 So, I'm Ilya Zverev, going to talk about some fun stuff, and, well, OpenStreetMap, I'll start, I guess, with the usual question, who here knows about the OpenStreetMap? Awesome. Now, who here does not like the OpenStreetMap? Oh, that's great. You're a terrible person, I will, we would need to talk. But then OpenStreetMap, well, gives plenty of reasons not to like itself, and it's a great project. It's open, under license, much like GPL, which must make Tillman happy. It was made by a programmer who didn't know anything about geography. It has great data model, which leaves room for some creativity. And when I explain OpenStreetMap, I usually, after that mention, that the moment you click the save button, all your changes are immediately uploaded to the server. And in a couple of minutes, everyone can see them. That's awesome. There is no moderation. Just map, and in real time, see your changes in the map. And interestingly, basically, the first question everyone asks, not about how to do something with OpenStreetMap, but what if I would like to draw something like this in OpenStreetMap? Or something like this? Or move some roads around? What is it with people wanting so much to break things? Or like this? Or maybe delete some stuff? Yeah. Will it break the OpenStreetMap? Well, for a bit, but there is communities that all the time watch edits. There are some very useful tools that, like who did it, in which you can subscribe to an area you're living, like Chinchet Analyzer, which highlights edits that are suspicious. And we have tools to deal with that. So it's not a long-time solution to breaking OpenStreetMap, because in a couple of days or if you're lucky in a couple of years, all your vandalism will be reverted, and the map will still be great. So we could look for another solution, because OpenStreetMap is open, it's based completely on OpenSource software stack. So maybe we could just sit and wait for it to break by itself, not OpenSource, you know. Will it work? Well, it did once. There was a time when some identifiers surpassed their four-byte signature limit and some core software broke, but in a couple of days, the community has fixed it and the project still works. So you might think that OpenStreetMap lasts and nothing can break it, but actually we found the solution. It's not the thing we're proud of. And the key to breaking the project is basically you need to not want to break it, you need to wish it well and to work every day on improving it. And then one moment you'll find that the common saying is that you heard the thing most that you love applies well to the open projects. A bit about me, I work at MapsMe. This is an app for Android IOS and it's basically a viewer for OpenStreetMap. It started a long time ago, just out of a need to have some offline maps on travels without requirements for internet, it's complete offline. And from that, it started to grow. Nowadays, it's got better UI and 3D buildings that you see and some useful stuff on the mainty cards, routing, sharing, whatever, a lot of stuff. It's an app store on Google Play, you should download it. And a lot of people over the time have liked it and installed, right now we have 11 million users who use the application monthly. To give a scale, that's exactly the population of Belgium. Well, in the time where there is no FOSDM around. So we thought that we really like OpenStreetMap. It has allowed us to make this great application and to get some money. So we want to contribute something back. But we're only a team of 20, so directly mapping won't give much value to OpenStreetMap. So we decided to share the thing we had the most, these 11 million users. So we implemented a simple editor in the application. It's really easy, just click a point and there are some, a simple form and stuff like that. No text, no geometry, just sends data to OpenStreetMap. No, you have to log into OpenStreetMaps through the application. And it works as well as we hoped. Not millions of users, but still many thousands. It immediately increased new contributors user base, as you can see on a lot of graphs. And we consider it as success. I've built a tool to monitor the edits. And as you can see, recently we surpassed 1 million edits with our application. It's around 100,000 of users. And the thing I'm proud of the most is that it's not regular people from regular countries that you can see on OpenStreetMap, like Western Europe and the US. You can see Iraq, Philippines in the top 10 and so on. So there are people from all over the world improving OpenStreetMap through the editor. And from the OpenStreetMap site, it is really, really simple. You can do basically two things with MapsMe. It's either create new points or change attributes. And that's all. And this is uploaded directly to OpenStreetMap without any processing stuff. So what could go wrong, basically? It's very simple. You cannot break any relations with it and stuff. But some people have started to notice some weird things. Like quite a lot of people, very many people, have started to mark their homes on the map. It's weird. You know, there are a couple of security tracks going on right now, somewhere around here. How to protect private information and stuff. But a lot of people don't care. They just say, this is my home. This is home of all my friends and there are roads which I take to get to them. It's weird. Then there were some Chinese tourists. They're okay, but they started renaming all the attractions, adding a name in Chinese or in Russian or in Korean, which also looked weird. And the mappers didn't like it. And there's basically a lot of stuff that you can't explain. Mostly it looks like people just mapping for themselves. Like adding these Peruvian sites and mobile phone shops, which are obviously not. They look... Yeah. And of course, mappers from OpenStreetMap came to us and said, you are breaking the OpenStreetMap. Please fix it. And we started to think, how do we do that? Okay, another month of work. Finally, we think fix that. So we put a big panel with warning that everything you map, you are set into a global database and billions of people will see that. So please don't add your homes. And we're forbidden to edit default