Millions that Don’t Care
      Ilya Zverev
      
      >> I think it's time to start.  So I am Ilya Zverev from Russia and I'm going to talk about my favorite topic, the inevitable demise of OpenStreetMap and what we are doing to, well, avoid that.  And I will start with resource.  I hear a lot of  OpenStreetMap, but the most used are the top ones.  It's pretty clear from statistics.  Mobile users get in person to about zero point whatever users.
      And what the operators, the more popular ones, what they are good for?  Basically, tracing imagery.  Everyone traces imagery in all the areas.  The first question a new person might ask is how do I get a Google satellite in my area because there is no  there isn't my OpenStreetMap.  So basically all that people do is trace imagery.  And we know missing maps project is very good for it.  And people like to trace buildings and roads in their own hometown.  And so 
      That's not a bad thing, but I assume that the world has been mapped pretty well by now.  And we have heard some talks about neural networks and how they do all that work for us.  So basically what we need to add in OpenStreetMap is local knowledge.  And I'll tell you that a resource operators are completed and suitable for aiding local knowledge to OpenStreetMap.  Because unlike  the local knowledge isn't inside your home.  You have to go out and somehow collect it and then get back to OpenStreetMap.  And there is very much of it.
      There are thousands, millions of shops, amenities, whatever.  And experienced mappers are not enough to collect all the data.  You have to  you have to find  make regular people contribute data to OpenStreetMap.  And it's obviously not easy.  Because in the current workflow you have to know about OpenStreetMap.  You have to learn the operators, you have to prepare field papers or make photos, and you have to run the security stuff taking photos, and you then you take that to home and you load it in the disc operator and then you fight timing and base issues and then, you know, you have to browse through features page to know context and so on and so on.
      So entering local knowledge points in OpenStreetMap is not easy.  But we do need it.  And the way to do it is quite obvious.  It's making people go out on the street and take knowledge.
      And there are already companies that make people go out of their homes.  Like this one.  Just two days ago, while I was waiting for a social event I have seen every 20 minutes people coming and catching some Pokemons.  And they're doing the mapping in the computer, but, of course, it can be leveraged.  And we'll know what company are the first to leverage these people.
      There was an announcement by Google just two days ago about adding a feature into their Google maps application.  So everyone, even without a Google account, can just edit any place on their map or at a meeting place.  And adding the local knowledge to Google Maps.  They don't have to register.  They don't have to do anything to learn any tagging system or whatever.  Just, well, collect the data.
      And that's the first part of it.  The second part is Google actually reversed such contributions.  They put in incentives like if you contribute a lot, then you'll become super wise in your area.  If you contribute you can then receive some gifts or maybe go to the mapper's gathering in your area.
      And it works very good.  Works quite good because a lot of people making Google maps have had a hard time not pressing submit button when I try this feature because it's great.  And you can  where OpenStreetMap comes into this.  And last year I said that this doesn't actually matter for OpenStreetMap.  That we are in no hurry.  We are not a business.  We don't have a deadline.  We can  in 50 years there will be no Google maps, but OpenStreetMap will finally get all this local knowledge.
      And I think I was wrong.  Because when a mapper in, say, Malaysia gets to that shop that hasn't been mapped, they won't find it because local knowledge have the feature that it appears and it appears quite quickly.  So if you don't map it now, there will be nothing to map later.
      And things we map now will disappear in a few years from Google, but will stay in OpenStreetMaps.  Our map doesn't reflect the real world unless we make other people contribute local knowledge a bit faster than that.
      And for that we need to reach out to millions of people with their mobile devices.  To not make them learn about OpenStreetMap, about taking schemes and so on.  And that is basically what we have been doing at MAPS.ME.  It's a pretty simple application for both iPhone and Android.  It displays maps.  It's basically everything you think when you think about option mapping of maps, for bicycle, pedestrian, car, bookmarks, searching and coding and whatever.  So pretty neat kind of thing.  Yeah.
      And what separates this app from other opposite map applications is that it has not only developers, but also a marketing team.  So our task isn't just to make an open source application, it also is to promote it so as many people as we can.  And right now it has  it's nearing 50 million users.  So it's basically the singlemost popular maprelated application.
