Google Maps Technology Thoughts

Whose map is it anyway..

View Larger Map

It’s mine all mine..

This is a map of recycling centres in Teddington, my local neighbourhood in London. I created it my looking up the locations of recycling centres run by my local council, the London Borough of Richmond, from their website and then added the points using the existing Google Map and Satellite image for context.

So who owns this new map ?

I do !

By publishing the map using Google Maps, I give Google a license to use my data but it’s “ownership” as such remains with me. The license is just an explicit statement of the implict intenention of publishing a map for consumption by the public using Google as a publication channel.

Google makes no claim over the intellectual property of the maps you create, they remain your maps and the data remains your data !

Written and submitted from home, using my home 802.11 network.

50 replies on “Whose map is it anyway..”


That’s an interesting perspective. Not supported by most people in this area, so far as I’m aware, so I’d be interested to see that stated in some official capacity. (I won’t go so far as to assume that a comment made in your personal blog counts as the official position of a legal nature. :))

I can’t tell if you simply disagree with the general legal consensus (actually won in court at least once) that basing your mapping efforts on existing maps is creating a derivative work. The case I’m thinking of was actually won against a company which wasn’t using the map as a basis for anything other than how to *get* to things they hadn’t mapped yet, and a court ruled that they were in violation, iirc. At the very least, the interpretation at the end of comment #18 is not the most common one among open mapping advocates; it’s specifically much more open than most people have thus far assumed.

I’d also be interested in Tim’s question in #20: How does this extend to satellite imagery? Clearly, there is interpretation in a specific cluster of buildings being in a specific place having some meaning — If recyling centers or whatever these things are are visible on the satellite imagery (and that’s how you got ahold of the locations), how does that affect things?

@ Tim,

I actually used a combination of the imagery which does show the location of some of the sites and local knowledge. Again as I have said before, because the recycling centres themselves don’t appear as features in any mapping, my map is not a derived work.

You use case sounds fine, you are not tracing a feature from the map, the operation you suggest is something you would do when brining in an image overlay into Google Earth for example.

@Chris.. You must understand the difference between intrepreting an image or placing a faeture based on your local knoweldge and tracing an exsiting feature that is already represented on the map. Tim question I think helps to clarify this, I can interpret the location of a recycling centre from the image and label its location, I can do this only because of my experience or additional local knowledge.

This is the key point to understand what is derived and what is not.

Evidently the UK IPR issue is restrictive and ambiguous to such a degree that it stimulate such discussion. Ed knows this and is using this medium to smoke out the culprits and stimulate change. Good call. The US model is the ideal model. We paid for the data, and it should be in the public domain without restriction to stimulate society. The OS model is nothing but a stranglehold. Need to go and put my bins out now. I will use my own interpretation of my spatial cache to position my grey and green bins based upon real world spatial correlation and the fact that council bin man will not walk too far from the truck. My knowledge is of course underpinned by years of OS map usage and thus, I am in debt.

Brief technical clarification:
The map that Ed has embedded above uses MyMaps and Google Maps – it does not use the Google Maps API. Therefore it is subject only to the Maps Terms of Use, and not to the Maps API Terms of Service.

So keep in mind that this particular discussion does not actually actually concern the Maps API ToS. Thanks!


Again, I would love to see comments on this issue from Google in a more authoritative location than your personal blog.

Specific use cases that I can’t understand based on your distinction:

I trace a footpath from satellite imagery. I know the footpath is there, because I’ve walked down it, but I don’t know if it has a name, just that it exists. (This is true, for example, for some foot paths in the hills of San Francisco, which I’ve walked down, but didn’t pay much attention to.) The local knowledge allows me to know this is a footpath, but the exact location of it is determined exclusively from the map.

One possible interpretation: Knowing that it is a footpath makes it not a derived work, because it required the local knowledge of walking down it to know what it was.

Another possible interpretation: The location of this foot path is traced exclusively from the map, and is therefore a derivative of the map.

Your statements above make it seem like you’re saying the former is true: You identified the locations of features exclusively from the map, using street/aerial imagery. These features are (in at least one case that I looked at) visible from the Aerial imagery, so the location is, at least possibly, derived. So, the local knowledge that you added is that it is a recycling center.

Another possible use case: I trace the same footpath, using the same imagery, but I *haven’t* walked down it before. However, I can clearly tell it’s a foot path because I see people walking down it.

In this case, the only difference is that I’m interpreting the image, instead of using a physical location as being the informative piece of context. I’m interpreting the size of the feature, the imagery of people walking down it, and making a judgement call about what it is. The points are exactly the same, and the end result is exactly the same. Is this one any more or less a derivative work?

