Talk:Cycle routes

From OpenStreetMap Wiki
Jump to navigation Jump to search

I've moved the discussions from the main page onto, well, this discussion page.


(The following was copied from Cycle layer)

  • route = ncn is only for UK, in germany we have Radfernwege so better route=bicycle???
    • ncn might be a UK specific thing, but I see no problem with using it for other nation-wide cycle networks, after all you use highway=motorway on autobahn, right?
      • ncn (LF-route in Dutch) is also known in the Netherlands. See [[1]]. Of course every country should try to find the equivalent for this (if existant).
  • how to tag increase((Steigung)is this the word?) of a route?
I'd prefer the suggested ele-tag at the start/end of a way/segment and then calculate the increase (or is it called slope?!?). But I'm not sure if the renderers are able to do this calculation. --Kumakyoo 17:21, 14 May 2007 (BST)
even if they are not able to do it now, they will be, so lets use this.
I don't see the point, why not just use SRTM-data and have the renderer or routeplanner or whatever take heights into account automatically? I think SRTM has good enough quality for most practical purposes.
SRTM data ends somewhere between latitude 60 to 62 depending on longitude. Alv 23:34, 1 August 2009 (UTC)
  • I've been playing around with creating a rule set specifically for road bike users - with no routing along tracks etc and more preference towards routing along minor roads. Would there be any interest in me hosting this somewhere or has it already been done? --JWSHale


  • Should the references contain 'RR' and 'NR', or just the number?
  • How do these tags map to other countries?
    • Depends on which country. Many of them have national cycle networks, and can use the ncn tags. Additionally, many european countries use the same white-on-red colour scheme, so that's even better! Gravitystorm 19:19, 29 October 2007 (UTC)
  • RCN and LCN refers to the 'regional cycle network' and 'local cycle network', but in reality (in the UK at least), the routes are all part of the 'National Cycle Network', and are classified as national or regional routes, which would suggest using ncr_ref and rcr_ref.
    • Geez, that's pretty nitpicking. It's not likely to change; there's no point. Gravitystorm 19:19, 29 October 2007 (UTC)
  • Do we really need both rcn_ref and lcn_ref?
  • What about a single way that has two or more routes on it?
    • You can use relations for this, but it's pretty new. More details will follow when the use of relations stabilises. Gravitystorm 19:19, 29 October 2007 (UTC)
  • How about an "international cycle route (icn)? See [[2]].
    • Show me some tagged routes somewhere, and I'll render them. But I'd imagine the EuroVelo routes will just be re-branded national cycle routes, so it's probably best done using relations rather than adding tags to 1000kms of ways. Gravitystorm 19:19, 29 October 2007 (UTC)
      • EuroVelo routes already exist, see relation 2938. -- Eckhart 16:56, 26 June 2008 (UTC)
  • I would like to map and tag some parts of the EuroVelo 6 around Vienna with relations. The question is what would be the best for the network tag? ncn, icn, ev or EuroVelo? I mean this route goes through about 9 countries so this is really a international route and i think ncn would not be appropriate. And for the ref key, EV6 or just 6? What do you think? --Nevermind 15:12, 31 October 2007 (UTC)
    • I use icn even though it's not agreed upon yet. For the ref: When using relations, this question becomes less important because changing the value for max 12 relations is no big deal, so feel free atm. -- Eckhart 11:14, 26 June 2008 (UTC)
    • On the Swiss page it was determined to use E6, as noted on the EuroVelo website. RickH86 17:00, 21 August 2008 (UTC)
  • How should we tag routes to a route? I've seen a few sign posts like "National Cycle Network 1 (logo of NCN 65)", as in it is 1 mile to NCN 65. The signs can make you think your on the route, but actually it is just a route(hopefully cycle-suitable) to get there. I believe this is like motorcar signs that say a road number in brackets, as in "take this turning to get on A-road, but your not actually turning onto the A-road yet.". -LastGrape/Gregory 20:59, 2 April 2009 (UTC)
    • I would say: Not at all. If you have the map or the navigation software, you don't need to know where the signs are placed. --Lulu-Ann 07:42, 3 April 2009 (UTC)
    • In Belgium and the Netherlands we make an extra relation for these kind of routes that for example connect a city center to the cycle network. This relation also gets an extra tag state=connection. I don't know if the situation is the same in the UK and if they really need to be mapped there. I guess the main issue for you to decide is whether the routes really are signed routes. If you see these kind of routes on the official maps the answer is clearly "yes". --Eimai 12:02, 3 April 2009 (UTC)
  • What's about including cycling routes which are by no mean official ? I live in an area, where there are some dedicated bike routes / lanes, but in any case for doing a day of cycling you need to find your own route, so combining the bike routes with low-traffic streets. I think it could be useful to be able to store this routes in OSM, but how to tag it ? They would be a kind of personal proposal, so maybe just adding "network=personal" or "operator=private" could be ok. In this case OSM would be in concurrence to the classical portals where people give access to the GPS tracks of their cycling tours. On the rendering site it would then be very nice to have an easy way to render the relations and additionally download relations (recursively !!) as GPX tracks, so you can add the tour to the GPS of the bike.
    • Don't put your personal routes in OSM. OSM can only be a place to routes that are signed in some way, and certainly not for someone's personal cycle trips. If you want your own routes, it should be stored somewhere else. --Eimai 16:01, 4 May 2009 (UTC)
      • Ok, I understand this. But what would you propose where to store such routes, if not in the OSM main database ?I would like to construct my proposals as relations out of existing OSM ways, that seems very logical. Can I do that somehow ? Such as making an osm file with JOSM and publish the file somewhere outside of OSM, such as in my blog, or whatever. Could people then still render them on top of OSM somehow ? Sorry for my questions, I just started with OSM. Is there any information on this somewhere ?
        • There are places like and to share personal routes. You do not want to construct your routes from OSM ways. You would have to split OSM ways at every junction where you make a turn. Imagine averyone and their dog doing this. Rrr 09:06, 22 May 2009 (UTC)
  • This wiki page contains contradictory information on how to tag the "propoed" state of a cycle toute: the first table describes the method of using the ncn|rcn|lcn tag in the form ncn=proposed, whereas the third tabel uses the older method with the (undocumented) tag state=proposed.

