I tried switching to OSM in an effort to get away from google maps. People always say 'OSM is consumer ready' but forget that OSM first and foremost is the dataset. And as far as i am concerned, there is no good maps app on the iOS appstore that even comes close to google maps. and i think i tried pretty much all relevant ones. skobbler (which i just found out is part of coast) seems to be the most polished. but as long as there is no unified search and you have to enter streetname/adress/city in seperate boxes which excludes searching for the names of buildings (for example university buildings) for which the names clearly exist in the database and are shown onthe map, i have no choice other than switch back to google.
i really hope this changes soon and the UI/apps catch up to the greatness that is OSM.
edit: i have an iphone and although i live in canada i still use the german appstore (CC requirement).
that's why i can't comment on the app the original post was about as it's only available in the US store. it was more of a general remark about my frustration with the state of apps using the osm dataset. I want to get rid of google maps and I feel the maps part of osm would be ready for that. I can tolerate if the commercial store data is not as up2date. But if the general usuability suffers, i rather opt for gmaps.
I tried using OSM a month ago when I was travelling around Bali, Indonesia. Google Maps is terrible there, so it wasn't like I had much choice. I looked on Google Play and found an app called "OsmAnd Maps & Navigation" that was well rated, it was also offline which was a plus as there is limited mobile coverage outside of the main parts of the island. I downloaded it then the dataset for Indonesia (200mb, it took a while), and tried finding a route. After a few minutes (!) it had calculated one, but the route definitely wasn't optimal. I tried another app "Sygic: GPS Navigation & Maps", it was premium but had a 7 day free trial (perfect for my 3 day road trip). The route it calculated seemed a lot better, so I went with that. I was actually really surprised by how well it worked, even in the north of the island which is quite undeveloped in places it was able to find the best routes.
TL;DR; Even though the dataset may be ready, you still need supporting consumer ready apps to use it.
(Originally posted this as a separate comment, but it is pretty relevant to this).
Have you tried the Scout App (which the article is about)? How does its search work?
I agree that geocoding has been a problem for OSM, especially for building numbers, where the data often isn't there. But on the software side there seems to be progress recently, for instance, here are two open-source geocoders released in the last few months:
On building numbers, we need geocoders which can interpolate between sparse data, and then use that system to highlight where the gaps are. People tend to add data to OSM when they're confident that it's actually being used, to its full extent. If there are a hundred buildings along a road, in theory you don't need to add many of them before you can get a very good idea of where an address might be.
The map should communicate that it is interpolating the geocoded address by a circle of probability. Or even show known addresses and then an alpha blended triangle between them.
As you are saying, assistance from a mapping service need not be binary, just add some extra information and communicate cleanly the probability of result.
Interesting projects! I have tried to use the Mapquest geocoder API but for anything outside the US it's worse than useless. Will definitely check those out. It's great that they're based on Elastic Search since obviously geocoding is about full text search more than it is about "geo".
not even Google Maps have full building number information. at least not world wide - I've been to several European countries where building numbers are ignored in Google Map searches (sometimes you actually get no result at all if you include a building number)
Agreed. Google Maps isn't just about addresses. For example I just searched for "museum about noodles in tokyo" and it correctly came up with the Shin-Yokohama Ramen Museum. I typed in "mona lisa museum" and it came up with The Louvre.
It's not perfect but it doesn't seem like a map based solely on map data is going to be able to compete with one that has all of the internet behind it.
This is my biggest complaint about Apple Maps on iOS. Searching doesn't work. And if I get an address or name partially correct, Apple Maps usually fails.
Google can find me places of interest, without an exact name or address. That's my most common use case when using navigation on my mobile device.
It's not just maps. No matter where you are searching, Google is very aggressive about trying to figure out what you meant. Almost no one else bothers to try. Amazingly, even Amazon didn't try to correct misspelled product searches until recently (a year or two ago, IIRC).
That's a huge and under-appreciated advantage for Google's mobile apps and services.
The screenshot they posted on top of the blog site is very illustrative. Although I may be used to the "Google Maps look", the OSM map displayed there is much harder to understand. I can't put my finger on it exactly, but there's too much color and detail I'm not interested in when looking at a map. This feeling does not occur to me when I use Bing Maps or Here Maps either (and I have used these quite a bit in the past).
