The real issue is geocoding and business listings. That's the primary user interface to getting your mapping endpoints for most people, and frankly, Google Maps is far and above most other services in this regard. Street View is indispensable for many tasks. I often use it to get a visual cue of the place I'm looking for, especially if it's a hole-in-the-wall restaurant, or to map out where I'm going to park if it's in a city.
That's awesome - I normally wear glasses and I wear contacts to ski and I'm now getting to an age where reading ski maps with my contacts in is getting to be a problem.
It has the huge ski areas in the Alps - I noticed the new "Vertical Experience" run at Meribel in the 3V which I'll need to add to the To Do list.
Yeah, there's a big U.S./European difference here. In many European cities, OSM has better business listings than Google. OSM is especially better about not continuing to list restaurants that closed 3 years ago, as Google likes to do. But in the U.S., Google is much better.
The UK is maybe the best case for OSM, unsurprisingly since it started there and has the most concentrated community there. There is barely any stone unmapped in the country, and stones that move get updated within days. France, Denmark, the Netherlands, Belgium, Germany, Austria, and Switzerland are also very well covered.
I usually plan my mountain-bike routes with a combination of Google Satellite and OSM. On the go, I only use the OSM, as it contains actual path data, but I haven't yet found an alternative to Google's Satellite View. There used to be tools which would scrape an area for you and use them offline, but they haven't been updated and I don't have the time do it myself.
If you want both OpenStreetMap and satellite imagery, check out MapQuest Open[0]. They provide OSM tiles and satellite imagery for anyone to use for free.[1]
I've done a lot of mapping of local mountain bike trails, followed up by making official PDFs of the trails for many local land managers. OSM is a great, great, great way to go with this, but you're right... Lacking the aerial imagery that Google provides makes things more complicated.
EDIT: That said, I have great luck with using something like OSMAnd+ to have offline sets of data that I'd contributed to. It works very nicely for electronic mapping of MTB trails.
If you live in an area where Google Maps is lacking, you can take the initiative to improve them — no coding required!
Most people think that Google Maps is based only on camera-equipped vehicles and satellite imagery. In fact, the service relies heavily on crowdsourcing.
People who live in remote regions can use Google Maps' community tools to describe how roads, trails, etc., have changed. For example, when the post office in my small town moved, I edited Google Maps to reflect that change. You can also change the position of roads, trails and other geographic elements.
Right. Now go back to 1994 and s/Google Maps/CDDB/g [1]
The lesson was clear: you contribute to proprietary databases at your great peril. People access Google Maps at the whim of Google. They could decide tomorrow to deny access, and you'd have no say in the matter.
I enjoy HN and have lots of respect for this community, but there's a strong contradiction in the way we discuss I.P.
On one hand, we're a community of people who found and contribute to software businesses. Most of these companies depend on selling intellectual property in the form of code and data.
Yet, as makers or hackers, some of us despise the companies that attempt to profit from selling I.P. We expect that all components of software be free for us to tinker with and repurpose.
"Well...wait a minute," one might say, "There doesn't have to be a dichotomy between totally closed and totally open."
Exactly! That's precisely what Google is trying to achieve with its mapping initiative. The firm is investing heavily in community tools (not to mention SDKs for developers) that give us an opportunity to participate: posting, querying and modifying data.
Being both entrepreneurs and makers, we should be a little more realistic about our relationship with IP. You can't always have your cake and eat it, too.
> We expect that all components of software be free for us to tinker with and repurpose.
OSM, along with its cousins-in-spirit projects Wikipedia and Libre Software are not about hacking on software. They are about building a society.
A society where no one has to depend on someone else. A society where everyone can learn and share with its co-citizens.
The number one foundation for building this society is that data be:
* accessible without needing explicit authorization,
* modifiable without needing explicit authorization,
* shareable with anyone without needing explicit authorization
in short, what the Free Software definition [0] does for software, but generalized to knowledge (for Wikipedia) or geographical position (for OSM).
By using and (worse !) contributing to Google Maps, you don't build this society. You trust Google to do it; but that's not the reason why Google exists, so you can't be sure about that. Plus, if you recall not-so-old history, you must remember how easy it is for Google to unplug applications even if they are used.
Google is giving you great, shiny toys to play with. But you can only build the present with them. I'd rather trust the future in a Foundation than in a company whose goal is to make money.
If you had to wrap projects together like Wikipedia, OSM, etc (open data sets that you could argue are the foundation of society), what would you call the collection?
Yes, of course, that goal is utopian (well, I hope we don't follow the definition too closely and actually reach the dream). But the good thing is that they don't depend on time, money or anything else. It's a belief, and the people's mind will change over time.
