Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
×
Open Source Businesses Technology

An Open Source Resistance Takes Shape as Tech Giants Race To Map the World (factordaily.com) 90

Posted by msmash from the keeping-the-fight-on dept.
Shadma Shaikh, reporting for FactorDaily: Chetan Gowda, 27, was speaking to a room full of students in IIIT Hyderabad for a workshop on OpenStreetMap for beginners organized by Swecha, a non-profit organization to support free software movement last month. There were close to 40 students in the room. Beginners often ask him: Why use open source maps when we already have Google Maps? For Gowda, it was the fact that Google Maps is a global, commercial product and did not capture local detail. Like the old banyan tree that was a major landmark in his hometown Hassan or public benches just outside the town where pedestrians could stop to catch a break or fire catchment areas in Bellandur lake in Bengaluru, India.

"It was fascinating to add little but important details of my town to open maps," says Gowda who was introduced in 2013 to OSM or OpenStreetMap, a global community of mappers formed as a collaborative project to create a free editable map of the world in 2004. Since then he has been an active contributor to OpenStreetMap and has conducted many workshops in colleges and institutes to induct more people in the community. Gowda has made 8500 edits in the OpenStreetMap, mainly covering areas in Bengaluru, Hassan and Hyderabad. Gowda and a few other contributors from India are part of a tiny yet growing resistance movement which doesn't want giant corporations to own all the mapping data. For the average consumer, this may not seem like a big deal. But mapping is big business.

The market opportunity for suppliers of mapping to the autonomous car industry is going to be worth over $24 billion by 2050, according to one estimate [PDF]. And that's just one industry. A study commissioned by Google in 2015 estimated that industries that run on top of the Global Positioning Satellite Systems and mapping generate nearly $73 billion in annual revenue. Worldwide, that industry is was estimated to generate $150- $270 billion in revenues. Although new research isn't available, with growing smartphone usage and the birth of companies such as Uber and many others it is safe to assume that the industry has only grown bigger. All the more reason why map data can't be held by only a few companies. With Google Maps beginning to charge small and medium-sized businesses and indie developers more for access to its platform, many have started to explore and switch to open source alternatives of Maps, and commercial services such as Here Maps.

Further reading: What OpenStreetMap Can Be, and Ten Years of Google Maps, From Slashdot to Ground Truth.
This discussion has been archived. No new comments can be posted.

An Open Source Resistance Takes Shape as Tech Giants Race To Map the World More

An Open Source Resistance Takes Shape as Tech Giants Race To Map the World

Comments Filter:

  • So, it will be like Wikipedia for maps? (Score:4, Insightful)

    by El Cubano ( 631386 ) on Monday October 01, 2018 @06:58AM (#57402734)

    For Gowda, it was the fact that Google Maps is a global, commercial product and did not capture local detail. Like the old banyan tree that was a major landmark in his hometown Hassan or public benches just outside the town where pedestrians could stop to catch a break or fire catchment areas in Bellandur lake in Bengaluru, India.

    "It was fascinating to add little but important details of my town to open maps," says Gowda who was introduced in 2013 to OSM or OpenStreetMap, a global community of mappers formed as a collaborative project to create a free editable map of the world in 2004. Since then he has been an active contributor to OpenStreetMap and has conducted many workshops in colleges and institutes to induct more people in the community. Gowda has made 8500 edits in the OpenStreetMap, mainly covering areas in Bengaluru, Hassan and Hyderabad.

    This sounds like a really neat idea. However, I quit contributing to Wikipedia because of the nonsense that comes with it. It was very frustrating to see the constant edit wars on even semi-controversial content. To say nothing of really controversial pages. Or of the people who think they are "experts" on some topic trying to "correct" people who are actual experts.

    Then there is legion of "we must add every detail of everything, no matter how minor" pitted against the legion "everything in Wikipedia must meet some arbitrary high standard of significance." And of course, everyone has to put up with the admins (I won't even get into everything that can go wrong there).

