28 ms·

Why you should not pirate Google’s geo APIs

309 points by spenrose 13y ago

TallGuyShort 13y ago
Here's another reason that I wish was more prevalently discussed: it's not okay to violate Terms of Service anyway. Don't accept the terms? Don't use the service. We'd all have a field day if "big brother" violated the terms of an open-source license, but the freedom we enjoy in the community comes with a responsibility to honor our agreements.
- Killswitch 13y ago
  I just wish people would stop being so self-entitled period.
  - TallGuyShort 13y ago
    Agreed. The tone of the article is correct, in my opinion, because this really does hurt the community. I think we'd have an easier time fighting things like SOPA, DRM, etc. if actual piracy wasn't such a big problem.
    - TillE 13y ago
      And if theft didn't exist, we wouldn't need police. This is a useless observation. You may be able to effect a limited degree of societal change, but you will never satisfactorily eliminate the problem at its source. Other solutions are necessary.
    - HNLogInShit 13y ago
      This isn't "piracy." It's easily shut down if the provider wants to.
  - baddox 13y ago
    I feel like I deserve to live in a world with fewer self-entitled people.
  - ianstallings 13y ago
    As my developer friend likes to say - this is why we can't have nice things.
- revelation 13y ago
  Violating the terms of a license is very different to violating the Terms of service you "sign" with a button click.
  - TallGuyShort 13y ago
    Actually I'd argue that you're more obligated to abide by something you explicitly checked "I have read and accepted..." than something that was listed somewhere in a file you downloaded with some other stuff. You probably just happen to like the terms of the GPL or ASL more than you like the terms of Google's API. I do too - but at what point are we just losing all integrity because we're too lazy to read a long document we have no intention of actually agreeing with?
    - _delirium 13y ago
      If there is an actual click-through and the license is understandable by a human, perhaps. Usually there is neither, though. I rarely even know how to look up terms of an API, or that there is an API. If I know the name of the API, I can look for "Google Foo API", but that name isn't always obvious. If I just see a URL with an obvious schema, and I change the HTTP GET parameters to retrieve different stuff, am I allowed to use that in, say, a small-scale art project, or not? The usual internet default is that public URLs are implicitly giving blanket permission to retrieve them, except that retrieval by spiders should abide by robots.txt. Google itself makes heavy use of this ok-to-retrieve default assumption. If I go to http://maps.google.com http://maps.google.com there is no license presented to me as a condition of using the service. I can however look for fine print and seek out the link "Terms of Service" of my own accord, at which point I get this: http://www.google.com/intl/en_ALL/help/terms_maps.html http://www.google.com/intl/en_ALL/help/terms_maps.html Those terms specifically forbid several things I'm definitely not doing (mass-downloading data, using it for turn-by-turn navigation, etc.), so my first read of it makes it sound like grabbing Maps content in my art project is okay. But then I'm not a lawyer, so maybe it isn't okay.
      - res0nat0r 13y ago
        Ridiculous. Playing dumb isn't an excuse. Can I stop paying my morgtage on my house because the 40 pages of documents I signed I don't 100% understand? No. If you want to enter a legally binding contract that you don't understand you hire someone to help you do so, or face the consequences if you signed something you were not smart enough to be signing.
- lmgftp 13y ago
  This really could use more exposure. People who believe that because they only had to click "Accept" they may claim ignorance or stupidity/absurdity of the terms the accepted ought to be corrected. It's simply silly to think you may violate the terms of any agreement made, especially when that agreement is necessary to your use of the service (hence "terms" of service). While it is clear that hacker culture embraces putting things from numerous sources together, in this case the Google Maps elevation API with crime data, it's one thing to make a cool presentation and another to make a cool presentation derived from someone else's work against their explicit wishes, expressed in the terms. The use of software implies (by necessity or otherwise) that you wish to make/create/analyze/do-X with the software, and by using it you've accepted the fact that you will do-X, but within the terms. It's a funny world we live in ;)
  - sillysaurus2 13y ago
    If they wish people not use it in that way, then they should make it impossible to use in that way, or ban those people from using it. Those are really the only two options, no matter how much we wish people would be honorable.
  - mikeash 13y ago
    Personally, I think it's silly to think that you may be bound to the terms of an agreement just because you clicked a button that says "Agree", without any of the fundamental attributes of a contract being present, and without even interacting with a representative of the other party in any way.
    - reginaldjcooper 13y ago
