Jump to content

Commons:Deletion requests/Data talk:Kuala Lumpur Districts.map

From Wikimedia Commons, the free media repository
This deletion discussion is now closed. Please do not make any edits to this archive. You can read the deletion policy or ask a question at the Village pump. If the circumstances surrounding this file have changed in a notable manner, you may re-nominate this file or ask for it to be undeleted.

Data talk:Kuala Lumpur Districts.map

[edit]

The following Data files are nominated as they are not CC0. Data media must be verifiably CC0. The data source for this map is under ODbL, see https://www.openstreetmap.org/copyright, which requires attribution and so is not reusable in a CC0 derived work. This DR is raised on a talk page as a work around as Data namespace files have no system for deletion. Refer to Phab:T178051.

  1. Data:Moscow Central Districts.map (based on the CC-0 work from Russian Wikivoyage, no evidence of the opposite was provided)
  2. Data:Singapore City Centre Districts.map
  3. Data:Madrid Inner Districts.map
  4. Data:Dubai Districts.map
  5. Data:Singapore Districts.map
  6. Data:Munich Districts.map
  7. Data:Kuala Lumpur Districts.map
  8. Data:Amsterdam Districts.map
  9. Data:Milan Districts.map
  10. Data:Saint Petersburg Districts.map
  11. Data:Cochabamba Comunas.map
  12. Data:Copenhagen Districts.map

This list was generated on a simple search for OSM as a source, there may be many other Data files with the same issue of not being based on CC0 source datasets. Unfortunately the JSON source files do not appear searchable using the conventional "insource:" parameter.

--Fæ (talk) 10:49, 12 October 2017 (UTC)[reply]