    I can't help but think that while this will in the macro sense be a good thing for society and for democratizing information, part of the cost will be that you have to actually investigate the history of every thing that actually matters to you. I can also see how over time it will become increasingly difficult for people who want to make just one or two small contributions when there are others with nothing better to do than to "police" their favorite content full time.

    • Re:So, it will be like Wikipedia for maps? (Score:5, Interesting)

      by Anonymous Coward on Monday October 01, 2018 @07:48AM (#57402966)

      I have been an active contributor to OSM now for about 1.5 years. I started these edits since neither Google Maps nor local public transit route planner recognized my home address, which resided on a "road" on top of a shopping mall, not connected to the actual road network. My edits concentrate mostly on my neighborhood which tends to be under constant flux due to massive construction sites, but also extends to the general capital area of a small European country. Amount of edits on this area hovers around couple dozen a day. Of these, maybe once a month I see an edit which might be considered somehow antisocial or unprofessional in nature, and usually they're corrected quite quickly.

      I think it works out quite well at least on my region. OSM has a stance to accept as much information as possible as long as it has general usefulness and it's factually correct (and not opinion-based, which shouldn't be that hard on maps!). Removing such information is not welcome; the product of OSM is data, not a specific visualization of it.

      Sometimes it *does* feel that too much is too much. Marking up every shop, every path, every tree and every traffic sign can be fine on most occassions, but I happen to live on a spot which way too many publicly traversable layers; there's a subway station, an underground logicstics hub, five levels of underground parking space, an underground bus station, one level of underground shopping mall, an underground street, a partially street-accessible, partially underground shopping mall level which extends to neighboring buildings, three over-the-ground shopping mall levels and a pedestrian path on top of the underground street, a jogging path with planting around it on top of the shopping mall, running partially under residential buildings which reside on top of two HPAC floors which also host a private plaza on top of them. There are public pedestrian paths of importance at least on half a dozen layers, streets on almost equal amount, and often they run on top of each other. What is the preferred detail to map all this?

      I have chosen to map as much as helps pedestrian, car and public transit, but leave rest of the details out for now. Map in such an environment becomes quickly so stuffed that a typical map user can't understand anything on it without a navigator interface. Why a shopping mall appears to have lawns inside grocery stores? What all these criss-crossing pedestrian paths mean? (It's typically not obvious that they're on different levels, but they are needed to provide public transit accurate and efficient walking plans.) I would actually want to stuff a lot more information on the map and nobody is really preventing me from doing it, but rendering would simply turn illegible if I did so. Thankfully for 99.9% of environments this isn't a real problem.

    • Then there is legion of "we must add every detail of everything, no matter how minor"

      This reminds me of stories a friend of mine used to tell me. He was one of the RC Patrol (I have no idea why) guys on Wikipedia 10ish years ago. He said they constantly had issues of articles being added for people who were basically nobody. It was almost always about Indian men in IT or computer science of some kind. Constant articles like: "Ambish Gupta, 32 years old, is an IT worker employed by Whatever Company since a date barely 1-2 years ago."

  • Have anyone actually used it for navigation? (Score:1, Interesting)

    by Anonymous Coward

    I have and albeit being eminently usable, it clearly (and not unexpectedly) lags behind other big players alternatives when it comes to map updates and consistency in general. That detail alone ruins it for end users not engaged with submitting corrections and updates to the maps themselves (I'd say that would be 99,9% of users).

    There's a reason Google and others want to charge for maps, and that's because having properly maintained maps is a valuable service. Navigation software can be frustrating with cor

    • Re:Have anyone actually used it for navigation? (Score:4, Interesting)

      by vtcodger ( 957785 ) on Monday October 01, 2018 @07:29AM (#57402870)

      "There's a reason Google and others want to charge for maps, and that's because having properly maintained maps is a valuable service."

      And one that involves significant costs if it isn't produced by volunteer labor. Realistically, for commercial products, either someone is going to be charged money for the map or the maps are going to come with ads, or both.

      • I would love for Google to charge $$ for Google Maps. Right now, they're completely unreliable, with API calls failing frequently. It's completely unreliable. We would love to have some mapping service that we could pay, so that we could have some sort of guaranteed level of service.

