Maybe also update the foot:backward wiki? It currently says:

In the past, some mappers have used oneway=yes for this purpose, but it goes against the definition of the oneway-tag which applies only to vehicles.

  • oneway=yes is still used like this (not only in the past)
  • it does not explain why oneway only applies to vehicles

Maybe something like the following (I’m not a native speaker):

Using oneway=yes for this purpose is problematic, because in the majority of cases it only affects vehicles (e.g. highway=residential+oneway=yes). It might be obvious that it is meant to affect pedestrians for something like highway=via_ferrata, but for highway=path or highway=footway this is not clear, as it could also be meant to affect bicycles, but not pedestrians.

And similarly for the foot:forward wiki, or just linking these wikis to each other.

Er, no. highway=footway; oneway=yes is not ambiguous.
It only starts becoming ambiguous when you start saying that “other modes of transport can also access that footway” which is very much the exception rather than the rule.

2 Likes

23 900 000 x highway=footway

  • 15996 (0.07 %) x highway=footway + oneway=yes (4410 x in Germany)
    • 5994 (37%) x highway=footway + oneway=yes + bicycle + bicycle!=no (3371 x in Germany)
    • 62 x highway=footway + oneway=yes + motor_vehicle + motor_vehicle!=no
    • 7 x highway=footway + oneway=yes + vehicle + vehicle!=no

So it is 37%, not very much an exception.

More statistics:

  • 8831 x highway=footway + oneway:bicycle=yes (3935 x in Germany)
  • 535 x highway=footway + oneway:foot=yes
  • 97 x highway=footway + foot:forward=no or foot:backward=no
  • 77 x highway=footway + bicycle:backward=no

14 500 000 x highway=path

  • 57112 (0.4%) x highway=path + oneway=yes (26424 x in Germany)
  • 16093 x highway=path + oneway:bicycle=yes (8869 x in Germany)
  • 235 x highway=path + oneway:foot=yes
  • 137 x highway=path + bicycle:backward=no
  • 97 x highway=path + foot:backward=no
  • 6 x highway=path + foot:forward=no

72177 x highway + oneway:bicycle=yes (16682 x in Germany)

  • 34420 (48%) x highway=residential|unclassified|tertiary|secondary|primary + oneway:bicycle=yes
    • 32096 (93%) x highway=residential|unclassified|tertiary|secondary|primary + oneway:bicycle=yes + oneway=yes (redundant)
  • 16093 (22%) x highway=path + oneway:bicycle=yes (8869 x in Germany)
  • 8831 (12%) x highway=footway + oneway:bicycle=yes (3935 x in Germany)
  • 6909 (9.5%) x highway=cycleway + oneway:bicycle=yes
  • 43170 (60% ) highway + oneway:bicycle=yes + oneway=yes (redundant)

1188 x highway + oneway:foot=yes
620 x highway + bicycle:backward=no
262 x highway + foot:backward=no

3 Likes

Well, I hope after 183 comments that it is clear to everyone that considering oneway=yes for pedestrian is most of the time wrong and that highway=* which are only allowed in one direction for pedestrians is a rare case.

Even if we find some situations where it might not be ambiguous, I have to ask if we want to make exceptions (plus probably exceptions from the exceptions) or if we want to make it clear and consider oneway=* only for vehicles (plus horses?).

Regarding editors, presets could include the information that oneway=* is not for pedestrians and offer another tag.

Regarding oneway:foot vs foot:forward/backward, I would prefer a clean cut and use the latter.

5 Likes

There are so many different opinions about how people should tag pedestrian one-ways, that’s why I avoided any language suggesting that people should tag a certain way or not in my Wiki edits, and just focused on documenting the current situation.

If we can agree how people should tag then maybe a proposal would be useful?

1 Like

Most people here seem to agree we need to find some solution, but have different ideas about what that tagging scheme should look like.

There are those who say that footway with oneway=yes (to mean a pedestrian one-way) is perfectly fine, and there are those who think it is problematic. Among those who think it’s problematic, some argue in favour of oneway:foot and others argue in favour of foot:backward. So I think we’ll need to find a compromise!

How would people here feel about a proposal along the following lines

  • oneway on footway is ambiguous: you can make a good guess what the mapper meant if you also evaluate other tags, but having to look at other tags to understand what oneway means is far from ideal
  • we should establish an unambiguous way of tagging that a highway is oneway for pedestrians. There are two competing tags for this (oneway:foot and foot:backward), but if we want editors to offer something other than oneway in their footway presets, the community should decide which alternative is preferred
  • even if we discourage the use of oneway on footway to mean one-way for pedestrians, this will never completely remove the ambiguity around the oneway tag (we don’t know if the mapper who put oneway has read the Wiki)
  • therefore it’s better to just not use oneway on footway at all. Better to be specific e.g. oneway:bicycle
  • given how similar footway and path are, it could make sense to include path in the proposal as well. For example, for separately drawn sidewalks with a one-way cycleway, it is clearer to use highway=path foot=designated bicycle=designated oneway:bicycle=yes instead of highway=path foot=designated bicycle=designated oneway=yes
  • therefore we suggest to mappers to avoid using oneway on footway and path, in favour of the more specific tags. We encourage editor developers that presets for footway should not include oneway as a key. Better to present the unambiguous tags to mappers

This proposed solution won’t make everyone happy:

If your view is that oneway already only applies to vehicles and a pedestrian one-way footway with oneway=yes has been tagged wrong, then you won’t be happy about effectively deprecating the tag on footway.

If your view is that oneway on highway=footway for pedestrian one-ways is perfectly fine, then you won’t be happy about it either.

