<div dir="ltr"><div dir="ltr"><br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Sat, 21 Dec 2019 at 14:43, Phake Nick <<a href="mailto:c933103@gmail.com">c933103@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="auto"><div>Reminder 1: There are loops within bus route doesn't mean the route is a circular or round trip route.</div></div></blockquote><div>Fully agreed. That's why I am saying we need to alok at this with a bit of calm. There plenty of diferent route toplogies</div><div> <br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="auto"><div dir="auto">Reminder 2: The roundtrip=* key is designed to use in combination with hiking routes or bicycle routes. A hiking/bicyle route that goes A→B→A which come back with the same start point with exact same alignment as the other direction doesn't really mean anything so I don't think a special value would be needed for such case. As for bus routes, whether or not it goes back along same road doesn't really mean anything either.<br></div></div></blockquote><div><br></div><div>Fully agreed.</div><div>We never looked at this and tagged different route types (bus and bicycle for example) independently folowuìing essentially different topological approaches.</div><div>Bus routes are (always?) tagged as unidirectional routes. The same bus line that connects A <> B is represented by two bus routes A > B and B > A.</div><div>But, by tradition, or whatever reason, a topologically equivalent bicycle route A <> B is represented by a single, mostly bidirectional route A <> B, where the few pieces where the A > B route differs from the B > A route (for example for roundabouts or cycle paths on both sides of a road) are handled by route segments whose ways are tagged with role=forward|backward within the relation.</div><div><br></div><div>By the way, these loop segments are, at least within the route network in Italy, tagged with role=forward|backward differently from the definition in the <a href="https://wiki.openstreetmap.org/wiki/Cycle_routes">cycle routes wiki page</a> <br></div><div>("
<b>Relation role</b>: Cycle routes sometimes have different paths
depending on the direction you are travelling. In this case, ways in the
relation should have a role of <i>forward</i> or <i>backward</i> as described in <a href="https://wiki.openstreetmap.org/wiki/Relation:route#Members" title="Relation:route">Relation:route#Members</a>. The direction is rendered on the cycle map (<a class="external gmail-text" href="https://www.openstreetmap.org/?lat=51.88885&lon=0.8972&zoom=16&layers=C">example</a>).")</div><div>in the sense that loop sections that are to be followed in the A > B direction are marked with role=forward and loop sections that are to be ridden in the B > A direction are marked with role=backward. <br>But that is a different story that needs sorting out as well.<br></div></div></div>