Hacker News Reader

Comment by hirako2000 | Hacker News Reader

hirako2000

no. I wish osm would adjust their licensing to prevent reuse of contributions to build non free (as in open) servicing.

It's the permissive vs. free software license debate. In practice there are people who won't be able to use the data for their use case if a non-permissive license is applied.

> In practice there are people who won't be able to use the data for their use case if a non-permissive license is applied

Well, of course. The whole point of a non-permissive license is to.... not permit things. It's frustrating to see people bring this out like some sort of gotcha. I use the AGPL because I think it's important that users have access to the source code the applications they interact with operate on. If people aren't okay with that, then they're completely welcome to not use my code.

(I am somewhat sympathetic to people who think that the GPL's definition of "Combined Work" is too broad—I've personally never viewed the inclusion of dependencies in a dynamic programming language with strong module boundaries like Node, Ruby or Python as creating a "Combined Work"—and I'd happily license my works with a theoretical LAGPL or an Affero-like Mozilla Public License if either of those were available. Unfortunately, the AGPL remains the only game in town for networked software.)

I'm not sure if you're aware, but the data on OSM is covered by their ODbL which is a "share-alike" license. From their FAQ[1]:

    What do you mean by Share-Alike?
    If you publicly use any adapted version of our database, or a partial extraction from it, or works (such as maps) produced from an adapted database, you must also offer that adapted database under the ODbL.

    In other words, if you improve our data and then distribute it, you need to share your improvements with the general public at no charge. A painless way to do that is to contribute your improvements directly back to OpenStreetMap.

    Share-Alike only applies if you distribute what you have done to outside people or organisations. You can do what you like at home, or in your school, organisation or company ... the following section does not apply to you.

It's true that people can use their software and data for non-free services built on non-FOSS code, but people using the data publicly do have to distribute modifications to the data-set under the same license

[1]: https://wiki.openstreetmap.org/wiki/Open_Database_License

Yes, I'm well aware. I think it maybe makes more sense to make this reply to one of the other comments I'm replying to.

I find the distinction difficult to understand. The way I interpret the OSM license, the data is really not freely available to use in a commercial context except as a base layer. If you use the OSM data in combination with any other data sources you have to publish both the resulting dataset and any processing steps or algorithms used to produce the dataset. So that means you cannot combine OSM data with any dataset which cannot be released, and it also means any innovation you develop in processing the data cannot be protected.

I can see why the license was chosen, clearly it's necessary to prevent Microsoft or Apple taking the data and closing it off for their use, but at the same time it seems to make most business models using the data unsustainable. It's a shame from a HN perspective, you'd hope open data would allow for the continual opportunity for the creation of startups to innovate on using the data, locking open competition in the same way that open software has done, but it doesn't seem to be working out that way. Most companies in the area seem to be creators and consumers of base layers, I don't see many doing anything with the geographical database.

Perhaps I'm misunderstanding the terms of the license.

It’s more complex than that. OSM’s licence isn’t a direct parallel of one single software licence, but it has an LGPL-ish permissive aspect for making things from the data (“Produced Works”).

OSM data is used a lot for routing (pretty much all bike navigation apart from Google’s uses OSM data) and increasingly for GIS-type analysis.

When I was looking at it, I got the impression a produced work was something like a t shirt with a map printed on it. Is that wrong? Or any idea where I could read about how the interpret that permission?

> If you make a map from OpenStreetMap geodata and publish it, you may publish the map under any license you like. In ODbL parlance, this is known as a "Produced Work".

https://wiki.osmfoundation.org/wiki/Licence/Licence_and_Lega...

But the second bullet point is:

> However, if you have added to or enhanced our data in order to make the map, you must make those additions publicly available without charge

So the produced work seems to be only a map, derived entirely from OSM data with no combination of data from other sources, except as a wholly separate layer. And any processing you do on the map has to be made available freely.

It's totally unclear to me how you can make a technology business based on the geographical database. For instance, I can't see how you'd read that section and conclude that the bike routing engine mentioned before would be a produced work (the section starts with "If you make a map from OpenStreetMap geodata..."). And if it did the algorithms you used to produce it would have to be openly published and made available to competitors.

> It's totally unclear to me how you can make a technology business based on the geographical database.

Routing engine using pure OSM data can be fully proprietary and in case of using additional data it is enough to make that data available. There is no requirement to open processing tools.

Google can make GoogleWithoutGoogleData Map app using OSM data and keep it fully proprietary.

They could take OSM data for roads and land use and own for shops.

Though if they would start merging OSM data and own they would need to publish own data. It would likely apply in case of taking traffic data and using OSM road geometries to process it.

warning: I am not a lawyer.

Since companies including Apple and Microsoft contribute a lot of data and funding to OSM, what benefits do you imagine would result from preventing them using it?

Reply on news.ycombinator.com