Proposal talk:Social path

From OpenStreetMap Wiki
Jump to navigation Jump to search

I oppose this tag. It breaks existing data users by elevating one minor attribute to control the entire tag. The OSM way for such attributes is to add an additional tag, such as access=no, or officially_maintained=yes. The former is already in wide use, and there will be far less objection to the latter, because an extra tag used in a private render doesn't harm those who don't understand it. Finally, removing unofficial paths from maps is not clearly the appropriate step; showing them and marking them prohibited conveys more information and makes navigation easier. --Gdt (talk)

Please re-read the proposal. The "paths" this tag refers to were never intended for navigation (they're not actually paths despite appearing as such due to repeated use; consider erosive shortcuts across switchbacks), thus official recognition / prohibition is secondary and used more to facilitate remediation than to control access (common sense should lead to people avoiding these "paths" when seen IRL). Tagging them as highway=path is actively harmful, as this further contributes to the illusion that they should be used.
Splitting these out from highway=path follows the rules of thumb set out in How to invent tags (each tag should stand on its own).
Having mostly hiked in the US/Canada, it's possible that social norms relative to trail use are different here.
--Mojodna (talk) 13:04, 25 March 2016 (UTC)
By "makes navigation easier" we mean not necessarily navigation of the path in question, but navigation of the official paths. The intersections between the official paths and these other paths are themselves landmarks. "Oh, I must be on the right path as my map shows an intersection with a path that isn't open to the public, and... here it is!". They would also be useful in the case if someone accidentally got lost and got off the official trail. This does happen to even experienced hikers, trail runners, mountain bikers and other trail users.Tekim (talk) 23:06, 25 March 2016 (UTC)

I am still opposed. If this were limited to very short short-cuts across switchbacks, a norm to not map them would be fine. But around me, there are real trails, that people really use, in conservation land -- except that they are not sanctioned by the authorities. On the ground, one can't tell. I think it's very telling that this tag is arising from a desire by authorities to control maps of their properties, not from a desire to solve a problem that mappers have conveying information about the world to map users. --Gdt (talk)

The trails you're describing do not sound like social trails (as defined here) and I agree that they should be tagged as you describe.
It feels that the fundamental issue is tagging the intentionality behind the existence of trails (which can sometimes be inferred on the ground, but not always). It also feels like the status quo is to map the symptoms, not the cause (which is great if we're informed by conditions on the ground, but with the inclusion of mappers affiliated with land management agencies, we gain the ability to benefit from their knowledge about intentionality).
--Mojodna (talk) 14:09, 25 March 2016 (UTC)
We map what is on the ground, not what someone intends, or intended, to be on the ground. If looks like a trail, has the physical characteristics of a trail, is used as a trail or could be used as a trail, we map it as a trail. If there is a sign or other indication stating "No Public Access" we capture that information with the "access=*" tag.Tekim (talk) 23:17, 25 March 2016 (UTC)

I should not that I have no objection to somehow marking trails that the authorities do not like, and indeed I am sympathetic to that notion But it should be an extra tag that if not understood is not harmful, not changing a primary key, and not breaking data for other consumers (e.g., those that don't care about authority).

a bunch of arguments/discussion on this topic occurred on talk-us mailing list : so far, everybody weighting in opposed changing an existing tag and suggested to use additional tags to mark the status (whether it's 'access' or anything else) --Richlv (talk) 22:47, 25 March 2016 (UTC)

I oppose introducing this tag for the reasons given by Tekim and Gdt.--geow (talk) 23:31, 25 March 2016 (UTC)