      I've noticed typically I agree with your views, but on this case we diverge. I think it's reasonable to assume that when the service cannot be accessed before the accessor clicks "agree", and the accessor clicks "agree", that constitutes a valid contractual agreement. It's no more silly than mailing in a credit card application and being bound by those terms. The fundamental attributes are present. There is a contract and you can agree to it or not. Your interaction with the company is viewing the terms the lawyers wrote and agreeing to them or leaving the website.
- ianstallings 13y ago
  I've said the same in the past and usually it's to the business managers or executives. Too often they are willing to conveniently overlook licensing or service restrictions in order to get what they want or need. A specific example - I had to convince executives that we needed to pay MPEG LA because we were using FFMPEG to encode and decode using 264 on our servers. I only convinced them after a long debate and contacting our legal department to explain the risk. It had very little to do with my personal views on licensing or patents and everything to do with professionalism. A lot of the other software developers thought this was wasting money and time, and paying off people that were somehow evil because they charge large sums for using their patented software. My stance has always been if yo don't like the terms use something else and avoid the ethical dilemma. It's just plain easier and avoids some risk.
chatman 13y ago
Everyone should try OpenStreetMap based APIs, e.g. MapQuest Open.
- potatolicious 13y ago
  I tried one part of the API recently - I'm building an app that's partially intended to give local cycling directions. Now I'm between a rock and a hard place. One the one hand some parts of the MapQuest Open API are extremely bad, on the other hand Google has an all-or-nothing license for their geo data, and their native iOS maps component sucks. MapQuest's cycling directions in Manhattan is worse than having no directions at all. They have a specific cycling flag you can set, but results will blatantly ignore bike lanes and send you down extremely wide avenues with no bike lanes and heavy traffic. IMO using MapQuest's cycling directions, at least in NYC, is an extreme hazard, and I can't in good conscience integrate it into my app.
  - graue 13y ago
    How do the directions compare to YourNavigation? http://www.yournavigation.org/ http://www.yournavigation.org/ Despite the clunky interface, I was pleasantly surprised by its biking directions: not as good as Google, but not terrible either. (I was in NYC when I tried using it for awhile.) It's fully open-source.
    - potatolicious 13y ago
      Just plugged in some of the test endpoints I used with MapQuest. YourNavigation seems to have two bicycling modes, "bicycle" and "bicycle (routes"). The former seems to disregard bike lanes entirely and generates some pretty hazardous recommendations (including cycling down 5th Ave, or busy sections of Park Ave in Midtown). The "bicycle (routes)" mode does better and seems to respect bike lanes, but not consistently. There are some endpoints where it still maneuvers you into dense traffic with no bike infrastructure, when a bike-laned street is nearby and takes only marginally longer. I just checked the OSM's dataset and the data is good - so it isn't a matter of someone inserting a bike lane where one doesn't actually exist. The app I'm working on is for NYC's new bikeshare system... Still not an acceptable alternative unfortunately. I may just give up and use Google Maps in the iOS app, but the performance of Google's native map component is extremely, extremely bad.
jacalata 13y ago
His site http://www.datasciencetoolkit.org/ http://www.datasciencetoolkit.org/ certainly looks interesting! One of the most impressive kinds of arguments - "I think you should stop doing x and here are the resources y I've created/curated to make it easier for you to do so."
- jack-r-abbit 13y ago
  I was thinking the same thing. Lots of people get the first part. But not so much with the second part.
dewitt 13y ago
FWIW, Google doesn't particularly like it when people use APIs outside the terms, either. Having to take down an app is never fun or enjoyable for anyone involved. And the cases in which the terms aren't programmatically enforced right away isn't necessarily a choice; often it is difficult to differentiate at scale between legitimate usage and disallowed usage. And while I don't know how every other company writes their developer Terms of Services, I do know that in Google's case we measure each word carefully and try hard to strike the right balance between offering as much as feasibly possible, while trying to create something long-term sustainable (so even if the usage grows far beyond our initial goals, we're still in a position to keep them available). We're hardly perfect at it yet, but we're certainly getting a lot better over time. Also worth noting, often times these terms are designed to honor upstream data contracts, so if for no other reason, it's worth trying to respect them because it's the right thing to do for the other relatively small parties, too.
- Mithaldu 13y ago