Certainly, I can look at images in the Google Map above and see what are clearly bins and so on. The fact that it is a recycling center is a piece of information that I can determine from the image. If I were to make the same determination from the image, or from walking by and taking a snapshot, it sounds like you’re saying that the latter is not a derivative work, and the former is.

If this is true — the attribute being the important creative work — there is still a key point that isn’t addressed: The physical location of the thing you are labelling — it’s latitude/longitude coordinates — is exclusively determined by looking at the map in *either* case. (Assuming that I don’t have a GPS with me when I take the snapshot.) In either case, the geographic coordinates that exist have to be owned by the same person, as far as I can tell, no?

Another use case: I drive a route to the airport with a GPS in my car. At one point, I go under some trees, and the GPS signal cuts out. I trace out a line over the GPS track, and then lay that down in MapMaker. I then notice that at the point where the GPS cut out, there is a small curve in the road.

I’ve been down the road. I’ve seen the curve. I know it exists, but not exactly where it is, because I don’t have the geodata collected from an independent source. If I take my line, improve it (by dragging it around in MapMaker) and then export the data again — who owns the geodata? I couldn’t have built that line without Google Maps, but I wouldn’t have built the same line without GPS either.

I’m willing to admit that the combination of the attribute and geographic information creates a piece of information that perhaps couldn’t be garnered from the map alone. However, the geodata itself — the placement of the marker — is clearly not based on local knowledge. The same data could be created — in at least one place — by simply looking at the map and selecting a point which is visibly a recycling spot. I don’t understand that the extraction of this piece of geodata can be said to be the result of local knowledge, in either case. (Local knowledge might put a recycling center at, for example, the intersection of Park Lane and the A309, but not in the specific parking lot where it is actually placed.) With that in mind, I’m interested in where the line gets drawn.

To expand the question: If I take imagery and put a point on everything that looks like a recycling center, I might get a similar map to what you have here. (Assuming sufficiently detailed imagery of the area in question, and enough free time to actually inspect every pixel.) The pixels would likely end up in a very similar geographic pattern — and I don’t understand if using the same source map, and getting the same output data, by using a different entry pattern, would mean that the restrictions on the result are somehow different.

The issue of what is / is not derived is complex, it simple terms you cannot trace any feature that already exists on the map.

Using an image and relative location of other features on a map to create and place a new feature is generally OK, and if creating a feature to be shared with others on a Google Maps is something we would urge you to do.

And back to the original point of my post.. for the new features you create, you provide a license to Google to allow the data to be used on other Google sites and by other users of Google products. The intellectual property rights remain with you as the creator of the new feature.

All semantics. Bottom line: OS claim everything is derived from their Godly Db in Soton. Accept it and pay your dues. All this talk is solving nothing. Rise up warriors of GI: lay waste to OSHQ. Free our data.


“We would urge you to do that on a Google Map” is clearly different from “The data belongs to you”: Since Google pays the licensing fees on all this data, what you can do with the data within Google Maps, Google MapMaker, or other Google related mapping products is clearly completely unrelated to whose data it actually is.

You claim the data that you’ve created here belongs to you, because you’re creating new features. I say that in at least some cases, I can observe the features that you created — and in no case was the creation of the features geographic location at any point separate from the proprietary/closed Google data. It seems clear that no one could take this data and, for example, integrate it into OSM: the locations are entirely based on the map vector and imagery data, not an independent data source which is not owned by Google.

You could say that you added information to the feature creation (the labeling of the features as recycling centers): I’m completely willing to grant that is local knowledge that you wouldn’t get from the proprietary datasource. But the locations themselves could not be generated without the map: being entirely dependent on Google Maps for creating the locations sounds an awful lot to me like it’s creating a derivative work.

If the point was to say “If you create data, we’re not going to start owning it just by you creating it” — for example, just by publishing/uploading a KML file of GPS waypoints into MapMaker, I haven’t given up any rights to it — that’s totally cool, and I’d agree 100%. But if that’s what you were trying to prove, creating data clearly derived from a proprietary source doesn’t sound like it’s making that case to me. 🙂


Please go back comment 1 and read them all again, if after the second reading what I am saying is still not clear I cannot help anymore, and my argument clearly needs recycling itself 🙂

As The Spatializer points out, the real target for this clarification could not care less anyway.

Leave a Reply to LIBRIS utvecklingsblogg // Angelologi för 2008 Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.