Hello Fæ, it was a mistake to state OSM as a data source. No data from OSM was used to create those district borders. I will delete the reference to OpenStreetMap now, since it is not correct.--Renek78 (talk) 11:46, 12 October 2017 (UTC)[reply]
How were they created? --Fæ (talk) 12:02, 12 October 2017 (UTC)[reply]
I used either Geojson.io or JOSM to draw the maps. No import from OpenStreetMap data whatsoever.--Renek78 (talk) 12:55, 12 October 2017 (UTC)[reply]
The maps used by Geojson.io rely on OSM, so if traced from those backgrounds still require attribution. JOSM is by its nature intended to start with OSM data, but I guess other sets may be imported. The boundaries are pinned on sets of imported data, even if as the author you traced these out by hand, you would have had to trace some other map or start with another dataset as nobody just guesses as to district boundary lines. If we cannot verify that the source dataset is CC0, then significant doubt must remain. --Fæ (talk) 13:13, 12 October 2017 (UTC)[reply]
The data file is only a boundary. There is no OSM background in these data. --Alexander (talk) 15:16, 12 October 2017 (UTC)[reply]
Strong oppose. The uploader states that OSM was not used, and any reference to OSM has been removed from the files. Indeed, OSM copyright would only hold if the whole map boundaries were extracted from OSM and pasted into the Data file on Commons. This was not the case, because all maps in question depict Wikivoyage districts, which are invented and decided upon by the Wikivoyage community. Such district boundaries do not exist at OSM. Therefore, they are essentially drawn by hand. The author might have perused OSM as well as other maps when drawing the district boundaries, but none of the boundaries are taken from any external source.
@Ymblanter, WhatamIdoing. --Alexander (talk) 12:15, 12 October 2017 (UTC) Moved from the talk page of the file--Ymblanter (talk) 14:56, 12 October 2017 (UTC)[reply]
Futher comment: someone here claims that tracing map boundary from the background creates a derivative work and requires attribution. This is not true. If I read a book and write a review on this book, I definitely rely on the book itself, on its plot and character. Nevertheless, I create an original work, which is not a derivative and does not require any consent from the author or publisher of the book.
Another example. As a non-native speaker, I often refer to dictionaries when writing here. I definitely rely on what I read in the dictionary, but I do not reproduce any part of the dictionary, and therefore I do not have to provide attribution.
Now, if I open OSM and click on a random point, I get the coordinates, two numbers which are trivial and not copyrightable. If I put tens or hundreds of such numbers together, this does not change anything. The OSM attribution would only be necessary if I export individual boundaries from OSM and compile them. That is not the case here, and we never do that on Wikivoyage. --Alexander (talk) 15:16, 12 October 2017 (UTC)[reply]
The statement on derived works is incorrect. If you trace any copyrighted map, then the tracing is a derived work. It makes no difference if that tracing is done digitally, even by selecting a series of single points along a map line. If you search through the deletion archives, there are plenty of past examples of these types of derived tracings being deleted.
The only scenario I can imagine of an uploader creating a map of (political) boundaries is if that dataset or image was derived from a preexisting map or database of geolocation data. It may well be that there are CC0 boundary maps released by government bodies or agencies, however in my experience most of these types of datasets have an attribution requirement (e.g. the default for UK Government open data) even if the data is released for free reuse. Again, the source needs to be verified. This may well severely limit what maps can be hosted on Commons, however the decision to enforce CC0 on Commons for the Data namespace was arbitrary and could be revisited with a new proposal to allow for CC-BY-SA datasets like OSM. --Fæ (talk) 15:38, 12 October 2017 (UTC)[reply]
To your first statement: I would like to see a legal (WMF or court) decision behind this.
To your second statement: we are not creating maps of political boundaries. We are not creating maps of any boundaries other than our own Wikivoyage boundaries. There is no source where such boundaries can be derived from. --Alexander (talk) 17:04, 12 October 2017 (UTC)[reply]
Fae is simply being scrupulous. The case law for database copyrights like this is not well developed, and Commons (specifically, and WMF in general) have an obligation to be err on the side of caution. That said, it seems to me that these datasets do not contain any copy of or modification of the data from OSM. A visual representation of the OSM data was used to guide the generation of these unique data, but I'm not convinced that counts as using the underlying OSM data as a "source". Powers (talk) 21:24, 12 October 2017 (UTC)[reply]
Keep Per Alexander comment: "Wikivoyage districts, which are invented and decided upon by the Wikivoyage community. Such district boundaries do not exist at OSM.". - Offnfopt(talk) 16:36, 12 October 2017 (UTC) (See updated comment below)[reply]
Sorry, that doesn't add up. Wikivoyage does not carve up the world into new regions, because that would make it incomprehensible to users. These districts are identical to normal political Boundaries like "Amsterdam", so come from somewhere. If they were fantasies, they should be deleted as out of scope. --Fæ (talk) 06:47, 13 October 2017 (UTC)[reply]
Wikivoyage does exactly that, see Wikivoyage:Geographical_hierarchy, and anything used in at least one Wikimedia project can not be out of scope per COM:INUSE. --Alexander (talk) 07:37, 13 October 2017 (UTC)[reply]
The page you link to indicates that districts are based on existing boundaries. It does not say Wikimedians invent them based on a blank sheet of paper. Without a real explanation rather than obfuscation, these files must be deleted per the precautionary principal. --Fæ (talk) 07:48, 13 October 2017 (UTC)[reply]
The page does not say this. It says But the others – continental sections, regions, and city districts – have fuzzier boundaries and definitions. How, then, do we decide where to define them? and later The legal divisions in the geography of the world – nations, provinces, and cities – don't necessarily make for reasonable travel divisions. When legal boundaries are used, there is no need to re-draw them. We can directly call such boundaries from OSM via Wikidata. An example is available in the Prague article.
What you see here are city districts invented by Wikivoyage. For example, Data:Moscow Central Districts.map is based on districts of Moscow that were traced following an extensive discussion here. If you find any administrative boundaries in these districts, or in the pertinent discussion, please, let me know. Otherwise, your claims remain unjustified. --Alexander (talk) 08:12, 13 October 2017 (UTC)[reply]
I am not going to analyse a discussion in Russian, neither do I have to in order to definitively prove that the statements made in this DR by several keen Wikivoyage contributors are blatantly misleading and untrue.
Closing administrator please note - Anyone can examine the OSM map here of Copenhagen and compare the Administrative Boundary of Copenhagen Municipality in OSM with the boundary of Frederiksberg in the Commons map of Data:Copenhagen Districts.map. My detailed zoomed in examination of the data points describing the North East part, which is very kinked and does not exactly follow any roads or other particular boundary that might be shown in, say, a satellite photograph, demonstrates that they are near as damn identical, or at least that one has been directly traced from the other. In fact the line that cuts through the children's playground has several datapoints clustered below 100m and is the most obvious give away that the Commons version is a derived work. This map is not CC0, and even my simple analysis would stand up in court as a violation of the right of attribution required by OSM as the Commons map is unambiguously a derived work.
