Talk:Proposed features/Transport modes on platforms and stations
Would not work
First, the issue with these transport modes is that they share keys with access values. And platforms, unlike stop positions, are routable. So by adding bus=yes on a platform, you are allowing buses to drive on that platform.
- While i must admit that i'm not a developer of a routeing engine, i think it shouldn't be a problem to override bus=yes tags on platforms, should it? Besides i can't imagine any situation of a platform being directly connected to two roads, so that a router would navigate a bus over the platform.
- Based on the reactions of your ideas to reform public transport mapping, i got the impression that the majority doesn't want to give up public_transport=platform. Therefore this proposal tries to make public transport mapping more efficient while changing as little as possible. --SelfishSeahorse (talk) 18:28, 12 September 2018 (UTC)
I prefer the direction Proposed_features/Refined_Public_Transport is taking. Mapping bus stops with ways as platforms seems very cumbersome for most cases and generates a high barrier for contributors to contribute to public transport data. --Datendelphin (talk) 06:48, 15 September 2018 (UTC)
- PTv2 doesn't demand that bus stops are mapped as ways. If there is no real platform, only a node should be mapped. And if there is a real platform, you only need to map that platform as public_transport=platform way or area. With Proposed_features/Refined_Public_Transport, you would need to map a highway=bus_stop node (this time on the highway), a highway=platform way/area and a public_transport=stop_area relation. In my opinion this is much more complicated than a single public_transport=platform object. --SelfishSeahorse (talk) 12:15, 15 September 2018 (UTC)
- You are right, it can be mapped as a node, sorry about that. I read Proposed_features/Refined_Public_Transport differently. Bus stop nodes should be besides the road, so one node is all that is needed for a bus stop (or rather two nodes on each side of the road if the bus travels in both directions). No stop position, no relation. But I am glad that we agree that a simple bus stop should be mappable by a single node :) That is my main concern. Though I still find the highway=bus_stop tag more appropriate than the platform. Some stops have a raised side walk, but even there a platform is kind of a stretch for the word. In many places the distinguishing feature is a post. So for me mapping the stop where the post is, is kind of the intuitive thing to do. --Datendelphin (talk) 15:23, 15 September 2018 (UTC)
- On several diagrams at Proposed_features/Refined_Public_Transport#Platforms and Proposed_features/Refined_Public_Transport#Stop_Areas_and_Groups, highway=bus_stop is mapped on the highway way instead of beside it which i find a bit confusing.
- In case there is a real platform, it would be best (both for mapping as well as for routeing) if there were only one single object. Unfortunately that's not possible with PTv1 tags, because highway=bus_stop isn't defined for areas and ways and because highway=platform and highway=bus_stop can't be combined (however one could combine public_transport=platform + highway=bus_stop, see this diagram.) With PTv2 tags this is possible: you just need to map a public_transport=platform area or way. --SelfishSeahorse (talk) 12:22, 16 September 2018 (UTC)
I strongly oppose referring to highway=bus_stop as "legacy"
- How should they be referred to instead? 'Old' tags? 'Legacy' is how JOSM calls them too.
- In my opinion public_transport=stop_position isn't needed because to route pedestrians to or from any public transport stop, public_transport=platform (that is the waiting area) is preciser. And for the rare cases where the stop position were needed, it should be possible to calculate it from the platform. However this proposal doesn't touch public_transport=stop_position because i think that further discussion on this matter is necessary. --SelfishSeahorse (talk) 11:50, 16 September 2018 (UTC)
Against deprecating highway=bus_stop, railway=tram_stop, railway=station, railway=halt etc.
I am not happy that this proposes to deprecate easier to use tags - "it would no longer be necessary to double-tag or double-map the older public transport tags".
This is quite fundamental difference, as one of main points of this proposal is to make deprecation of such tags easier and I understand that it is not easy to resolve. But I wanted to just note my opposition to public_transport=* that attempts to represent everything at once and in my experience it is not making tagging or processing easier. Mateusz Konieczny (talk) 06:50, 17 September 2018 (UTC)
While the old tags might be easier to understand, the old scheme is less powerful. For example, one can't map a train route that routes passengers to the appropriate platform, they can only be routed to the station.(Edit: theoretically this is possible, although undocumented.)
- More important to me than the names of the tags, however, is that we find a consensus to simplify the current mess. But if some users only want to keep the old tags and other users only want to keep the new ones, and the rest of them reject both, we will never find a majority -- let alone a three-quarter majority. :-( --SelfishSeahorse (talk) 18:32, 17 September 2018 (UTC), edited 20:54, 26 October 2018 (UTC)