You could draw a parallel with ecology. 50 years ago, no one would care about the environment; resources were basically free, pollution was seen as "minor". A world were everyone cared about the planet was seen as utopian. After half a decade of fights, the minds have now switched to a different mind (at least in the western world): pretty much every single piece of electronic hardware you can buy has some notion of "we-are-green" to it. While it's not 100% eco-friendly (just take a look at how electronics are "recycled"), people's mind certainly have changed. The guys behind the Libre software movement certainly don't expect anything big happening in the next 10 years (remember that RMS has been fighting for the last 30 years), but things will change, you can be sure of it. Just take a look at the recent plans to to "take control back" on the internet infrastructures initiated by Europe and Brazil. Sure, technically the things will more or less remain the same; but the mindset is changing, and it will keep spreading.
> How about a less grandiose goal like "making software better and more accessible"
This is a really noble goal in itself, and Google can be praised as one of the biggest enablers in this field, seeing how much they have changed the domain. While I may not sound like it, I actually thank them for all they did to the tech world, and the end-users in general (with an extraordinary amount of stuff available as OSS). Actually, that goal is more or less what OSS tries to achieve, except that the target is not the end user but the developer: the goal is to make sure that developers don't reinvent the wheel again and again and again.
But that goal is orthogonal to what the Libre movement (FSF, OSM, Wikipedia) fights for. That goal can very well be accomplished with private companies: in the majority of cases, they help society by building better and more accessible "things", because that's a sure way to make money (with the notable exception of US telcos), and money is what they're here for. But that doesn't mean you are independent; it's quite the contrary in fact, because their tool is so easy and powerful to use that you just use them. This is what many people call a walled garden: you can live in a complete ecosystem, where everything works perfectly together... but once you start wandering around, you quickly realize that there are boundaries to what you can do [0]. The tech world is even worse than this description, because everything can change in 10 years; you can't expect all your belongings to remain accessible at the end of your lifetime.
The FSF doesn't want to build better software. There even was a post recently mocking their logos. They want to build software so that you don't need to depend on anyone else. If you don't depend on anyone else, you can do whatever you want.
>some of us despise the companies that attempt to profit from selling I.P.
If I despised Google, or any company that profits from selling IP, I would have said so.
If you can either contribute to an open dataset or a closed one, you contribute to the open one because the benefit no longer relies on the goodwill of any dataset owner. This principle underlies not only Open Street Maps, but OSS in general, and with a little finagling, things like Bitcoin. If the last few years have taught us anything, it's that we cannot rely on the basic decency of any institution.
I'm not happy about it. I wish I could trust institutions to be decent. But time and again they act indecently, and weather the (short) media storm, and keep going, business as usual, and we must protect ourselves from that. To do otherwise is to ignore the lessons of the past.
There isn't really anything contradictory about being cautious of doing free work for a huge for profit corporation. If Google shared the data you would be getting different responses, but they only use the data to improve their products.
(I can totally see submitting a change to get rid of an annoyance; I don't get warm fuzzies from 'helping Google make maps a better product'.)
Yet, as makers or hackers, some of us despise the companies that attempt to profit from selling I.P. We expect that all components of software be free for us to tinker with and repurpose.
Why should we trust Google with this data? Why should we supply google with this data when open alternatives exist? Will their incentives in using this data always be aligned with ours? You don't have to despise Google or want all things to be totally open to ask these questions. You don't have to be hostile to IP or companies selling IP to decide it is more in your interest to contribute to an open project.
I'd rather contribute data to an open project and watch a thousand mapping projects bloom (including Google's) based on that data, than help a corporation like Google or Apple take over every corner of our digital lives. OSM has led to a lot of great mapping services based on their data, and no-one resents those.
It'd be wonderful to see a similar open web search project based on open data too.
Being both entrepreneurs and makers, we should be a little more realistic about our relationship with IP. You can't always have your cake and eat it, too.
I don't think people hate google for offering proprietary mapping, but nobody wants to be subjected to unnecessary leverage.
For example, I wanted to use google in a commercial application, but only for backend services like geocoding and place search. I wanted to pay them for these services, but they refused the business because I did not want to display the data on a google map and use the google js api.
That is the sort of thing a company with lots of leverage can do. I wanted to pay full price or more to use a feature (10s of thousands per year), but due to a monopoly on the data, they can turn me away. Sorry, but I already have to deal with one Comcast.
> Being both entrepreneurs and makers, we should be a little more realistic about our relationship with IP. You can't always have your cake and eat it, too.
Why not? We can always demand more, and what is wrong with that? And if we set out to make and eat that cake, chances are we can do it. It is not unrealistic at all. So far crowdsourcing has worked really well for many projects, and I can only see it getting bigger as we as a community realize the potential of open data and our ability to collect & manage it in numbers that few companies can afford to challenge.
The point made by GP is a perfectly valid one, and many of us have been screwed in some way or another by corporations that hold the keys. I do not see why playing along with such schemes has anything to do with being realistic.