Let's hope the increased use of the incredibly detailed and free OSM data will one day lead to better applications that make use of it. User experience comes first.
I just tried the OP's app (Scout Maps), which is free, and it looks pretty good so far for where I'm at. Not as clear or concise as Apple or Google maps, but the road data seems good so far.
Isn't apple based on OSM? Which basically means we got more players here. All that's left is for google to jump on the bandwagon and we're in business.
Apple used some of the old data, before OSM changed its license to share-alike. That data is probably still in there somewhere, but I think it's very unlikely they are incorporating any OSM updates, past about five years back, because the license would mean they'd have to release their combined dataset.
Android really needs to do something about the permissions model. This app asks for every permission in the book, but I have no way to tell if the app will use these permissions for a good reason.
I agree. Anyone know whether there's an Android distribution which will give you more granular control over permissions? Cyanogenmod seems to be a little better, but I haven't used it myself.
There was a feature named "app ops", once released accidentally and disabled afterwards [1].
There are also other 3rd party solutions [2].
I remember that some ROMs go even further and prompts you each time an app uses a specific permission (there is a "remember" option of course...). MIUI seems to be one of them, though I'm not sure.
Granularity isn't really the problem. The problem is there's no connection between something I want to do in the app and the permission it requires. It might be obvious why this app wants to use my microphone at the instant I tried to invoke some feature, but at install time it's not apparent at all.
I'm just gonna point out something: OSM is 10 years old, and after countless hrs and investing its finally ready for prime time. People who think a new internet is coming via some disruptive technology need to read this article and realize that real, long-term, awesome change takes time.
yes, and part of the reason this took so long could be that google maps was free/ very cheap for a lot of that 10 years. Once they got more expensive people had more incentive to switch....
I would argue that really depends on the subject. Maps are to some extent limited. Yes, they are never 100% up-to-date, but looking at my small village if I had mapped that 10 years ago all I would have had to change up until today would be maybe 15 new homes and 2 new streets a few stores and some small adjustments, but that's pretty much it. On the other hand if I wanted to build a search engine that's a completely different story.
It's still not really usable for navigation, has all kinds of accuracy problems and the worst is that they don't even want help scaling up the data pipeline, they just want armies of people to click by hand on their weak editing app.
Labor is expensive, technology is cheap, why insist on such an insanely labor-intensive data pipeline?
Things are changing in a way. There is Slide from Strava ( http://labs.strava.com/ ). Where you just roughly click where a path is and it creates nice path from its database of gazillion GPX tracks. It can be used as an iD editor plugin already. There are scripts for automatic lake and building creation from satellite images but most interesting thing seems to be coming from Digital Globe.
In this year state of the map USA ( http://stateofthemap.us/session/mapping-the-world-in-raster/ ) they said that this year SDK is coming where they would automatically draw buildings and streets from satellite images and you would get APIs to play with this and insert into OSM. (They already provide images to OSM through Bing and MapBox). So yes drawing streets and buildings isn't fun but you need people to know what is a street/cycleway POI and etc.. There are also projects for figuring out one way streets, max speeds.
OSM prefers user created data. Mass imports often contain errors all over the place (many going back unnoticed for years). OSM holds a general mindset that data should only be included if someone is fairly certain the data is correct (which they would not be on a mass import). "When in doubt, leave it out." That said, it is more of a guideline than a rule, and there have been mass imports in the past.
One solution might be for them to maintain their own error-checked database, and then provide a website and map overlay ontop of OSM, pointing out where the discrepancies are between the two. Then local mappers could go around and confirm the corrections.
I do think in the long-run OSM will need some error-checking processes if it aspires to be a definitive database. At the moment, I could quite happily subtly mess up a motorway junction in Japan, and most of the OSM mapping software would have no way to knowing not to package it up and send it out. Vandalism is not an issue now, but as the dataset becomes more popular, I think some mechanisms will need to go into place to protect its integrity (particularly for major infrastructure).
