Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
Am 06.05.2014 21:40, schrieb Rob Myers: On 05/05/14 09:16 AM, Simon Poole wrote: We have raised the question of Dynamic Data in a dedicated guideline given that a number of things are not so clear and even while, using the example from the guideline, the occupancy of a parking lot is an observable fact it is questionable if we would want to require that anybody that creates such or similar application has to provide a real time feed of the data on ODbL terms. If a loophole for this case is inserted, expect to see a sudden increase in realtime (and realtime) generation of data. ;-) I hope we (as in the LWG) are not creating the impression that we are trying to assemble as many loop holes as possible, it is more identifying some of the edge cases and trying to document the community consensus on the interpretation. In the case at hand we are simply saying: here is a potential use case that is in a grey area, what do you think? In the case of the parking lot occupancy, dynamic data case: assume we have a proprietary source for that data. The owner of the data will already know which parking lot and where it is by some means, the dataset would clearly have value on its own. One way to combine the data in a useful way would be an app that routes you to the nearest parking lot that still has space (underlying assumption is that the app some kind of live feed of the occupancy data). Does this trigger SA? Does it depend on the internals of the app? Does it depend on how the match OSM parking lot id - proprietary parking lot id is done? Simon signature.asc Description: OpenPGP digital signature ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On 07/05/2014 08:20, Simon Poole wrote: [..] Does it depend on how the match OSM parking lot id - proprietary parking lot id is done ? In this thread, we have seen a few mentions of the implementation as the ultimate factor in discriminating the resulting database between derivative and collective. Isn't that going to result in hypocritical implementations exploiting the letter of the license but not its spirit ? Shouldn't the decision be taken by looking at the data itself and how it combines at a functional level ? For example, would asking whether the external dataset can stand on its own be a relevant question in clarifying a situation ? ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On Tue, May 6, 2014 at 11:20 PM, Simon Poole si...@poole.ch wrote: I hope we (as in the LWG) are not creating the impression that we are trying to assemble as many loop holes as possible, it is more identifying some of the edge cases and trying to document the community consensus on the interpretation. In the case at hand we are simply saying: here is a potential use case that is in a grey area, what do you think? It is difficult to provide clarity and consistency/reliability without also introducing loopholes. Richard Fontana (co-author of GPL v3) has written about this issue in the software copyleft context here: https://lists.fedorahosted.org/pipermail/copyleft-next/2013-April/000639.html Richard's email may be useful reading for folks thinking about how OSM should approach this very difficult problem. Short version: balancing between clarity and loopholes is a well-known problem for lawyers (sometimes called the rules/standards problem), and there is no good answer when trying to write general-purpose legal documents, like laws, constitutions, or copyleft licenses :) I would suggest that OSM is better off creating some clarity (and thereby encouraging more contributions) and risking some loopholes, since the people interested in the loopholes are likely to not contribute back anyway. But that is a judgment call and there is no 100% right answer. Luis -- Luis Villa Deputy General Counsel Wikimedia Foundation 415.839.6885 ext. 6810 NOTICE: *This message may be confidential or legally privileged. If you have received it by accident, please delete it and let us know about the mistake. As an attorney for the Wikimedia Foundation, for legal/ethical reasons I cannot give legal advice to, or serve as a lawyer for, community members, volunteers, or staff members in their personal capacity.* ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On 05/05/14 09:16 AM, Simon Poole wrote: We have raised the question of Dynamic Data in a dedicated guideline given that a number of things are not so clear and even while, using the example from the guideline, the occupancy of a parking lot is an observable fact it is questionable if we would want to require that anybody that creates such or similar application has to provide a real time feed of the data on ODbL terms. If a loophole for this case is inserted, expect to see a sudden increase in realtime (and realtime) generation of data. ;-) If we change realtime to ephemeral, mapping data is ephemeral in geological time. If we stick with Dynamic, well, SQL queries and views are dynamic aren't they? If I want to e.g. combine parking data with littering data for a study or translate it into audio so I can consult it safely while driving, access to the data is useful. This may not amount to a moral imperative, but the debate is currently framed in terms of utility... Deciding which data is and isn't useful (to users, not OSM) has two problems: 1. It requires an Oracle. We cannot know which data is and isn't useful. For example the locations of taxis in London may not seem all that useful after the fact, but: http://wiki.openstreetmap.org/w/index.php?title=Merchandiseoldid=1838 2. People will always push to avoid the license, and any exception will be abused. That isn't a reason to not add an exception or clarification, but it is a reason to be wary of pressure for them. - Rob. ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
