Is it correct that Navigator does not consider lanes:backward and turn:lanes:backward?
I am asking because I tagged a street section that way: http://www.openstreetmap.org/way/289588330 - and Navigator does not display a lane assistant. (In this case, the assistant is not really necessary, but my question is a more technical/theoretical one). This street is actually a "simple" street you can drive on in both directions. Hence lanes:backward must be used.
From "select way, node_to, tag_value from osm_lane where tag_key = 'turn:lanes:forward' OR tag_key = 'turn:lanes'" in the year(?) old experimental conversion script I would guess that "lanes:backward" conversion is not implemented yet. Note, that a year ago there were not many lanes data in OSM so it was rather "experimental work" (for fun ;-)). Now it could be different (I was actually surprised on how many places in Germany I saw this lane assistant last weekend) and we may use your case as reference example (I will check it at work). thanks Martin
using overpass turbo http://overpass-turbo.eu/s/5qV (some parts of France, Germany, Italy and Switzerland in this example) you see how often "turn:lanes" an the variations are tagged.
Feel free to change the area of the map to get an overview on other areas and you will imagine, that it could be worth to integrate "turn:lanes:backward" .
As a little side step: I consider the "destination:lanes" and their variations just as important (here and here: maybe superfluous but to stimulate others :) ). The destination:lanes give you the blue/green/yellow motorway signs in the top-left of your screen (from Google Play) . So your query slightly modified: http://overpass-turbo.eu/s/5r7
For motorways I always add both the turn:lanes and the destination:lanes (and destination:ref:lanes and destination:int_ref:lanes to be entirely complete). For lower order roads I add the turn:lanes (and variables), but not always the destination:lanes (and variables)
Hi Martin, in Germany we actually begin to start a "job for a week" in OSM. The target is, that as many mappers as possible put as much data as possible including turn:lanes=* and the variations of it into OSM-database. There is also announced, that your Navigator free is one of the routers using these data. I think it would be nice to expand usability in using informations from most of keys "turn:lanes:*" and mainly the *:backward one. turn:lanes: forward and -:backward is mainly in use on non-dual-carriageway-roads and so often inside of towns, where informations about turns is more needed than on motorways for stranger's orientation as I mean. So I please to set priority in implementing the use of all the offered turn:-data of OSM. Thanks Michael
Hi hurdygurdyman and others, in December 2014 OSM data are already included lanes:backward. I received bug report that this example in Hungary does not work properly: https://www.openstreetmap.org/way/212796596 "turn:lanes:backward"="left|none" I was looking for some other examples (in Luxembourg, small map, but contains most examples) and it seems to me that there is some misunderstanding, for example: https://www.openstreetmap.org/way/225665043 also "turn:lanes:backward"="left|none" but it should be rather "none|right|?? Can you confirm that second example is not correctly mapped? thanks Martin
Edit: The second example is correct (based on satellite images), but in OSM is missing road to the industrial area.
It is correct to set the values in an order from driver's perspective, and from left to right, see OSM Wiki: "Lanes should be ordered from the leftmost one to the rightmost one." [..., then regarding "backward"] "doing so view the road in the opposite direction of the osm-way for this."
Interesting, thanks - for help see http://josm.openstreetmap.de/wiki/Styles The December 2014 conversion bug is probably only (?) reversed order of displayed lanes for tag (turn:)lanes:backward. The selected line should be correct. This should be fixed to January 2015 data. thanks Martin
@rab Meintest du mit recommended wirklich "empfohlen"? Wenn ja, verstehe zumindest ich nicht, was du damit sagen möchtest :-) Ich tippe, dass du meinst, dass merge_to_* nicht dargestellt wird. Das ist mir nämlich auch aufgefallen.