The problem is that corporations are not just satisfied with making a profit but tend to progress towards squeezing the last penny out of users at all costs. For example, Google is rolling out tracking which stores Android phones visit, and if you turn location services off, a lot of things including Google Now stop working[1]. Why would I want to work for free on improving maps so they can track where people go for advertising purposes? It is the difference between contributing to Wikipedia versus Encyclopaedia Brittanica.Not to mention doing user hostile things like making ads look like search results[3]
>That's precisely what Google is trying to achieve with its mapping initiative. The firm is investing heavily in community tools (not to mention SDKs for developers) that give us an opportunity to participate: posting, querying and modifying data.
So is OpenStreetMap, Navteq(commercial again) [2] and other mapping companies. Atleast Navteq/Nokia allows you to download complete offlline maps for their Here maps on supported devices unlike Google which wants to force people to be online for tracking/advertising purposes.
Precisely. I remember when the track names for some of my CDs vanished because CDDB had taken the data I created for for them private and I was no longer able to see the information I typed in.
I tried to edit Google Maps once and had someone ask me for supporting documentation to prove that the road across the street from where I live really was closed. Total nonsense. This heavy-handed moderation was what pushed me to Openstreetmap in the first place.
Don't you think that kind of caution is warranted for a popular service like Google Maps? If/When OSM gets really popular, I am sure there will be efforts to spam/vandalize the data. At that point, "heavy-handed moderation" is essential to maintain quality. (Another example - wikipedia articles are sometimes heavily moderated).
No I don't think it's necessary. I think changing 200 ft of road across from where I live to "closed" while construction happens for the next year is a minor enough edit that it should be accepted no questions asked. I was tired of having navigation send me down a closed road.
You can imagine how this can easily be gamed to change how Google routes people right?
If I was a shady business, you can imagine how I can mark all the roads in front of competitor as closed. Or I can mark the streets in such a way that traffic gets routed to go in front of my business.
So what would prevent vandalism? There is an enormous amount of effort in Wikipedia to repair the damage that vandals constantly do to it, and I'm sure that OSM/Google have run into the same issue.
There's an oke who I caught trying to change random listings in a small township in South Africa into something related to him. One small edit that was allowed was for him to map an existing road, but he named the street after himself. Map Maker has its flaws, many of them (like bots that go around changing information for whomsoever knows reason, the heavy-handed moderation in itself) but the moderation is sometimes necessary. We (or whoever is contributing and editing) also have to use common sense when making and approving edits. For me "documentation" could have meant going outside and taking a geotagged photo with a time stamp, if the moderator still wants more then to hell be with it.
Most of the edits that I have made have been for areas where I'm interested in walking paths, because I use the Maps API quite a lot, but going out of my way to map a city or something, I'd rather do that on OSM where it's free for all.
Just as a sidenote - this is not a remote area, this is Vienna, the second biggest city in the German-speaking world. That being said it is strange that Google doesn't have more data on it
Austria had a ban on Google street view cars due to privacy. That's why you see no street view anywhere in Austria. I think they removed the ban, but since then, Google hasn't exactly been in a hurry to send any cars that way.
It seems a little strange because I'd assume mapping Vienna, one of the world's more well known cities, would be of high importance. I wonder if they're now avoiding Austria on purpose, simply to send a message.
Edit: It looks like the ban was lifted in 2011. Three years later, and there's still no street view in Austria. I'm scratching my head on this one.
When governments pass asinine laws like "Nobody can take photos in public except us," there are consequences. Maybe the citizens of Austria will think about why they don't have Street View coverage during the next election.
There was no law passed in this setting, but a temporary ban specifically targeted at the Street view cars <http://www.google.com/hostednews/afp/article/ALeqM5gqQOUMhV0... in the wake of the discovery that they were collecting private wireless data, in order to investigate.
I do not live in Austria, but, in my opinion, this does not seem an entirely unreasonable reaction to that discovery. (Incidentally, I think I would be in favor of laws controlling the massive untargeted acquisition of information in public space, for privacy reasons... and I would not mind at all if this meant there was no Street view coverage in my country; quite the contrary.)
No, the principle is that you can have overreaching privacy protections or convenient geolocation features, but not both.
(That said, a3_nm's post indicates that this case had more to do with Google's collection of private data over WiFi than the photographs themselves. It's hypocritical for governments in the age of Snowden to criticize a private company for doing that, but at least it's understandable.)
The problem I have with editing Google Maps is that the tools are so much harder to use than OSM's. I really, really enjoy using JOSM and can really improve things in a short time.
Google's web-based tool is much slower, and the dependency-chain approval process can make what should be a quick edit into a multi-day process. [1]
One example is a local park's hiking trails which are closed to cycling traffic. I'm do a lot of local biking and regularly use Google Maps to plot routes, so when I saw that the hiking trails had been somehow marked as cycling routes (which Google was directing people along!) I went to change them. Two hours later and I still wasn't done with simply unflagging two square miles of hiking trails.