Am 05.05.2014 06:38, schrieb Rob Myers: .. But the license doesn't exist to collect data for OSM. .. True, but our immediate, admittedly egoistic, interest is that we are free to use any improvements (in a wide sense of the word) to OSM data and that derivatives of OSM remain free. It exists to ensure that all the users (or in the terms of the license, all its recipients if you Use it Publicly) of that data, in combination with whichever other data and in whatever form and wherever they encounter it, are free to use it. The licence is substantially more restrictive in what it effects (original OSM data and derivatives of it) than you are portraying it. In particular it explicitly allows combination with other data without effecting the legal status of such. IF the ODbL had the effect you attribute to it, then we would really have a problem (and likely OSM usage would go through the floor). Simon signature.asc Description: OpenPGP digital signature ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
2014-05-05 14:05 GMT+02:00 Tobias Knerr o...@tobias-knerr.de: *And share-alike only applies to what we collect.* Let me first say that this is a brilliantly clear way to put it. I like this a lot. I believe this is somehow more limiting than what we actually might want. E.g. we don't collect traffic data, but if there was a company which used our data as basemap and associated average speeds for time spans to our graph (e.g. automatically from the analysis of their users / smartphones) I think we would be interested to get this data to improve our routing. cheers, Martin ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On 05/05/2014 16:32, Martin Koppenhoefer wrote: 2014-05-05 14:05 GMT+02:00 Tobias Knerr o...@tobias-knerr.de mailto:o...@tobias-knerr.de: *And share-alike only applies to what we collect.* Let me first say that this is a brilliantly clear way to put it. I like this a lot. I believe this is somehow more limiting than what we actually might want. E.g. we don't collect traffic data, but if there was a company which used our data as basemap and associated average speeds for time spans to our graph (e.g. automatically from the analysis of their users / smartphones) I think we would be interested to get this data to improve our routing. Usefulness to Openstreetmap is orthogonal to the ODBL. That traffic data might relate to way identifiers, but it does not improve or even modify Openstreetmap data - so I fail to see how Openstreetmap might lay claim on a map showing traffic over Openstreetmap highways. ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
Hi, On 05.05.2014 16:39, Jean-Marc Liotier wrote: I believe this is somehow more limiting than what we actually might want. E.g. we don't collect traffic data, but if there was a company which used our data as basemap and associated average speeds for time spans to our graph (e.g. automatically from the analysis of their users / smartphones) I think we would be interested to get this data to improve our routing. Usefulness to Openstreetmap is orthogonal to the ODBL. That traffic data might relate to way identifiers, but it does not improve or even modify Openstreetmap data - so I fail to see how Openstreetmap might lay claim on a map showing traffic over Openstreetmap highways. True but the use case sketched here went far beyond simply displaying an overlay; this use case was about snapping speed recordings to OSM street data to find out which street the recording was for in the first place, thereby creating a derivative database. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On 05/05/2014 16:47, Frederik Ramm wrote: the use case sketched here went far beyond simply displaying an overlay; this use case was about snapping speed recordings to OSM street data to find out which street the recording was for in the first place, thereby creating a derivative database. In data modeling terms, displaying a speed data overlay requires snapping speed recordings to OSM street data (giving ways a 'speed' attribute) - both operations are complementary, not alternative. According to your message, a graphical rendering of Openstreetmap ways and traffic data mashed up in a browser by joining on an OSM ID implies a derivative database. If it is, then I understand how some may avoid using Openstreetmap data for fear of virality. What the ODBL says: Derivative Database -- Means a database based upon the Database, and includes any translation, adaptation, arrangement, modification, or any other alteration of the Database or of a Substantial part of the Contents. This includes, but is not limited to, Extracting or Re-utilising the whole or a Substantial part of the Contents in a new Database - See more at: http://opendatacommons.org/licenses/odbl/1-0/#sthash.2YNYZwRi.dpuf Derivative Database -- Means a database based upon the Database, and includes any translation, adaptation, arrangement, modification, or any other alteration of the Database or of a Substantial part of the Contents. This includes, but is not limited to, Extracting or Re-utilising the whole or a Substantial part of the Contents in a new Database - See more at: http://opendatacommons.org/licenses/odbl/1-0/#sthash.2YNYZwRi.dpuf Derivative Database -- Means a database based upon the Database, and includes any translation, adaptation, arrangement, modification, or any other alteration of the Database or of a Substantial part of the Contents. This includes, but is not limited to, Extracting or Re-utilising the whole or a Substantial part of the Contents in a new Database - See more at: http://opendatacommons.org/licenses/odbl/1-0/#sthash.2YNYZwRi.dpuf Derivative Database -- Means a database