      And it can 
      [ Applause ]
      It can seem at first that this is just another case of some business people coming and trying to take business advantage out of open source data, but they didn't forget about OpenStreetMap.  And this April we added editing to this application.  And we were trying to bring our 7 million monthly active users to edit OpenStreetMap.  Yeah.  So that's what it looks like.  It's pretty simple.  It's mostly like the one in the Google Maps.  Adding features, adding text.  It's  well, from OpenStreetMap, it's also very simple just  it adds nodes and edit text on objects.  So it's pretty trackable.  You can revert it at any time.
      Yeah.  So we opened it to 7 million active users and started to wait for results.  And in the first month we get 10,000 users registered to adding things to OpenStreetMap.  Not as much as I'd hoped.  I hoped for a million, actually.  But because I submitted this talk before we launched editor, hence the name, a million users.  But still.
      Yeah.  And still there's a good way just to master it with marketing.  And there are  we decided to bring every user directly to OpenStreetMap, not through some proxy or through moderation.  So every of these users have been registered in OpenStreetMap with their actual verified email and they could contact us.  But the registration process in OpenStreetMap is not simple.  I think one of the reasons that there are so few people is that it's very hard to register on the OpenStreetMap Website.  Nobody from OWG wants to help with that.  But whatever.
      Yeah.  About some statistics.  Three months have passed and you can see some surges on the charts.  Like this contributor per month.  You see that it was around 8,000 before maps .ME and almost doubled after.
      [ Applause ]
      Yeah.  I'm just finished.  So active contributors, so we have 10,000 more.  Each month this graph  this chart is about adding notes to OpenStreetMap.  You can do a search and you can see that there were spikes before.  That's the air, that's probably ebola, MAPS.ME is clearly better than ebola.  Oh, sorry.
      Here is  it's called  well, percent of users for initiator.  You can see that it was around 70% users were using it before MAPS.ME.  Now MAPS.ME is slowly going to be the most popular in OpenStreetMap.
      But, of course, in the number of edits, MAPS.ME is not so high.  Because most users only make a couple of edits.  But these are important couple of edits.  They add things that are near their homes.
      Of course, users can inundate the map as all MAPS.ME users during the month, but that's different.  So we published and waited for a response from the map community.  And at first it was, well, pretty good.  Longstanding mappers found out about these and tried it.  And said it's quite useful.  You know how it goes.  There are some shops near your house and you want to map them, but you postpone it because there are  well, near your home, so you can do it tomorrow or next week.  So I had some that were unmapped for like a year.
      But with this application I did it quite fast.  And there were users that registered to OpenStreetMap in like 2007, 2005.  And didn't map.  But with MAPS.ME they basically made their first edit to OpenStreetMap.  So all this time they were waiting for the correct editor to appear.
      All right.  And last month mappers have finally started to notice there's something happening in OpenStreetMap.  And this is part of the post in the mailing list which basically was about how MAPS.ME users destroy the map.  It was very strange because nothing  there was literally no response for two and a half months.  And then suddenly people started, well, to write some stuff.
      Yeah.  Of course, MAPS.ME edits have their flaws.  Mostly because the editor is just three months old and it's not perfect.  So you'll note, for example, China's tourists traveling around the world and changing names of landmarks to Chinese.  Or MAPS.ME, the data.  You don't have to be online to do anything in the MAPS.ME.  So, of course, there are duplicates of points of interest because the data is old.  And some users may add the same thing twice or more.
      A list of types in MAPS.ME is limited, so some people choose incorrect types.  Or maybe some users add, like, their home as a tourist attraction and name, I live here.  And that's not very good, I think.  Yeah.  And some users, some mappers, said that people editing maps with MAPS.ME don't answer.  But I think that it resembles the pattern with other users.  So it's not really MAPS.MErelated.
      And the MAPS.ME data team has tested their suggestions that most updates through MAPS.ME are bad.  They interviewed 200 people and found out that only a few of them were actually can be considered vandalism.  For example, this one is basically a department store.  So over  you can see that one is open.  So at least you know where to go. 
      Yeah.  And then another person wrote into us and I found out that I'm the cause of all that's bad in OpenStreetMap.  I'm very proud.  And I expected somebody else in the mailing list to answer that.  You are not  well, you shouldn't say that.  But the next reply was, well, it turned out that they were right, really.