  I would have some sympathy with this if you made it easy for single people to pay you to be able to use your data in a manner outside of the ToS of the free APIs. However, for the Elevation API there are two choices: * free and heavily restricted * business api, which looks extremely sketchy † and is blocked to most people The gaming world has learned this with Steam and friends already. People will pirate the hell out of digital offerings that are not available in a reasonable manner. Steam recognized this and made games available in the most convenient manner humanly possible and are making money hand-over-fist. Sure, there are still pirates in gaming, but those are the people who are either paranoid about Steam or those who truly couldn't afford it anyhow. You're still in this pre-realization phase and have only one single entity to blame if people "misuse" your APIs: Yourself. Make the APIs available for normal people in a reasonable manner and you will find adherence to your ToS becoming the norm, not the exception. † By sketchy i mean the facts that there is zero pricing information, it's a buzzword-slaughterhouse, and the only way to get at it is to fill out a form with a lot of information, an indication of what one might want to do and which offers no further information or even actual flesh-and-blood human contact.
  - 13y ago
lhnz 13y ago
You would have thought San Francisco's crime rate would have been higher with all of these unauthorised API usages... (Yep, the data was pirated by a guy in San Francisco. Ironic.)
jimmyrocks 13y ago
If you're looking for elevation data, the USGS provides this for free for the US. Example: http://gisdata.usgs.gov/xmlwebservices2/elevation_service.asmx/getElevation?X_Value=-104.88938886666669&Y_Value=39.59397308786357&Elevation_Units=feet&Elevation_Only=&Source_Layer= http://gisdata.usgs.gov/xmlwebservices2/elevation_service.as...
- picea 13y ago
  Also worth noting, Adam Schneider's GPS Visualiser website is excellent for quick elevation lookups globally. Handy for me as Geoscience Australia (another good resource) unfortunately charge for their DEM data. http://www.gpsvisualizer.com/elevation http://www.gpsvisualizer.com/elevation http://www.ga.gov.au/ http://www.ga.gov.au/
iamleppert 13y ago
It's rate limited for a reason and every other so called "better alternative" will have some sort of limitations in place, if they want to stay in business. Nothing is for free. By saying people will just magically make a free, unlimited API service as long as people stop using Google's is dumb.
blackmad 13y ago
I've been meaning to put together a more technical blog post about the various geo pieces that we've released at foursquare. But for now, here's one that's an overview of the things we've been working on: http://engineering.foursquare.com/2013/06/11/quattroshapes-a-global-polygon-gazetteer-from-foursquare/ http://engineering.foursquare.com/2013/06/11/quattroshapes-a... The biggest piece is http://twofishes.net/ http://twofishes.net/, our coarse (city/neighborhood level) geocoder based on geonames data. It also serves as our coarse reverse geocoder when we combine it with polygon data, which we've released at http://quattroshapes.com/ http://quattroshapes.com/, as a combination of open government data and synthesized data. http://zetashapes.com/ http://zetashapes.com/ is a side project of ours, where we're seeing if people want to draw their own crowdsourced neighborhood boundaries in US by coloring in a map of basically city blocks. I'm always eager to chat about open geodata as well as complimentary commercial solutions.
ubasu 13y ago
From the article: "I’m not RMS, I love open-source for very pragmatic reasons. One of the biggest is that I hate hitting black boxes when I’m debugging!" RMS started the free software movement because he wasn't able to adapt a black box printer.
- mkopinsky 13y ago
  Even if RMS started from a pragmatic standpoint, his position has clearly moved far beyond pragmatism.
  - reginaldjcooper 13y ago
    Has it? I think it's very pragmatic to want control over your hardware and the ability to repair problems or add functionality as desired. If you consider it from what is best for society as well, it's very restrictive to have hardware where the software cannot be modified.
ChuckMcM 13y ago
I wonder how much of this is an outgrowth of 'hackathon culture' which is to say "Hey we can do this thing ..." and get to some demo without having the foundation bits worked out. (or more importantly the future foundation bits). There are a lot of people who used to make 'kit cars' which was they took a VW bug chassis, removed the bodywork, and installed some fiberglass body on it. Really "cool" looking sports cars. But you couldn't really say "We're going to be the next Porsche selling our new "Wangdingo" sports car which, as it turns out, is a fiberglass body on top of a VW bug chassis. The difference between kids making a prototype which looks cool (professional even, some of these cars were really really nice) but as a 'new car' it was always a cheat since the other bits weren't engineered, they were 'acquired' and outside the plans of the original engineers for the chassis. Compared to the 'motorhome' business where the company buys a Chassis from Ford or Chevy as a chassis to build a new vehicle on. Makes for an interesting conundrum. You can fail fast and iterate prototyping like this, and you can arrive at a saleable product for which there is no route to production because the underlying tech you rely on cannot be used 'legally' the way you want to use it, and the amount of money the product might conceivably generate is less than the cost of re-creating the crucial bit of technology from scratch.