based upon the Database, and includes any translation, adaptation, arrangement, modification, or any other alteration of the Database or of a Substantial part of the Contents. This includes, but is not limited to, Extracting or Re-utilising the whole or a Substantial part of the Contents in a new Database Collective Database -- Means this Database in unmodified form as part of a collection of independent databases in themselves that together are assembled into a collective whole. A work that constitutes a Collective Database will not be considered a Derivative Database. - See more at: http://opendatacommons.org/licenses/odbl/1-0/#sthash.2YNYZwRi.dpuf Collective Database -- Means this Database in unmodified form as part of a collection of independent databases in themselves that together are assembled into a collective whole. A work that constitutes a Collective Database will not be considered a Derivative Database. According to these definitions, because no alteration of Openstreetmap data takes place, adding a 'speed' attribute to Openstreetmap ways produces a Collective Database, not a Derivative Database. Therefore Share Alike does not apply. Collective Database -- Means this Database in unmodified form as part of a collection of independent databases in themselves that together are assembled into a collective whole. A work that constitutes a Collective Database will not be considered a Derivative Database. - See more at: http://opendatacommons.org/licenses/odbl/1-0/#sthash.2YNYZwRi.dpuf ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On 05/05/2014 17:26, Martin Koppenhoefer wrote: no, this was not about overlaying 2 graphical layers but about joining the data into one layer (necessary I guess, in order to perform routing). [..] Usage may be different, but the data is the same: ways with an hypothetical 'speed' attribute added to them in the persistent database of your choice. Whether you use that joined data to perform Dijkstra stunts or just render it graphically does not change its nature. In either case, no Openstreetmap data is altered in any way - only extended thus meeting the definition of a Collective Database. To give a counter-example, if the speed data was used to guess maxspeed=* or change some of the highway=* classification according to detected speeds, then Openstreetmap data would be altered and there would be a Derivative Database. ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
2014-05-05 17:42 GMT+02:00 Jean-Marc Liotier j...@liotier.org: Usage may be different, but the data is the same: ways with an hypothetical 'speed' attribute added to them in the persistent database of your choice. Whether you use that joined data to perform Dijkstra stunts or just render it graphically does not change its nature. In either case, no Openstreetmap data is altered in any way - only extended thus meeting the definition of a Collective Database. The requisite for a collective Database is that the parts are independent, which they aren't I think when you add information to our graph. Maybe one has to look into the details of an actual case in order to see whether the dbs are independant or not. IMHO an extension of OSM data is already an alteration. If your point of view was correct, users would hardly ever have to comply to share alike provisions. cheers, Martin ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
While I think the case of the traffic data is interesting, it really very much depends on implementation details if and when a derivative DB might be created. For example if weights were calculated from the data and associated directly with OSM ways then likely you would have a derivative DB, but the question would still remain if the weights are actually useful information. But if you used the coordinates of the OSM way to do a search in the traffic database and use that on the fly, you likely would not have an issue. Definitely there are dozens of ways to skin this particular cat. And I would note again as I've pointed out before, my interpretation of the ODbL is that only the data in derivative database is what is subject to the ODbL not the original data source. We have raised the question of Dynamic Data in a dedicated guideline given that a number of things are not so clear and even while, using the example from the guideline, the occupancy of a parking lot is an observable fact it is questionable if we would want to require that anybody that creates such or similar application has to provide a real time feed of the data on ODbL terms. Simon signature.asc Description: OpenPGP digital signature ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On Mon, May 05, 2014 at 05:42:37PM +0200, Jean-Marc Liotier wrote: On 05/05/2014 17:26, Martin Koppenhoefer wrote: no, this was not about overlaying 2 graphical layers but about joining the data into one layer (necessary I guess, in order to perform routing). [..] Usage may be different, but the data is the same: ways with an hypothetical 'speed' attribute added to them in the persistent database of your choice. Whether you use that joined data to perform Dijkstra stunts or just render it graphically does not change its nature. In either case, no Openstreetmap data is altered in any way - only extended thus meeting the definition of a Collective Database. I cannot imagine any more wrong interpretation. ODbL was created just to stop activities like this. with your interpretation you could easily create a private fork: just extend osm data with tags like myproject:name/myproject:maxspeed/... ridiculous. so, for the avoidance of doubt I wanted to say that is definitely a derivative db (from information provided) and not a collective db. michal -- michal palenik www.freemap.sk www.oma.sk ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On 03/05/14 08:51 AM, Michael Collinson wrote: Geocoding: So I have to share a patient's medical record because it is geocoded against OSM? Who with? Dynamic Data: So if I use OpenStreetMap car park location data, I have to share the real-time occupancy data? Who with? Algorithmic transformations: So I thought of this clever idea to pre-format OSM data for fast loading into my game. Now I have to share my that or my algorithm? Who with? General maps: I want to use OSM to show locations of restaurants on my restaurant review site. Now I have to share the reviews? Who with? *And share-alike only applies to what we collect.