Had this been in a JOSM-like interface it'd be 15 minutes at most, including committing the change back.
I understand that Google doesn't want to make changing things too easy, lest there be all manner of vandalism and personal-motive edits, but it's currently hard enough that those of us serious about making substantive changes have the tools getting in our way.
Hmmm. Maybe it's better now, but there was once a local pub on the wrong street corner that I corrected. It took almost a year for the change to show on google maps. I don't know what the validation process is but I remember looking and feeling like I was totally ineffective.
Google created something great with google maps. But I'll always feel like it's there to be consumed, not to be part of.
Do you have any ideas why? Did google simply deem it not profitable enough to send their camera-vans round bratislava, or are there particular reasons that make it more expensive there (regulations of some kind, a fuck-tonne of bureaucracy to get permission?)
Ever notice why you don't see Street View in China, or some Euro-areas? Some of it is probably regulation. Apple was recently denied overflight rights on some areas in Germany IIRC to do 3D mapping.
I actually disagree. Google maps has a link to Google Earth in the bottom left corner, which immediately gives a satellite view of the place without having me navigate elsewhere.
I tried various layers on OSM (transport map layer is cool, btw) but no layer was as informative to me as what Google Earth showed. YMMV.
It's funny, Google doesn't seem to be using OSM data directly, but I think they might be looking periodically at OSM activity to know where to focus their map generation.
Google never had any real super high quality detail in my tiny home town. When I first got into OSM, I did what lots of people do: I filled in a bunch of details I know about where I grew up. Kind of fun. Well a month or so later, I checked out Google maps and, lo and behold, their data was much better for that little town. It was NOT the OSM edits I made--they didn't steal anything. But I found it odd that they'd all of a sudden take interest in a little town of less than 1500 people in the middle of nowhere. I don't know. It could have been a coincidence--It's possible they had a big world-wide map update during that month that updated everything.
Oh, and it RAISES the question, not "begs" the question.
Quoting Wikipedia:
In modern vernacular usage, "to beg the question" more frequently is used to mean "to raise the question" (as in "This begs the question of whether …") or "to dodge the question".
They can't, unless they're willing to free their own datasets as well:
"You are free to copy, distribute, transmit and adapt our data, as long as you credit OpenStreetMap and its contributors. If you alter or build upon our data, you may distribute the result only under the same licence. The full legal code explains your rights and responsibilities."
Does that "build-upon" clause also apply for the case where a provider is mixing OSM data with other third-party licensed map data ? For example, it is mentioned/known that Apple makes some use of OSM data[1] to improve their maps in some places. Are they then legally obliged to release the full data behind their maps for those regions?
If so, that's very nice but I am not sure that's what is actually happening: the third-party source whom they are licensing data from is unlikely to be happy with a scenario where their data ends up in the public domain as this considerably reduces the commercial value of their map data.
What I see people saying is that Apple is using OSM data from earlier in the project, when the license was somewhat different.
But Apple simply cannot force someone else's data into the public domain, if they were mixing the data without permission, they would be violating the licenses and legal action could force them to fix that (but it wouldn't directly change the licensing of any of the data)
OpenStreetMap's ODbL license has a sharealike clause that would legally require Google to share the dataset they combine OSM with: http://www.openstreetmap.org/copyright
No-one said they should take the data and not contribute back anything. Except OSM in their license of course, http://www.openstreetmap.org/copyright (well it's attribution anyway).
Either a bit a fraction of a percent of a days profit or some data for OSM, sure either would be most acceptable.
One thing us hackers should be especially aware of is that it is a violation of the free Google Maps API terms of service to embed Google Maps in anything that makes money.
Read that again. If you make a classifieds website but charge businesses to list, it's a violation of the free Maps API TOS and you need to purchase an Enterprise license. Enterprise licenses start at 5 figures/annum for low-volume usage.
This is not some abstract ideological nitpick. Google has ramped up its attempts to monetize Maps ever since they introduced quotas a couple years back, and they are actively pursuing businesses to force them to buy an Enterprise license. So there is a very real risk that you could be targeted if you embed Google Maps, especially if your website gets big. This is no doubt one reason why companies like Foursquare, Craigslist, etc. use OpenStreetMap rather than Google Maps.
I work on the Google Maps API team. I am not a lawyer, but my reading of the terms does not equate with yours.
Here's the excerpt:
(a) Free Access (No Fees). Your Maps API Implementation must
be generally accessible to users without charge and must
not require a fee-based subscription or other fee-based
restricted access. This rule applies to Your Content
and any other content in your Maps API Implementation,
whether Your Content or the other content is in
existence now or is added later.
My reading is that in your example, charging businesses to list is okay, but charging users to access the site is not.