- cinquemb 13y ago
  It even gets more complicated as `dewitt mentioned: "And the cases in which the terms aren't programmatically enforced right away isn't necessarily a choice; often it is difficult to differentiate at scale between legitimate usage and disallowed usage." This sets up a situation where people/companies can effectively 'get away' with it. So what is an aspiring entrepreneur to do when they are able to take advantage of a situation like this in a way that the other company isn't in a position block it (without throwing the baby out with the bath water) nor are they even trying to do what you are doing when you combine such usage with other "legal" means or data one collects themselves that is providing some value to people they tested their service/'hack' with?
tegansnyder 13y ago
If you are looking for a good mapping alternate checkout Leaflet.js. You can load from a variety of tile servers. I think a while back someone had a tutorial an setting up your own tile server on AWS. Actually http://tiledrawer.com/ http://tiledrawer.com/ provides a base AMI. Another good commercial solution for geocoding is http://smartystreets.com/ http://smartystreets.com/ if you need to do batch processing.
nick2021 13y ago
You guys can stop linking to stupid blog posts yesterday please.
nraynaud 13y ago
just to clear up a misconception, the SRTM is just unusable in it's normal form, you need a lot of time to clean it up and cross it with other sources. This guy does a lot of curation: http://www.viewfinderpanoramas.org/dem3.html http://www.viewfinderpanoramas.org/dem3.html (and I also suck google's elevation data for a test because that's the simplest way to do it: http://nraynaud.github.io/webgcode/test_threejs.html http://nraynaud.github.io/webgcode/test_threejs.html )
Sukotto 13y ago
[Edit to add] Yes, you're right. When I compare their map side-by-side the same area in maps.google.com they are clearly different. I should have looked more carefully before asking my question. So I guess my question should be "If they had the exact same page except using a Google map, would they be in compliance?" - It's not clear to me in what way the "Crime doesn't climb" page violates the quoted terms of use. Would someone please explain it to me? The Elevation API may only be used in conjunction with displaying results on a Google map; using elevation data without displaying a map for which elevation data was requested is prohibited. They used the API to display results on a Google map. Their site used the elevation data to display a map for which elevation data was requested. - I'm not being facetious. It seems to me that their use matches the requirements. So what exactly is the problem?
- bdon 13y ago
  The map in question looks like it's actually from Yahoo! Maps: http://maps.yahoo.com/ http://maps.yahoo.com/
  - X-Istence 13y ago
    I don't think so, it says Nokia in the bottom right, along with CartoDB: http://cartodb.com http://cartodb.com
    - bdon 13y ago
      Indeed, it looks like Yahoo! uses Nokia "here" maps as a data source. CartoDB is not a map provider, it's a web-based GIS suite that lets you import different base layers.
nucleardog 13y ago
Here's my problem: 1. The TOS aren't heavily enforced; thus 2. No one I work for cares about complying with the TOS; thus 3. No one is particularly supportive of my spending additional time to work within the TOS. Using and contributing work time or resources to an open source project is going to be a hard sell until there is some pressure on the business end to change. Having our data sources cut off, or having to reimplement on different providers would help.
javajosh 13y ago
Okay, so I have a question: how can open source geocoding services sustain themselves? Doesn't it cost money to gather all of that data? Where does that money come from if the service is free? I would expect an OSS geocoding service to be a) seeded with data that is out of copyright (basic geography and placenames, b) spruced up with any well-regarded modern free datasources (probably government funded), c) spruced up even more from volunteers donating GPS data. Would a geocoding service be a good candidate for a state-run utility?
timClicks 13y ago
Does anyone know of a good guide to DEM (Digital Elevation Model) formats? New Zealand has some great open data[0], and I expect that it would be a fun exercise to create your own elevation API from it but I don't know how to get started [0] http://koordinates.com/#/layer/1418-nz-80m-digital-elevation-model/ http://koordinates.com/#/layer/1418-nz-80m-digital-elevation...
jetd 13y ago
I have one comment and one question. Q: Are OSM/Nominatim good for most countries now? When I tried it was very good to good enough for developed countries like US or UK but for others clearly not. C: I really dont get Google mindset. They sell geoservices. OK. How much money that would bring 1mil, 10mil or 100mil in over 20bil company. What is the logic? Creating barriers? I dont know.

