Open database licensing makes that kind of thing clear legally and it's just been going on for a long time. It's caused many of us to become lawyers when we haven't wanted to be, but it'll just make it a little bit easier for more people to use the data. So this is an image that was created early in the project. The project was started in London, England, and there was a taxi, a courier company, which had a unit in the vehicle with a GPS and with a GPRS and they were transmitting their positions to the control center for this courier company every minute. They had to feed the 50 cars and as a result they had a heck of a lot of GPS data and they early on just gave that to OpenStreetMap to do something with because it's hard to collect this much GPS data. And this was an early visualization. This is just three days worth of GPS tracks over London and even in that little time it's already almost a usable map. You can see the River Thames and there's things like the Royal Parks and different parts of the city and roads are already visible just from GPS tracks. And so this was really inspirational to convince us that this was, yes it was still a crazy idea, but maybe it was actually possible to make a map with the entire world by running around the GPS units and using OpenStreetMap software to collaboratively build the map. Let me go with this stuff. This is actually an animation of the history of OpenStreetMap in Bangalore. Hopefully the net will cooperate. So the database keeps a full revision history. So all changes over time are available. And so this is one interesting thing that someone did with those traces. You see up around along the bottom is a kind of a time slider of history. Here we're in the middle of 2007 and this is still the map of OpenStreetMap. And then around 2008 the name maker appears, some of the major roads. This was around the time in early 2008, Schauer and I came and visited for the first time. And it's a little bit washed out actually. Here we are approaching 2010 and then I think this was around the time that Arun Ganesh actually moved to Bangalore. You see some of the parks are coming in. There's a lot of detail in the center that you can't see in the projection. Yes, it's swelling. The music is like these strings. Ah, the map, it's cut me. So it's actually, I mean if you look at OpenStreetMap for Bangalore right now, it's an incredible amount of mapping data. One of the key events in the history of OpenStreetMap was actually the Haiti earthquake which happened in early 2010. Massive earthquake, massive loss of life and immediately following that very little geographic data. It's very useful if you are coming in as a search and rescue team to a city in a place which has pretty much been completely destroyed to actually do your job, you have to know where to go. You don't have the photo of the National Mapping Agency? No, the National Mapping Agency tragically was completely destroyed and most of the staff perished in the quake. And so now it was a very little data to share before then the capacity for the Asian government to even share what they had was completely demolished. Google Maps have not lowered the idea. What's that? Google Maps have not lowered the idea. And Google Maps was not very detailed. And OpenStreetMap wasn't very detailed in Haiti. Google Maps wasn't very detailed. They were not very detailed either. But there were a lot of people online who were working on this notion of crisis mapping and came together in, especially in OpenStreetMap, we got access to free imagery, up-to-date imagery of Haiti, of border prints. And it's too bad that this is kind of washed out from this project. But the bottom image here just has these few major roads and this is much more filled out. And the remote response, which was several hundred people, maybe 750 people in the OpenStreetMap community, tracing aerial imagery, happened in just a week or two. And as a result, OpenStreetMap being a freely available database became the default base map in Haiti from all the response agencies. Pretty much all of the UN agencies and Big NGOs will have some kind of GIS capacity, any data to work with. And so they were downloading OpenStreetMap and making maps with it. This is the kind of level of detail we were collecting. All the X's like here are actually destroyed buildings, which you can see clearly from the imagery. These are IDP camps, which you could also see gathering. And of course this is all the rest of the city and border prints. And this was immediately useful. This is actually the same map on a Garmin device, just like we'll be using, which a search and rescue officer from Fairfax County, Virginia, loaded the map onto his GPS before he went to Haiti. He'd never been to Haiti before. There was a lot of American presence because Haiti is so close to the States. And they were using OpenStreetMap right away. Did you leave out my photo of the Belgian guy? I think so. I have a story to tell about it. Yeah, yeah, that's good. About the immediate aftermath of the quake in Haiti. I was about two, maybe three weeks after the earthquake and I was working for the World Bank in the GIS, the Office for the Coordination of Humanitarian Affairs, GIS 10th at the Airport and Border Prince. When this guy from the European Commission, Humanitarian Aid Team comes in and says, I have truck drivers that are going to Grand Oav and Petit Oav to verify a food distribution scheme and I need OpenStreetMap to give them. Well, Google Maps would be okay, but really what I wanted was OpenStreetMap. And I was like, oh my God, this is a person who is completely non-technical. He's not a mapping person. He's a humanitarian aid worker. But he had heard, it turned out, that