LWN: Comments on "Microsoft helping OpenStreetMap"

OSM mapping and (in)consistency

smurf — Mon, 29 Nov 2010 21:07:50 +0000

Yes, OSM is inconsistent. There's a simple reason for this: Mapping is dead simple in theory, but looking further into the nitty gritty details reveals a whole lot of interesting problems.

And if Brazil, as a culture, solves the real-world problems with traffic, naming things, writing maps, etc., in some way that's almost, but not quite, entirely unlike the way the UK or Japan or does, that's just life. Deal with it.

The other side of the problem is that the OSM creators themselves were somewhat naïve when they started the project (and some contributors still are ). So yes, the whole set of tags is somewhat inconsistent at times.

But then, the very idea to go and map the entire world is so damn daunting that anybody who does not have that bit of naïvety wouldn't even be able to get started. They'd throw up their hands when confronted with the sheer magnitude of the problem, and leave mapping in the hands of some data-hoarding corporations who are at least two years behind the Real World.

That kind of contribution economy just plain works. It works for Wikipedia, and it does for OSM. This, more than anything else, IMHO is the whole point of having this huge thing called "The Internet".

OK. I'll get off my soapbox now. Sheesh. ;-)

Microsoft helping OpenStreetMap

dlang — Mon, 29 Nov 2010 09:11:04 +0000

it's Tomtom that has the option for users to edit maps.

Microsoft helping OpenStreetMap

cmccabe — Mon, 29 Nov 2010 07:37:58 +0000

This is a smart move for Microsoft. And hopefully it will further the cause of openness for map data.

However... my experience from the GPS industry is that users quickly become frustrated with even minor faults in map data. It leads to the routing engine giving you back weird routes, like routes that wander through multiple parking lots, zigzag around through residential areas, or just take streets that are always busy. We had a lot of trouble with TeleAtlas' map data (their US data is terrible).

The worst part is that the people writing the routing algorithms start to put in fudges to "clean up" the map data (scare quotes intentional). These are tested by looking at a few areas on the map that the routing engineer knows well. What he doesn't see is that his fudge usually makes things worse in a lot of other parts of the map. The right solution is to fix the data.

It's hard to crowdsource map data. One of the big GPS vendors-- I think it was Magellan-- rolled out a system that let ordinary users edit the maps on their devices. They also rolled out a way that users could contribute the data back very easily. Unfortunately, what they found is that the biggest use people found for the map editing system was to insert fake blockages on roads they didn't want the routing engine to take! People often prefer to do things the quick way rather than the correct way.

Unfortunately, unlike code, it's hard to audit map data contributions from the community. Linus can check a kernel patch to see if it make sense. But if someone from Podunk, East Texas tells you that there are 5 stop signs on Main Street, how do you evaluate that? Nobody back at HQ has ever been to Podunk. We had this problem all the time when working with map data. Unfortunately, it was usually resolved by shrugging and saying, "well nobody lives there anyway."

The silliest part about all of this is that the data for where the roads are, the signage, speed limits, and so forth is usually known to someone in the local government. It's just that the municipalities don't do a good job of putting that data in a standard format and making it available. Well, ok, TIGER data is available, but it's of poor quality-- much worse than TeleAtlas or NAVTEQ. Hopefully these guys will get more accurate, and better at coordinating with the community, in the future.

Map recognition

cortana — Sun, 28 Nov 2010 17:30:25 +0000

That was an April Fools Day prank, right? :)

Map recognition

smurf — Sun, 28 Nov 2010 09:34:54 +0000

Not only the commercial mappers do that. OSM has an entire fake Bavarian village hidden in its database.

OpenStreetMap licence?

johill — Sun, 28 Nov 2010 09:23:42 +0000

The "beatosu" one also has "goblu" -- according to "How to Lie with Maps" (ISBN 0-226-53421-9)

OpenStreetMap licence?

jhhaller — Sun, 28 Nov 2010 00:35:32 +0000

In general, the commercial map companies have identified copies of their maps by adding information for things that don't exist, such as fictional cities or cul-de-sacs that don't exist (depending on map scale). The maps might not be quite to scale, or other things that allow the map company to identify verbatim copies without disturbing people looking for legitimate things. My favorite was the Beat OSU in a Michigan map (a local sports rivalry). This is the same type of thing that phone books use to identify copies.

OpenStreetMap licence?

zotz — Fri, 26 Nov 2010 16:07:59 +0000

"You are right that a CC licence is quite restrictive in some cases, preventing a map from being included as part of a larger work unless that work is also CC-licensed."