309 points by spenrose 13y ago

Here's another reason that I wish was more prevalently discussed: it's not okay to violate Terms of Service anyway. Don't accept the terms? Don't use the service. We'd all have a field day if "big brother" violated the terms of an open-source license, but the freedom we enjoy in the community comes with a responsibility to honor our agreements.

Killswitch 13y ago
I just wish people would stop being so self-entitled period.
- TallGuyShort 13y ago
  Agreed. The tone of the article is correct, in my opinion, because this really does hurt the community. I think we'd have an easier time fighting things like SOPA, DRM, etc. if actual piracy wasn't such a big problem.
  - TillE 13y ago
    And if theft didn't exist, we wouldn't need police. This is a useless observation. You may be able to effect a limited degree of societal change, but you will never satisfactorily eliminate the problem at its source. Other solutions are necessary.
  - HNLogInShit 13y ago
    This isn't "piracy." It's easily shut down if the provider wants to.
- baddox 13y ago
  I feel like I deserve to live in a world with fewer self-entitled people.
- ianstallings 13y ago
  As my developer friend likes to say - this is why we can't have nice things.
revelation 13y ago
Violating the terms of a license is very different to violating the Terms of service you "sign" with a button click.
- TallGuyShort 13y ago
  Actually I'd argue that you're more obligated to abide by something you explicitly checked "I have read and accepted..." than something that was listed somewhere in a file you downloaded with some other stuff. You probably just happen to like the terms of the GPL or ASL more than you like the terms of Google's API. I do too - but at what point are we just losing all integrity because we're too lazy to read a long document we have no intention of actually agreeing with?
  - _delirium 13y ago
    If there is an actual click-through and the license is understandable by a human, perhaps. Usually there is neither, though. I rarely even know how to look up terms of an API, or that there is an API. If I know the name of the API, I can look for "Google Foo API", but that name isn't always obvious. If I just see a URL with an obvious schema, and I change the HTTP GET parameters to retrieve different stuff, am I allowed to use that in, say, a small-scale art project, or not? The usual internet default is that public URLs are implicitly giving blanket permission to retrieve them, except that retrieval by spiders should abide by robots.txt. Google itself makes heavy use of this ok-to-retrieve default assumption. If I go to http://maps.google.com http://maps.google.com there is no license presented to me as a condition of using the service. I can however look for fine print and seek out the link "Terms of Service" of my own accord, at which point I get this: http://www.google.com/intl/en_ALL/help/terms_maps.html http://www.google.com/intl/en_ALL/help/terms_maps.html Those terms specifically forbid several things I'm definitely not doing (mass-downloading data, using it for turn-by-turn navigation, etc.), so my first read of it makes it sound like grabbing Maps content in my art project is okay. But then I'm not a lawyer, so maybe it isn't okay.
    - res0nat0r 13y ago
      Ridiculous. Playing dumb isn't an excuse. Can I stop paying my morgtage on my house because the 40 pages of documents I signed I don't 100% understand? No. If you want to enter a legally binding contract that you don't understand you hire someone to help you do so, or face the consequences if you signed something you were not smart enough to be signing.
lmgftp 13y ago
This really could use more exposure. People who believe that because they only had to click "Accept" they may claim ignorance or stupidity/absurdity of the terms the accepted ought to be corrected. It's simply silly to think you may violate the terms of any agreement made, especially when that agreement is necessary to your use of the service (hence "terms" of service). While it is clear that hacker culture embraces putting things from numerous sources together, in this case the Google Maps elevation API with crime data, it's one thing to make a cool presentation and another to make a cool presentation derived from someone else's work against their explicit wishes, expressed in the terms. The use of software implies (by necessity or otherwise) that you wish to make/create/analyze/do-X with the software, and by using it you've accepted the fact that you will do-X, but within the terms. It's a funny world we live in ;)
- sillysaurus2 13y ago
  If they wish people not use it in that way, then they should make it impossible to use in that way, or ban those people from using it. Those are really the only two options, no matter how much we wish people would be honorable.
- mikeash 13y ago
  Personally, I think it's silly to think that you may be bound to the terms of an agreement just because you clicked a button that says "Agree", without any of the fundamental attributes of a contract being present, and without even interacting with a representative of the other party in any way.
  - reginaldjcooper 13y ago
    I've noticed typically I agree with your views, but on this case we diverge. I think it's reasonable to assume that when the service cannot be accessed before the accessor clicks "agree", and the accessor clicks "agree", that constitutes a valid contractual agreement. It's no more silly than mailing in a credit card application and being bound by those terms. The fundamental attributes are present. There is a contract and you can agree to it or not. Your interaction with the company is viewing the terms the lawyers wrote and agreeing to them or leaving the website.
ianstallings 13y ago
I've said the same in the past and usually it's to the business managers or executives. Too often they are willing to conveniently overlook licensing or service restrictions in order to get what they want or need. A specific example - I had to convince executives that we needed to pay MPEG LA because we were using FFMPEG to encode and decode using 264 on our servers. I only convinced them after a long debate and contacting our legal department to explain the risk. It had very little to do with my personal views on licensing or patents and everything to do with professionalism. A lot of the other software developers thought this was wasting money and time, and paying off people that were somehow evil because they charge large sums for using their patented software. My stance has always been if yo don't like the terms use something else and avoid the ethical dilemma. It's just plain easier and avoids some risk.