You are not a lawyer, sure, but you are also, apparently, not Google's "compliance team". Section 10.1.2 says you can't charge anyone a fee for the "implementation". What is the "implementation" and what isn't? You might have one answer. Someone else in the company might have another. That's typical of big companies, where not everyone may enforce policies different, but still pretty scary for anyone who falls in this gray area.
That's not how I read it. It looks like you can make money just not by charging for maps API. i.e. your maps implementation should be in your free SKU.
"10.1.2 Restrictions against Commercial Use.
(a) No Fees. You must not charge users or any other third party any fee for the use of the Maps API Implementation, the Service, or the Content, except as permitted under Section 9.1.2"
Had to make an account and comment - my friend went through this whole thing and the terms of service get pretty fuzzy pretty quickly.
For his listings website that charged business owners fees for "premium" listings, Google[0] considered the embedded Maps to be part of what he was selling. They argued that the service as a whole (incl. maps) was "the product".
We disagree, apparently, but our interpretation doesn't mean diddly squat until push comes to shove.
[0] Edit: by "Google", I mean whatever sales drone is trying to get you to pay. I assume their salespeople are on commission like any other company.
> I assume their salespeople are on commission like any other company.
With one important difference: they're the most arrogant, uninterested and rude sales people i've ever had to deal with. "Because, what are you going to do, use Bing?"
Huh? This is definitely not true. As long as the map itself is not behind a paywall, you're fine, and charging for being listed in your system (not for visiting it) would pretty much easily pass the terms of service requirement listed above. Lots of people make money on sites with google maps embedded in them. The only thing you have to worry about is being more popular than 25k loads per day.
One thing where OSM really shines is the ability to downloaded maps for offline use. When i was travelling in South East Asia for two months i had an iPhone with an app that allowed me to download a part of any OSM map. Because every hotel had Wi-Fi access i just downloaded the parts where i was travelling and used that instead of finding a local SIM card. Even in pretty remote parts the maps were usually good, and in some cases they were even better than the Google Maps equivalent (e.g. Laos).
FYI on android's Google Maps I've been able to do that for at least a year and a half, and it works great for me (I used it extensively in SEA too, coincidentally). Not sure if it's available on iOS's version.
It's incomparable. Google Maps only caches around 10km x 10km, and in offline mode search and routing doesn't work. Also, in Google Maps it's impossible to tell if/what is cached and what not.
OSM maps can be for any area, and applications do provide search, including POI search and routing.
Does OSM have bike paths? Does it have bus routes, train schedules? Can it tell you the route to take to get from point A to point B on public transit? Does it know about current closures? Does it search based on info other than address?
I ask because I've been using Google Maps to travel and those features have been invaluable. An offline map is usually missing those features. The schedules are not up to date and searching by context requires data not usually part of map data.
Actually, where I live it's OSM which has excellent bike paths and Google Maps which hasn't (posted an example above).
And yes, of course it has bus routes and can route using car mode, walking, bike, public transit. Using applications like Osmand you can do much more complex queries than Google Maps, optimize by time, by length, prefer means of transportation X, etc.
Of course every offline app potentially has the stale data problem, but for me, it's actually OSM with the better data.
It has the routes, but no it does not have train schedules. Why? Because OpenStreetMap is just a basic map. It is not supposed to have such information.
Someone has to write an app that uses the OSM data and then add the schedules.
Btw. I guess you are from the US? Because good luck finding those schedules in Google Maps outside the US. Germany for example only has trains. No Buse and Tram data.
For what it's worth, Google has extremely good public transport data for Perth, Australia, and it worked quite well for bus & train times when I was in Sweden last year (especially Malmö). So it's not just the US. Google's support probably depends in part on the transport authority providing a timely & accurate data feed of their schedules that Google can use.
Unfortunately, for off road, even though the classification of whether bikes are allowed on a trail can be input, I haven't seen any viewers that show this information. I find it difficult to determine the trail type (single/double track), condition (dirt, pavement), and whether bikes are allowed.
Agreed. I live in NYC, and every corner has been completely mapped, and it's a level playing field for almost every mapping tool that's out there. Google isn't competing on completeness of maps, but rather on transportation data. Subway schedules, route changes/delays, traffic, and construction data all matter a lot more to me than actual geolocation. And in NYC, streetview is enormously helpful.
In many parts of the world, basic mapping data is incomplete, and one tool will have fresher data than the other for cherry picked examles. In another 10 years, we can imagine that mapping will all be relatively similar, and integration is what we'll compare against.
Although, that has its own drawbacks! I got used to google maps telling me about service changes on the subway (it's pretty good about warning about planned service changes), and I found myself getting angry at maps when I found out the PATH had a planned service change that maps didn't warn me about.
Unreasonable, yes, I know - but something worth sharing I thought, since if you get into the realm of "transportation data" it's kinda gotta be all-or-nothing, I think.