But I don’t think there is a solution that everyone is happy with.

Quick opinion poll:

  • Good plan
  • Bad plan
0 voters

Thanks for the nice summary

You may add also to
encourage renderer to draw the prefered oneway:foor or foot:backward instead of oneway on footways and paths

I still think it’s enough to have a clear solotion ro express “oneway:foot”

How about a poll before starting defining the compromise. This may shorten the discussions during the proposal.

From the numbers of usage I think foot:backward hasn’t proven successful and oneway:foot seams to be more intuitive

I suggest to move this proposal of a proposal and poll to a separate thread. I (and maybe others) mostly stopped reading here when there were half a dozen responses per hour which made it impossible to follow and answer.

2 Likes

It only starts becoming ambiguous when you start saying that “other modes of transport can also access that footway” which is very much the exception rather than the rule.

While it is not the most common case, it is not really uncommon actually, there are more than 150.000 highway=footway with bicycle=yes in Germany alone.
https://taginfo.geofabrik.de/europe:germany/tags/highway=footway#combinations

23 900 000 x highway= footway

  • 1132199 (5%) x highway=footway + bicycle + bicycle!=no
    • 845523 (75%) x highway=footway + bicycle=yes
    • 108777 (10%) x highway=footway + bicycle=designated
    • 105455 (9%) x highway=footway + bicycle=dismount
    • 35182 (3%) x highway=footway + bicycle=permissive
    • 21190 (2%) x highway=footway + bicycle=private
  • 22732 x highway=footway + motor_vehicle + motor_vehicle!=no
  • 2724 x highway=footway + vehicle + vehicle!=no

13825 x sidewalk(:both/left/right):bicycle=yes

In Germany:
2 024 000 (8.4%) x highway=footway

  • 184128 (9%) x highway=footway + bicycle + bicycle!=no
    • 153458 (83%) x highway=footway + bicycle=yes
    • 16614 (9%) x highway=footway + bicycle=designated
    • 8789 (5%) x highway=footway + bicycle=dismount

950 x sidewalk:right:oneway=yes / sidewalk:left:oneway=-1 (only occurs in Germany) (refers to cyclists)

2 Likes

16000 oneway=yes with footways according to taginfo, 441 of them have a motor_vehicle=* tag according to overpass turbo, but only for 62 ways it is not no.

I also checked for footways with oneway=yes and motorcycle not no, there are 12 in the world.

Then I checked for path with oneway=yes and motorcycle not no, there are 513 globally, or with motor_vehicle not no there are 151.

These numbers suggest we can review all cases in a relatively short time frame, provided there are mappers in the area.

1 Like
  • 54878 x (0,23%) highway=footway + oneway=*
    • 38 704 (71%) highway=footway + oneway=no
    • 15 996 (29%) highway=footway + oneway=yes
  • 11 436 (0,05%) highway=footway + oneway:bicycle=*
    • 2 570 (22%) highway=footway + oneway:bicycle=no
    • 8 831 (77%) highway=footway + oneway:bicycle=yes

  • 54878 x (0,23%) [highway=footway][oneway]
    • 22022 (40%) [highway=footway][bicycle][bicycle!=no][oneway]
    • 658 [highway=footway][motor_vehicle][motor_vehicle!=no][oneway]
    • 379 [highway=footway][motorcycle][motorcycle!=no][oneway]
    • 70 [highway=footway][motorcar][motorcar!=no][oneway]
    • 49 [highway=footway][vehicle][vehicle!=no][oneway]

Thanks, I’ve made a new thread to decide between the two unambiguous alternatives.

I imagine a cycling map will want to show whether the way is oneway for bicycles, a hiking map will want to show whether it is oneway for hikers. With clearer tagging we’ll make it much easier for both to interpret the tags. Therefore I am hesitant to include a rendering suggestion.

What I mean is that a renderer should not show deprecated or unclear tags.

osm-carto draws arrows for oneway - even on footway where it is not clear what is meant.
If we discourage to use some tags they should not be drawn any further. Instead oneway:foot (or foot:backward) or/and oneway:bicycle should be considered on footways

1 Like

… but there’s no way to enforce that. You’d have to try and persuade whoever looks after the renderer that “it’s a good idea not to show X”; if they want to keep showing that tag, they can.

As an example, have a look at https://taginfo.openstreetmap.org/projects/someoneelse_style#tags. Some of the low-value tags supported there are labelled as “obscure synonym for…” - they aren’t the best or even a vaguely sensible way of tagging something, but it was clear what the mapper meant.

2 Likes

you also can not force a editor programmer to change something but this is included in the proposal.
You can try to influence it, especially for the default map shown on openstreetmap.org

1 Like

The idea is to provide some certainty to developers of editors and QA tools that the community is happy with oneway:foot= being included in a preset, and happy with oneway= triggering a validator warning. Of course, any suggestions are just that, suggestions.

The formal RfC has now started.

2 Likes

The idea is to provide some certainty to developers of editors and QA tools that the community is happy with oneway:foot= being included in a preset, and happy with oneway= triggering a validator warning. Of course, any suggestions are just that, suggestions.

they would also have to support foot:forward/backward anyway, as it is standard tagging, so basically the proposal asks to introduce/confirm an additional variant for the same thing, correct?

1 Like

the “standard-tagging” foot:backward=no is rarely used (only 262 x) compared with oneway:foot=yes 1188 x and a much higher number of simply oneway=yes on pedestrian infrastructure. So oneway:foot will not be introduced by the proposal, it is is already in use.

3 Likes

Both already exist. The proposal just clarifies that using one of them is mandatory for pedestrian oneways.