chatman 13y ago

Everyone should try OpenStreetMap based APIs, e.g. MapQuest Open.

potatolicious 13y ago
I tried one part of the API recently - I'm building an app that's partially intended to give local cycling directions. Now I'm between a rock and a hard place. One the one hand some parts of the MapQuest Open API are extremely bad, on the other hand Google has an all-or-nothing license for their geo data, and their native iOS maps component sucks. MapQuest's cycling directions in Manhattan is worse than having no directions at all. They have a specific cycling flag you can set, but results will blatantly ignore bike lanes and send you down extremely wide avenues with no bike lanes and heavy traffic. IMO using MapQuest's cycling directions, at least in NYC, is an extreme hazard, and I can't in good conscience integrate it into my app.
- graue 13y ago
  How do the directions compare to YourNavigation? http://www.yournavigation.org/ http://www.yournavigation.org/ Despite the clunky interface, I was pleasantly surprised by its biking directions: not as good as Google, but not terrible either. (I was in NYC when I tried using it for awhile.) It's fully open-source.
  - potatolicious 13y ago
    Just plugged in some of the test endpoints I used with MapQuest. YourNavigation seems to have two bicycling modes, "bicycle" and "bicycle (routes"). The former seems to disregard bike lanes entirely and generates some pretty hazardous recommendations (including cycling down 5th Ave, or busy sections of Park Ave in Midtown). The "bicycle (routes)" mode does better and seems to respect bike lanes, but not consistently. There are some endpoints where it still maneuvers you into dense traffic with no bike infrastructure, when a bike-laned street is nearby and takes only marginally longer. I just checked the OSM's dataset and the data is good - so it isn't a matter of someone inserting a bike lane where one doesn't actually exist. The app I'm working on is for NYC's new bikeshare system... Still not an acceptable alternative unfortunately. I may just give up and use Google Maps in the iOS app, but the performance of Google's native map component is extremely, extremely bad.

jacalata 13y ago

His site http://www.datasciencetoolkit.org/ http://www.datasciencetoolkit.org/ certainly looks interesting! One of the most impressive kinds of arguments - "I think you should stop doing x and here are the resources y I've created/curated to make it easier for you to do so."

jack-r-abbit 13y ago
I was thinking the same thing. Lots of people get the first part. But not so much with the second part.

dewitt 13y ago

FWIW, Google doesn't particularly like it when people use APIs outside the terms, either. Having to take down an app is never fun or enjoyable for anyone involved. And the cases in which the terms aren't programmatically enforced right away isn't necessarily a choice; often it is difficult to differentiate at scale between legitimate usage and disallowed usage. And while I don't know how every other company writes their developer Terms of Services, I do know that in Google's case we measure each word carefully and try hard to strike the right balance between offering as much as feasibly possible, while trying to create something long-term sustainable (so even if the usage grows far beyond our initial goals, we're still in a position to keep them available). We're hardly perfect at it yet, but we're certainly getting a lot better over time. Also worth noting, often times these terms are designed to honor upstream data contracts, so if for no other reason, it's worth trying to respect them because it's the right thing to do for the other relatively small parties, too.

