One more thing: the distinction between bicycle=no and bicycle=dismount has made its way to this important article for various countries around 2015. [14]
[14] https://wiki.openstreetmap.org/wiki/OSM_tags_for_routing/Access_restrictions On Wed, May 27, 2020 at 10:12 AM Fernando Trebien <fernando.treb...@gmail.com> wrote: > > On Tue, May 26, 2020 at 8:55 PM Mateusz Konieczny via Tagging > <tagging@openstreetmap.org> wrote: > > May 27, 2020, 01:35 by fernando.treb...@gmail.com: > > > > On Tue, May 26, 2020 at 1:48 PM Mateusz Konieczny via Tagging > > <tagging@openstreetmap.org> wrote: > > > > May 26, 2020, 18:04 by fernando.treb...@gmail.com: > > > > Bikes may "pass" in two different ways: riding > > (bicycle=yes/permissive/destination) or pushing (bicycle=dismount). > > Bikes are only completely forbidden if bicycle=no/private. > > > > bicycle=no does not mean that you cannot push bicycle > > > > The wiki defines bicycle=no the same as access=no, which means no > > access. If you have foot=no, that means no access by foot. > > > > and if you have bicycle=no that means no access by bicycle > > It says nothing about access with bicycle (pushed/carried). > > I went back to this edit [1] before the wiki was changed recently. > Back then, bicycle=no was simply defined as "where bicycles are not > permitted." If nothing else is said, then nobody can conclude that > "riding bicycles is not permitted but carrying/pushing is," it said > "bicycles." This has been pointed out before. [2][3] In the same > table, a distinction is made for values such as bicycle=use_sidepath > and bicycle=dismount. If misunderstanding is common [4][5], the only > solution is to create new values and deprecate the old ones, as was > done for surface=cobblestone. [6] > > > bicycle=no and bicycle=dismount are de facto equivalents > > > > How can you conclude that? > > > > Based on my experience of how people map such restrictions? > > Based on my experience how tags in such situations are processed by data > > consumers? > > > > And it is not just me, see > > https://josm.openstreetmap.de/ticket/9158 > > The answer on this ticket is specific to Germany. > > > https://lists.openstreetmap.org/pipermail/tagging/2013-October/thread.html#15135 > > Looks like this thread did not reach a conclusion, resulting in no > changes to the wiki. Interpretations appear to be divided. To me, it > looks like this proposition [7] would have solved all situations, but > I see no usage of the proposed tag or other alternatives proposed > throughout the discussion. > > Since 2014, GraphHopper [8] and OSRM [9][10] implement the > interpretation of bicycle=dismount as pushing but no riding and > bicycle=no as no access whatsoever, not even pushing. No questions so > far regarding this interpretation. Bicycle routing using GraphHopper > and OSRM has been offered in OSM's main website for a very long time. > The UK-based CycleStreets journey planner also implements this > interpretation. [11] Some guys on brouter [12] agree with you, but > brouter profiles still assign a very high cost when bicycle=no [13]. > You should probably note that those remarks were made 3 years after > brouter has offered bicycle routing with the current interpretation. > > If there really is widespread agreement that bicycle=no should be > treated like bicycle=dismount (plus, perhaps, some treatment when > foot/access=destination), I would expect more requests to correct this > in applications such as OSRM, GraphHopper, brouter and others. > > That said, I do not oppose changes to clarify this situation. First we > need either proper tagging scheme or a change of definitions that > embraces all situations mentioned so far, then we need to ask > developers to change their routing profiles to avoid confusion among > mappers and users. > > [1] > https://wiki.openstreetmap.org/w/index.php?title=Bicycle&oldid=1965874#Bicycle_Restrictions > [2] https://lists.openstreetmap.org/pipermail/tagging/2013-October/015308.html > [3] > https://wiki.openstreetmap.org/w/index.php?title=Tag:bicycle%3Ddismount&oldid=1919911 > [4] https://lists.openstreetmap.org/pipermail/tagging/2013-October/015315.html > [5] https://lists.openstreetmap.org/pipermail/tagging/2013-October/015356.html > [6] https://forum.openstreetmap.org/viewtopic.php?id=61042 > [7] https://lists.openstreetmap.org/pipermail/tagging/2013-October/015276.html > [8] https://github.com/graphhopper/graphhopper/issues/695 > [9] https://github.com/Project-OSRM/osrm-backend/issues/78 > [10] https://github.com/Project-OSRM/osrm-backend/issues/5072 > [11] https://www.cyclestreets.net/help/journey/osmconversion/#toc9 > [12] https://github.com/abrensch/brouter/issues/79 > [13] https://github.com/abrensch/brouter/issues/226 > > -- > Fernando Trebien -- Fernando Trebien _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