OpenStreetMap had the most detailed maps of Haiti at that time, which we did. And so he knew to ask for it by name. And that was just incredible to me. I felt like that was a real watershed, you know, just out of nowhere. I was like, I want OpenStreetMap. Oh, Google Maps would be okay, but I really want OpenStreetMap. Yes! Ask for it by name. Ask for it by name. So we've continued to work in Haiti. The humanitarian OpenStreetMap team formed around that time. I can talk about that more later. But there's actually a Haitian mapping team on the ground now that does just the same stuff we're going to do today. Who are mapping all over Haiti. This is important. A lot of emergencies, there's a huge rush right after something happens. There's a lot of attention, and then as time goes on, organizations lose interest. But with OpenSource and OpenData, there's a way you can continue, even though the exact people and personnel might change. Another big project, OpenStreetMap, that I've been personally involved with is called Mount Kibera. And Kibera is a large slum in Nairobi. Some of you were here for Erika's talk last night. Basically, regarding mapping, it was a blank spot. It's a very famous slum in some places. It's a movie star slum. It's sort of like Darabee. And despite 250,000 people living there in two and a half square kilometers, it was a blank spot on OpenStreetMap, on Google Maps. And there's a lot of information, a lot of research about this place, but it's not openly available. There are maps of Kibera, but it's on someone's desktop somewhere, and they can't access it. And this is a very poor, desperate place, but also a very interesting place. And a very place full of people who are eager for opportunities. So I am an Erika. We went in 2009. We worked with local community organizations, recruited young people who had very little computer skills, like they could not really use the mouse or actually save a file properly. But they're very interested in gadgets and GPS units. They don't have phones, and they learned very quickly and actually made the initial map in three weeks of the slum, something that no one thought was possible. But they're from the slum. They're from that place. And so they're basically the best position to know what's there and to know how to make the map. There's the map, a little bit washed out, but you can see this is the main railway that goes through Kibera. And there's lots of features. Everything from churches, pharmacies, clinics, water points, toilets. There's lots of shops. It's actually a lot of commercial activity within Kibera. And that whole got map. There's thousands of points of interest just in that 2.5 square kilometers. And the project continues in various ways. One, particularly it's being used in a Ushahidi instance called Khoisa Kibera as a community reporting platform. So this is a place where reports come in pretty much daily and get placed exactly where they happen on the map. And it's a very interesting archive and ongoing reporting site for what's happening in Kibera. Do you guys know Ushahidi? No, it's an open source piece of software developed in Kenya after the 2007-2008 post-election violence. The basic idea is that you can send an SMS in. It goes into this web platform. That text can be expanded upon and geocoded and everything gets displayed on the map. So it's a very easy way to, say, crowdsource information from any kind of event. So those are two examples. And I think to talk about what distinguishes OpenStream app from Google Maps or from other mapping projects, to me, there's three main points. One is an open license, which we talked about a little bit. Basically, the license allows anyone to do whatever they want with it as long as they contribute back to the data itself. There's technical freedom. So there's an API. There's downloads of all the data that you can access. And you can use that information however you wish. You aren't restricted to one platform or another. You can use ESRI tools with that OpenStream app. There's a plug-in in ArcGIS 10 to edit and use OpenStream app. There's lots of open-source tools. And so it integrates with pretty much anything that you want to use. Lots of web platforms. We'll go into all of this. And there's a community. There are tens of thousands of people who are really passionate about mapping, who spend their free time mapping. And there's a lot of active discussion. And having that community around data is crucially important because if no one really cares about the data, it will quickly soon go stale or quickly lose its value. So having tens of thousands of people who are interested in the map is important. This is one thing that can happen if you do not have those kinds of things. I think that is an actual map of Holden McDonald's in Los Angeles. At least as a few years ago. There's probably a few less by now actually. They're trying to do a closing down. But if you go with commercial providers of maps, then you are subject to their terms, which might result in advertisements being placed on the maps or charges being made on the maps. So as a commercial user of mapping data, you might not be interested in the ideology. Open source is great, but you're not necessarily running a business to support an ideology. You're running a business. You want to be successful. And so these are the kinds of considerations. There's definitely trade-offs that you make. And this is one thing to consider using these commercial providers is that they can change their terms of service at any time and things can change around. This is a photograph of a Deccan Airlines magazine. The airlines you can still around. Yeah. Oh, Kingfisher's going down the toilet. Well, when we were here last