Mithaldu 13y ago
I would have some sympathy with this if you made it easy for single people to pay you to be able to use your data in a manner outside of the ToS of the free APIs. However, for the Elevation API there are two choices: * free and heavily restricted * business api, which looks extremely sketchy † and is blocked to most people The gaming world has learned this with Steam and friends already. People will pirate the hell out of digital offerings that are not available in a reasonable manner. Steam recognized this and made games available in the most convenient manner humanly possible and are making money hand-over-fist. Sure, there are still pirates in gaming, but those are the people who are either paranoid about Steam or those who truly couldn't afford it anyhow. You're still in this pre-realization phase and have only one single entity to blame if people "misuse" your APIs: Yourself. Make the APIs available for normal people in a reasonable manner and you will find adherence to your ToS becoming the norm, not the exception. † By sketchy i mean the facts that there is zero pricing information, it's a buzzword-slaughterhouse, and the only way to get at it is to fill out a form with a lot of information, an indication of what one might want to do and which offers no further information or even actual flesh-and-blood human contact.
- 13y ago

lhnz 13y ago

You would have thought San Francisco's crime rate would have been higher with all of these unauthorised API usages... (Yep, the data was pirated by a guy in San Francisco. Ironic.)

jimmyrocks 13y ago

If you're looking for elevation data, the USGS provides this for free for the US. Example: http://gisdata.usgs.gov/xmlwebservices2/elevation_service.asmx/getElevation?X_Value=-104.88938886666669&Y_Value=39.59397308786357&Elevation_Units=feet&Elevation_Only=&Source_Layer= http://gisdata.usgs.gov/xmlwebservices2/elevation_service.as...

picea 13y ago
Also worth noting, Adam Schneider's GPS Visualiser website is excellent for quick elevation lookups globally. Handy for me as Geoscience Australia (another good resource) unfortunately charge for their DEM data. http://www.gpsvisualizer.com/elevation http://www.gpsvisualizer.com/elevation http://www.ga.gov.au/ http://www.ga.gov.au/

iamleppert 13y ago

It's rate limited for a reason and every other so called "better alternative" will have some sort of limitations in place, if they want to stay in business. Nothing is for free. By saying people will just magically make a free, unlimited API service as long as people stop using Google's is dumb.

blackmad 13y ago

I've been meaning to put together a more technical blog post about the various geo pieces that we've released at foursquare. But for now, here's one that's an overview of the things we've been working on: http://engineering.foursquare.com/2013/06/11/quattroshapes-a-global-polygon-gazetteer-from-foursquare/ http://engineering.foursquare.com/2013/06/11/quattroshapes-a... The biggest piece is http://twofishes.net/ http://twofishes.net/, our coarse (city/neighborhood level) geocoder based on geonames data. It also serves as our coarse reverse geocoder when we combine it with polygon data, which we've released at http://quattroshapes.com/ http://quattroshapes.com/, as a combination of open government data and synthesized data. http://zetashapes.com/ http://zetashapes.com/ is a side project of ours, where we're seeing if people want to draw their own crowdsourced neighborhood boundaries in US by coloring in a map of basically city blocks. I'm always eager to chat about open geodata as well as complimentary commercial solutions.

ubasu 13y ago

From the article: "I’m not RMS, I love open-source for very pragmatic reasons. One of the biggest is that I hate hitting black boxes when I’m debugging!" RMS started the free software movement because he wasn't able to adapt a black box printer.

mkopinsky 13y ago
Even if RMS started from a pragmatic standpoint, his position has clearly moved far beyond pragmatism.
- reginaldjcooper 13y ago
  Has it? I think it's very pragmatic to want control over your hardware and the ability to repair problems or add functionality as desired. If you consider it from what is best for society as well, it's very restrictive to have hardware where the software cannot be modified.

ChuckMcM 13y ago

I wonder how much of this is an outgrowth of 'hackathon culture' which is to say "Hey we can do this thing ..." and get to some demo without having the foundation bits worked out. (or more importantly the future foundation bits). There are a lot of people who used to make 'kit cars' which was they took a VW bug chassis, removed the bodywork, and installed some fiberglass body on it. Really "cool" looking sports cars. But you couldn't really say "We're going to be the next Porsche selling our new "Wangdingo" sports car which, as it turns out, is a fiberglass body on top of a VW bug chassis. The difference between kids making a prototype which looks cool (professional even, some of these cars were really really nice) but as a 'new car' it was always a cheat since the other bits weren't engineered, they were 'acquired' and outside the plans of the original engineers for the chassis. Compared to the 'motorhome' business where the company buys a Chassis from Ford or Chevy as a chassis to build a new vehicle on. Makes for an interesting conundrum. You can fail fast and iterate prototyping like this, and you can arrive at a saleable product for which there is no route to production because the underlying tech you rely on cannot be used 'legally' the way you want to use it, and the amount of money the product might conceivably generate is less than the cost of re-creating the crucial bit of technology from scratch.

