Stage=* and name=* in long-distance hiking routes and superroutes

General talk Tagging general discussion
1

This question arises from a tagging discussion about the Sentiero Italia long-distance hiking route (example relation: OSM Relation 6925687), where stage codes were added to name=* even though stage=* exists. I am using this as a concrete example to clarify how wiki guidance should be interpreted.

The wiki page on names states: https://osm.wiki/Names

Name is the name only. The names should be restricted to the name of the item in question only and should not include additional information not contained in the official name such as categories, types, descriptions, addresses, refs, or notes.

The stage=* key exists specifically to store stage identifiers:

https://wiki.openstreetmap.org/wiki/Key:stage

Including stage codes in name=* appears to introduce redundancy and reduces the semantic distinction between object names and structured identifiers, which are already handled by dedicated tags.

Question:

According to wiki guidance, identifiers and codes should not be included in name=* and should be placed in dedicated tags. In this case, the stage code is present in name=* while stage=* is already set; should the stage code be removed from name=* to comply with this principle?

2 Likes
Namen von Wander/Radrouten und Routen-Etappen
Mappatura Sentiero Italia CAI
2

I’ve been following the discussion on the Italian channel and would like to add a few considerations (TL;DR: I don’t have a definitive answer).

First, the wiki page on names also states:

However, if a street is officially named “East 110th Street”, this should be the name notwithstanding the fact that the “street”, the “110” and “east” might be deducible from some other information; likewise, “Ho Chi Minh City” should remain intact even though the place is tagged place=city.

So adding something like “Stage XXX” is not automatically wrong in absolute terms, if that is effectively how the stage is identified.

Second, the name of a specific stage XXX is certainly not just the name of the larger route it belongs to, “Sentiero Italia”. Otherwise we would be confusing the part with the whole, and all stages would end up being indistinguishable. The purpose of a “name” is precisely to distinguish an item from others of the same type in the same area.

Furthermore, according to the page on Relation:superroute, “Sentiero Italia” should probably be mapped as a superroute. While this is slightly off topic, the page gives as an example the superroute “Euroroute R1”, whose components are named “Euroroute R1 - XXX”.

Similarly, the page on Roles for recreational route relations shows an example where the parent route is “Floris V-pad” and the components are named “Floris V-pad XXX”.

In summary, I would not exclude a stage name in the form “Sentiero Italia XXX” (with XXX to be defined appropriately), nor would I leave name=* empty. What I would definitely exclude is calling every single stage simply “Sentiero Italia”.

I think we should look at how the stages are actually named on the ground or locally. If the stage code is part of the way the stage is publicly presented - for example on signage or in official material - then including it in name=* would be consistent with OSM principles. If it is merely a structured identifier, it should probably remain only in stage=*.

have a nice day ~ Nicola

2 Likes
3

I have noted the same issue with the well known Camino de Sanitago which has the name Jakobsweg in Germany. This construct is a huge network of pilgrimage trails covering all of Europe, being split into hundreds of stages.

The name of the whole network in Germany is Jakobsweg and nothing else. Anyhow mappers started to invent names for the stages like

Jakobsweg from Mytown to Yourtown being part of a superroute
Jakobswege in East Germany being part of a superroute
Jakobswege in Germany

I have not touched any of these relations so far, just noted the misuse of the name tag and agree that the stage information should be placed in the stage tag only. Having said that I am not sure if this works well with a nested network like in the example:

name=Jakobsweg + stage=Mytown to Yourtown as part of the superroute
name=Jakobsweg + stage=East Germany as part of the superroute
name=Jakobsweg + stage=Germany

being finally is a part of the entire network (which is not mapped as a superroute so far).

Would that fit?

No, the purpose of name is to document the real name only, nothing else.

In case of route relations for motorways the “name” is represented by ref usually and even if such a motorway covers hundreds of kilometers we do not add any stage information to the ref. Same with roads having a name like the Great Ocean Road in Oz. The name is Great Ocean Road and nothing else.

2 Likes
4

In some European networks of routes in at least the UK, from and to are also used a lot (see e.g. here, which was a just a random click in Waymarkedtrails).

2 Likes
5

Oh yes, from and to are additionally tagged for each of the partial routes … probably to confirm that this section is REALLY leading from “A” to “B”.

3 Likes
6

The name of the route is not something you make up yourself, it needs to be a verifiable fact like everything else in OSM. A path being called “South West Coast Path” is a verifiable fact - there are signs saying this. A path being called “South West Coast Path, Mousehole to Penzance section” is not.

We do occasionally admit glosses for ease of editing. The from and to tags are good examples of this. But the name tag is user-facing, not mapper-facing.