names that show on maps. So users can't add Chinese names to this. And edit some monitoring and validation just to be sure. Okay, so this should have worked, but not as great as we hoped. First, since we forbidden editing default names, there are multi-language names, which started to differ from default names, as you can see here and here, is just in one hour. So it was a little frustrating. And people were still adding their homes. But now we could say that that is not our fault. Every user have seen that warning, so contact this user, it's not our problem. And then the code of dealing with OpenStreetMap of converting our field types or field values to OpenStreetMap has grown big. And some of our mistakes started tipping on the map. Like some of our internal tags were added to OpenStreetMap. Not much, like 100 tags on the home world, but some mappers made a big deal out of it with posts and a lot of diaries, with discussions on forum, how MapsMe is destroying their map with those tags that can be easily cleaned up and stuff. Then we found out that we broke in some addresses in some parts by translating the street names. It is really a big deal, and we're working on fixing that. We fixed it in the application, but there are some left. But nobody in OpenStreetMap, virtually nobody, has noticed it and which I think proves that mappers are not conserved with data quality and mostly with MapsMe, but still. And the licensing issue, we see a lot of, well, some people that are adding thousands of addresses onto the map using our application, but we're not sure they're using proper sources, open license sources. But we can check it, not sure what we can do with it. So to go a bit back, first, this is really a great application, and it allows you to add a lot of places to update open hours and stuff if you like to improve the map and a lot of people are using it. So when we added the editor, I finally mapped a shop near my home, which I was too lazy to map for previous two years, and OpenStreetMap is not only great, in most cases, the only choice you have when you need a map. And of course, we love it very much, and I, myself, being on the board, I adore it very much. But underneath all the success stories, all the great data, the data model of OpenStreetMap is not as good as it could be. So if you try to do the simple editor for people who don't care for OpenStreetMap, and it's not for 10,000 of people who know how to choose tags from 10,000 different tags they use, you know, that tagging wiki pages are like 50 screens long, and it's impossible to select. So we wanted just to make an editor for millions of people, just to make it easier. And when you try to marry that with the OpenStreetMap data model, most times you feel like this. There is always something that you need to fix. But we're still working on that, we're improving validation, and other companies are looking for quality of data, so yeah, that are not big problems, just things we're working on. And the last thing, from the start, OpenStreetMap project, like any Open project out there, wanted to have more editors. It was its goal since day one, I guess. And in the last year, the number of its daily contributors has doubled from three to six thousand editor daily. And it's great, and MapsMe has contributed to it a lot, and Pokemons have contributed to it, but that's another funny story. And somehow, Mappers are not happy, they've got thousands new contributors, but, well, they're not as great as they hoped. So I guess the lesson is be careful what you wish for. Yeah, that's all. Thanks. Right? I guess we have a lot of time for questions. Do you think OpenStreetMap could survive a financially backed, coordinated attack, like a distributed button that attack was somehow financed when it survived it? Well, first, do I think that OpenStreetMap would survive a coordinated big attack on it? And first, the API is slow enough that you can add a lot of data in the short time. And if you do, there are always backups from the previous day, from the previous hour, and if it's really bad and the map is broken, then you can just revert it. All the history is stored, so it's not a big deal. And well, our sysadmins are pretty active, and they tend to notice this stuff early on. Yeah, I have a question. I've been using MapsMe on FDroid, but I wonder why there isn't a brand-new version on FDroid? Right. About MapsMe on FDroid, that's not for us. We don't plan to support an FDroid version, but all the code is open. It's on GitHub, so anyone can build an Android version. And the main barrier for publishing it on FDroid was that nobody could find a server to host their own version of Maps. So if you have a server with just 40 gigabytes of disk drive, you can go to FDroid Ticket in MapsMe issue tracker and offer your help, and that way we can have an FDroid version. Right. I can answer questions for OpenStreetMap, for MapsMe, whatever. Yeah. Right. A question about MapsMe slowing down in the past months, I guess. Well, obviously, when you improve an application, it gets slower with time, but of course it's all for reason because, for example, just a month ago we added real-time traffic information, and to render it is to get some processor time from other things, so it gets more slow and stuff. So yeah, it gets slower, but we have really great graphics team, graphic engine team. So they're working on improving things a bit. So, yeah, it will be, just you have to wait. Yes, it is a company that was bought by a Russian company, Mail.ru. It has around 20 people, of them around 13 programmers, and yeah, it's not a separate company, it's subdivision, but we're pretty independent. The business model of MapsMe currently, there isn't one that's sustainable. We still spend more money than we earn, but last year we started to make a partnership with other companies, like you can book hotels from MapsMe via booking.com, and you can call an Uber and some other stuff. So we have this partnership, and we have some plans to make it sustainable, but it's absolutely not an experience of the core functions, it will always function offline, it will not send some private information to servers, so the core functions and security stay the same.