Hi, I think the biggest problem with such tags is that they would be very "subjective" (as can already be seen in this thread), and thus not a good fit for OSM (see https://wiki.openstreetmap.org/wiki/Verifiability).
It might be better to try to break down "why" you do not consider a certain route desirable, and then see if related facts about the route could be captured instead (e.g. is the street too narrow, speed limits too low, is it a residential area, etc), and then routing engines could add some weighting based on such verifiable facts. Best regards, Christian On Mon, Dec 11, 2023 at 10:18 PM Dave F via Routing < [email protected]> wrote: > HI > Could you clarify why you "don't want to use this this road as a > through road, even though it's legally allowed " > > DaveF > > > On 11/12/2023 11:53, Florian Lohoff wrote: > > Hi, > > > > Prolog: > > Discuss a generic tag for reducing/increasing route-weights/costs > regardless > > of other tags. > > > > > > For 10 years i am running Routing Quality Assurance by calculating > > routes every 30 minutes and watching their changes over time. > > > > When i add regions to that monitoring i typically go through > > and fix "static errors" where road taggings cause small side roads > > to be used as shortcuts. > > > > Most of the time fixing tagging on the roads and the shortcut fixes > > the issue. I typically tag maxspeed, lanes, surface, lane_markings etc. > > > > But sometimes i have cases where these tags are not sufficient and i > > dont get the relative weight of the routes to the higher class > > road network. > > > > I am missing a generic tag to influence routing in ways like "dont use > > this road as a through road although legally allowed" e.g. make it > > more expensive in routing. > > > > Examples: > > > > The route uses "Im Kracht" where it should stay on K22 and L775: > > > https://www.openstreetmap.org/directions?engine=fossgis_osrm_car&route=52.2340%2C8.5138%3B52.2381%2C8.4837#map=15/52.2354/8.4993&layers=N > > > > The route uses "Westerweg" but it should stay on K27, L876, L803: > > > https://www.openstreetmap.org/directions?engine=graphhopper_car&route=52.2803%2C8.6965%3B52.2890%2C8.7094#map=15/52.2825/8.7106&layers=N > > > > > > Dont get into discussions why one engine gets it and the other doesnt. > > We are not talking about routing engine/profile comparisons. I can > > create such a problem in ANY engine and profile just depending on the > > geometries and tags understood in one or the other engine/profile. > > > > > > So my idea would be to create a tag like **relative_route_cost=** or > > the like which is a float between 0 and 2 where "1" is "use the route > > weight calculated from your route engines physical tag analysis". > > > > So > > > > 0 -> decrease cost of this road to "0%" > > 1 -> Let the cost be the cost of the engines tag analysis > > 2 -> double the cost of the engines tag analysis > > > > This would allow us to directly influence the usage of "shortcuts" or > > through roads used "accidentally". > > > > I would like to not create "routing engine specific" tags like > > **osrm:relative_weight=** or something. > > > > Thoughts? > > > > Flo > > > > _______________________________________________ > > Routing mailing list > > [email protected] > > https://lists.openstreetmap.org/listinfo/routing > > > _______________________________________________ > Routing mailing list > [email protected] > https://lists.openstreetmap.org/listinfo/routing >
_______________________________________________ Routing mailing list [email protected] https://lists.openstreetmap.org/listinfo/routing