If a map printed in a book would be treated like a photograph printed in a book then I don't think this worry is true. See the discussion in the cc-community archives. Same for a newspaper article and what have you.

I don't like that this is true and keep trying to get BY-SA changed to close what I consider to be a loophole, but that hasn't happened yet.

drew

OpenStreetMap licence?

epa — Fri, 26 Nov 2010 12:45:09 +0000

Pretending it isn't copyrightable is right up there with pretending the world == America.

I completely agree with this and it is the point I was trying to make. Much of the justification for dropping CC-BY-SA in favour of the more complex ODbL is the assertion that 'copyright doesn't apply' for some reason. This is simply false.

I don't believe that listing the names of all contributors is necessary even under the strictest interpretation; it's not like the BSD advertising clause where there is some fixed text specified by the licence. You are right that a CC licence is quite restrictive in some cases, preventing a map from being included as part of a larger work unless that work is also CC-licensed. This is a good argument for offering a more liberal licence as an alternative. It does not argue for revoking CC-BY-SA permission altogether.

OpenStreetMap licence?

epa — Fri, 26 Nov 2010 12:40:43 +0000

The worst possible state in my opinion : there is a contribution attribution agreement. I was willing to contribute, but when I saw that, I refused.

This 'agreement' is a new development and applies only to new contributors - existing ones are grandfathered in for the time being, where you need only agree to license your contributions under CC-BY-SA. The terms are still being discussed, so you have a chance to influence the debate. I also hope that the one-sidedness of the contributor agreement will be fixed.

OpenStreetMap licence?

smurf — Fri, 26 Nov 2010 09:41:04 +0000

I totally agree.

Correction: The "Accept" button also places your contributions in the public domain only applies if you also click the appropriate button below the agreement. It defaults to Off.

OpenStreetMap licence?

kleptog — Fri, 26 Nov 2010 07:37:27 +0000

Well, it's a vote in the sense that if you disagree you don't do anything. If enough people disagree, well, then it's obvious something else needs to be done.

The real problem is that CC-BY-SA is a completely inappropriate license and ignoring that fact isn't helping anyone. Under the current licence if you make a map from OSM data you need to technically list the name of every contributor who ever edited anything in that area. Nobody does this of course but it indicates the issues. And that licence applies to the resulting map, meaning you can't include it in other works with a conflicting licence. CC-BY-SA is a too restrictive licence, you need something with less scope. Pretending it isn't copyrightable is right up there with pretending the world == America.

BTW, The "Accept" button also places your contributions in the public domain (wherever the concept exists and is applicable) if that makes you feel any better.

Yahoo satellite images -> Bing satellite images?

robert_s — Thu, 25 Nov 2010 14:07:53 +0000

"with abutters=residential"

People haven't been using abutters= for years.

You don't have to worry so much about the classification of roads. It's far more pragmatic than you seem to think. The general rule is that you are classifying the road relative to the other roads in the surrounding area.

OpenStreetMap licence?

epa — Thu, 25 Nov 2010 13:11:09 +0000

The problem is: CC-BY-SA and the actual and/or intended use of the data conflict in a few places. Many people (around 85% of the OSM contributors, AFAIK) think these are important and have therefore voted for replacing the license with one that's tailored for databases, while essentially granting the same freedoms.

There has been no such vote. The OSM contributors have not been given the opportunity to vote - although there is a page with a single 'accept' option, with the threat that your data will be deleted from the project if you don't agree.

Personally, I'd be happy to add 'a licence tailored for databases' as an option, but not by dropping Creative Commons compatibility. That is a big step backwards.

OpenStreetMap licence?

ballombe — Thu, 25 Nov 2010 12:36:26 +0000

The ODL try to enforce EU-style sui-generis database right in jurisdiction where they do not exist in law.

OpenStreetMap licence?

smurf — Thu, 25 Nov 2010 12:13:01 +0000

True, facts can't be copyrighted. That's not the problem.

The effort to collect and describe these facts, and therefore the resulting work (i.e. the database) can be. The only question is which license one puts them under.

OpenStreetmap has traditionally used CC-BY-SA, but that's inconsistent. Databases are not a work of art, and the "Oslo is the capitol of Norway" entry in the OSM database is not a Wikipedia article.

The problem is: CC-BY-SA and the actual and/or intended use of the data conflict in a few places. Many people (around 85% of the OSM contributors, AFAIK) think these are important and have therefore voted for replacing the license with one that's tailored for databases, while essentially granting the same freedoms.