It may be that the community at Wikivoyage are less aware of how copyright works, or maybe you think that creating misdirection and tangents about how these maps were created will make the problem go away, but if you want this problem solved permanently, you should lobby to get the way that the Data namespace works on Commons changed to allow attribution rather than attempting to deliberately mislead Wikimedia Commons community about what is going on. --Fæ (talk) 15:12, 13 October 2017 (UTC)[reply]
Addendum: Here are three matching data points, cut and paste from the Commons Data file and the OSM map, these are part of the NE corner described above which define an obvious right-angled corner to the political boundary and for which the two end points are entirely arbitrary and unrelated to any physical feature, such as a road:
  • 55.6977123/12.5383257 vs. 55.69771/12.53833
  • 55.6968181/12.5382111 vs. 55.69682/12.53821
  • 55.6966702/12.5366677 vs. 55.69667/12.53667
To give context these are within an area of about 200m across and the correspondence is near to 5 decimal places, i.e. the same locations to around 1m or 2m. It is not in the least bit credible to call this a coincidence. --Fæ (talk) 15:50, 13 October 2017 (UTC)[reply]
First, you have to read the discussion in Russian in order to understand where these districts come from, how they were created, and why. As long as you do not want to do this, I cross out Data:Moscow Central Districts.map. I traced these boundaries myself and released them at Russian Wikivoyage under CC-0, see here and other template pages of this kind.
Second, the north eastern part of the Frederiksberg boundary has no kinks, and it does not follow any administrative boundary. It sometimes coincides with the railway line, and in other places it is completely arbitrary. I do not even understand what you are talking about.
Third, you refer to three matching geo-points as an evidence of the derivative work. FYI, any geo-point coincides with some geo-point on some other map. Administrative boundary of Copenhagen is not copyrighted. Otherwise, you should probably request its deletion from OSM, because a very similar boundary is also available on Google maps.
Fourth, you did not provide any serious legal argument to support your statement that tracing map boundary using OSM creates a derivative work.
Fifth, your blank accusations of the Wikivoyage community are grossly offensive and prove that your sole intention is damaging sister projects, not license concerns, which in fact you fail to prove. --Alexander (talk) 23:02, 13 October 2017 (UTC)[reply]
You do not have permission to strike text in my nomination, so don't do that. I have to presume that the source you are now stating you traced the Moscow map from was Open Street Map, please state if you used a different source. In that case the nomination stands as stated, as it is a derived work and is not CC0. My understanding of copyright law with regard to tracing of copyright works is based on multiple past deletion requests on this project, if you want to overturn that established understanding, the onus is on you to present a case, per the precautionary principle. I have not spent time checking the precise data points for that map, but if you imported any of the data and rounded up the digits, rather than doing your tracing by hand, please state so.
With regard to untruths, the understanding above was "The uploader states that OSM was not used", and votes were made in this DR fully relying on that assertion. The precise opposite was later stated by the uploader, but only after I invested significant time in a detailed analysis to force that statement to be made. Both statements cannot be true at the same time. Certainly both Wikivoyage and Wikimedia Commons are damaged by deliberate copyright violations, however the actual party being damaged is the copyright holder, i.e. Open Street Map contributors. My action in raising this for scrutiny and discussion is not a healthy reason to attack me. Thanks. --Fæ (talk) 23:36, 13 October 2017 (UTC)[reply]
I think it's your task to provide evidence for copyright violation on my side. I have to presume that the source you are now stating you traced the Moscow map from was Open Street Map means I can nominate all your uploads for deletion because "I have to presume that Fæ's uploads are derived from copyrighted works".
The boundaries of Moscow districts have been drawn by hand based on my knowledge of the city and the pertinent discussion at Russian Wikivoyage. We have long-standing experience of drawing "static" city and regional maps by hand, many of them created in the period before OSM had any decent coverage.--Alexander (talk) 08:40, 14 October 2017 (UTC)[reply]
I withdraw my earlier claim, because data points in Data:Moscow Central Districts.map are not the same as in my CC-0-licensed boundaries available on Russian Wikivoyage. --Alexander (talk) 13:31, 14 October 2017 (UTC)[reply]
Alexander since Fæ provided no evidence in his initial request I had good faith at what you were saying but I overlay'd the boundaries for the Zuid and the Noordshown neighborhood on the Amsterdam map and it does match up pretty well with the administrative boundaries shown in OSM. I do see some minor variances in the line from time to time, so it does appear the boundaries may have been traced but in the case of these two neighborhoods it is not a invention of wikivoyage. You can see the overlay here: https://imgur.com/a/RrmjG (dark line is the commons data, orange line is OSM/CBS data)
That specific boundary in OSM originally came from Centraal Bureau voor de Statistiek as part of their StatLine opendata initiative which does have the stipulation of Creative Commons Attribution (CC BY). Now since there is slight variance since it was traced I don't know the legalities since they are GPS coordinates and not a image, so the coordinates don't match up 100% the same since it was traced. I don't fault you not knowing this since you aren't the one to upload the map data but as I said above in the case of these two neighborhoods they weren't the invention of wikivoyage. Now there very well may be areas that wikivoyage defined, but the above two cases do add doubt that all of the areas were a invention of wikivoyage. I'll be receding my above vote until we get some input from someone with knowledge of the legalities regarding the tracing. - Offnfopt(talk) 19:46, 13 October 2017 (UTC)[reply]
Fæ, Offnfopt, as mentioned above the OSM map was used as a visual reference to draw the boundaries. If the official district borders of Frederiksberg are also used in Wikivoyage then of course I draw as close as possible to the official district boundary. By the way: This boundary for sure was an import from the city of Copenhagen. Shouldn't we check the license there to be perfectly compliant?--Renek78 (talk) 21:18, 13 October 2017 (UTC)[reply]
 Comment Renek78 I'm no expert, but what I gather is since it was traced from OSM data it seems these count as derivative work, so would need to be licensed under a compatible license. I think it would be good to use this experience for us to see if we can get wikimedia to allow other licenses under the Data namespace. If we were able to use other licenses other than just CC-zero then this information would be able to stay and I think everyone would be happy. - Offnfopt(talk) 22:17, 13 October 2017 (UTC)[reply]
Sounds good! Let's hope for the best.--Renek78 (talk) 22:28, 13 October 2017 (UTC)[reply]
Offnfopt, I am not sure what you are talking about. Some district boundaries do match administrative boundaries, but each district is original and is not contained in OSM or in any other map available so far. Otherwise, it would not make sense to trace this district, we could simply use the OSM boundary. As for the administrative boundaries, you will find same boundaries of Amsterdam, Copenhagen and any other city at OSM and Google, one under free license and the other one not. How do you infer that this boundary should be attributed to OSM under CC-BY-SA-2.0 and not to Google? --Alexander (talk) 23:08, 13 October 2017 (UTC)[reply]
  •  Delete It would be a misuse of Wikimedia Commons to host the above example as CC0 when it is at minimum a derived work of a CC-BY-SA map and based on the data appears to be a direct data import with a minor bit of rounding up of digits. As we are seeing misdirection about this one example, we must insist that all other uploads by the same account (Renek78) require a full explanation of how they were created. "I made it up one day" is not credible and is easily debunked. --Fæ (talk) 15:15, 13 October 2017 (UTC)[reply]

