Not a bad idea at all, I think. It doesn’t help with the gpx-export, though.
1 Like
I’d think it makes sense to have an interruption in the gpx track, whatever solution we choose.
I just found route Relation: European long distance path E2 West, part Belgium (alternative) (14005445) | OpenStreetMap. Should we add it somewhere in the relation structure?
Several years ago, I checked this with the Flemish operator. They have a roundtrip through Belgium, which serves as the GR5 / E2 in Belgium. E2 lands on the roundtrip en halfway round leaves it to go on to Luxembourg and France. So hikers just pick one half of the roundtrip. The operator said both should be kept, but as a main route, the anti-clockwise trip is recommended. So we decided to include that one in the main route E2 west, and kept the other one as a stand-alone alternative within Belgium. If there were a toplevel collection relation containing all the alternatives of E2, it would be included.
But a. this would be a collection rather than a route (not recommended in OSM) and b. We have no OSM-based software that offers extra value for such a collection relation. For rendering, a collection relation is not necessary; renderers render individual route relations, and some top those with the superrelations, which together covers everything. Routing doesn’t use supersupercollections either.
PS I forgot to mention: I did this together with the Belgian mapping community.
I would describe it more as a network than a collection. It is similar to node networks, in its own way: no named or numbered nodes, of course; but a deliberately designed set of interconnected sections. If the operator designed them that way, who are we to say differently?
The bright side of seeing it that way is that it would help reintegrating node networks in the management of hiking routes, whereas today they are set aside.
No Nodes, means no Node Network. Node Networks are defined by Nodes referring to each other in pairs; that enables the hiker to create an itinerary defined by the labels of the Nodes visited. For long distance routes, many alternatives are alternative connections between two locations; both go from A to B via different routes. In Node Networks, that shouldn’t happen. In those cases, an extra Node with its own label is placed in one of the alternatives. That way a string of node labels always represents a unique itinerary.
I know some “long distance paths” in Nederland which would be better described as Node Networks. E.g. one is like a wheel with 4 spokes, or a four-leaved clover. The other one wanders about a region, where the sections seem to cross and interconnect willy nilly. We have suggested to label the intersection points, but the operator does not want to place labeled Nodes. They say it’s too expensive and too complex (which it isn’t, but it’s their call…). So we mapped them as collections of routes, recognisable as one “path” by the ref and the name. Ground truth prevails again!
We all know what makes node networks different from other types of hiking networks or routes: they support a specific routing method. Nevertheless, I don’t see why this would put them 100% aside of other hiking networks or routes to the point that, e.g. OsmAnd offers the choice between rendering them or rendering the others. On the contrary, I am convinced that we should strive to make them as consistent as possible with other relations, isolating what makes them special.
I am also convinced that there are other solutions for creating that consistency than asking operators to switch to named or numbered nodes everywhere. For instance we could have anonymous nodes, and decide that their presence in a “node network” is an error but that they are allowed in “networked routes”.
The routes in a Node Network are regular route relations. One extra tag is available to signify that they happen to belong to a Node Network. Data users and renderers can treat them as special routes or as regular routes, as they see fit. OsmAnd has separate layer switches for the two; Waymarkedtrails just changes the rendering of the route line to a much thinner line, regardless of the network tag. Knooppuntnet handles only Node Network routes.
So, the information is there, it’s the app that decides how to use it.
If I was granted a wish, I would wish for an application that lets the user pick connecting route relations at will then combines them to one gpx for export. With Knooppuntnet Planner you can, but only with Node2Node routes.
(The next wish would be an application that splits routes or combined routes into user defined sections)
We saw this, hiking through Nederland on a national hiking trail which is part of the E9
And this:
Where the E2 uses the same ways as the E9.
No consistency, and rare, but yeah, somewhere on the ground you can spot the E-routes.
I will try and get the operator to associate there routes with the E-route more often, and more obvious.
2 Likes
what is (F)?
France (Bay of Biscay)
That’s what I feared. There’s so many ways this sign is wrong 
Well, https://hiking.waymarkedtrails.org/#route?id=1254604&type=relation&map=4.0/50.3133/24.1295 suggests “not quite the whole story” rather than actually “wrong”. It does actually go to (and beyond) the places mentioned
1 Like
True. But my point is, there ís a sign.
3 Likes
I know where the place named Gdansk is. I don’t know about the other. I would not like to have to map this sign
But Peter is right, this is beside the point.
Apologies if I’m stating the obvious here, but as @rhhs said, “Biskaje” is just “Biscay” in Dutch (the sign is in the Netherlands).
European Coastal Path E9
Biscay (France) - Gdansk (Poland)
“LAW” is presumably an acronym for section - @Peter_Elderson will probably be able to fill us in on that. It’s this bit, but following the coast
Let me try an example. On Peter’s other picture there’s also a sign about E2. Imagine if the other sign had said something about E2 going to “Brittany (UK)”.
It’s one of the marginal problems of international routes: confusions about names. Here, Biscaye is a Spanish region; the French name of the bay of Biscaye is “Golfe de Gascogne” but E9 more or less does not go through Gascogne.
It’s pretty common for the signs in a country to match the main language of that country, and sometimes those names will be pretty different to the actual name in the foreign country.
Therefore, if there’s some language in the world where a UK placename renders as “Brittany” in that language, that would make perfect sense! It doesn’t happen i Dutch with Brittany (en) / Bretagne (fr), as that’s also “Bretagne” in Dutch. Personally I see “local, especially English, versions of foreign names” as a bit of an anachronism, but I can understand why they exist. Google reckons Golfe de Gascogne == Golf van Biskaje and Bay of Biscay, and it seems plausible to me.
(and apologies for veering even further offtopic here)
“Lange-AfstandsWandeling” = Long distance hike=Grande Randonnée
Dutch LAW are reused (in part or as a whole) as the national sections of the European hiking trails. All the national and international trails use the white-over red symbol. All are bidirectional, except for the Westerbork trail, which is trailblazed in one direction with the Westerbork camp and monument as the unavoidable end point.
European route signs are rare, hikers are supposed to pick the right LAW then follow the white-red symbols.
The Camino is not systematically signposted. Probably because there isn’t a fixed Camino route.
1 Like
Here in France, more and more Camino routes are signed, but there is some complexity:
- some routes have been turned over to FFRP for management, and they are signed in white-over-red.
- some are managed by local associations and signed mostly in yellow-over-blue…
- … including variants of the white-over-red routes. See along GR 655 for examples of that. We are in the process of mapping those yellow-blue routes independently.