I'm also hesitant to bring apps to the discussion, since we don't tag for apps, but things like Nominatim (a fundamental piece of the ecosystem) have been around for a long time and I doubt something much different will emerge. One could bring satellite navigation apps into the discussion and the logic would apply equally (and be even more restrictive).
Please do not understand my attempt to figure this problem out as a criticism to your solution. I think our solutions are both equally capable of representing the same ideas, they simply do so in different ways. For instance, both solutions allow the representation of multiple conflicting parties, of dejure and defacto/controlled borders (if we decide that's what defines the actual administrative border), and also allow proper rendering (yours would require checking the role in the boundary relation, mine would require checking for one tag). In Antarctica they would all be de jure borders, right? Unless I'm missing something, our solutions are equivalent in theory, but I suspect (not being entirely sure) that mine might create less friction for widespread adoption. To be sure, I'd need more time to think and also to consider more cases of disputed borders. One related problem is that of assigning a value for the "name" tag (in addition to "name:[country code]") for things that lie on the border of two countries. In the case of disputed territories, if we choose to remain neutral, that problem would extend to anything falling within a disputed territory, including its borders. But that's a minor problem. In Brazil and Uruguay we chose to copy European countries and put both names separated by a slash, like "name in spanish / name in portuguese". Since we didn't find any information to how it was done in Europe, we decided the order of the names would be the sorting order of the strings, so as to eliminate the perception that any of the countries takes inherent precedence. On Mon, Dec 23, 2013 at 7:29 PM, Michael Krämer <ohr...@gmail.com> wrote: > First of all I think there is no really perfect solution to the problem. To > me this is inherent to the dispute as different parties have a different > view of what's right and wrong. So I think this conflict will show up in the > data anyway. > > > 2013/12/23 Fernando Trebien <fernando.treb...@gmail.com> >> >> Today, from a practical perspective, a letter to anyone in Abyei would >> probably be addressed to Sudan, a phone call would dial Sudan's area >> code, a visitor would go through Sudanese immigration. So maybe it's >> in Sudan. > > I agree. > >> >> Maybe it's part of both countries then. In that case, tools like >> Nominatim should reflect that - but Nominatim currently thinks that >> Abyei is in South Sudan (possibly making some non-technical Sudanese >> users a bit uneasy). > > Ok, but I am personally quite hesitant to tag things in a way so Nominatim > gets correct results. After some checking with places in Israel and the > Palestinian territories I think Nominatim might not be the best benchmark to > use. > >> >> Overlapping administrative borders should then be >> a basic assumption of every app - and they're not, since they almost >> always are administrative "subdivisions". An overlapping >> administrative border of equal admin_level would make more sense if >> both parties were friendly to each other and collaborating within the >> area. But then I think nobody would describe these as "disputed" >> territories. > > I do not consider adding the region to both countries to be a great solution > myself. It's just an pragmatic approach to get things covered 'good enough'. > Ideally the disputed borders were marked as disputed and rendered > accordingly e.g. dashed instead of solid. Or event the disputed area could > even be hatched or so. > >> >> But is the area faring independently? No, so we represent the claim >> conflict using the "claimed" role in each country's defining relation. > > If we add 'controlled' or 'dejure' and 'defacto' this sounds good to me: So > the northern border of Abyei should be 'dejure' in Sudan and the southern > border 'defacto'. The South Sudan should have things just the other way (if > I got the legal situation right). Basically there is a section where the > border of both countries is not really well defined and for me the data > should reflect this. > >> >> Extending that to Antarctica, one may ask: are the overlapping >> territories claimed by Argentina, Chile and UK part of any of these? > > I agree that for Antarctica my strategy of being part of both definetly does > not work. This is different and probably needs special treatment. > > When working on the border beween Sudan and South Sudan I spent some time > looking into disputed borders. Some key observations for me were that (1) > there are more than one would expect and (2) there currently is no > established method of handling this in OSM. So I am happy if we find a > better solution but I guess this easily requires changes to some other tools > like the rendering or perhaps Nominatim. But this shouldn't keep us from > looking into it. > > Michael > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging > -- Fernando Trebien +55 (51) 9962-5409 "The speed of computer chips doubles every 18 months." (Moore's law) "The speed of software halves every 18 months." (Gates' law) _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