@CKoerner (WMF), the opinion of the Discovery team would be highly appreciated. Someone has to explain Commons, once and forever, that geoJSON data created using the visual editor are under CC-0. Otherwise, the Data: namespace on Commons becomes useless for storing any map data. --Alexander (talk) 15:16, 12 October 2017 (UTC)[reply]

Won't be legal team a better fit for giving copyright related statements? --Base (talk) 16:23, 12 October 2017 (UTC)[reply]
I'm asking around to see what's going on. CKoerner (WMF) (talk) 16:45, 12 October 2017 (UTC)[reply]
@Base, if you know how to reach them, please do. Thank you! --Alexander (talk) 17:04, 12 October 2017 (UTC)[reply]
I've just created phab:T178210 hopefully it will draw some attention. Good way to contact them was by email I believe, but I do not fancy using it atm. --Base (talk) 23:39, 13 October 2017 (UTC)[reply]
@Base, thank you. Previously, WMF legal advised us that they only give legal advice to WMF, not to the editors. Therefore, I deemed the Discovery team the most natural point of contact in this situation. --Alexander (talk) 12:54, 14 October 2017 (UTC)[reply]
Just a quick head's up. There's a meeting scheduled with the WMF Legal team tomorrow (16 Oct) to discuss. CKoerner (WMF) (talk) 16:28, 16 October 2017 (UTC)[reply]
 Keep if that were up to me Syced (talk) 06:39, 13 October 2017 (UTC)[reply]
 Keep, because no map data from OpenStreetMap was used and it was a s...load of work ;) --Renek78 (talk) 07:18, 13 October 2017 (UTC)[reply]
It is worth noting your canvassing on Wikivoyage: "All of them are going to get deleted soon, because the district borders were traced using OpenStreetMap as a background image." (diff) By your own statement that makes these files derived works and legally not CC0. It is a pity that you are not so direct in explaining how you created these files on Wikimedia Commons. --Fæ (talk) 16:26, 13 October 2017 (UTC)[reply]
 Keep, I do not find Fae's arguments convincing.--Ymblanter (talk) 13:16, 13 October 2017 (UTC) Withdrawn the keep vote per discussion below.--Ymblanter (talk) 10:17, 14 October 2017 (UTC)[reply]
  •  Comment It seems clear that the underlying maps are from OSM and annotated as such. To claim that OSM data wasn't used is unconvincing. On the other hand, why can't OSM data be used to create derivative works, as permitted by the CC BY-SA‎ license? How does the copyright attribution requirement vitiate its free license, any more than our {{Attribution}} license, for example? JGHowes talk - 18:20, 13 October 2017 (UTC)[reply]
 Keep, because the location of a political boundary is a simple statement of fact, not a copyrightable creative work. Nations tend to use soldiers and guns to protect their borders against attempts to be "creative" in moving them. That said, this data should be moved back to en:voy for the reasons stated in phab:T154908. Hosting page-essential content across three wikis (Wikivoyage, Wikidata, Commons) for the same article makes the guide more difficult for new users to maintain and risks breaking Wikivoyage destination guides if any of these external wikis delete components which are needed to correctly describe a destination. Wikivoyage needs to be self-contained. K7L (talk) 12:48, 16 October 2017 (UTC)[reply]
The Court... found that compilations of factual data are copyrightable if the author's selection, coordination, and arrangement of the information depicted are sufficiently creative to qualify as a creative expression. I find it difficult to imagine how a political boundary can be "sufficiently creative". --Alexander (talk) 13:33, 16 October 2017 (UTC)[reply]
This is precisely how Open Street Map is created, with users using their own creative judgement on where to place nodes and create connectors. Hence attribution is required and why all actual legal reuses you can find, like apps, include copyright statements of attribution. 13:44, 16 October 2017 (UTC)
  •  Delete Tracing an OSM-map makes the data a derivtive of OSM, even if no data has been directly exported from OSM. However, if that derivative reaches a threshold to warrant OSM to claim any kind of copyright (and demand attribution) over it is the question at hand here, not whether or not the districts are made up by our editors. I believe these data files weere traced over OSM maps, and therefore followed the lines (streets, land, areas etc.) on these OSM maps. The traced data would therefore be dervivative of the maps made avalible by the OSM, and require atribution and be licened on the same way as OSM. --Jonatan Svensson Glad (talk) 20:19, 13 October 2017 (UTC)[reply]
    •  Still delete Striking the above reason for deletion, since it appears more than normal tracing was done from OSM, and rather obvious import (read stealing) from OSM was done, and then lied about by the Wikimedia users, as apologised for below in the discussion. This is not a question wheter or not it was traced nomore, but instead about if the data imported from OSM (without proper attribution - which is not possible or "allowed" in the data space) are above COM:TOO, and I still maintain that it does. --Jonatan Svensson Glad (talk) 22:53, 14 October 2017 (UTC)[reply]
  •  Comment A friendly coment to my admin colleagues, remember that we d not care about consenus (or !votes) here on Commons. We care about copyriht laws, and our policies such as COM:L and COM:PRP. It is up to the users that want to keep these files to prove that they are different enough from the traced maps to not warant attribuion (see COM:TOO). --Jonatan Svensson Glad (talk) 20:22, 13 October 2017 (UTC)[reply]
    I guess we are waiting for the comment of the legal, but Fae apparently wants it to be deleted before the legal has a chance to comment. If this gets deleted, likely all Wikivoyage contours on Commons will have to be deleted, meaning that on the Wikvoyage we will likely stop using Commons and go to local uploads.--Ymblanter (talk) 21:24, 13 October 2017 (UTC)[reply]
    • Such considerations, of what other projects may or may not do, should be secondary to anything to do with the copyright status and discussion related to it on this page. Consequences (good or bad) should not be considered at all if such thoughts in anyway are in violation to our policies about COM:L or copyright laws. --Jonatan Svensson Glad (talk) 21:41, 13 October 2017 (UTC)[reply]
      I agree with that. I, however, see for a long time that many Commons administrators interpret the precautionary principle like where they are unable to understand the keeps arguments should be deleted, which obviously makes me pretty much disappointed in Commons. I stopped transferring free files from the projects to Commons and advise against such transfer everybody who asks me. This is just another example.--Ymblanter (talk) 21:52, 13 October 2017 (UTC)[reply]
I am not God of Commons, neither did I have any say on the license rules for Data namespace. How about pushing for honesty from Wikivoyage uploaders and a change from CC0 to attribution allowed, rather than pretending I am guilty of something? --Fæ (talk) 21:46, 13 October 2017 (UTC)[reply]
I do not know, I did not draw the maps. However, so far you argument is "The contours have the same coordinates as OSM, consequently they are derivatives of OSM". You can say as well they have the same coordinates of Google maps and therefore they are proprietary. Administrative borders are administrative borders, and they show in the same way on every map. Most (but not all) of Wikivoyage region borders coincide with some administrative borders, and as such, are the same as on OSM, on Google maps, or if I just walk around with a GPS and write coordinates on a piece of paper.--Ymblanter (talk) 21:55, 13 October 2017 (UTC)[reply]
I limited my comments to pure political boundaries, not natural features or where anyone would naturally walk. If you examine the sample data above, this is not even derivative, it is the OSM data directly imported and rounded to 5 decimals. Calling it CC-zero is deliberate copyright violation. Please do not pretend that is a good thing. --Fæ (talk) 22:02, 13 October 2017 (UTC)[reply]
This is up for the creator to comment, not up for me. I agree that importing data is not acceptable, However, political boundaries (in this case, municipal district boundaries) often coincide with natural features including streets, and my above argument stands. I am not quite sure why you assume bad faith towards me, but I do not particularly care at this point.--Ymblanter (talk) 22:08, 13 October 2017 (UTC)[reply]
Ymblanter, I replied above to Fæ already. I used the OSM boundaries as visual reference - just like roads and railway tracks - to create the polygons. By the way: What is so despicable about the ODbL license that it cannot be used here?--Renek78 (talk) 22:28, 13 October 2017 (UTC)[reply]
We do accurate tracing up to the fifth digit, and yes, we do put quite some work into these boundaries. You may also check that OSM and Google show same administrative boundaries not only to the fifth but even to the sixth digit. It does not mean OSM stole their boundary from Google, or vice versa. It simply means that the administrative boundaries are well defined. --Alexander (talk) 23:21, 13 October 2017 (UTC)[reply]
Jonatan Svensson Glad, COM:L explicitly allows the CC-BY-SA-2.0 license. What's the problem then? --Alexander (talk) 23:21, 13 October 2017 (UTC)[reply]
This is a message I want everyone to read. It seems this issue was caused by a misunderstanding of copyright law, which to be fair is understandable. It is a very complicated subject and changes widely based on the type of content.
I think it would be best for us to use this experience to push for more flexibility from the Data name space regarding licenses.
The Data namespace was rolled out only allowing CC-zero contribution, if we were to allow other licenses that are compatible with OSM data we would not have to delete this information and would just have to fix the licensing/attribution information. OSM data is licensed using Open Data Commons Open Database License (ODbL) - {{ODbL OpenStreetMap}}. If we were to allow this license this contribution would be allowed to stay. If we were to allow other licenses like CC-BY and CC-BY-SA in the Data namespace this would also allow use to expand out data set. For instance I noted in a previous comment that some of the initial boundaries used by OSM origonally came from Centraal Bureau voor de Statistiek as part of their StatLine opendata initiative, that data is released under the Creative Commons Attribution (CC BY) license. I hope everyone is on board with this idea and hope we can push to allow this change which would allow us to better take advantage of the information available at OSM to further our own projects. - Offnfopt(talk) 23:09, 13 October 2017 (UTC)[reply]
Offnfopt, I hope you understand that the deletion request is not the right place to discuss policy changes. It is not up to the Wikivoyage community to propose policy changes on Commons. All of us believe that these map boundaries are trivial and thus licensed under CC-0, and we have not seen any serious legal argument to the opposite. Instead, we see blank accusations and the general inability of the Commons community to handle free data, which these map boundaries clearly are. Both CC-0 and CC-BY-SA-2.0 are free licenses. --Alexander (talk) 23:15, 13 October 2017 (UTC)[reply]
Commons can handle free data just fine, however these files currently claim they are CC0. Republish with the verifiable correct license, say by converting to SVG format, and there will be no issue with hosting.
Here's a few past DRs you can review if you want to understand how Commons normally handles tracings and derivatives of maps:
Commons:Deletion requests/Image:Map FFVII (vierge).png
Commons:Deletion requests/Image:Australien--Qld--Townsville+Thuringowa--UreinwohnerStämme--RS03a.jpg
Commons:Deletion requests/Image:Petersburgmetromap2005.png
Commons:Deletion requests/Image:Mapa Europa despres Tractat Westfalia 1648.jpg
Commons:Deletion requests/File:Pepper Lunch Philippine map.JPG
--Fæ (talk) 23:58, 13 October 2017 (UTC)[reply]
I have not found any serious legal argument in any of these previous deletion requests. Some Commons users without real-life identities assert that something should be deleted. This is no different from the ongoing deletion request. --Alexander (talk) 08:47, 14 October 2017 (UTC)[reply]
It's fascinating that you are convinced that the Commons community has been wrong for at least the last decade. Try this specific guidance on tracing maps in the legal FAQ on OpenStreetMap. The community there interprets the law for derived works in the same way. --Fæ (talk) 10:39, 14 October 2017 (UTC)[reply]
The Commons community is not a legal entity. Its opinion and experience in legal matters are no more valuable than, for example, mine.
As for the legal FAQ of OpenStreetMap, I think they describe a different situation. If you use a copyrighted source to trace the full length of a particular street or the whole political boundary, you refer to an object that exists on other maps. Here, we do not refer to any such object. We simply draw polygons. They may or may not be close to objects existing on other maps. --Alexander (talk) 13:07, 14 October 2017 (UTC)[reply]
The truth for this DR is that there is no may or may not, the objects were taken from OSM, and as there was deliberate obfuscation by declaring them own work and CC0, that's stealing from the OSM contributors. Everyone should take that seriously, not fudge around it or obscure it behind hypothetical debate. --Fæ (talk) 22:21, 14 October 2017 (UTC)[reply]
The question here is whether copyright on geo-points can be supported by a legal argument, such as a previous court decision, or any other authoritative source. So far it seems to be merely an opinion of a group of Commons editors, who are not experts in legal matters. As you know, WMF legal was contacted, and it's best to wait for their comment before throwing further accusations. --Alexander (talk) 23:38, 14 October 2017 (UTC)[reply]
  •  Info Apologies everyone, I made a mistake in my earlier detailed analysis of the three example geolocations, as the data I was taking from OSM was getting rounded down by the method I used of examining OSM hyperlink data. Here are the direct comparisons between Renek78's uploaded map of Copenhagen, and then as an extra test Madrid, and the full OSM data (the latter published 5 years ago on OSM):
Copenhagen
Point 1
OSM: 55.6977123, 12.5383257
COM: 55.6977123, 12.5383257
Point 2
OSM: 55.6968181, 12.5382111
COM: 55.6968181, 12.5382111
Point 3
OSM: 55.6966702, 12.5366677
COM: 55.6966702, 12.5366677