Terrible route names are the reason Nominatim doesn’t find routes, and also the reason that cycle.travel (and no doubt many other data consumers) has to have many lines of code to screen out this sort of gloss from the information presented to users.

6 Likes
7

I was referring to “name” as a general concept, not specifically to the name=* tag in OSM.
What I mean is: the name should match the on-the-ground usage. And on the ground, people give names to things in order to identify them.
Sorry if I’m overcomplicating this - just trying to clarify my point.

name=Jakobsweg + stage=Mytown to Yourtown as part of the superroute

I don’t think the “real name” of the route from Mytown to Yourtown is Jakobsweg.

“Jakobsweg” refers to a much larger network. That specific segment is just a part of it. So either that stretch is actually called something more specific by people (for example, “Romantic stage of the Jakobsweg” or “Jakobsweg stage 651”), or it simply has no distinct name.

And that’s not a problem. The name already exists at the superroute level. Not every segment necessarily needs to inherit or duplicate it unless that is how it is referred to on the ground.

3 Likes
8

I’d say the difference in between the general concept and the use in OSM is not that much. The OSM rule is to use real names only, not descriptive terms nor any other additional information. And that is not much different in real life imo.

People use different terms to identify objects. Sometimes an object has a name, sometimes people use merely a description. In a village people will say “I drop my kids at the school” or “I’ll be at the cemetary in the afternoon”, but neither “school” nor “cemetery” is a name. Anyhow it serves to identify the object. Only in case there are 5 cemetaries around, having distinctive names like “King Kevin cemetary” or the like etc. it becomes a name, in reality as well as in OSM.

Not at all. The name of all the segments is “Jakobsweg” in Germany and people using it will just say “I’ll have a walk on the Jakobsweg tomorrow” which is quite specific as in most regions there is only 1 “Jakobsweg” and not 5 or 6. If someone wants to add details they will say “I’ll have a walk on the Jakobsweg near Point Alpha” or anything like that, adding a description to the name.

I’d say it is very much the same with the “Sentiero Italia”. Someone might say “I’ll have a walk on the Sentiero Italia over the weekend, I’ll give section xxx a try”, and as such adding some descriptive details to the name “Sentiero Italia”.

Another quite common case are ways having a separate local name. If the Jakobsweg near Point Alpha has the local name “Mousehole trail” people will probably use that name. In such cases we can tag the name “Mousehole trail” directly to the path or track (or a separate route relation), wheras the route relation of the Jakobsweg keeps the name “Jakobsweg”. Btw. all the sections of the Jakobsweg are marked with the same symbol (St James mussel) representing the “Jakobsweg” without further details. To my knowledge you would not find a sign anywhere showing a distinctive section or area name.

1 Like
9

Regarding the name Sentiero Italia - Tappa XXX it seems to me that it hasn’t emerged as an official name.
What I notice most often is the use of the code “SI” which corresponds to the value of the ref=* tag

On the official website, for exampe (https://sentieroitalia.cai.it/tappa/?id=SI+A01), the visible texts are:

  • Tappa

  • SI A01

  • SI A01: Rifugio Calvi - Rifugio Lambertenghi

On this sign ( File:20220716-lemma tartano valleve-101.jpg - Wikimedia Commons ), the texts visible are:

  • SI tappa n.D17S

  • Rifugio Balicco - Rifugio Dordona

On the example image in Key:stage - OpenStreetMap Wiki , the sign shows:

  • SI

In the FVG (Friuli-Venezia Giulia) region I have personally observed signs, and CAI FVG (official organization) confirmed that signage uses:

  • SI
1 Like
10

It depends? As an example, here someone has stickered route markers with “Camino Ingles to Santiago” (sic). That literally is the name of that route, one of a couple that runs along the river there. The local names for the actual ways vary - one part has a name of Cherry Tree Avenue, but that is different to the name of the route.

In the example I gave earlier, the segment of the John Muir way is part of both the complete John Muir Way (which I suspect will be signed there as “John Muir Way”) and part of the E2 (which I suspect won’t be signed as such but documentation will be available that says “follows the John Muir Way”)

11

I don’t know about that but the website is named “sentioriitalia” and SI stands for Sentiori Italia as well.