reusing routes

I just walked part of a cycle route. In fact we walked three cycle routes at once. I would like to create a relation, type=route and then copy this to the two other relations containing the same ways. Or is there a smarter way to go about this? Could relations contain other relations? I have the same problem with bus routes. They also tend to overlap quite a bit. Oh, I think for a bus route I actually tried another trick. Open route in JOSM, select all ways of it, then close and add those ways to a another route. The problem with that approach was that backward/forward information was lost... --Polyglot 22:58, 20 January 2008 (UTC)

Route tags on ways

Why the rewording of the message about tagging ways for routes? They *are* deprecate right? You shouldn't tag new routes with that scheme since you can't tag more than one route on the same way properly and you keep the data together when it's in a relation, decreasing chances for errors dramatically. --Eimai 15:24, 7 May 2008 (UTC)

It's generally harder to use relations for new users, there's no plans to drop support for the basic way tags by any of the current applications, and so it's a bit much to call them deprecated. Plus I'm a bit wary of calling anything deprecated when we have no actual official tag set to start with (just agreed suggestions). I agree fully that it's often better to use the relations as they have all the benefits you mention, and that their use should be encouraged, hence modifying the wording rather than just deleting the sentence entirely. -- Randomjunk 16:59, 7 May 2008 (UTC)
But new users have to get familiar with the concept of relations anyway. You split a way - you possibly break a relation. -- Eckhart 22:02, 18 July 2008 (UTC)
This is not true when you edit using the latest version of JOSM. The 2 broken parts are automatically added to the relation.--Itserik 20:28, 19 July 2008 (UTC)
Okay, better example: You append some nodes to a way - you break a way. Especially easy because JOSM tries to re-use ways. -- Eckhart 21:50, 19 July 2008 (UTC)

Named routes and references?