time, they had this map of the route map. It actually had a warning at the bottom. This pictorial presentation does not purport to be the political map of India. It sounds like a tongue twister, but basically they're saying, don't get upset if you don't agree with the borders of Kashmir. Well, I hope somebody will correct me if I'm wrong about this. Is it actually illegal to publish a map in India that does not display the official borders of the line of control in Kashmir? So if you publish a map in India that has different borders, you can get a serious trouble for that. And I think that's why there's this disclaimer. But the point that we're getting at here is that OpenStreetMount as a one-tool or one dataset gives you the opportunity to depict the world in your own way. If you want to depict the world with the borders that India officially claims to have, the data lets you do that. And if you wanted to depict the world with different borders, I mean, if you might want to do that, I don't know why. You can't. I mean, if you are using another provider, you do not have that control. It's up to them to decide. And there's several cases with Google Maps where they've become the arbiter of geographic truth. And I know even the Indian government has a lot of Google Maps to change how they show the borders. I feel like that's something that's important enough to be in everyone's control. This is an example from Hurricane Katrina in 2005. A bridge not too far from New Orleans was wiped out completely by the hurricane. That's the image here on top. Despite that, for two years, if you went to Google or Yahoo, which at that time all relied on Naftech or Tele Atlas data, these are the big data providers, if you ask for routing from this point to this point, it actually puts you over the bridge. But if you went to that bridge, it's gone. And so it turns out that it takes a very long time for those data companies to notice that something's changed, package it up in their system, distribute it to their customers. Their customers need to load it up into their system. It takes a very long time for that to come back into the extra applications. Do you want to tell us your story for today? Sure, I'll tell you. So we actually made aware of this by... Are you hearing me okay, or should I use this? So a good friend of ours was actually... He's a firefighter and also a geeky who used to work for Amazon. And he went down to Katrina to help out. And he was actually camped out somewhere near here. Do you notice a lot of people coming up to this bridge and then getting out of their car and getting really confused and looking around and he kind of started asking like, why are you here? The bridge is out. And they were told, well, I called the Red Cross and they told me to evacuate this way. And so he called the Red Cross and was like... They had a huge call center at that point. He asked, why are you sending people this way? The bridge is out. There's no reason to send anyone here. And their answer was, well, are you sure? Because I'm looking at Google Maps right now and that bridge is still there. He's like, no, I'm pretty sure because I'm standing right here and there's no bridge. And that person said, okay, okay, well, stop. I won't recommend that anymore. But people kept coming because other people at the Red Cross were looking at Google Maps and they believed that was like the truth. So that's when we started to investigate this. And eventually they did change the routing. You can see here, this is the bridge and they started route run. But actually two months after this changed the bridge reopened in one direction. And so you need to be... The world change is really fast, right? And so that's one thing I think is important. One thing that OpenStreetMap excels at is that changes can be brought into the database and propagated really quickly. So here's just a bunch of examples of things that have been done with OpenStreetMap data. This is what you can't necessarily do with other systems. This is a tap-tap map. Tap-taps are like the bus system in Port-au-Prince, but it's much less formal. There's no bus company. People just get a cab or a truck and start driving the route. So there's no formal system. There's no maps, but using OpenStreetMap, using GPS and data, all of the routes and all of the stops were entered into OpenStreetMap. And then this is a visualization of the route of the entire system on tap-tapmap.org. This is a tourist map. It's actually a printed map that's printed out and distributed. And printing is something that you definitely can't do with any other dataset. And it's actually a tourist map of the Gaza Strip, of Gaza City. It's kind of a tongue-in-cheek thing about the tourists in Gaza, but there are a lot of international succumbs, so they're sort of reimagining Gaza under a different scenario. So first of all, you cannot get maps of Gaza anywhere besides OpenStreetMap because of political restrictions. If you go to Google, it's not there because Google actually has a pretty big Israeli office, and so there's political tensions there. OpenStreetMap doesn't have that kind of restriction. Anyone who wants to map can map and do what they want with it. This is an example of it. You can use any sort of tool to bring data into OpenStreetMap. This is aerial imagery, which has been warped using the OpenSource map where Scala worked on. And it's recent imagery from flooding in the Philippines, which floods almost every year. And it was used to help create data for the response. And so this data was able to be georectified and then brought into the various editing tools in OpenStreetMap. You don't always have that technical freedom. This is a photograph of a couple of these guys who