I would tag this as

name=Sentiori Italia
stage=A01 or alternatively SI A01
from=Rifugio Calvi
to=Rifugio Lambertenghi

name=Sentiori Italia
stage=D17S or alternatively SI D17S
from=Rifugio Balicco
to=Rifugio Dordonna

1 Like
12

The name sentiero Italia and the SI code are correct and refer to the entire route:

name=Sentiero Italia
ref=SI
type=route (maybe "type=superroute" would be better)
1 Like
13

Yes, I also think that’s correct.
Since ref=SI is also present, I think this is more correct:

name=Sentiero Italia
ref=SI
stage=A01
from=Rifugio Calvi
to=Rifugio Lambertenghi

2 Likes
14

I’m a bit confused by the description of stage=* in the wiki:

it can be used to indicate a reference number (and other stage specific data) for the stage.

This sounds like a duplication of ref=*, plus some vaguely defined “other data”. It’s not very clear what problem stage=* is meant to solve that isn’t already covered by ref=*.

On the other hand, the How to map section seems to suggest a different model:

It is suggested to map each stage as its own route relation, with name, from, to, ascent, descent, duration:forward etc. referring to the individual stage, and then combine the stage routes with a superroute relation into the whole route, which will get the attributes of the complete route.

Following that logic, I would expect something like:

Superroute:

type=superroute
name=Sentiero Italia
ref=SI

Individual stages:

type=route
ref=A01
name=SI A01 or Sentiero Italia A01, ... (*)
from=Rifugio Calvi
to=Rifugio Lambertenghi

(*) Let’s take a step back: why do we have these stages in OSM in the first place?

As far as I understand, they are not (completely) signposted on the ground. Do they exist in OSM because they are defined on sentieroitalia.it?

If that is considered a valid reason to map them, then logically the stage names should also be taken from the same source.

2 Likes
15

I see. And in your example the complete John Muir Way has a different name=John Muir Way to its members

name=John Muir Way (Helensburgh to Balloch)
name=John Muir Way (Balloch to Strathblane)
...

That’s the point of the discussion: if a superroute and all its members (route stages) should have the same name (that of the superroute) or individual names.

16

Keep in mind that the wiki says a name must be just a name, with no codes or anything else, so in my opinion, your example is correct, except for the name, which I don’t think is a name.

Regarding the rest, yes, the stage should be mapped as a separate object with any stage-specific tags, if any. For example, in this case, the stage name doesn’t exist, in my opinion, and therefore shouldn’t be set.
Therefore, the current name=Sentiero Italia, according to this interpretation, shouldn’t be set because there’s no stage-specific name; only the name of the entire route exists. And in the wiki for stage, it specifies that missing tag values ​​are inherited from the parent route; in this case, the stage would inherit the name of the parent route.
In fact, in the Italian discussion, one mapper proposed leaving the name blank. I don’t know if this is perfectly valid, but it’s plausible.

I partially agree, but the A01 is a stage, not a ref. Furthermore, for me a name should be just a name, with no codes or anything else, so I think this is more correct:

type=route
ref=SI (empty would be correct?)
stage=A01
name=Sentiero Italia (empty would be correct?)
from=Rifugio Calvi
to=Rifugio Lambertenghi

17

The names should be actual, attested names, not something made up by OSM mappers. Whether they are the same as the parent route is not necessarily the issue.

For example, the Stratford Greenway is a cycle route from Stratford-on-Avon to Long Marston. It is absolutely valid to create a route relation for that with name=Stratford Greenway. It is also part of National Cycle Network route 5, so you can put that relation into a parent route relation with network=ncn, ref=5.

What isn’t valid is to create a relation with name=National Cycle Network route 5 (Stratford-on-Avon to Long Marston), because that isn’t a name used anywhere on the ground.

3 Likes
18

I agree with you - I personally don’t think that naming members things like name=John Muir Way (Helensburgh to Balloch) is correct because it doesn’t match the actual name on the ground. I haven’t jumped in and edited that example because it’s not one that I have surveyed recently and it’s clear that there’s still a discussion (this topic!) to be had - some people want “made up names” so that they come through in the name field in JOSM**. iD combines the name with things like from and to to make up the words it shows to the user.

Like @Richard above I have consume this data and have lots of code to deal with silly names - special-casing some high profile offenders and chopping off the name at a certain length (for results see here)

** yes, one person really did say exactly that. Other people make up names “because the wiki tells them to”.

2 Likes
19

Sorry for mutilating the name from “Sentiero Italia” to “Sentiori Italia” … my fingers sometimes make up their own life on the keyboard … :flushed_face:

Yes, ref in addition to stage is possible imo but one could argue that the ref for one of the states should include the stage numer, like ref=SI A01 for instance. I think this is a matter of agreement among the mappers engaged in mapping this trail.

20

In reality, both ref=* and stage=* have been mapped for a long time, and there’s no one contesting the mapping of these tags. The dispute only concerns the name=* tag.