* But the license doesn't exist to collect data for OSM. It exists to ensure that all the users (or in the terms of the license, all its recipients if you Use it Publicly) of that data, in combination with whichever other data and in whatever form and wherever they encounter it, are free to use it. If that leads to patients having better access to their medical data, people being able to find somewhere to park, players of games being able to maintain and modify them creatively to build communities around them and drive sales, and people being able to check the actual rankings of the restaurants they're being directed to that's definitely a win for Open Data. - Rob. ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
[OSM-legal-talk] The edges of share-alike on data Re: Attribution
I've renamed the subject because it has gone way off topic, but I wanted to come back on Tobias' comment because it struck a chord and I would like to share a personal research topic. I am curious to evolve the idea further to see if there is any positive value. Open data is a different animal to software source code and highly-creative works and I suspect it will a few more years yet until we understand it all fully. I personally see this unwanted data is an underlying theme under many of the issues the LWG has been looking at under the Community Guidelines process :- Geocoding: So I have to share a patient's medical record because it is geocoded against OSM? Dynamic Data: So if I use OpenStreetMap car park location data, I have to share the real-time occupancy data? Algorithmic transformations: So I thought of this clever idea to pre-format OSM data for fast loading into my game. Now I have to share my that or my algorithm? General maps: I want to use OSM to show locations of restaurants on my restaurant review site. Now I have to share the reviews? And so on. Now many of these issues may be resolved, and in some case have been resolved, in other ways which remain within the scope of the current ODbL version. But a very simple way of dealing with everything in one go is to say: *The OpenStreetMap project collects long-lived geospatial data as a set of intelligently or machine-made physical observations only.* [Wording needs improving!] And then to say: *And share-alike only applies to what we collect.* Again, it just a research topic. I see it as benefiting the OpenStreetMap project enormously but at the same time potentially debasing the whole concept of share-alike for the wider open data community ... perhaps those restaurant reviews should be shared? Mike On 30/04/2014 23:35, Tobias Knerr wrote: On 30.04.2014 19:37, Rob Myers wrote: On 30/04/14 03:18 AM, Tobias Knerr wrote: But we have to judge a license based on its actual effects, not the original intention. What annoys me, for example, is when we require people to publish data that we wouldn't even want if they offered it. The users of the data may want it. The license exists to benefit them, not (just) OSM. If the actual effects worked against this then yes there would be a problem. I think there is quite a bit of data that will, with high likelihood, never be of use to anyone. That's especially true for byproducts of the creation of a produced work. But your argument about also shows that there are mappers who ask for a lot more than just giving data back when you fix things. Thus it would be foolish for a data consumer to assume they only have to follow that spirit, as much as I wish that was enough. ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk
Re: [OSM-legal-talk] The edges of share-alike on data Re: Attribution
On Sat, May 3, 2014 at 11:51 AM, Michael Collinson m...@ayeltd.biz wrote: Open data is a different animal to software source code and highly-creative works and I suspect it will [be] a few more years yet until we understand it all fully. Sure. Of course, we are part of why it is a big deal now, and we are also part of why it is still evolving. It should be no surprise that Old Model Entities have a hard time keeping up with us. :-) [ ... ] Again, it just a research topic. I see it as benefiting the OpenStreetMap project enormously but at the same time potentially debasing the whole concept of share-alike for the wider open data community ... perhaps those restaurant reviews should be shared? One benefit of the community guidelines, and the recognition of them by ODC-ODbL[1] is that they apply to our community, and not to another Open Data project community unless they deliberately adopt them. Another benefit is that the community guidelines are trivially easy for the OpenStreetMap community to amend and adapt compared to waiting for an updated version of ODbL. That gives us the ability to revise the wording of the community guidelines should that be required to adapt to our changing world. That's not a hypothetical situation; we're changing it. Super to have your thoughts on this Mike. [1] http://opendatacommons.org/faq/licenses/#What_is_8216Substantial8217 ___ legal-talk mailing list legal-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/legal-talk