The "maps are facts and therefore not copyrightable" idea may be cute in theory, but the real world doesn't work like that. The commercial mapping companies out there have repeatedly prevailed in court against people who simply placed a "here's how to find me" map image on their homepage.

Yahoo satellite images -> Bing satellite images?

cortana — Thu, 25 Nov 2010 11:05:19 +0000

Maybe it's part of the British heritage of the project. In the UK, "primary routes" are A-roads, which are the trunk roads that pre-dated the creation of the motorway network.

OpenStreetMap licence?

epa — Thu, 25 Nov 2010 08:59:11 +0000

Feel free to, for example, copy the fact that Oslo is capitol of Norway from a map, from a encyclopedia, from any other copyrighted work. That fact is not, and can not be, copyrighted.

Yes, and it's absolutely bananas to treat that as a bug, and attempt to restrict distribution of such 'mere facts' by putting magic text in a licence document, turning it into an EULA (which is nonetheless still not enforceable, since nobody need agree to it).

This misguided effort has been driven by the idea that since facts like the capital of Norway are not subject to copyright, somehow the entire map is not subject to it, simply because it is stored electronically in machine-readable form.

OpenStreetMap licence?

ekj — Thu, 25 Nov 2010 07:45:16 +0000

There's a difference between map-data, and maps. That is, a map is something more than a collection of data.

The individual datapoints don't enjoy copyright, it's the creative and artistic expression of them, including the selection of which facts to show, that may.

Feel free to, for example, copy the fact that Oslo is capitol of Norway from a map, from a encyclopedia, from any other copyrigthed work. That fact is not, and can not be, copyrigthed. But a text that writes about Norway and mentions the capitol, can be. And a map that shows Norway, and displays the capitol, also can be.

Yes it's tricky.

Yahoo satellite images -> Bing satellite images?

cesarb — Wed, 24 Nov 2010 22:49:20 +0000

> Nobody wants to buy map data from sixty different countries, each in a different format, with inconsistent rules about what terms mean ("Oh, in Germany, this distinction does not exist, but instead ...")

Instead, you get a single inconsistent set of rules as to what the values of the tags mean. Is it highway=residential or highway=tertiary with abutters=residential? Which of the two sets of explanations as to what the values for highway= mean on Brazil should I use, or should I use a third one from somewhere else? If I have a two-way road with continuous lines in the middle (no overtaking allowed), should I use a single oneway=no way, or a pair of oneway=yes ways? Should buoys be marked with the simple set of tags, the complicated set of tags, or both at the same time?

(Sorry about the rant, but I fear at least half of the highway= ways I added are tagged incorrectly. And I still cannot see how the the six levels of highway= tag values, two of them being more primary than "primary", match the kinds of roads on the major urban area I live in.)

OpenStreetMap licence?

epa — Wed, 24 Nov 2010 22:34:24 +0000

When version 1.0 of the licence was first announced I posted a comment on LWN with my thoughts.

Since then, I would add that the 'copyright doesn't apply to map data' meme which is so often repeated does not appear to have any basis in fact. If you doubt it, please try wholesale copying of Tele Atlas or Ordnance Survey or any other copyrighted map data set. Indeed the OSM project itself recognizes this fact: copying data from other maps is prohibited unless they are out of copyright (even if they are old enough to predate the existence of any database right).

But, for the time being, OSM is still using the free CC-BY-SA licence, and it is unlikely to switch before achieving consensus on any replacement. (A move to public domain or attribution-only would make more sense than the big hairball of legalese that is the ODbL.)

OpenStreetMap licence?

robert_s — Wed, 24 Nov 2010 21:49:34 +0000

"It goes as far as forbidding use that would be legal according to copyright law."

The problem here is that copyright law is _very_ complex and nuanced when it comes to geodata. Remember - depending on your region, there may be more than copyright that is applicable to the data - e.g. "database rights".

"We wondered why OSM moved to self-destruct mode with the ODL and copyright assignment. Now we know: Microsoft is involved."

This is nonsense. Whatever you think of the license, the motivation behind it was with the best of intentions.

The "self destruct mode" is also very exaggerated. What we have is _relatively_ few dissenters who are making a lot of noise and are tiring out most of the sane voices. Remember - keeping the data CC-BY-SA is effectively the same as making it public domain because CC-BY-SA is not really applicable to map data.

OpenStreetMap licence?

vblum — Wed, 24 Nov 2010 21:47:06 +0000

What exactly does it prohibit?