Madrid
Point 1
OSM: 40.4234006, -3.7108900
COM: 40.4234006, -3.71089
Point 2
OSM: 40.4234149, -3.7107671
COM: 40.4234149, -3.7107671
Point 3
OSM: 40.4248480, -3.6906559
COM: 40.424848,  -3.6906559
It is irrefutable logic that when Renek78 (talk · contribs) stated in this DR that "No data from OSM was used to create those district borders" and most recently that "I used the OSM boundaries as visual reference - just like roads and railway tracks - to create the polygons", and took action to delete previous visible statements about sources (diff on Madrid map), these were deliberate untruths intended to deceive the Wikimedia Commons community, and probably the Wikivoyage community, that these were maps made by manual "tracing", hence probably derived works. The truth is that they include direct data imports from Open Street Map.
Creating deliberate copyright violations on Commons is known as vandalism. Failing to comply with license attribution requirements should be treated as seriously as any other form of copyright violation. Due to persistently attempting to mislead the community, none of Renek78's uploads to Wikimedia Commons should be trusted and should be deleted as copyright violations unless fully verified otherwise. Thanks --Fæ (talk) 06:21, 14 October 2017 (UTC)[reply]
I don't even know how to use OSM data to draw those Wikivoyage maps. Maybe you can teach me. Okay, have other things to do than discussing about this stuff. Bye.--Renek78 (talk) 07:37, 14 October 2017 (UTC)[reply]
@Atsirlin: : Since, as usual, nobody is interested in trying to understand what happened, the request is likely to be closed as delete. Is there any way we can host the data locally and forget about Commons? (We can still host here maps as images, but I will not even try to relicense them as CC-BY-CA if the creator says they did not import Wikidata).--Ymblanter (talk) 07:49, 14 October 2017 (UTC)[reply]
Yaroslav, yes, we can store map boundaries locally. It won't be as nice as in the Data: namespace here, but we will surely use that option. --Alexander (talk) 08:47, 14 October 2017 (UTC)[reply]
Well, one degree (latitude) is 111 km, meaning 10^{-7} of a degree is 1 cm. No map ever gives this resolution. This likely means that by clicking on any point of the area around the point, OSM (and possibly other maps) will provide the same result, with the precision specified in OSM. I do not know what algorithm is used in OSM, but it means that if one redraws a map just by clicking on the control points (such as street crossings) at OSM (and this is exactly what Renek78 claims to have done), it is just not possible to give a different value. If this is correct, (i) I am inclined to believe Renek78; (ii) we still have a question of licensing (I am not sure how much just clicking on the map makes it a derivative - probably not, but I would like to see a comment from the legal), but it is definitely much more complicated than Fae wants to imply.--Ymblanter (talk) 08:01, 14 October 2017 (UTC)[reply]
@Ymblanter: I find it astonishing that you are allowing yourself to continue to be hoodwinked by blatant deliberate untruth. At the start of this DR it was stated: "I used either Geojson.io or JOSM to draw the maps. No import from OpenStreetMap data whatsoever.--Renek78"
Examining that claim, I have now tested how this would work in geojson.io. It can be discounted as if points were selected using the point tool (not using any import), then the points that would be used to match nodes give 14 decimal figures, not 7. The tool does not "click" to a node in a way that would give you the identical data.
Now taking the JOSM tool, the only way I can see that it would be used would be to import the desired boundaries from OSM (this has been very specifically repeatedly denied as having happened). You then get a direct clone of the OSM data with the exact same decimal precision. I have tested this for the Copenhagen boundary line link, and it would give exactly the same data that Renek78 has released on Commons as "own work". By the way, it is not credible that Renek78 does not know how to do this, it is fundamental to the way JOSM works. Though I am a long term contributor to OSM, this is first time I have used JOSM and it took me about ten seconds to do this test.
I stand by my statement above as unambiguously based on the testable facts. The maps were not traced, they were cloned/copied/imported from the OSM data. Pick your preferred word, it makes no difference to the copyright violation. --Fæ (talk) 10:04, 14 October 2017 (UTC)[reply]
I checked how these particular data points look like in OSM. The geoJSON data were extracted using polygons.openstreetmap.fr with the relation ID 2192363.
Copenhagen
Point 1
OSM (asserted by Fæ): 55.6977123, 12.5383257
OSM (geoJSON): [12.5383256,55.697712299999999]
Point 2
OSM (asserted by Fæ): 55.6968181, 12.5382111
OSM (geoJSON): [12.538211,55.696818100000002]
Point 3
OSM (asserted by Fæ): 55.6966702, 12.5366677
OSM (geoJSON): [12.536667599999999,55.6966702]
Therefore, Fæ's statement on the "direct data imports from Open Street Map" is not justified. --Alexander (talk) 09:36, 14 October 2017 (UTC)[reply]
The tool you have used was not used by the uploader, based on their own statement. As for how I got the data, I did this by editing OSM online using the built-in editor. Exported data will be precisely as I stated and can be reproduced by anyone that bothers to go to OSM or clones the data in JOSM, this being a tool the uploader has stated they use. The fact that the geocoordinates are identical between OSM and Commons cannot be explained any other way than being direct copies.
Making this really simple to check, go to http://www.openstreetmap.org/node/1763583923#map=19/55.69771/12.53833. You will see Point 1 in your list, with the identical numbers as I have "asserted" in the information box on the left. When you are in doubt when faced with two people, only one of whom can be telling the truth, pick the one that provides proof rather than the one that nobody can actually pin down with an explanation. --Fæ (talk) 10:21, 14 October 2017 (UTC)[reply]
I never used JSOM, and I am unable to set it up on my computer. Nevertheless, I find your link convincing, and I agree with your argument. I find it unfortunate that Renek78 imported the data from OSM, but failed to inform the community about it.
However, I still see two major flaws in your reasoning. First, you seem to infer license requirements on individual numbers without considering their threshold of originality. This parallels to license requirements on individual words. Second, and even more importantly, your deletion request is a clear violation of COM:POINT, If you feel that a policy or guideline should be changed, and others disagree, do not attempt to enforce the existing rule with the aim of provoking opposition to it. I am sure no Wikivoyager will ever discuss changes in license requirements when someone is holding a knife at their throat. We will rather use other places to store our data. --Alexander (talk) 13:48, 14 October 2017 (UTC)[reply]
These cases are more than single points, and the analysis I have done is just a sample from those maps. In fact entire map features were imported and deliberately and falsely declared as own work. It should concern everyone that these same data sets are being copied back to Wikivoyage without regard for correct attribution.
With regard to pointy, this DR was created after discussion at the Village Pump which had nothing to do with proving anything about policy. In good faith I have raised the issue of how to use housekeeping tools on Data files in Phabricator, which was my main concern, which is a technical matter not a Commons or Wikivoyage policy one.
Throughout this DR there have been repeated lies about the copyright of works from OpenStreetMap and multiple presumptions of bad faith against me as the nominator. Even now the related thread on the Wikivoyage Pub appears to put the fault for this fiasco down to the stupidity of Wikimedia Commons contributors, rather than the actual actor who has misled both communities. In my view that person should have been blocked by an administrator for deliberate vandalism. I hope that in future DRs where there are clear copyright concerns about files created to support Wikivoyage, that the nominators are not treated as an enemy of Wikivoyage simply for attempting to apply Commons policies and guidelines.
--Fæ (talk) 22:13, 14 October 2017 (UTC)[reply]
Personal squabble. Does not add anything to the DR in terms of if this should be kept or deleted. Collapsed as admin action by --Jonatan Svensson Glad (talk) 23:21, 14 October 2017 (UTC)[reply]
:::::: Fae, you are again making ungrounded accusations and then surprise that people do not take you seriously. Where have you seen that "these same data sets are being copied back to Wikivoyage without regard for correct attribution". Please point us to it. I am a Wikivoyage administrator, and I am not aware of any copying going on. These accusations are unhelpful, both for you and for your standing in the community.--Ymblanter (talk) 22:52, 14 October 2017 (UTC)[reply]
There has not been a single unfounded accusation from me, perhaps you should stop saying that. As for my "standing" I have none and I'm not here to play politician. The action to copy these files to Wikivoyage is as stated in the Pub discussion. Instead of griping at me, go and investigate it. Thanks --Fæ (talk) 23:01, 14 October 2017 (UTC)[reply]
This is actually the second one, after the canvassing accusation. I was asking about "these same data sets are being copied back to Wikivoyage without regard for correct attribution". I remember, however, that arguing with you is rather pointless - we all still remember what happened to you on the English Wikipedia "for numerous violations of Wikipedia's norms and policies", even though you were later unbanned.--Ymblanter (talk) 23:06, 14 October 2017 (UTC)[reply]
Commons is unknown as a highly uncivil project, so I am not surprised. --Alexander (talk) 23:38, 14 October 2017 (UTC)[reply]
FYI, Wikivoyage content is released under CC-BY-SA and thus fully compatible with the OSM license, should the attribution be required. We are waiting for comments from WMF legal and will follow their advice on the license/attribution, but there is no reason not to host these map boundaries on Wikivoyage. If you disagree, you can go to Wikivoyage and discuss this matter. You can also propose administrative actions against the uploader. Wikivoyage is open for anyone who is reasonable and respectful to others. --Alexander (talk) 23:38, 14 October 2017 (UTC)[reply]

