In my proposal, expressway=limited describes a road in the same way that the current expressway=yes does, while expressway=controlled describes a road as having the same physical characteristics that have typically been associated with highway=motorway. expressway=controlled differs from highway=motorway in that it does not imply any legal restrictions or importance. Additionally, highway=motorway would now by default imply expressway=controlled, which is usually correct. However, in the countries where there are roads that should be highway=motorway because of legal classification despite lacking some access control, expressway=limited may be used alongside highway=motorway to convey the physical situation.
I can do that. I am mentioning it here because there is some discussion of the physical construction of a road versus its importance classification. If you think my proposal has issues, I would appreciate hearing your concerns so I can improve it.
I have created a thread for the proposal, for those who are interested.
I really hope Americana is featured on the OSM frontpage, too.
2 Likes
Well the paper maps I do know and I do remember treat trunks and motorways in just the same way as we do in Germany and our neighbours in France 
Because of that Iâm very very satisfied with the way we do that in OSM. I wonât have it changed.
You donât know what a trunk [road] is, do you?
@Adam_Franco this is what Iâm talking about. Itâs going to be very difficult to change the meaning of a tag which is already so solidly encoded in the minds of mappers.
The best part is when a map publisher attempts to communicate a foreign land to their domestic audience, usually in partnership with an overseas publisher. California for Swiss travelers:
France and Benelux for American travelers:
Have fun everyone!
3 Likes
I do believe that we should completely decouple any physical characteristics from highway=trunk, but it seems like it will have to be done on a regional basis, at least for the time being. As more regions adopt this meaning, it will become more compelling for other regions to follow suit.
1 Like
high_speed_road= / highspeed= can have a different criteria in urban areas, and a loose speed limit doesnât mean itâs designed and attainable (cf maxspeed:practical= etc ). In a sense, itâs similar to the argument for expressway= , in how maxspeed= alone doesnât reliably convey this. The urban vs rural difference is something I asked about expressway= too.
In Poland we applied the same criteria for both but since the maxspeed threshold was to be over 50 km/h, it included both urban and rural roads of very good quality.
In East Asia, there are many widely-spaced at-grade or grade-separated roads at 50km/h, to even 40km/h. Thatâs one reason for my question.
Thatâs why I think there should be consensus among each country what they consider to be an expressway. Initially it should be what they use for highway=trunk right now, if itâs defined based on parameters. If not, and there isnât even something such as an âexpresswayâ defined administratively, they should discuss about what they consider something a âsemi-motorwayâ but perhaps there isnât even anything like that and they donât have to use this tag.
1 Like
I would want to see this problem solved for cycleways as well.
1 Like
The solution is importance=national/regional/urban/suburban.
Then itâs too subjective.
1 Like
Generalizing a ton, the most common difference between rural and urban expressways is historical: a rural expressway is designed to be an expressway, possibly with a view towards a future upgrade to a freeway, whereas an urban expressway was state of the art for its time but never got upgraded as highway standards improved. Of course there are plenty of exceptions, but if the main difference is where the highway is located or when it was built, then OSM data consumers have much less need to make this distinction via tags.
This raises a good question. For limited-access roads, do you purely look at maxspeed= for the speed criteria of expressway=yes now, or make a âsanity checkâ on turns and gradient to exclude curvy or steep roads?
Anyway, the main issue I wanted to raise is how 50km/h to even 40km/h can be considered high in urban areas, on limited-access roads. Therefore itâs not as simple as a 70km/h or some number, as found in rural areas.
So I made a thread about the implementation of official classification tagging:
And if someone needs a reason for why importance is needed, I have an example.
Here is the the route between Warsaw and Prague when driving a passenger car: GraphHopper Maps | Route Planner
And hereâs the same route for lorries: GraphHopper Maps | Route Planner
The first route would be tagged as importance=supernational and the second one as importance=national + importance:hgv=supernational.
There are countless examples like these which result from restrictions for lorries or better conditions for them on other roads. With the current tagging, a road will be treated the same for both passenger cars and lorries despite one being more important than the other. An importance=* tag solves the issue.
3 Likes
And, since GraphHopperâs suggested roads are completly correct, why do we âneedâ another tags?
1 Like
Why do we need ANY âcolouringâ of roads? Why not make it all gray like in Google?
But when we acknowledge that people have spent time making these, letâs not delete all that data but instead try to add a little bit more detail to it. A road is treated the same for all 4+ wheeled vehicles so letâs add a distinction so that one can make a vector map with personalisation for each vehicle type where it wouldnât display some roads so brightly and a regular car layer also would have lorry-only roads less noticable.