Modeling multiple Eurovelo route alternatives

Hi,
In country of Serbia, there is 3 Eurovelo routes (6, 11 and 13). Local NECC placed fingerpoint signs, but for each of these routes, we have lot of “alternative” routes and “detour” routes. When I look how Eurovelo is tagged in other countries, I am not seeing these “alternatives”. @Florange_Grimoire told me that this is not per Eurovelo guidelines, but we want to map those “alternatives” (this all started from topic in Serbian community where we are surveying those signs). If you don’t know what I am talking about, here is info table explaining it (it is in the upper right corner, there are 3 different colors in fingerpoint signs, green is alt). Here is one such sign in the wild.

On-the-ground truth map of signs looks like this:

(official/pink route is on embankment following Danube with ground surface, while “alternative”/blue route is following asphalt road, and signs are in green circles). Here is overpass query with signs and routes. There are around 5-10 examples of these alternative parts on these 3 Eurovelo routes.

We are now wondering what is best way to tag them:

  • One relation (per Eurovelo route) - all alternative ways and detours are in same relation.
  • Two relations (per Eurovelo route) - one relation for main route and one relation with “alternative” route. Relations would share same ways when they do not differ, and would split in places where they differ
  • Multiple relations - one main relation and one additional relation for each “alternative” detour (“name” tag would contain some suffix to differentiate them).
  • Something else?

To answer some questions in advance:

  • No, our NECC is not giving us data for routes and for fingerpoint signs, we are surveying them manually (overpass query above is for that purpose)
  • No, we cannot take data from Eurovelo website, as what is there is having lot of errors and differences than on-the-ground truth (funnily enough, it is more wrong than right in all cities we surveyed :smiley: )
  • Yes, we do want to map all these alternatives, even though it is not official guideline - they are sometimes faster, sometimes with more sightseeing and IMHO, they are worth visit more than official routes (and signs on the ground could confuse cycling if alternatives are not on the map).

We don’t know what data consumers are using those Eurovelo data and what would be best for them to model those routes? If it is just for “visual” aspect on cyclist maps, we can go with either approach (one relation seems easiest to maintain), but we are not sure what else there is. What are your thoughts?

2 Likes

While mapping the Appennino Bike Tour, a mere > 3000km in total in 44 stages, did come across route parts of other relations that had the ‘alternate’ role. Of course JOSM does not like it when you touch such a way and then run the validator, but it is a role in use.

What are your thoughts?

if these alternatives are signposted you should add them, the question is if they are just local routes or alternative parts of the eurovelo route? If you decide for the latter, you could have a superroute where these alternative parts are added with appropriate role. Although the description says “ Used to group multiple continuous routes into a parent relation” you can see that “alternative” is among the used values: https://taginfo.openstreetmap.org/relations/superroute#roles

1 Like

They are.

Alternative parts

That’s also a good option (I hope no data consumer will complain that there is now another superroute inside main EV superroute :smiley: ). Though it doesn’t help me with original question - how many relations should be inside that superroute - 1, 2 or many?

Relation:superroute#Alternatives points to Creating super-relations for routes and that (page) clearly mentions alternative routes and has also something about data-consumers.

For type:route alternative is used much more often and it looks to me superroute is just a route with the limitation that it has only sub-routes.

I expect every data consumer that supports routes with subroutes/alternatives and that also supports superroutes als will support alternatives for superroutes.

I hope no data consumer will complain that there is now another superroute inside main EV superroute :smiley:

IMHO you don’t need a superroute inside another one, the alternative could be a normal route (unless there is good reason for structuring it with more complexity)

I suspect not. At least in the bits of that network that I am familiar with, there’s already quite a hierarchy, for example:

Alternatively, you can use a single relation and describe alternatives and excursions with roles:

https://wiki.openstreetmap.org/wiki/Roles_for_recreational_route_relations

That was voted on relatively recently (2020?! how the time flies!), so it’s probably not in widespread use, but it’s straightforward and logical.

As I mentioned earlier, came across ‘alternate’ as route roles, this wiki link to recreational new to me and documenting the more frequently used ‘alternative’ so will certainly keep that in mind should I come across this ‘alternate’ role use again and update.

For what it’s worth, alternate (as an adjective) is chiefly American usage, while alternative is British (and usually better understood internationally, since alternate may be confused with alternating [e.g. current]).

Since the observations were in Italy I go with British and how it’s written in the wiki ;o)

Confusingly, there is also state=alternate, documented as a tag on the route relation rather than a role for member relations. Yes, “alternate” not “alternative”.

It’s possible that if you come across the “alternate” spelling for roles it’s because someone had seen the state tag previously - I think I did that myself a couple of times when I first mapped alternative roles.