On some maps I used JOSM to extract the boundaries from OSM. I am sorry for not being honest from the very beginning. Just wanted my maps to survive. Please accept my apologies - especially to the guys who wasted time on this.--Renek78 (talk) 10:31, 14 October 2017 (UTC)[reply]

@Renek78, could you confirm that Data:Moscow Central Districts.map is based on my district boudnaries from Russian Wikivoyage and has nothing to do with OSM? --Alexander (talk) 12:51, 14 October 2017 (UTC) Upon a visual inspection, I see they are different.[reply]
  •  Comment I think that this deletion request raises several general questions:
i) should CC-BY-SA licenses be allowed in the Data: namespace?
ii) for a map boundary traced from an existing map, how can I verify the map source, so that other users trust it?
iii) for map boundaries (partially) copied from an external source, such as OSM, where is the originality threshold that results in license and attribution requirements?
I think there is no clarity on any of these issues, which puts all map files into the gray zone. I took the liberty to launch a discussion of these general questions at the Village Pump. Regardless of the outcome of this particular deletion request, I believe that one needs clear policies for map geo-data. Otherwise, no map data can be and will be ever stored on Commons. --Alexander (talk) 15:14, 14 October 2017 (UTC)[reply]
You may believe this, however property law allows maps of all types to be copyrighted. This includes maps just showing boundaries. --Fæ (talk) 19:33, 14 October 2017 (UTC)[reply]
Of course, maps can be copyrighted but data that they contain can not be. Ruslik (talk) 19:39, 14 October 2017 (UTC)[reply]
If the data is above COM:TOO data can be copyrighted. If code is copied verbatim, it can be copyrighted. If a database of data is copied it may be protected by database copyright. Even map data can be copyrighted. --Jonatan Svensson Glad (talk) 22:55, 14 October 2017 (UTC)[reply]
Jonatan Svensson Glad, could you advise me on how you define originality threshold for maps? COM:TOO does not even mention map geo-data. This parallels to the questions that I raised in the Village Pump, and your comments there will also be appreciated. --Alexander (talk) 23:14, 14 October 2017 (UTC)[reply]
No it is true that our current guidelines about this deos not mention this, and that is failing I agree. It is my understanding that most data aren't copyrightable in the US, but I also know that SVG code may or may not (discussions have gone both ways on Commons) be copyrightable, since it is a piece of code which dictates where a path should be and is therefore classified as computer code. I see similar things here, where code (data points) been copied from OSM, and I therefore believe this is a similar situation. The boundery data could be made in other ways (not using OSM) and would then get different results (less/more precision, slight variations etc.), and therefore it becomes a verbatim copy of where OSMplaces their data points, and since OSM claims copyright of ther data (you can't i.e. scape their datbse and call it "it is just data")...(more to come later, it is late and I can't formulate myself atm)... But I am open to being proven wrong. --Jonatan Svensson Glad (talk) 23:25, 14 October 2017 (UTC)[reply]
Jonatan Svensson Glad, to my understanding, OSM can not claim copyright on individual geo-points, so the question is how many geo-points build up a licensed object that requires attribution. In my opinion, OSM can claim their copyright on map objects, which are full map boundaries. On the other hand, part of a boundary does not make up a meaningful map object, and the situation may not be as acute as some people here try to present it.
I also deem it essential to build a consistent policy on tracing boundaries from existing maps. --Alexander (talk) 08:25, 15 October 2017 (UTC)[reply]
Unless it can be shown that some of the files listed in the nomination have no more than one or two geocoordinates taken from non-CC0 sources, a discussion about copyright of data points is irrelevant to a decision of whether to keep or delete. --Fæ (talk) 11:42, 15 October 2017 (UTC)[reply]
So you mean that 3 points is the originality threshold. Why that number? --Alexander (talk) 12:21, 15 October 2017 (UTC)[reply]
A number is irrelevant and probably not of any legal meaning. The files nominated have probably copied a magnitude more than any 'small' number anyway. If not, pick out the file you think is below any copyrightable threshold. --Fæ (talk) 12:28, 15 October 2017 (UTC)[reply]
They don't work. Try adding a category to a map. Ref Phab:T155290. --Fæ (talk) 19:46, 17 October 2017 (UTC)[reply]
I see, tnx.--Ymblanter (talk) 19:49, 17 October 2017 (UTC)[reply]
  • Along with highlighting possible copyright problems, creating this category highlights a couple of maps where the Commons uploader is clearly the same person that created the data on OpenStreetMap. For example Data:Leibi.map. Once there is the technical ability to include a release with the map on Commons, these can have some housekeeping to ensure there is a verifiable release which confirms these are the same person - not just that an account has been created on Commons with the same name. --Fæ (talk) 09:34, 18 October 2017 (UTC)[reply]
  •  Info I discussed this issue with Ilya Zverev, one of the OSMF board members, and he pointed me to the relevant OSM policy. According to him, all boundaries in question should be considered insubstantial extractions, because we use by far less than 100 features, and in fact none of these features is reproduced in full. If you are worried about the fact that we do many such extractions, the policy covers this as well, More that 100 Features only if the extraction is non-systematic and clearly based on your own qualitative criteria for example an extract of all the the locations of restaurants you have visited for a personal map to share with friends or use the locations of a selection of historic buildings as an adjunct in a book you are writing. We do exactly that, because we use our own qualitative criteria when drawing these district boundaries. We never aim at extracting substantial datasets that would replicate OSM content.
There are also some useful links at the bottom of this page, including an e-mail discussion with the WMF representative, and I think they provide additional context, which renders this whole deletion request moot.
@Offnfopt, Ymblanter, Jonatan Svensson Glad, Jeff. --Alexander (talk) 09:41, 19 October 2017 (UTC)[reply]
The fact that they have a policy on what they do not believe is a violation and what they do not act on does not affect the underlying issue that we have to consider. Is it legally copyrightable or not. If you could get OSM to make a note that any use of less than 100 units are cc0/not copyrighted by them would be what we need, or some legal evidence of that affect. They may change policy, and we should care enough to discuss it for ourself. (On phone atm.) --Jonatan Svensson Glad (talk) 10:15, 19 October 2017 (UTC)[reply]
Well, their ODbL license only says substantial, and they have to provide their own definition of this term, which, together with the definition contained in the license ("Extraction" is an extraction of the whole database or its substantial part), explicitly says "any use of less than 100 units are cc0/not copyrighted". I think that's as much as you can get. If you want to provide your own definition of substantial in the context of the OSM database, that may sound a bit funny and will not have any legal effect at all. --Alexander (talk) 10:31, 19 October 2017 (UTC)[reply]
It's an interesting reference, thanks. It does give some grounds to accept maps on Commons as CC0 that contain some OSM data. However this would mean that we need a process for testing whether 100 data points or 100 features have come from OSM, one issue with this being that by "feature" in the ODbL guidelines, this is not intended to mean the technical definition of a feature and may specifically exclude a county or country border. Note that in this deletion request there are maps with multiple features extracted from OSM, and with significantly more than 100 nodes extracted from OSM.
At the current time project administrators have no tools for verifying any of this. There is also the issue that the intention of Commons in hosting Data files is to serve as an open database for as much reuse as possible, and this may be argued to fail 6.2 of ODbL.
This would also not address the issue of accepting maps as CC0 that may rely on other databases with different licenses, including data which has first been imported to OSM and then imported to a map hosted on Commons. A real example in this DR is Data:Amsterdam Districts.map, which contains data from OSM which has a declared source of CBS (Netherlands) Central Bureau of Statistics and so requires attribution to CBS as the source based on my reading of their generic terms. Consequently the attribution requirement supersedes any terms in the ODbL, and the resulting map hosted on Commons cannot be CC0. Again, we cannot expect Commons administrators to police map files to this level of fine detail. In comparison, actually allowing an attribution license with the map file, like ODbL, will avoid any possible doubt that we are respecting the copyright of the source data, even if there can be arguments that a particular case may have fewer than 100 data items in it. --Fæ (talk) 11:47, 19 October 2017 (UTC)[reply]
Addendum I have written to rullzer on OpenStreetMap who did the original CBS import several years ago, hoping to confirm their precise source. However as so many years have passed, we should not rely on a reply. --Fæ (talk) 11:58, 19 October 2017 (UTC)[reply]
I think we are talking about a small number of features here. OSM policy says In the text below a "Feature" is a street, restaurant, park, cemetery, bus stop, mountain top etc., so one administrative boundary is one feature, and I don't think there are more than 10 such boundaries involved in each map. And maps with over 100 features can also be justified, unless they deliberately copy part of the map, which is easy to figure out.
As for the verification, I think there is no Commons policeman who is sitting and verifying all new images for not being duplicates of something already available on the web, and there is no need in such a policeman here. But a clear policy on what should be done and what should not, would be, of course, a great advantage. --Alexander (talk) 14:02, 19 October 2017 (UTC)[reply]
