In England and Wales, a good distinction between *=yes and *=designated occurs on most Public Bridleways. These are typically countryside routes that were originally used on foot and horseback, and the public has gained the legal right to continue to use them like this. It used to be unlawful to ride a bike on Public Bridleways, but a few years ago that offence was removed, so you can now legally cycle on a Public Bridleway. But while land-owners have a duty to maintain the route in a suitable state for walking and horse-riding, there is no equivalent duty for them to maintain the route in a suitable state for cycling. Hence the usual access tags for a Public Bridleway would be foot=designated, horse=designated, bicycle=yes - i.e. itâs legal to cycle there, but the route is not specifically designed/maintained for cycling.
Is that so out of principle or might such a bridleway merit a bicycle=designated, when in appropriate shape?
If a route along a Public Bridleway has also been signed as a cycle route and/or clearly designed as one, then yes, Iâd tag a bridleway as bicycle=designated. Depending on the physical appearance it might well also get highway=cycleway, rather than one of the other values.
In fact thereâs an example of this in the town where I live. An historic bridleway runs through what is now a housing estate, and the route has been made up as a shared-use cycleway, complete with the appropriate cycle route signs: Way: âȘGreen Lane⏠(âȘ138307275âŹ) | OpenStreetMap
You are of course free to think itâs not a problem. Perhaps itâs not for the use case you are thinking about. For the use case I am thinking about it is a problem. Iâm imagining making a map displaying walking trails, bike trails, and combined walk/bike trails each in a different color. So with trails mapped the way you suggest:
highway=footway: red linehighwy=cycleway: blue linehighway=path+foot=designated+bicycle=designated: purple line[1]
For publicly accessible trails this works just fine and displays each type in the expected color. But I also want to indicate accessibility with a different line style on this map. I want trails where only customers are allowed to be highlighted in light yellow but with the centerline still having the same red/blue/purple color for walk/bike/shared. Ok so:
highway=footway+foot=customers: red line, yellow highlighthighway=cycleway+bicycle=customers: blue line, yellow highlighthighway=path+foot=customers+bicycle=customers: purple line, yellow highlight
No problem so far, but what should I do with highway=path + access=customers + foot=designated + bicycle=designated? Some mappers would intuitively say this combination means the trail is for customers only and it is designated for walking and cycling. However, according to a strict interpretation of the access tagging hierarchy it actually means the trail is designated for walking and cycling by the general public, but customers have full access using any transport mode (car, horse, foot, etc). If I take the intuitive interpretation, I would give this trail line a yellow highlight for âcustomers onlyâ. If I take the strict interpretation, I would consider it a public path and leave off the yellow highlight. This ambiguity is not good.
I also want to make private trails dashed lines, but keep the same color scheme. What should I do with highway=path + access=private + foot=designated + bicycle=designated? Is it a fully private trail designated for walking and cycling but only for invited guests? Or is it designated for the general public to walk and cycle on, but the private group may use it with other transport modes as well? If the former, the line will be dashed. If the latter, solid. If the latter is correct, then a truly private trail must not be tagged with *=designated. But then information about what the trail is for is lost.
As long as permissive is a value worth having in the database, then it is also worthwhile to combine it with what the way is designated for. Otherwise, all you know is that walking and cycling are tolerated on a way, not that the way is specifically designed and designated for walking and cycling. Only a problem when using highway=path of course.
In reality the logic would be more complicated than this to account for various other combinations like
highway=path+foot=designated+bicycle=no. â©ïž
Most of it is possible now. Also, tagging entrances and the area a private carries information, which could help such a specialised application.
The result would be a map with an improvement wish. Once this map is there and the usage is overwhelming, Iâm sure this will trigger a support movement for better access tagging!
Seriously, I think access tagging packs different aspects into one value set, and one encounters situations where you would want to assign multiple values. Practically, I think for now we have to make do with the current access values, which already carry more information than most general purpose applications use.
If this becomes unacceptable for shared foot/bicycle ways mapped as paths+designated tags, it is up to the communities tagging this way, to come up with the issue and with solution. After all, it is their territory that is rendered/routed suboptimally.
Personally, I am happy with the tagging and the map results for cycling and hiking in Nederland, and I think most other countries still have a long way to go before the bicycle / foot tagging starts to fall short. Most of the improvements I think of are at the data userâs end.