At some point it becomes difficult to mechanically integrate the data, and having the data nicely integrated is sort of one of the selling points of OSM. There is also a repeated pattern of imports planning on 'fixing it later' and then not really following through. Another thing is that existing OSM data will tend to be more recent, so 'trampling' it for better coverage isn't all that desirable.
One thing that seems to get overlooked is that data that is in a state where it could be imported into the main database is also in a state where it can be mixed with a snapshot of the main database. So whoever has that data can make use of it to increase their coverage without worrying about how other users would want the data integrated.
Not really related to the post but they are talking about making OSM cosumer-ready and every time I try to use OSM instead of Google Maps I realise it's just not on the same level yet.
It doesn't have to be on par in terms of error correcting input or location based search result because they just don't have the data on me like Google does (locations close to my work place, home,...) but even just the basic task of searching for my street on OSM [0] will give me a bunch of information the end-user doesn't really need to know like:
- tags
- created by
- version
- changeset #
- location ID
Is there a "cleaner" more consumer friendly web interface for the maps, maybe with a prettier mapstyle like the ones used in apps like Foursquare (afaik they are using OSM)?
You're really not meant to be using OSM.org directly as a drop-in replacement for Google Maps or Bing Maps. It's primarily a back end dataset, and that site is the front door to contributing to the data set.
This is something I think OSM really needs to communicate more clearly. You're not supposed to use the main OSM site for actual navigation or mapping purposes. It lets you contribute, and see all the contributions, but the OSM website is not for the end-user. Even though it appears to be.
When people say OSM is ready for the consumer, they're lying. What they really mean is OSM is ready for people to wrap into something for the consumer.
> Feel sorry for how proprietary maps are currently built. When there’s a new road built, they all have to scramble to add it.
I'm not into the business of maps, but don't the major players pay Nations/Cities or third parties that have the latest data about roads to be promptly updated? (whoever build the roads must have the data, and they'll be happy to sell it I suppose).
(ok, probably every city uses a different format or has the maps on paper...)
Then we can talk about if this data should be free or not.
http://regio-osm.de/listofstreets/index.html uses streetname lists from local governments to see if anything is missing, spelled wrong etc. There are usually always a few errors.
I agree with all that you said. But I fear our civilization will never become advanced enough (politically & socially) to bring that. We can share map data today though.
The marginal cost of all of those is dropping. If you can automate mobility (Self-driving EVs), the production of housing (3D printed homes), and farming (automated planting, maintenance, harvesting), you can indeed provide it all for free.
It depend on which country/state you are talking about.
In Sweden, there is a central government entity that do mapping, which then sells the map (created from tax money). Given the old thinking of that entity, I would just assume that updates are slow if they exist at all, and that the "official" map is compiled once a year rather than being updated live.
Well, skobbler did all this for iOS and Android. And TeleNav bought skobbler for $24M in January [1] -- so it's obvious why TeleNav might be able to do this now.
But will skobbler support stop? Why switch from skobbler to TeleNav's own products at all? With skobbler, you can get your whole continent's maps for roughly $7.50.
"Personal information does not include anonymous or aggregate information, which after processing may not be associated with a specific person or entity."
But I find this a dishonest definition of personal information.
The concept of the app depends on personal journeys being recorded and stored ("indefinitely"), which would very often and very easily identify a person via their home or workplace. There should be consideration for this in the privacy policies.
What's a good navigation app for Android that works outside the US? (Scout doesn't)
I have a couple of apps in my phone (Navigator and Waze), but the UI is really cumbersome compared to Google Maps. I use them because OSM maps are better in small tows in my area than Google Maps, plus you can keep offline maps.
OsmAnd? I don't really use it for navigation (why not, see below) but I know that it is a feature and I'm 100% sure of its offline maps capability.
Why not: It doesn't have bus/train/streetcar transit routing, something I don't expect it to have, though I'm feverishly adding all of the bus stops around me just in case.
I use it for a general idea of where to drive. But I don't follow the directions blindly. It seems to equate all "highways", meaning it gives equal weight to interstates and state highways, which results in routes which are direct, but slower than a slightly more indirect route making more use of interstates.
In Settings > Navigation, you can change the routing engine, incidentally, which might solve that problem. Although the other options require an internet connection for the initial calculation.