cinquemb 13y ago
It even gets more complicated as `dewitt mentioned: "And the cases in which the terms aren't programmatically enforced right away isn't necessarily a choice; often it is difficult to differentiate at scale between legitimate usage and disallowed usage." This sets up a situation where people/companies can effectively 'get away' with it. So what is an aspiring entrepreneur to do when they are able to take advantage of a situation like this in a way that the other company isn't in a position block it (without throwing the baby out with the bath water) nor are they even trying to do what you are doing when you combine such usage with other "legal" means or data one collects themselves that is providing some value to people they tested their service/'hack' with?

tegansnyder 13y ago

If you are looking for a good mapping alternate checkout Leaflet.js. You can load from a variety of tile servers. I think a while back someone had a tutorial an setting up your own tile server on AWS. Actually http://tiledrawer.com/ http://tiledrawer.com/ provides a base AMI. Another good commercial solution for geocoding is http://smartystreets.com/ http://smartystreets.com/ if you need to do batch processing.

nick2021 13y ago

You guys can stop linking to stupid blog posts yesterday please.

nraynaud 13y ago

just to clear up a misconception, the SRTM is just unusable in it's normal form, you need a lot of time to clean it up and cross it with other sources. This guy does a lot of curation: http://www.viewfinderpanoramas.org/dem3.html http://www.viewfinderpanoramas.org/dem3.html (and I also suck google's elevation data for a test because that's the simplest way to do it: http://nraynaud.github.io/webgcode/test_threejs.html http://nraynaud.github.io/webgcode/test_threejs.html )

Sukotto 13y ago

[Edit to add] Yes, you're right. When I compare their map side-by-side the same area in maps.google.com they are clearly different. I should have looked more carefully before asking my question. So I guess my question should be "If they had the exact same page except using a Google map, would they be in compliance?" - It's not clear to me in what way the "Crime doesn't climb" page violates the quoted terms of use. Would someone please explain it to me? The Elevation API may only be used in conjunction with displaying results on a Google map; using elevation data without displaying a map for which elevation data was requested is prohibited. They used the API to display results on a Google map. Their site used the elevation data to display a map for which elevation data was requested. - I'm not being facetious. It seems to me that their use matches the requirements. So what exactly is the problem?

bdon 13y ago
The map in question looks like it's actually from Yahoo! Maps: http://maps.yahoo.com/ http://maps.yahoo.com/
- X-Istence 13y ago
  I don't think so, it says Nokia in the bottom right, along with CartoDB: http://cartodb.com http://cartodb.com
  - bdon 13y ago
    Indeed, it looks like Yahoo! uses Nokia "here" maps as a data source. CartoDB is not a map provider, it's a web-based GIS suite that lets you import different base layers.

nucleardog 13y ago

Here's my problem: 1. The TOS aren't heavily enforced; thus 2. No one I work for cares about complying with the TOS; thus 3. No one is particularly supportive of my spending additional time to work within the TOS. Using and contributing work time or resources to an open source project is going to be a hard sell until there is some pressure on the business end to change. Having our data sources cut off, or having to reimplement on different providers would help.

javajosh 13y ago

Okay, so I have a question: how can open source geocoding services sustain themselves? Doesn't it cost money to gather all of that data? Where does that money come from if the service is free? I would expect an OSS geocoding service to be a) seeded with data that is out of copyright (basic geography and placenames, b) spruced up with any well-regarded modern free datasources (probably government funded), c) spruced up even more from volunteers donating GPS data. Would a geocoding service be a good candidate for a state-run utility?

timClicks 13y ago

Does anyone know of a good guide to DEM (Digital Elevation Model) formats? New Zealand has some great open data[0], and I expect that it would be a fun exercise to create your own elevation API from it but I don't know how to get started [0] http://koordinates.com/#/layer/1418-nz-80m-digital-elevation-model/ http://koordinates.com/#/layer/1418-nz-80m-digital-elevation...

jetd 13y ago

I have one comment and one question. Q: Are OSM/Nominatim good for most countries now? When I tried it was very good to good enough for developed countries like US or UK but for others clearly not. C: I really dont get Google mindset. They sell geoservices. OK. How much money that would bring 1mil, 10mil or 100mil in over 20bil company. What is the logic? Creating barriers? I dont know.