There are lots of points here, so the issue is confusing enough to remain a concern under COM:PRP even if any one data file can be shown to only contain OSM sourced data with no other original sources with different license. From OSM's own summary as you previously linked to, a "feature" is intended to be of modest size, so a boundary for a country, county or even a city region is not going to fall under that definition. RP88 below makes a differentiation between database rights and content rights, which is yet another layer of complexity that should be of concern for anyone wanting to verify the copyright of the file.
Commons may not literally have police, but simple verification of the copyright of files must be made possible and realistic for our unpaid volunteer administrators and license reviewers. At the current time I don't see how "proving" that map data files being created on Commons are CC0 when they are reusing data from other sources is realistic to do as a housekeeping task. I write as a regular OSM contributor and someone who has played around with the API (and several related APIs), and just tracking down sources is hard and actually unrealistic as the sourcing statements in OSM are highly unreliable. --Fæ (talk) 14:19, 19 October 2017 (UTC)[reply]
Thanks for the pointer to the OSM's interpretation of "substantial" as it relates to their ODC Open Database License (ODbL) licensed database. Something that I think should be clarified with regards to this discussion is that the ODbL is a license to the database right for a database, not a license to the contents of the database. The text of the ODbL is very clear about this: "the ODbL only governs the rights over the Database, and not the contents of the Database individually. Licensors should use the ODbL together with another license for the contents". Under OSM's contributor terms, individual contributors agree to their contributions being used under the ODbL for the database and the ODC Database Contents License (DbCL) for the contents of the database. The DbCL is extremely permissive and does not require attribution, but it does not appear to be compatible with CC0. However, not all content in the OSM database is licensed under the DbCL, some of the content is CC-BY or otherwise requires attribution (as OSM makes clear at https://openstreetmap.org/copyright ). A new database at Commons created by extracting content without alteration from the OSM database where the sum of the extracted content is below the threshold of originality is not protected by copyright and can be used under a CC0 license. A similar database created by extracting content from OSM below OSM's interpretation of the ODbL "substantial" threshold does not trigger the ODbL share-alike clause (section 4.4) but would still have to have to honor the terms of the license (or licenses) associated with the content extracted. Lastly, a new database at Commons created by extracting content without alteration from the OSM database above OSM's interpretation of the ODbL "substantial" threshold must be ODbL licensed as well as honoring the terms of the license (or licenses) associated with the content extracted. —RP88 (talk) 13:57, 19 October 2017 (UTC)[reply]
@RP88: This seems a long tangent that may be avoidable. The content license at https://www.openstreetmap.org/copyright makes it clear that the data is under ODbL, and so the data that users may import into a JSON file and host here as a Data namespace pages, only needs to comply with ODbL (and may or may not be judged as CC0 depending on content). The rest on that page about CC-BY-SA is in relation to reuse of map tiles, these are used in the Data namespace to help display the map, but are not part of the data. The embedded display feature gives a correct credit back to OSM as required. My understanding is that the CC-BY-SA requirement is therefore irrelevant to any discussion about extracting data from the OSM database, as in this DR, so long as we are not extracting the images of map tiles. --Fæ (talk) 14:33, 19 October 2017 (UTC)[reply]
I did not mention the CC-BY-SA cartography/map tiles, and I agree those are irrelevant to our discussion. I was referring to the portion of the contents of the OSM database that are CC-BY licensed. OSM has been careful — anyone who complies with the ODbL and links to the OSM attribution page will be compliant with the licenses to the content in the OSM database. Much of the contents of the OSM database will be factual, and as such public domain in the US, but perhaps not in other jurisdictions. Individual contributors to OSM license their contributions as DbCL, which requires them to surrender almost all their rights in the data. However, a small portion of the contents of the OSM database are CC-BY, such as the content from Stadt Wien, while other content is under other licenses, such as the NLSFI License. The only reason I brought this up at all was that it is “substantial” use that triggers the ODbL share-alike restriction — less than “substantial” extractions from OSM can potentially be used under the terms of less restrictive licenses. —RP88 (talk) 15:44, 19 October 2017 (UTC)[reply]
Okay, I think I understand, but if we have to create guidelines we'll need a picture :-). My conclusion for this framing of the licenses that apply is that we definitely have to remove the CC0 restriction, if Commons is going to continue to host maps in the Data namespace. Even the OSM defined "substantial" argument when treated liberally cannot tidy away these remaining non-trivial copyright concerns. --Fæ (talk) 16:26, 19 October 2017 (UTC)[reply]
  •  Info In follow up to the above comment about nodes in Data:Amsterdam Districts.map coming from CBS, further investigation shows that other nodes in that map are originally sourced to BAG. The data is available on an open license, per https://maps.amsterdam.nl/open_geodata/TermsOfUse.html, however this is equivalent to public domain + attribution (similar to the OGL). Consequently the reused data on OSM cannot be reused on Commons as CC0. The relevant attribution requirement is "You are required to mention the City of Amsterdam as the source holder in any publication or distribution of the dataset (wholly or partly) in your future use. In case you distribute the dataset as a dataset, with fee or for free, in its original or in modified form, you are required to mention that the original dataset is freely available from the City for everyone, under these Terms of Use, which also apply to the distributed dataset." Consequently we can firmly state that the Data file on Commons must be removed unless attribution can be displayed with the file and it is neither claimed as 'own work' or as CC0. --Fæ (talk) 14:51, 19 October 2017 (UTC)[reply]
Fæ my comment about CBS was for the administrative boundaries for the Zuid and the Noordshown neighborhood I didn't check the other boundaries. OSM claims those boundaries came from CBS (which can be seen on the screenshots I provided above which can be cross referenced with the Contributors list). Where did you get information that says otherwise? - Offnfopt(talk) 15:46, 19 October 2017 (UTC)[reply]
I'm running some automated point by point analysis, which is very slow due to API limitations, even though the analysis is deliberately incomplete. An example is line 457 in the JSON file, matching [52.4228633, 4.9529256] node 2194159287, examining this changeset in the node history gives a source statement as BAG. As said previously, source statements may not be entirely reliable, but where users have given source statements there is no reason to doubt them. As also said, this type of forensic analysis is not realistically repeatable as a means to verify the licensing of map files, certainly I am not going to invest much more of my volunteer time in this way nor for other good reasons will I turn my method into a general tool. --Fæ (talk) 16:00, 19 October 2017 (UTC)[reply]
 Info A second detailed case is available on Data talk:Copenhagen Districts.map where 100 sample matches from OSM are listed. At least some of the data is originally sourced to Geodatastyrelsen who include very specific terms for attribution, which must include the date of the original dataset used. The use of subsets of their data and map presentations of it are clearly not CC0. --Fæ (talk) 08:06, 20 October 2017 (UTC)[reply]
 Info A third detailed case is available on Data talk:Kuala Lumpur Districts.map. Some or all of the data is originally sourced to data.gov.my. The terms of use for the data include a specific attribution requirement, so reuse such as the map hosted on Commons can not be presumed CC0. --Fæ (talk) 12:00, 20 October 2017 (UTC)[reply]
 Info A fourth detailed case is at Data_talk:Munich_Districts.map. A source used for many of the OSM imported data is http://www.geodatenservice-muenchen.de. As far as I can work out, the default license for reusing any of their data is non-commercial use with anything else needing permission. Keeping this file hosted on Commons would need further investigation to demonstrate that either the data was not used directly, or than some alternative permission existed or was specifically granted when imported to OSM. This investigation was limited, though past versions back to 2014 on archive.org were examined. The city maps exist as a special data service, not the embedded OSM city maps that are available on the site, and it may be possible that the site has additional terms of use that I was unable to discover on my surfing through it.
Update Further examination of other nodes shows sourcing that gets to https://www.opengov-muenchen.de/pages/nutzungsbedingungen. If this license applies to all the imported data (which my analysis does not guarantee, though the same OSM user appears to have contributed all these), then a standard attribution license has to be used. This makes the derived map not suitable as CC0, but may be hosted if that condition were relaxed to allow attribution. --Fæ (talk) 14:39, 21 October 2017 (UTC)[reply]
I am afraid we really need help of the legal.--Ymblanter (talk) 17:41, 19 October 2017 (UTC)[reply]
We've posted a response to the CC0 licensing update in Phabricator and on the Commons Village Pump discussion. DTankersley (WMF) (talk) 01:45, 7 November 2017 (UTC)[reply]

