PT v2 says you CAN map stops using 2 objects. People reading that understood that you MUST use both a stop_position node and a platform way/node.
Then it was interpreted as: both of those HAVE TO be added to the route relations. To make things look consistent in the route relations, then some mappers started adding platform ways where no actual platforms exist. All in all the scheme, when interpreted like that, is too complex. Information is duplicated across several objects, creating a need to keep them in sync afterwards, whichj is, of course absurd. Personally I don't see a need for the stop_position nodes, so for the few ones that I do add, they don't get details like name, and I don't add them to the route relations. Also the route relations, it's not one per direction of travel. It's one per variation of itinerary, which often comes down to 2 relations per route_master relation, but it can also be just 1, or up to 50 ( https://www.openstreetmap.org/relation/3944387). For "telescopic" lines, I only create route relations for the longer variants. The whole thing is messy enough as it is. Polyglot 2018-03-29 9:37 GMT+02:00 Topographe Fou <letopographe...@gmail.com>: > Hi, > > Your intent is to simplify but I don't understand how replacing one tag by > three or more with different syntaxes key/value according to the type of > transportation and their introduction in OSM can make things easier. > > I share your view when you say that two schemas is too much to maintain > but I would rather jump to the conclusion that it is PTv1 which should be > dropped if we want to drop one as it has limitations. PTv2 is not that > complexe, it is public transportation which are complexe to map. > > One thing I never understood was why we have to maintain two schemas > (probably because consensus was not reached). I guess this is the main > reason why some people (especially new mappers) can be lost. And also why > wiki can sometime say one thing or the opposite. I think it delayed the > evolution of renderers and tools instead of pushing them to evolve with the > schemas. I have the feeling that your analysis dropped those factors. Once > the map on osm.org will render PTv2, I'm sure it will be a huge step and > that all the work done will pay. > > Then I don't see where is the difficulty to map two different features > instead of mapping a single approximate one. It means more work, right, but > it adds more values to the project as a whole. And since nobody is forced > to mapped everything on its own... > > To conclude: I disagree with part of the analysis and with the whole > conclusion. > > Yours, > > LeTopographeFou > > Message original > De: i...@zverev.info > Envoyé: 28 mars 2018 3:54 PM > À: tagging@openstreetmap.org > Répondre à: tagging@openstreetmap.org > Objet: [Tagging] Still RFC — Drop stop positions and platforms > > Hi folks, > > A while ago I've made a proposal to deprecate some public_transport=* tags: > > https://wiki.openstreetmap.org/wiki/Proposed_features/ > Drop_stop_positions_and_platforms > > The discussion was very slow, and in general mappers seemed to accept the > change. I'd like to push this to voting in a few days, but first I want to > know if somebody has anything to say about the proposal. Like, why we > should not. I'd prefer to discuss that before the voting has started. > > Ilya > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging >
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