Just a short roundup: access=designated (that is how it is documented) is just an alias for access=yes, when only concerned with access, i.e. the legal matter.
Lots of people though read more into *=designated. E.g. to them it conveys, that the legal right is also signposted, or that the feature is managed or somehow better usable than if merely tagged yes, that there comes an obligation from it, &c.
One consequence of this: Only the same body that can hand out yes can also hand out designated. In Germany or Austria e.g. that leaves us with the road code (StVO). In MUTCD countries?
Did I comprehend that fully?
access=designated is documented as without meaning, I think? Designation requires a specific mode of transport. A cycleway is a path designated to cycling.
If you look at access in a given situation, access is yes or no. You are either legally allowed to use the path or you are not.
bicycle=designated does not necessarily mean it must be signposted, managed, better quality or mandatory. It just means the path is officially and verifiably assigned to cycling: it is a cycleway.
There must be something to tell the cyclist that it is a cycleway, but it is not necessarily a signpost; it could also be lineage, surface colouring, bicycle logo on the surface, or any kind of design that indicates to people that it is a legal cycleway.
Since it is a cycleway, you may expect it is cyclable, but you cannot tell anything about the actual quality, surface, smoothness, width, maintenance, signage or if itâs mandatory or advisory.
With bicycle=designated, the cyclist has legal access AND you know that there is a much better chance that it can actually be cycled on than a non-designated highway=path, which just conveys you are allowed to use it, but gives no information that you could even actually cycle on it.
Who designates and who allows pure legal access? In Nederland, it is rather complex and involves several layers of authorities. I donât think itâs always the same authority. A path without signs can be used by cyclists, thatâs the legal access. Another authority can decide to add a cycleway sign, which adds the designation.
I think this goes over the top of my head. First you say, it has nothing to do with usability, then you say, designated conveys better usability over yes, or probably so â since when is openstreetmap about probabilities?
There is no âfunny thingsâ category, so I post the image here â higway=path;roe_deer=designated â look for the PoI in the back of the scene, a game feeding:
PS: In the German subreddit a topic was: Can we map something a footway even though it is not signed? It was about footways in residential housings. As far as I know, they are permissive â I have no legally enshrined law to use them, when not a resident.
It is the presumed identity of highway=footway and highway=path;footway=designated" that was introduced with path that breaks access tagging worldwide. A retroactive redeclaration of what a footway is.
PS: Possibly, in a residential area, paths to reach the entry to the building where you live are not even legally enshrined, they are just by contract. The conclusion in the German topic curiously was: If it looks like a footway, then tag it a footway, no matter the signage.
Welp, there goes my flimsy rationale for distinguishing between bicycle=yes and bicycle=designated in the U.S. based on different non-regulatory signs. 
Itâs hard to say with any certainty, since we donât have the tags to express this situation. I suppose a large share of these 20,674 foot=private bicycle=private ways would involve some sort of designation for pedestrians and/or cyclists, let alone other combinations.
Anecdotally, POPOs are quite common in my area, particularly at apartment complexes, due to zoning regulations. Any path inside these spaces would have permissive access based on signs that specifically allow certain modes of transportation while cautioning that permission can be revoked by the owner.
The private or quasi-public shared use paths winding through the campuses of Apple, Google, and Meta have bedeviled local mappers for years. As far as I know, there arenât any signs specifically reserving these paths for both pedestrians and cyclists, but the presence of company-provided bike racks and benches tells us all we need to know. Thereâs nothing particularly special about these tech companies; any campus of sufficient size can provide similar facilities. However, these are mapping companies; maybe we should go the extra mile and perfect their access rules. 
Does *=designated reliably communicate this information? This topic has turned up several different interpretations of the same access values:
yes |
designated |
|---|---|
| Presumably allowed | Allowed according to signs etc. |
| Allowed | Allowed regardless of signs etc. |
| Allowed | Allowed and suitable |
| Allowed | Allowed with priority over others |
| Tolerated | Welcome |
| Not prohibited | Allowed |
I really wonder how many of these interpretations hold up when applying the same values to features other than paths, such as amenities or protected areas.
It almost sounds like we need to pair highway=* with official_highway=*, just as we pair name=* with official_name=*. The designation=* key is still patiently waiting to have its moment in the limelightâŠ
In the U.S., the MUTCD only regulates signs, signals, markings, and road layout along public thoroughfares. The type of paths weâre talking about are regulated not by any traffic code but rather by trespass laws, plus any rules set by the park manager or private property owner. The property owner is free to mark a trail with a sign stipulating that you may only proceed when youâre in a good mood, hopping on one foot while patting your head and rubbing your tummy. 