      • Maps are becoming more important with the rise of autonomous vehicles. That's why Nokia was investigating the use of cellphones as means to keeping maps current, and accurate. One would also think with all the drones flying around the citizens mapping movement would be in high gear.

    • Good for navigation (Score:5, Interesting)

      by Kludge ( 13653 ) on Monday October 01, 2018 @07:40AM (#57402914)

      I use OSMand, which is an Android app that uses Open Street Map. It works pretty well for navigation.

      The best part is, unlike Google Maps, I can preload the entire maps onto my phone, like an actual GPS device, so I do not need a data connection to navigate.
      Also great is the Wikipedia feature, which automatically pre-downloads Wikipedia articles related to points of interest. On vacation I can walk/drive around, click on interesting things on the map, read the Wikipedia article, and appear amazingly educated, without a data connection. It started to drive my family nuts in Athens as I described the historical significance of everything.

      • Re: (Score:1)

        by Anonymous Coward

        The data also works on Garmin GPS systems.

      • Re: (Score:1)

        by Anonymous Coward

        I'd love to ditch Google spyware completely and use Osmand only. But it's still slow as molasses, memory hungry, uses modal(!) popups(!) to prompt for downloading maps, shows useless administrative borders that look like paths, keeps forgetting waymarks, and the search function is horrible (no search as I type, no fuzzy, no proximity/frecency sorting). When I use a nav app I need it to get a fix and show me a route within seconds -- Gmaps does that, Osmand doesn't. Gmaps also has public transport timetables

      • The best part is, unlike Google Maps, I can preload the entire maps onto my phone, like an actual GPS device, so I do not need a data connection to navigate.

        You mean exactly like Google Maps. This has been an option for some time, or at least you can select regions to download.

        • Re: (Score:2)

          by GuB-42 ( 2483988 )

          The regions you can select in Google Maps are limited in size, you can't select entire countries.
          Not good it your plan is to use it for a long road trip.

    • Re: (Score:1)

      by Anonymous Coward

      It depends. My regional public transit (maybe 1-1.5M inhabitants in the region) uses it as their map and especially pedestrian/bicycle routing data backend. For these regional purposes it's unbeatable (especially as the public transit authority is quite active at maintaining anything they get as a feedback from public transit route planner users).

      Then again, if I look 80 km to the countryside, to my parents place, I can see that maps lack lots of detail and are very rarely edited in comparison the the city.

  • For Gowda, it was the fact that Google Maps is a global, commercial product and did not capture local detail. Like the old banyan tree that was a major landmark in his hometown Hassan or public benches just outside the town where pedestrians could stop to catch a break or fire catchment areas in Bellandur lake in Bengaluru, India.

    I'm sure that matters to that person personally but that's a TERRIBLE argument in favor of open source mapping projects. If it proves important to enough people then Google could add that capability in a matter of seconds and then what is his complaint? There are excellent arguments why not having all your mapping data controlled by a few large companies is probably a good idea. They are similar to the arguments why having a decentralized internet not controlled by single entities is a good idea. There i

    • Re:Terrible argument (Score:5, Insightful)

      by cascadingstylesheet ( 140919 ) on Monday October 01, 2018 @08:54AM (#57403356) Journal

      The problem open source mapping projects are going to have is funding and resources (especially people) unless they can get one or more big companies with deep pockets to fund such projects. You need satellite map, an army of people to pore over and process them, a huge amount of hardware, a well coordinated team to oversee the whole thing and write the relevant code, and a shit ton of money to make it all happen. Not saying it's impossible but it's going to be very challenging and such a project is already years behind what Google, Apple, and others are already doing. You're talking about a project that rivals the linux kernel or other major software project in complexity but requires a lot more people for data processing and hardware to actually function.

      Well, you've just explained why Wikipedia "can't" work ... yet it does.

      Not saying that will be true of everything, but clearly it's possible.

      Not to mention that OSM has been around for quite awhile, and clearly does "work" fairly well.