I vehemently oppose introducing this tag for the reasons given by Tekim and Gdt. The access=no tag means what it says, or access=(something_else) might be proposed. This is not about the PATH itself as a DIFFERENT kind of "highway" distinct from a plain-old garden-variety path: THAT is what this is, a PATH. Yes, I agree, it has a distinct additional semantic that we want to capture. The existing tag for access does this, as ACCESS (=no) is what makes this path different from an otherwise free-to-use path. Without seeming like I mean to take a cheap swipe at Stamen for having done this (I don't wish to do so), it seems that the introduction of "social path" was a quick, dirty and unabashed shortcut to having a renderer (one Stamen seems to be producing or co-producing) rather simply scan for this tag, and render it differently. That sound sneakily, suspiciously and uncharacteristically like "tagging for the renderer," something OSM wants to heartily discourage. While I deeply appreciate the wonderful plasticity of free-form tagging inherent in OSM, I encourage PROPER tagging using EXISTING tags like access=no in addition to the accurate tag of highway=path as its primary identifier of what it is. If access is so subtly-defined in the cases of what is meant by "social path" then let us have new values introduced for the key access: perhaps "emergency_only" or "permit_required" or "trail_building_while_erosion_control_is_underway" (well, I exaggerate a bit with that last one). But let's load up the values, not invent vague keys, of which nobody or very few know the meaning. I have never heard of "social path" and I listen for these sorts of things. The mountain bike trails on UC Santa Cruz' Upper Campus, and as they extend into the western portion of Henry Cowell Redwoods State Park are a good example of "extremely closed trails -- don't you dare ride a mountain bike here, as the erosion and destruction you wreak is terrible." Tagging these access=no causes mapnik to add red dashes, and that seems to keep OSM in good stead for now. These examples should be used where the same semantics are true. ("Social path:" BAH!) In vernacular, this is a "closed path." --Stevea (talk) 23:54, 25 March 2016 (UTC)

From the proposal "Social trails are considered unauthorized by managing agencies": not always! There are many social trails in RMNP and Roosevelt National Forest in Colorado and the relevant land managers generally have no objection to these being used. In fact this is how many of the peaks and climbing cliffs within those lands are accessed. They are covered in many guidebooks (some of which the park service sells in the visitor center).Tekim (talk) 23:59, 25 March 2016 (UTC)

Yes. Here's an example in Ilkley, UK where OpenStreetMappers have detailed a lot of the little paths which have formed around these interesting rocks. Good details of what we see on the ground. They're mostly social paths. Would we want to see these retagged and then dropped from most map displays? I don't think so. Would the managing agencies want to see this? I doubt it. I think on the whole they would either find the maps of real paths quite interesting and like to see them stay, or they wouldn't care either way. So I think there's a problem with the proposal assuming that social paths should be hidden from most map displays.
Maybe the proposal should state that this tag is specifically for social paths which are problematic and which managing agencies are actively discouraging the use of (maybe it does state that. In which case it could be clearer and use a clearer tag name). I can see an argument that hiding such paths from map displays is the useful thing to do for most map users.
A few people have suggested access=no. It's a path that you're not allowed to use. I notice our standard style currently shows this with the pink splodges example. Maybe it should hide such a path. Or maybe some additional combo tag might trigger it to be hidden, but then there's the "useful for navigation" point some people have made.
- Harry Wood (talk) 02:50, 30 March 2016 (UTC)

From the proposal: "and it can also lead to dangerous situations where casual hikers find themselves in steep or otherwise hazardous terrain." The appropriate tag for this situation is sac_scale, e.g. sac_scale=difficult_alpine_hikingTekim (talk) 00:02, 26 March 2016 (UTC)

The proposal talks about bootleg trails which appear to be Tag:highway=path. Bootleg trails don't appear to be this tag, they are this tag, or Tag:highway=footway. Pnorman (talk) 02:12, 26 March 2016 (UTC)

I oppose introducing this tag because social_path has something positive and is not discouraging people from using the path. illegal_path, unwanted_usage, or something similiar coudlbe used in addition to access=no --Fx99 (talk) 10:29, 26 March 2016 (UTC)

1st: it's the right way, not to delete those social path but to use special tagging. 2nd: I prefer using existing tags over inventing new tags. That means that I prefer access=* instead of highway =social_path 3rd: assuming auch way,which is not allowed to use the first thing you see is a hw=path therefore I tag this. Then for whatever reason it is refined with access rights. 4th: non drawing does not help, as the way is still on the ground. => ich oppose--Hfst (talk)

American vs British English

Is "social path" or "social trail" much used in American English? "Social Trail" is mentioned on as a synonym for "desire line" (the usual "British English" term. OSM usually uses British English where there's a choice (on the "you might as well pick one standard and stick to it" principle), and it's worth noting that "path=desire" is already in use: . This is a separate issue to whether the features in Marin that were tagged as "highway=social_path" that started the ball rolling here really are desire lines - at first glance they seem not to be. --SomeoneElse (talk) 18:58, 25 March 2016 (UTC)