The assignment as a cycleway is a legal thing. The physical suitability does not define a path as a cycleway. Many ways are suitable for cycling; some of those get to be a legal cycleway.
Once a way is a designated cycleway, data users can be reasonably sure that it is suitable for cycling. In fact, most of the time, once a path is deemed to be a cycleway, the path is upgraded to get the legal signs and markings and the physical characteristics of a cycleway.
We might as well be living on different planets. 
Phrases like âlegal cyclewayâ, âlegal sidewalkâ, âlegal streetâ, or âlegal buildingâ will only be met with blank stares among the authorities over here, because that isnât the role of the law. In fact, the idea that we must be so dependent on the authorities to classify real-world phenomena for us would be anathema in this society. Itâs the main reason I and other Californians were so rankled by someone hiding behind California law in defense of their notorious service=driveway2 proposal. We only treat young language learners like that:
You cannot be fully sure and you still need to check tags.
I have something that is basically MTB trail (narrow, next to ravine, unpaved, step slope, curves with R=1.5mâŠ) signed as cycleway, with official traffic signs (that blue circular one).
Perhaps a designation tag would work in NL, to extend the normal definition in OSM?
No. This was an example of how something that is special and unique for a country, is modeled for OSM using OSM-wide mapping and tagging, so that all OSM-based applications can use it and relay it to the users, without having to know the exact signs and traffic laws of the country.
yes |
designated |
|---|---|
| Legally allowed | Legally appointed as the way for this transport method |
Consequence (example bicycle)
yes: you are allowed to cycle there, no matter what type of highway this road is. Suitability for regular bicycles is not certain.
designated: this path is legally and practically meant for cycling. You may expect it is suitable for cycling on regular bicycles.
Thereâs also the highway=residential + bicycle=designated usage which is common in some areas to indicate a âcycle streetâ. This is the only[1] scenario in which cycle.travel treats =designated as anything different from =yes.
Ok, ok, not quite the only scenario, but the others are really very niche. â©ïž
Huh? I donât see how you can read this in my text?
Bicycle routing should always honour the bicycle=yes tag. Because it means that you are allowed to cycle there.
If your countryâs expressways allow cycling and are tagged as an OSM-type that allows cycling, there is no need to tag bicycle access as yes, because it is already implicit/default.
If the expressway does not allow cycling, and is tagged as an OSM-type that does not allow cycling, again there is no need to tag bicycle access as no.
If the expressway does allow cycling, but it is tagged as an OSM-type that does not, bicycle=yes indicates the situation and routers can route this road.
If the expressway does not allow cycling, but is tagged as an OSM-type that does, bicycle=no indicates the situation and routers can route this road.
If the expressway has an accompanying side path especially for cycling, then that path is a cycleway, and bicycle=designated is implicit.
What did I miss?
Per request I am splitting this thread and moving message #27 and replies to it to a new thread.
In NO, we treat the first sign as foot=yes (should be treated by data consumers as the national default value for cycleway if foot isnât specified) and the second sign as foot=designated. The pedestrian on the sign is what makes it designated. Since the bicycle is on both signs, itâs a given that bicycle=designated (which follows from cycleway).
We donât have this sign, but this would be the same as the second sign, except the way is segregated=yes, sidewalk=left|right|yes, or maybe even two parallel ways:
NL has the first sign: cycleway, so bicycle=designated is implicit. We tag mofa=designated with it, thatâs a Dutch rule so explicit tag needed. Also, foot=yes is Dutch national default for cycleways, if there is no sidewalk or separate footway, but the worldwide access table says that foot=no is the world wide default, so we tag foot=yes explicitly when there is no sidewalk or separate footway.
(We donât expect applications to know and apply the Dutch traffic regulations.)
The second and third signs we do not have as a rule, but the situation occurs, and local authorities and land owners will improvise.
Many cycleways have their own sidewalks. In some countries this is tagged as highway=path + bicycle=designated + foot=designated + segregated=yes. We prefer highway=cycleway + sidewalk=right|left.
Example:
The cycleway sign on the picture is a variant that includes mopeds (moped=designated).
You are not alone; Rewording that:
(We donât expect applications to know and apply anything but global defaults.)