There is support for capturing bus routes and a reasonable scheme for bike paths (and bike access, lanes and such). But the data is not necessarily complete. In some areas it will probably beat Google, in other areas not.
I like to keep a complete world map on my phone, since the only time I'm ever going to need it, is very likely to be the most important time (i.e. possibly life or death).
It's over 400GB uncompressed, so currently this person would have to settle for a subset on a 128GB MicroSD like http://www.amazon.com/dp/B00IIJ6W4S/. I think there are larger SD cards. And certainly there are small 1TB hard disks. I'm not sure what sort of device would be best for carrying around the world viewing these things, but probably what this person wants is possible, if expensive and inconvenient.
The uncompressed xml is not the right point of reference. The planet extracts in the pbf binary format are about 20 gigabytes. Looking at the OsmAnd map files, data for most of the world, in device ready format, would be several dozen gigabytes (even high density areas like California only end up being a few hundred megabytes).
Good point. My estimate was a naive upper bound, ignoring things like efficient storage and assuming all data is included, and was focusing on how much storage is available on portable devices (that sometimes stretch the bounds of portable).
Then the question is what does one do with all the extra space in this portable world device? Wikipedia information? Languages? Fun times.
Tablets with 512GB of flash in the next 2-3 years is definitely possible; I could see a small niche market for a "world atlas" Android/iOS app that stored OSM data and indexes compressed, and read portions the compressed files into memory for the relevant map data.
The OsmAnd people translate most of the world and I just checked, all their maps would fit into ~25 gigabytes. That's device ready, indexed data with quite a lot of overlap (they provide smaller extracts to help people save cell data and larger extracts for people that want them).
That even includes a bunch of POI data from wikipedia.
On the other hand, as a for instance, that total includes more data for Germany than for Africa.
Wikipedia (text articles alone) is about 10GB compressed. I'm in awe that I can carry around most of the world's mapping data and a lot of its knowledge on an SD card or my phone.
"including POI search" (if the data is present in OSM)
I guess some map providers might be pulling in other data sources, but in a lot of areas, OSM data doesn't have many POIs to find.
(the minimal effort way to improve this is to drop an anonymous note on the map, using the not very prominent button at the bottom of the right panel: https://www.openstreetmap.org/ )
That's why I love open source or hackable products. You don't need to whining to a shitty support forum to grind a kpi driven product manager to make empty promises.
I find this applies even semi-locally. I take a lot of trips out east into the Anza Borrego desert near San Diego and my phone coverage out there is spotty at best, and often non-existent. Google Maps' offline caching is a pretty terrible bolted on feature that they obviously don't care much about; Having OSM maps (I use the OsmAnd client) available out there is insanely convenient.
Due to the offline deficiencies, the terrible user experience regression from Maps 6.x->7.x, etc, I find myself using Google Maps less and less to the point where I'm not even sure my next phone will be Android based (until recently Google Maps/Nav was the killer app for Android for me, now I barely enjoy using that app at all).
This is good enough to use as a paper map (a zoomable paper map, with GPS and starred locations). HOWEVER, there are some huge shortcomings: No offline search, no offline navigation, and in certain regions, it's not available. I'm not sure exactly what countries are off-limits, but since I'm also backpacking through SE asia, I can tell you now that all of Singapore and Vietnam are downloadable, but at least some parts of Thailand and Indonesia are blocked. While this feature is incredibly useful, it's limitations are making the switch to OSM tempting.
Calling that a ‘feature’ seems somewhat strange to me – the native maps applications on both my E63 and N9 allow me to download whole countries (or parts of them) onto my phone.
It's hard for google to know where you've been (or where you are) if you use their maps only in offline mode. Them being an ad company, it is unlikely this feature will ever be fully available.
I find myself increasingly leaning towards OpenStreetMap due to the utterly infuriating changes in the Google Maps interfaces, and that's even knowing the GM privacy issues.
OSM has definite holes in the data, but that's fixable, yet also suffers from a too much information by default situation for their standard representations. The map data is there, but the resulting maps just aren't quite as good, though I'm sure someone will use it to get there.
I know I hate it too. Use https://www.google.com/maps?output=classic. Google's become so extremely disrespectful, they don't even ask you to opt in. Seems like every time they hire a new designer they re-design either YouTube or Google Maps when they didn't need a single thing changed.
- I have the new Google Maps UI forced on me. The slow loading one that's un-intuitive.
- I'm no longer allowed to see a "satellite view" of my city. I'm forced to use "3D view" or normal map view.
- They removed the underlined text and increased the font in Google Search for me. Making the text harder to read.
- Forcing me to use Google+ to comment on YouTube videos.
- Redesigning Analytics so it looks like garbage and I can't open multiple sites in new tabs.
There's no better time to compete with Google. At this rate they'll piss all their users away.
Yes. Google Maps used to make it easy to cache an area and know which areas you'd cached. It was also integrated with My Maps, which I found useful. One of their updates a while back took both of these away without warning.