As a Canadian growing up with Americans, and having never heard the term "Social Path" before; I doubt it's all that common. I have heard of "cow path" before, but not in common use --Alan Trick (talk) 22:28, 25 March 2016 (UTC)
To clarify, "social path" is not commonly used in English (again, my experience is with North American English), it's actually "social trail". But we chose "highway=social_path" as a compromise because of the similarity with "highway=path". Perhaps "highway=social_trail" would've more accurately reflected the commonly-used phrase. --Alan (talk) 23:26, 25 March 2016 (UTC)
It's true that path=desire is already in use, although just barely, and it never went through the approval process. Also, they appear to be almost exclusively in urban areas (and mostly in the city of Nottingham, UK). This fits with my experience of American English, that "desire lines" is the term usually used for the unofficial paths through city parks, while "social trails" is the term used for rural parks and for hiking. But, if the existence of "path=desire" is evidence that this concept is acceptable OSM tagging practice, we would be happy to switch to using that tag, following OSM's British English convention. --Alan (talk) 22:18, 25 March 2016 (UTC)
note that "approval process" is not required, doesn't really change anything and is actually frowned upon by most long-term project members (not relevant for the current discussion, but a useful thing i had to learn myself :) ) --Richlv (talk) 22:44, 25 March 2016 (UTC)
As someone who probably added path=desire in the Nottingham area a few comments: a) we have been fortunate in Nottingham that several of us share a common vision of what we want to map & a strong desire to be completist about it, a side-effect is that I believe the city is a useful proxy for gaining an understanding of how likely one is to come across a feature in any mid-sized city in NAm & Europe; b) I did not invent the tag, I noticed it on something mapped by 80n so I suspect the tag was used rather early in OSM's history; c) I document tags I use I NEVER submit them to the proposal process. SK53 (talk) 10:35, 4 May 2016 (UTC)

key:informal is used for these

No need to reinvent the wheel, there's the informal=* that is already used for paths like these. --06:19, 26 March 2016 (UTC)

+1 --ChristianSW (talk) 07:23, 26 March 2016 (UTC)
+1 --De vries (talk) 08:26, 8 April 2016 (UTC)


The proposal not to render paths that should not be used is nothing but shooting in one’s own foot IMHO (there was a discussion recently on the German user forum about one user deleting paths within a natural reserve). Walkers still will arrive at the beginning of this trail – wonder where it leads to – look it up in the map rendering – on not finding anything, decide to explore – there you are. You are causing traffic instead of avoiding it. Better tag the trails as access=no or similar, rendering that information too, to have a clear statement in the maps and keep those trails from being used in planned trips. --Kreuzschnabel (talk) 08:31, 26 March 2016 (UTC)

++1! --ChristianSW (talk) 12:48, 26 March 2016 (UTC)
+1 --Hobbesvsboyle (talk) 20:52, 26 March 2016 (UTC)


By the very definition on the proposal a highway=social_path looks the same as a highway=path. How do you tell the difference when doing an on-the-ground survey? --Peter Mead (talk) 10:57, 29 March 2016 (UTC)

Related schemes

There are (at least) two proposed schemes for path quality which might relate to this. Neither are in proper use yet as far as I can tell - but they both are attempts to try to capture the same kinds of information you are looking at. I agree that there is a problem that we need to deal with - but I think that this proposal might be too narrow. Please see...

Proposed_features/pathtype Proposed features/pathtype(alternative) --Rostranimin (talk) 17:33, 6 April 2016 (UTC)

This is terrible case of tagging for broken renderers

It is pure tagging for renderer - what worse, there are already access tags covering this issues. highway=path, access=no is covering case of paths illegal to use. One may add widely used tags like surface, smoothness etc to describe in more detail. One may also add use/propose new tags giving extra detail, for example reason for access=no. Maybe social_path=yes may be a new one.

I see no point in creating a new tag for paths/footways illegal to use, given that current tagging covers it and among interested data consumers there is support for access=no/private.

For more discussion of this horrible idea see and Mateusz Konieczny (talk) 08:07, 8 April 2016 (UTC)


Thanks everyone for your feedback. Instead of proceeding with a vote on this proposal, we decided to withdraw it. We will use highway=path plus access=no and informal=yes instead. --Alan (talk) 23:43, 2 May 2016 (UTC)