Thanks. I guess I've shied away from those because I'm often out of cell service when I need directions :) But I should do some test runs to see how they compare.
How is this such a big news? There is no navigation option in default openstreetmap.org experience. And there have been several other navigation/routing websites based on OSM data before. I sense nothing new here.
For those looking for a good Android app that uses OSM. I use Mapfactor Navigator and think it is great (I've used it in Italy, Austria, Slovenia, Crozia). It let's you use it with OSM data or Tom Tom, you can use Google for address search and use it offline if you need.
Like I said - consumer ready? I do not think that word means what you think it means. A consumer consumes. Assume this consumer isn't me and they want to visit my house and can't find a street that is 8 years old - they will be cursing their "consumer ready" map and use something else. They won't show up late to my house and ask me why I haven't added my neighborhood to some map they use that I haven't heard of.
If there is some milestone for the project I'm happy for ya, but "Go map your neighborhood" isn't "Consumer Ready".
Well, OK, it (still) isn't there yet... but it would be if everyone would only donate large amounts of manual labor to reproduce data which is probably available somewhere else, and just isn't allowed to be imported because of OSM policies against automation
Was some import from you "rejected" or why are you so aggressive? There are no rules against automation. There are just guidelines for not screwing up imports. You can import lots of data perfectly fine if the license allows it and you invest time into making it work flawlessly.
The situation in my area is nowhere near as good. I've got a couple Garmin GPSes as well as a GPS capable phone, and I would just love to map every road and path in a 100km radius from home. Unfortunately all the mapping software I've tried is so horribly slow I'd have to buy a new computer to be able to do much at all with the data I can collect.
> So for those of you who live in the US, congrats. The situation in my area is nowhere near as good. I've got a couple Garmin GPSes ...
This smells of FUD. First, we are talking about OSM. Not whatever tiny dataset is loaded onto a Garmin. Second, you did not name your location.
Any time someone says "oh, you can't use online maps in $country" I pick some little town in the middle of nowhere and OSM consistently has three times as many roads as the Big Name maps.
> Unfortunately all the mapping software I've tried is so horribly slow
You don't need mapping software. Just wander around, export to GPX and upload the GPX to OSM. Someone else can do the editing from your traces.
OSM is cool. Like I said, I would love to help with it, and I've got the tools I need to collect data. So what is my motive to lie about it? Why would I spread FUD?
Any time someone says "oh, you can't use online maps in $country" I pick some little town in the middle of nowhere and OSM consistently has three times as many roads as the Big Name maps
Okay, let me give you a couple examples I am familiar with. Plenty of stuff is missing or just plain wrong, right in town (now Google isn't perfect either but the coverage is much better). It only gets worse once you get out of town.
Most of the time the country in question was in central/south america or africa. I would never have expected Finland to be a counter-example. Thank you.
But pretty much everyone else in your thread here disagrees about the tools part.
What mapping software did you use? Just curious, because when I lived in Bangalore, India, I used to use one of the many android apps that record a gpx file, whenever went cycling. And, I used this mostly when I cycled in parts of the city that weren't mapped well on OSM. Then, all I had to do was upload the gpx files and if I found time I used to annotate street names, landmarks etc. And, nothing in this process seemed resource intensive to be slow.
You could try the JOSM editor[1], but iD is about as lightweight as it gets, so I am not sure if it would be much better. For what its worth, I use iD regularly without performance issues, but YMMV on older machines. Unfortunately, vector rendering and editing will tend to push browsers to their limit. This may be alleviated somewhat by the use of WebGL in coming years.
iD lightweight? It has trouble rendering a good block of houses when there is a lot of data whereas josm only gets into trouble rendering a small city or so of equally dense data.
Yeah, iD can be slow. If you're just adding roads, the Potlatch 2 editor (available in the dropdown menu next to the edit button) is less polished, but perfectly adequate, and much faster.
I just gave it a shot during lunch on Android and the app is missing the OSM data I added to my neighborhood, so I am assuming it's still using their old data and that they'll kick over to the OSM data as part of an update.