 Keep The underlying map is attributed to OSM. The overlying shapes and colors were created by the user in question. Seriously... Regardless CC BY SA is an open license and there is no reason why we cannot have CC BY SA licensed data. While I respect Fae position I simple consider it to much of a stretch at this point. Doc James (talk · contribs · email) 03:40, 30 December 2017 (UTC)[reply]

I suggest you push to get an administrator to close the proposal to extend data files to all acceptable Commons licences. With that in place this DR is an automatic keep and we can move on with the technical change needed to display all types of valid license. --Fæ (talk) 04:47, 30 December 2017 (UTC)[reply]
Thanks User:Fæ. Will do. Doc James (talk · contribs · email) 05:51, 31 December 2017 (UTC)[reply]

Kept: After reading this discussion, examining the maps and cross-referencing them with the OSM maps, reading Stephen LaPorte's statement at phab:T178210, and reading m:Wikilegal/Database_Rights, I have come to the conclusion that these maps are not copyright violations. The maps are not based on OSM data, even if OSM maps may have been used as a reference and occasionally points from OSM maps conincide (most likely to a "snap to point" feature in JOSM or other editors). This is mostly limited to cases where the boundaries in these maps follow administrative borders and OSM can not claim any creative control or even sweat of the brow protection over those, and the database rights clearly do not extend to them. --Sebari – aka Srittau (talk) 23:42, 1 March 2018 (UTC)[reply]

Retrieved from "https://commons.wikimedia.org/w/index.php?title=Commons:Deletion_requests/Data_talk:Kuala_Lumpur_Districts.map&oldid=290001592"