For a fully featured Android map app, with many options for map sources, try Locus.
OSM is the data. That map is meant for mappers. There are countless other maps based on the data. One of them might be more for you. Try mapquest's or the humanitarian style.
I hope this movement toward community owned not-for-profit internet utilities is a trend, not just mapping but also for search and social networking. These functions are so key to the user and give so much power to whoever operates them they belong in the hands of the community. There is too much conflict of interest when in the hands of Google, Apple, Facebook or the like.
I'm using OSM data in a fun little app I'm building (mapping out the tunnels in downtown Houston, TX) and I've been very impressed by the tooling. Everything from the OSM site and editors themselves, to related tools like MapBox, Leaflet.js, TileMill, etc.
There are some very beautiful and functional geo-related products out there, and OSM works amazingly well in all of them.
Glad to know there's interest in it! I'm making it primarily because _I_ need it--not easy to find what you need down there. I was always surprised there wasn't one yet, so I decided to make it as a fun side project.
Both iOS and Android. I'd love beta testers, you can get a hold of me through my bio.
It sets up and imports OSM data into a postGIS-enabled PostgreSQL database. From there you can use TileMill to create whatever type of map you fancy or just access the raw data.
Can anybody recommend a good OSM-based maps app for iOS? Currently I'm using Apple maps as it offers a great user experience. It's extremely fast, fluid and has turn-by-turn directions. Unfortunately the data is sub-par for my area.
I've tried Google Maps' latest version and it feels laggy; When I zoom in it seems to jump to an image at the end of the animation. Nobody is getting vector maps right except for Apple. I wish it could just use OSM for the data.
I downloaded some maps but panning / zooming results in the map flashing while it's displaying the next tile. As I said, I use Apple maps because it has a great user experience. This has the same issue Google maps has.
Google Maps is a great product. OpenStreetMap is a great dataset.
Google clearly know how to make their data support features they want in apps. They have been known to use computer vision techniques to read street signs in order to work out the logic of junctions, just so that they can do better route finding.
OpenStreetMap on the other hand focuses on just having a huge amount of data. Their online editing is much more discoverable, and as a result their data is often better. There's someone at my university who has an interest in it, and consequently, the main campus is modelled perfectly down to the level of individual trees being correctly placed.
Just recently evaluated Google Maps, MapQuest, OpenStreetMaps, MapBox for a client project. I tried to locate a home address among a tract of addresses in a busy part of Northern Virginia. Only Google Maps and MapQuest could map the address. As long as there isn't some sort of comprehensive address data for valid US addresses, I would hesitate to use it or recommend it for any business applications.
I understand OpenStreetMap is crowdsourced but there are commercial vendors such as MapBox that rely on that data. I couldn't use those resellers either unless they had other sources of address data.
Hi, I'm working on an application that need to do Geocoding, and it's a problem that's definitely not solved.
We've tried countless solutions but hadn't found any satisfying one until now. We cannot (and will not) use Google's geocoder becuse we wont put the information on a google map (part of their TOS).
We recently tried decarta and seems to be working really really good! http://decarta.com/ Unfortunately some house numbers are missing, but still, the quality is pretty high!
(I dont have any affiliation with them, we're just a customer now.)
Also, I would like to bash yahoo placefinder! It's realatively decent in finding addresses, but, but, but! The pricing is very high, the documentation ludicrous, customer service/forums nonexistent and you'll get terrible uptime, from now once a day or two, you'll just get mysterious 503 for which the official answer seems to be wait for a while, and hopefully it will fix itself, Just effing great!
We are leaning towards MapQuest because their pricing was much better than Google Maps for the kind of volume that our client was looking at.
We had to go with the business license approach because of the TOS involved.
BTW, MapQuest does have a "Community License" which unfortunately for us was using data from OpenStreetMaps and not the data behind its commercial product.
This seems to me to be something of a hipster argument. I use Google Maps (1) because it is usually better than the competition, (2) it usually integrates well with everything I have, and (3) Doesn't show me a bunch of useless information (such as walking paths).
Basically, Google Maps works better than OpenStreetMap from what I've seen (I tried using OSM, but it was hard to tell what things were being clustered and what not).
I'm surprised at how quickly people advocating for OSM make the comparisons to Google maps. The user facing stuff at openstreetmap.org isn't comparable to maps.google.com or the various Google apps.
But OpenStreetMap isn't just the tiles you see at osm.org, it is a data product that you can adapt to whatever you need, for the simple price of attribution. And while the data is not complete, the global coverage is increasingly reasonable.
I think the reason Google maps works so great now is why a lot of people support OpenStreetMap. They simply don't want a private company to have some kind of almost monopoly on good online maps, which then can dictate the rules how it is used.
