That is an irritating case. 1) with you assumptions it is possible to argue that it refers to case where there is a separately mapped sidewalk that nevertheless is inaccessible (some technical/escape route in a tunnel or on motorway?)
2) in practise it is far more likely to be used in case where user tags each way with own access tags. Basically > that foot=* applies to "the whole of the road" including the roadway, > shoulders, verge, sidewalks, and so forth has added at the end > "unless represented with separate geometry" 3) to avoid problem from (2) foot=use_sidepath was invented to mark "yes, on carriageway you cannot walk, but you can walk on separately mapped sidewalk" ----------------------------------------------------------------------------------------------- practical advise: - if there is no separately mapped footway then sidewalk=separate is wrong and should be fixed - if pedestrian cannot walk on carriageway and can on sidewalk and it is separately mapped then use foot=use_sidepath instead ----------------------------------------------------------------------------------------------- No idea: - how to tag situation from (1) that distinguishes it from case (2) Not sure: - should we treat foot=no instead of foot=use_sidepath in case (2) as invalid data > Would folks regard that as accurate data modeling? I.e. should I change my > software to treat streets tagged in this way as pedestrian-accessible, or > would folks regard this combination as a tagging error? > For start, is there a separately mapped sidewalk geometry there? If no then tagging is invalid. If yes, then I personally would be treating as invalid data and assume road to be pedestrian accessible, maybe flag is worth reviewing and ask users for feddback what is going on there and ask for photo. 18 gru 2022, 21:29 od zelonew...@gmail.com: > Hello, > > I am the author of a data consumer which generates a list of streets that are > accessible to walkers and joggers. The idea is that a user would have a map > of the streets in their town and can challenge themselves to walk/jog down > every street, and they can look at statistics on which streets they've > completed. I use a 25-meter rule, so if a user can walk along the shoulder, > or on a sidewalk/pavement, or in the verge, that's acceptable. > > I recently came across an unexpected tagging combination and I would like to > understand how folks in various places would interpret this: > > highway=<whatever> > foot=no > sidewalk=separate > > In my software's logic, I've made the assumption that foot=* applies to "the > whole of the road" including the roadway, shoulders, verge, sidewalks, and so > forth and thus excluded any roads that include that tag, regardless of other > tagging. I came to understand that this tagging was used by a mapper to > indicate that "pedestrians are not allowed on the roadway, however, they are > allowed on the sidewalk" > > Would folks regard that as accurate data modeling? I.e. should I change my > software to treat streets tagged in this way as pedestrian-accessible, or > would folks regard this combination as a tagging error? > > >
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