      What happened  and then the next day 
      [ Laughter ]
      The next day independently users have also found out about MAPS.ME in another way.  And I think they all wrote even worse  they said that we should regard all MAPS.ME, we should ban all MAPS.ME users who did things through MAPS.ME.  We should ban the MAPS.ME editor and until the operation working group is installing that.  Or the working group.  We should install auto for all coming through MAPS.ME.
      >> [Question away if microphone]
      >> I don't know, actually.  I  and between these constructive suggestions they also  well, they were calling names, put in obscenities, and the forum has no moderators sometimes.  So they're still there.  Yeah.
      And some of this has leaked into OpenStreetMap diaries.  So there  the question is, why do they write it?  Why do they feel that way?  And it's not because I break OpenStreetMap.  I don't do that.  It's because mappers in opposite map found out that they have no control over these edits.  There are now around to 40,000 users that have contributed something through MAPS.ME.  And we don't have any tools to deal with that.  We don't have any monitoring, analytics and whatever.
      So we often say that we need a lot of new mappers in OpenStreetMap.  We say that, well, a million or two will be nice.  Yeah.  But when it comes to actually seeing new mappers to train, and new mappers  well, it doesn't really work.  Because OpenStreetMap members want all new people to learn everything about OpenStreetMap.  How to use editors, how to tag things.  But you know, OpenStreetMap turns 12 in a few weeks.  And you know map updating was all notes, relations and texts.  And in 12 years nobody has written a tutorial on how to choose text.
      All manuals basically say there's highway, there's amenity, and the rest you can read on map features page.  Have you seen map features page?  That's not a tutorial.  Right.
      So what to do with these all new contributors?  In the past few months, years, I have made some  there's of course who did it.  This is edits for Seattle for the past week.  This is a simple tool that we can subscribe to changes in your area.  And see if anyone deletes anything.  You should use it.  It's the only such tool on OpenStreetMap.  And then of course I wrote a reverter.
      When I  before this you had to install a plugin and this was quite hard.  Now we can have ways to change that link.  And when I announced it on the Russian forum, people said, oh, great, now we can install and automatically do all changes from MAPS.ME.  That's when I left the Russian forum for good.
      Yeah.  And then there's monitoring and statistics tools for MAPS.ME edits.  You can see that there is 300,000 edits by this time.  It's not changed  it's changes.  Concept of change stats is very bad in OpenStreetMap.  So there are  well, top users, countries.  So you can filter by anything.  So, for example, see recent changes for your country.
      And also right from this tool you can open this in level zero.  It's the greatest editor ever made.  And you can revert to them.  So if you see any vandalism, you can revert.  Yeah.  But when we made the editor I really expected the community to make more analytics, monitoring and other tools to deal with a large number of changes.
      And things that were released in the last few months were quite  not as good.  Because monitoring  some monitoring tool was just counting changes in the application div.  An analytics tool was basically just displaying buildings that appeared over a few days.  Yeah.
      Well, the problem is  the issue is definitely not in editor, but in our tools and in our community.  Yeah.  And I'd like to remind you about Google maps and that we actually really have competition.  Yeah.  Thanks. 
      [ Applause ] 
      Do we have time for questions?  Yeah.  Any questions? 
      >> How did you handle  [away from microphone]  
      >> Well, MAPS.ME actually doesn't give object versions, so there are conflicts not possible because it looks at what has been changed in an object, it downloads the latest version and then just changes with whatever is in the database.  Yeah.
      >> [Question from microphone]  
      >> Well, whoever is last wins.  So  yep.  Sorry.  Yep?
      >> So you don't feel that there should be a call to ban  I mean, visual imager on MAPS.ME?
      >> Well, it doesn't hurt to call to ban MAPS.ME, at least now.  But the issue wasn't in spotlight, like the issue wasn't in maps.  It's our tools our data model.  I think it's fixable.  I don't remember anybody calling to ban spotlight.  And MAPS.ME doesn't break anything in OpenStreetMap.  It just adds a lot of data that we need to process.  Yeah.  They say we have to  but you can ask me questions later.
      [ Applause ]