>Enter Telenav, where I work. We’ve spent approximately a zillion man-years to fix these issues and today we’re announcing navigation using OSM within Scout, our consumer navigation app. We’re starting in the US and on iOS with the rest to follow.
I believe that they might be referring to using OSM data. The Scout app for Android has been around for a while. But the copyright notice in the lower right says "year-year TomTom". So I'm thinking that, yes, there are apps for other platforms but they don't use (at least not completely) OSM data yet.
EDIT: yup, on Android it's missing a bunch of edits I did to OSM in my neighborhood. Some of those edits are probably a couple of years old, so it looks like they haven't switched over on all platforms yet.
Maybe I overlook? Are there turn-by-turn navigation api yet? Over the last year, every time I check it seems to be not available yet. I believe most or all streets in OSM only support up to street level rather than street NUMBER level. Well, please point me to some counter-examples if I am wrong.
Try out OSRM[1]. It is open source[2], so you can run it yourself, or you can use a public api for limited usage. The api has a JSON response with turn-by-turn[3], and it is remarkably fast.
Wow, speed is great. I'll definitely check it out. However, the second statement still seems to be valid: OSM is still at the street name level rather than street number level.
I read some MapQuest's document saying that Mapquest actually took points on a street and interpolate all the numbers. Maybe OSM can do the same. Eagerly waiting...
OSM does have addresses, but coverage is far from perfect. Nominatum (the geocoder used by OSRM) will interpolate addresses if it does not have the exact parcel info, which is surprisingly accurate in many cases. Work is also being done to create full address coverage as well[1], which you should check out if you are interested.
Routing is coming to osm.org real soon now, and as morganherlocker has pointed out, the software is already widely available (and used) elsewhere. But yes, house number data in OSM is partial at best.
I installed this app and it seems fairly gross. Is it really just for driving through places and not stopping? Because it seems to have close to zero points of interest. In my town it is incapable of finding any restaurant or other eatery which is not part of a huge national chain like McDonalds.
That's all well and good for people who enjoy doing that. But I'm not interested, and I suspect a lot of other people aren't interested.
And you can turn it around and say, "Well then it's your own fault OSM isn't better," but the fact of the matter is, I don't care, and I'll just use Google Maps, or Apple's maps, or some other map that has the info.
I like OSM, and I think it's neat, but trying to guilt people into helping out is going to turn away more people than it's going to bring in.
If Telenav is attracting them, getting them to pay money and using some of that money to improve OSM, the community probably shouldn't be too hostile to honest evaluations from those people.
Now all we need is for millions and millions more people to do this, and then we have a bunch of data which is just as likely to be inaccurate as if we had actually tried to automate any of the process rather than continually demanding unfeasibly huge amounts of free labor.
I suspect that points of interest would be fully up to date in places like New York City, but lacking in smaller towns and villages for obvious reasons.
OSM has better maps than Google in my 200k pop town in the north of Germany. Recently google even tried to have me walk through a closed company to get where I wanted to go.
Sadly searching sucks (unless you know the specific address) and the reviews for the international scout app make me very wary of using it.
Will the routing data be fed into OSM, or will telenav be building their own closed routing database that they use in conjunction with OSM? I didn't see an answer to that on the linked page.
I am glad to see OSM getting more popular because Google Maps is just getting worse by every day. The new WebUI is absolutely terrible, I had to permanently disable it because I could not even use my laptop while a GMap tab was open. They removed features I used and added many I don't care. On the top of these, the routing algo in GMaps is just laughable. It does not have the updated version of no left and no right turns, if you have 2 routes one with no highway one with highway it is going to pick the highway one even though it is 2 times the distance and so on. The traffic information is pretty much useless, only got better when they merged in Waze information but still bad.
i really hope this changes soon and the UI/apps catch up to the greatness that is OSM.
edit: i have an iphone and although i live in canada i still use the german appstore (CC requirement).
that's why i can't comment on the app the original post was about as it's only available in the US store. it was more of a general remark about my frustration with the state of apps using the osm dataset. I want to get rid of google maps and I feel the maps part of osm would be ready for that. I can tolerate if the commercial store data is not as up2date. But if the general usuability suffers, i rather opt for gmaps.