OpenStreetMap licence?

ballombe — Wed, 24 Nov 2010 19:54:18 +0000

The Open Database License is a pretty bad license. It goes as far as forbidding use that would be legal according to copyright law.
<tinfoilhat>
We wondered why OSM moved to self-destruct mode with the ODL and copyright assignment. Now we know: Microsoft is involved.
</tinfoilhat>

OpenStreetMap licence?

micka — Wed, 24 Nov 2010 18:52:52 +0000

The worst possible state in my opinion : there is a contribution attribution agreement.
I was willing to contribute, but when I saw that, I refused.

Microsoft helping OpenStreetMap

iabervon — Wed, 24 Nov 2010 17:40:27 +0000

It seems like Microsoft doesn't see much value in having their own data these days (assuming that the point of bing is searching the data rather than the data itself). They canceled Encarta last year, for example. I think it actually makes a lot of sense for them to skip trying to create and maintain the database, which is likely to never be outstanding, and instead focus on writing software to make use of a freely-accessible database that they can contribute a bit to. If you can't have notably better maps than everyone else, you can at least make sure that the best maps are ones you can use, and differentiate your products on something else.

It's not really a new strategy for Microsoft, even; consider how much encouragement they gave to PC clones starting in the 80s, including sharing a lot of information on how the hardware would have to act in order for Microsoft's OSes to run, when IBM might have preferred not to have an open standard. Microsoft wasn't going to get big margins on hardware, so they made sure nobody else would either, letting the competition drive down hardware prices and leaving people's budgets free to spend lots of money on software from Microsoft.

Yahoo satellite images -> Bing satellite images?

tialaramex — Wed, 24 Nov 2010 17:27:26 +0000

The big picture is that map data is a natural monopoly. And even Google probably agree that it doesn't make sense for that monopoly to be run by a for-profit company.

In a joined-up world it also doesn't make sense to try to divide the monopoly geographically, as happened historically. Nobody wants to buy map data from sixty different countries, each in a different format, with inconsistent rules about what terms mean ("Oh, in Germany, this distinction does not exist, but instead ...") and licensing conditions that bury novel uses in tortuous legal process.

In the past OSM had access to Yahoo images (and I believe that Yahoo may have been one of the sponsors which bought me a pleasant meal after a day surveying a maze-like housing estate for OSM). I'd be interested to hear how that relates to this new announcement. Is it the same images under a new name, an expanded range of images, or ...?

OpenStreetMap licence?

smurf — Wed, 24 Nov 2010 16:48:18 +0000

CC-BY-SA 2.0.

They're switching to the ODbL (Open Database License), but I don't know the current status.

Microsoft helping OpenStreetMap

SEJeff — Wed, 24 Nov 2010 16:47:50 +0000

You must have not gotten the memo. Microsoft hired the openstreetmap founder, and added a OSM layer to bing maps:

http://www.bing.com/toolbox/blogs/maps/archive/2010/08/02...

Microsoft helping OpenStreetMap

smurf — Wed, 24 Nov 2010 16:32:04 +0000

What? Both the Pope and M$ do something that actually makes sense? In the same week?

OK, it's somewhat offset by the Novell-sells-its-patent-pool s**t (they should have contributed them to OIN, dammit), but still

Microsoft helping OpenStreetMap

drag — Wed, 24 Nov 2010 16:22:07 +0000

Microsoft already has their own map product. http://www.bing.com/maps/

I doubt they see openstreetmap as a potential competitor to Google maps.

OpenStreetMap licence?

coriordan — Wed, 24 Nov 2010 16:19:15 +0000

What's the status on OSM's licence?

This is when it becomes important.

Microsoft helping OpenStreetMap

butlerm — Wed, 24 Nov 2010 16:17:33 +0000

I agree, this is outstanding. The Yahoo imagery in the United States is excellent, it is just about five years out of date.

Microsoft helping OpenStreetMap

ccchips — Wed, 24 Nov 2010 15:56:05 +0000

Well, in the process of fighting Google, maybe they'll learn something about freedom. And maybe not. We'll see.

Microsoft helping OpenStreetMap

woooee — Wed, 24 Nov 2010 15:49:01 +0000

There was a typo in the article, it should have read
"Microsoft has announced that it will be contributing to any project that is an alternative to anything Google."

Microsoft helping OpenStreetMap

svena — Wed, 24 Nov 2010 15:40:45 +0000

This is exciting news.

Especially here in Sweden (and the rest of Europe?), the Yahoo imagery is quite poor, so Bing could make for a big improvement in coverage.