    • Re:Terrible argument (Score:4, Informative)

      by petermgreen ( 876956 ) <plugwash@p[ ]ink.net ['10l' in gap]> on Monday October 01, 2018 @09:23AM (#57403520) Homepage

      You need satellite map

      Fortunately they have found companies prepared to donate access to Aerial imagary (afaict actual sattelite imagary is usually too low resoloution to be much use), the two big ones seem to have been Yahoo and Microsoft.

      such a project is already years behind what Google, Apple, and others are already doing.

      Depends on the area.

      In the first world openstreetmap has as you say had to start from a position of trying to catch up. In some areas they have caught up and even overtaken the propietary mappers, in others they are still behind.

      OTOH my understanding is that there are places in the world where openstreetmap contributors were the first to produce a map of the streets in the area.

    • Not saying it's impossible but it's going to be very challenging and such a project is already years behind what Google, Apple, and others are already doing.

      You do realize that Apple was years behind Google when they started their mapping project, and Apple used OpenStreetMaps as their base map to catch up.

  • Open source resistance (Score:1)

    by Anonymous Coward

    I've tried OSM and I may have developed an OSM resistance. Their database is great but the search is totally unusable.

  • Maps have always been inherently public property (Score:4, Interesting)

    by rbrander ( 73222 ) on Monday October 01, 2018 @09:38AM (#57403608) Homepage

    I don't mean that a privately-made map is inherently public, of course, just that almost the only people who found it worthwhile to map were governments. Better put would be "inherently of low value, but to large numbers of people so that cheap access for everybody was the only way to pay for it".

    A map has huge value when you need to find someplace new, but the huge majority of travel is to already-known locations. Cab drivers are an exception, but consider London, where "The Knowledge" required for cab drivers, is a memorized map learned on the job.

    So there are very few indeed private companies mapping - the paper maps of your town for decades were just purchased data from the city government, sold for a tiny fraction of what it cost the city to make, because the city had to map every pipe and street anyway to maintain them. Indeed, to know where the heck the property lots were. (Land titles are generally a higher level of government, but where I worked, the Province had an agreement with the City to let the City map all lots inside its borders and provide that to Provincial Land Titles).

    Google changed that with their cool car-with-8-cameras mapping, but generally also buys the City data because it's sold so cheaply - and is maintained every year, whereas you can see on the Google maps that the photos are only refreshed after multiple years.

    For non-commercial use, City data is mostly free these days - "open data" initiatives became common years ago and they post up files in ESRI's "shape file" format (ESRI is the Microsoft of GIS, their formats are like MS office formats). There are also free standards like "KML" files.

    Bottom line, there is no reason to let any private companies take over this space. The government mapping efforts have not ceased; the "value added" from information about business and services is *easily* exceeded by the OSM editing described here: people who live there will always have an advantage at highlighting local interests. (Also, the value of a location depends on who likes it, not "who pays google" to flag it.) The streetview is one of those features that's more cool than actually useful.

    OSM is available for your phone, by the way, and works almost identically to google: uses your GPS to just show the map around you. Give it a try!

    • I recall back in 1988, I was working for the US Census Bureau doing something they called a "Pre-Census Survey". They had maps with all their blocks configured. These were compiled from a large number of sources. City maps, State Dept of Highway maps, Planning maps, etc, etc. Was about 85% accurate. I was in rural WV and some of the things they listed as roads had not been roads for 100 years. Other areas were where the Dept of Highways had originally planned to re-route roads to, but never actually did it.

  • I yet to find something on F-Driod that is as good as Maps.
    I don't know if this an app thing or a OSM, but the apps that use it cannot find my address.
  • The problem is that autonomous cars need maps with this level of detail in the first place. A truly autonomous car should be able to integrate data from a basic map (i.e. line drawing of how streets connect converted to a node diagram in a database), GPS, street markings, and the outside environment in order to get from points A to B. After all, non-autonomous drivers do this with a set of less-than-perfect cameras and microphones, not even with IR cameras or microwave radar.

Slashdot Top Deals

Avoid strange women and temporary variables.

Close