are mappers in Afghanistan. Did you see any structure? Scala was just in Afghanistan. Did you see any amazing structures like this one? No, unfortunately. Probably just stippled up in your hotel a little bit hard to move around in Afghanistan. But these guys are locals in Afghanistan and they can move around, and they've been mapping in Eastern Afghanistan around Jawa and done a really amazing job. And the map in Afghanistan is being used in those sorts of ways. This is a map of the forbidden city in Beijing. You can see it quite well, but it's in absurd detail, obsessive detail. This is... Sorry, you can't see this well. This is actually the Vatican City in Rome. This is another example of printing. This is a book on cycling in London and OpenStreetMap was used in print. So you're free to use the images in this way, as long as you give attribution. Did you see that? You've gone through it, and you're the only one here. Does that mean you can't print off all the maps? Not legally. I know a lot of people do. I've even seen it, but it's... Yeah, it's not... And certainly if you put it in a book, you see advertisements sometimes, like small-scale stuff. They're not going to be able to catch everything, but you can even make blankets. These are map blankets. I think this is San Francisco. This is also San Francisco. Oh, an office was right there. So if you really, really love maps and you really want to feel more cozy with them... You know how this was like right here? Right on the blanket. Yeah, useful in so many ways. That's what I think they have. You can even make cake out of OpenStreetMap. There's something called the cake test. This is actually a cake from OpenStreetMap Data. And if you're free to use a map for making a cake, you're pretty much free to use it for doing anything. So that's the test, by the way. Free map data, ask yourself, is this license allowed me to make a cake from this map? If the answer is yes, then you have a delicious map. It is also free. This is a piece of software which is 3D rendering and it's brought in OpenStreetMap data. There is some, especially in Germany, where there's an obsessive amount of detail, there is some 3D information that's been added into OpenStreetMap. This is the C-Fact of a train in, I think it's great western trains in the UK. It has a map, it shows your location. It's all offline. It's not actually requesting from the net. But you can store all of the maps offline on whatever server is running on this train and it just uses the GPS on the train to show your exact location. It's what? Well, so we'll get into how you actually, how data is represented in OpenStreetMap. And I can talk a little bit about how you would tag things so that you can talk about 3D data. But then once the data is there, it then gets transformed into other data formats. So this piece of software probably has, I don't know what data format in order to represent 3D data, but basically you get this chunk of XML out of OpenStreetMap and write a transformation to put it into this program. So... It's not connected directly. It's just another project which is making use of OpenStreetMap data, which there are many. OpenStreetMap at its core is basically a database and an API and a website. And then everything else is built, the editors, all the visualization, everything else is built around them. That's completely washed out and it's really beautiful, by the way. This is a map of France, land cover. Check the slides on one. This is New York as a video game. So this is an 8-bit rendering, Mario-style rendering of the map of New York City from OpenStreetMap. So you can do all sorts of crazy things with it. And increasingly, big commercial users are starting to switch. Part of the reason is that there's no charges associated with Google Maps. But yesterday, Foursquare, which has some penetration and switched to using OpenStreetMap as their default mapping player. And here's the community. I think that's really what... This is the core of the community. Several hundred people who attended our yearly conference state of the map. But in India, there's a very active community and there's thousands of people all over the world who are just passionate about sharing OpenStreetMap and helping other people. So that's the introduction. How are we doing on time? Not bad. We're catching up. Any questions before we move to the next session? So the core of the project is maintained by the OpenStreetMap Foundation, which is a UK-based nonprofit. I'm on the board. We have no paid staff. Our only expenses really are servers and some bandwidth and lawyers and other things like that. So it's actually very low budget at this point, and most of the funding comes through donations. We get a lot of sponsors for the conference and sometimes there's fundraisers for specific things, but the overall budget's very low. If you compare it to the Wikimedia Foundation, which has a budget of $10 million, I think we're below 100,000 pounds per year, which doesn't mean we couldn't use more money or spend more money. It's just sort of an in-open-source project. There's often kind of a bit of tension with bringing money into the project, especially when everything's all volunteered. We have no paid staff whatsoever, including our system administrators, who are, I think, heroic, but I don't know how long we can go on with people spending their... having two jobs, having their day job and then their job up running our servers. But that's the core. It's quite federated. There's companies which invest a lot of time in developing software for OpenStreetMap, and the kind of work that I do is there's a lot of development sector interest in OpenStreetMap, so it just comes from lots of different ways.