The search function is not impressive. I put in the address of a hotel that I know exists (I have a reservation): "Finnegårdsgaten 2A, 2003 Bergen, Norway". It says "no results found". In the map window I have the OSM map of Bergen displayed with Finnegårdsgaten in the center (not as the result of the search) so I know the street is in the data.
Open GMaps, enter the same address, it instantly displays it plus on the left, a Street View of the building front, and links for three crucially useful things that OSM doesn't appear to have: Directions, Search nearby, and Save to [my] maps. I use all three of these features heavily when planning a trip.
Bottom line: they may have the map data (the Bergen data looks pretty much the same, modulo the different graphic styles) but they lack a lot in the features built on it.
I hope they are successful, especially with the most recent update to Google Maps which is pretty poor in my opinion.
That said, its extremely difficult to do really good maps over a wide area. Having seen, from the inside, some of the stuff Google went through to get the 'truth' about what is on the ground, and continues to do, its quite an intense undertaking.
This is badly needed. Mapping is one of those areas where I still have to rely on Google, even though I really try to exit their services. The internet needs more of its core services being based on open services and open data.
Last time I tried OSM, I had a very bad experience, but looking at Locus (as recommended elsewhere in this thread) things are definitely looking better.
Consider contributing yourself. If everybody would just add his house with his address and make sure the street is there and maybe some houses around you that would already help a lot.
Google Maps' killer feature for me is getting directions for public transport. Finding out how to get from point A to point B using buses and subways is invaluable. It doesn't seem like any Android OSM client has that yet... too bad. I would much rather use OSM.
I still wonder how do you really make sure there is no spam and data vandalizing in OSM. I still wonder about it, since there can be some entities that would just erase competition for a price, but I guess that's the same way wikipedia is dealing with it.
If Google used the OSM data, they'd risk being forced to share their mapping data - OSM is distributed under a share and share-a-like style license. There are probably ways around it, but Google regards their map data as being core to their business & they've spent $billions on it. There's no way they're going to risk that.
Apple is using OSM data from before the change to the new licence if theirs is from 2010. Google could do the same, but what would be the point?
it is not missing from osm. osm is about vector data. you probably mean it is missing on openstreetmap.org, the project's homepage. I disagree! The website does not do a great job communicating osm's purpose.
Any company could setup their own osm-based map and include a layer with aerial images (which they would need to buy).
Not buying satellite images was the point of my question.
To follow the OSM crowdsourcing model, we'd need some way to gather ~street level pictures and reconstruct[1] vertical views, limited of course, no roofs etc, but better than nothing.
You wouldn't have enough images to generate anywhere near the data that Google, Bing, Yahoo has. This may change if amateur automated drone imaging takes off.
Now what could be done is crowdfunding an imaging run by a satellite imaging company, and licensing the data under an open license.
For isolated or risky areas I agree, nothing can replace high altitude views. But for lively places ... are you sure ? millions of cellphones, a few snapshots with geoloc and orientation, that amount to a lot of data.
I wish for drone imaging too, but automated cars aren't legal, and I think drones won't be for even longer (unless Amazon unlesh some nice lobby-fu).
About crowdfunding satellite licensing.. I just remember a lot of people are sending high altitude balloons, maybe that would be a cheap middle ground.
> About crowdfunding satellite licensing.. I just remember a lot of people are sending high altitude balloons, maybe that would be a cheap middle ground.
Want to work on an autonomous dirigible project? I'm up for it. We could build a home made Argus for image collection: http://en.wikipedia.org/wiki/ARGUS-IS
"The three principal components of the ARGUS-IS are a 1.8 Gigapixels video system plus two processing subsystems, one in the air and the other located on the ground.[17]
The sensor uses four lenses and 368 cell phone cameras, 5 megapixels each.[18]"
What are you missing? I always just used the one in the browser and that did most things just fine. Just recently used JOSM which of course has some advantages, but the basics seems to work everywhere.
I might just be more patient than you, but I spend more time waiting for imagery to come in from servers than I spend waiting for Josm (on a 5+ year old laptop with 2 GB or ram...).
Yeah, I was just trying to establish something like 'modest hardware works well enough for me'. I think Josm wouldn't work very well with any less ram (swap is an occasional pain point).
Do you think the style of interface presented by iD is more in the right direction? I don't mean the in-browser stuff, I mean the more simplified editing model and such.
I really can't say. UIs are something I might be able to judge after actually using them for a while. But as long as these editors freeze for seconds whenever I try to do anything, and crash with OOM after a dozen or two minutes, I just cannot get far enough to say anything about the UI and workflow.
On Windows with up to date Java it wants ~800 megabytes of virtual memory, but I never get out of memory crashes, just the occasional pause for swap (the pattern is very clearly swap).
(I'm not really trying to express an opinion on whether this is hoggish or not, just putting the numbers down)
OSM: http://www.openstreetmap.org/#map=15/48.2509/16.2533
Google Maps: https://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q...