On Mon, 24 Aug 2009 23:32:46 +0200, Renaud MICHEL wrote: > Le lundi 24 août 2009 à 15:25, Lester Caine a écrit : > > Adding an extra node does make sense, but probably needs a 'relation' to > > the intersection as well? In any case the direction through this new > > node is the critical piece of information? Tagging ways would require > > that every section of a way is broken up. I'm thinking of some route > > around here that have several intersections along them, many but not all > > of which are compulsory stop along that single way. Simply adding nodes > > on the correct side of each intersection would be somewhat easier to > > implement, while currently these restrictions are not recorded. > > How about simply creating a relation with those two nodes? > > You have an intersection of two (or more) roads and when you come to that > intersection from one particular road (in one particular direction) you have > a stop. Then you add a node on that way before the intersection, then create > a relation (let's say of type=stop, or any more self-explanatory value) > where you have the stop node with role "stop_from" and the intersection with > role "stop_at".
Sorry, this is useless. We could just use highway=stop and teach routing software that they're valid only if coming from a certain direction, decided on a distance-from-nearest-junction basis. This is what we're (I am, at least) trying to avoid: arbitrary placement of real-world nodes -- if we all had 1cm-resolution GPS units, then we would just use highway=stop. IMHO, YMMV. > Such a scheme can be easily interpreted as: when you pass over a node > highway=stop that is a member of a relation type=stop with role stop_from, > then you must stop at the node of that relation that has the role stop_to. No, you must NOT. In this case, you would stop in the middle of the intersection. A stop signal instructs you to stop *before* the intersection. :) Kindly, David -- . ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino : :' : Linuxer #334216 --|-- http://www.hanskalabs.net/ `. `'` GPG: 1392B174 ----|---- http://snipr.com/qa_page `- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174
signature.asc
Description: PGP signature
_______________________________________________ talk mailing list [email protected] http://lists.openstreetmap.org/listinfo/talk