Near me there is a local cycle network(?), where the entire route is named, and is also composed of several named sections. How should the network's name be indicated? How should the individual section names be indicated? --Hawke 20:12, 3 July 2008 (UTC)

Create a relation of the type route for each section. Use name for the name, ref for the abbreviation or reference number. If you think the network needs a name, too, you can create another overall relation and add all sections, but I don't think that is necessary. --Lulu-Ann 11:53, 4 July 2008 (UTC)

Proposed routes in relations

What is the best method of marking proposed sections of routes in relations? The NCN 22 through Banstead is of propsed status, but the route through the borough of Epsom is fully implemented. Should two relations be used, one of each status, or should one relation be used with proposed or something similar be placed in the role for the specific ways? BlueSpecs 21:50, 18 July 2008 (UTC)

Ten years later, I have the opposite problem - I have portions of routes that are now obsolete (as they've been upgraded to be a newer kind of route - for example, bits of London Cycle Network which are now London Cycle Superhighways. On the road, their LCN signage is gone and replaced with CS singage. This ends up looking messy on OpenCycleMaps and WaymarkedTrails, as they still display the old route overlapping under the new one, when only the new one is signed. The simplest thing would just be to delete these bits of that LCN from their relation? But that would lose this historical info. Could we have a tag such as "state=obsolete" for those bits?--Feline1 (talk) 19:10, 13 June 2018 (UTC)

Mountain bike routes

In Switzerland we have designated National and Regional Mountain Bike Routes. How are we to differentiate them from the other bike routes? The names are different, but the numbers duplicate. Suggestions? RickH86 17:02, 21 August 2008 (UTC)

Use route=mtb, that's what others are using for mountainbike routes. I don't think it's documented somewhere, but it gets rendered on the cycle map IIRC. --Eimai 17:31, 21 August 2008 (UTC)
Hmm, I'm now actually wondering whether it should be route=mtb, or network=mtb... --Eimai 17:48, 21 August 2008 (UTC)

I would think network=mtb would be more accurate, but if it renders them differently with route=mtb, that is better. That still doesn't help my ref, unless I use MTB#.. RickH86 17:26, 22 August 2008 (UTC)

For reference if you havent noticed: Mountainbiketours, I am in preference of route=mtb instead of network=mtb because network=*. Discussion about this is better continued here though if there are valid arguments. To me network could be used as network=mtbnn/mtbrn/mtbln (mtb national network, mtb regional network, mtb local network). On the other hand I only know of one kinda national mountainbikeroute (Alpentour Austria) and to me it's hard to see the difference between local and regional mtb routes. Maybe a better classification would be to organise mtb routes by daytours and tours with more stages that are usually completed during several days.--Extremecarver 13:03, 15 December 2008 (UTC)--Extremecarver 13:03, 15 December 2008 (UTC)


I have been trying to come up with a way to do a cycle route without stopping and recording every turn, etc, especially when you are in area without much of a map on OSM. I found is helpful for keeping track of where you were. RickH86 17:30, 22 August 2008 (UTC)

Marking access roads

In the city I live in, Göteborg, we have a network of cycle paths, but there are a number of places where there is no dedicated cycle path, instead there is a recommended route on a low-traffic street. It makes sense to capture this information and display on a cycle-centric map. There is a map published by the city that shows these recommended routes. The dotted lines are the recommended routes and the filled lines are cycle paths. I'm not planning to copy this verbatim, but instead find good such routes and mark them on the map. The question is how to tag them. Can anyone suggest a good way to tag them, and to have them rendered on the bicycle map? Here is the area in OpenStreetMaps. Norpan 11:48, 15 November 2008 (UTC)

bicycle=designated combined with lcn=yes is the way to go, in my opinion. You may need to use relations as described on the main page, as well. --Hawke 00:59, 18 November 2008 (UTC)
Well they are not routes per se. At least I don't know of any route numbers or route names. It's just a bunch of bicycle paths and lanes, those are relatively easy to tag (although I sometimes feel the need for more expressive tagging, for instance on what side of a road a bicycle lane is, and the fact that a bicycle lane can be two-way even if the road is one-way). Tagging the recommended access roads as bicycle=designated is maybe a the best for now, but the access roads are not designated for bicycles, they are just recommended. So it's not a physical attribute of the road but a routing attribute. So maybe there is a need for a new tag. Something like bicycle=recommended that can be put there for the purposes of the bicycle map and for automatic bicycle routing. Norpan 13:16, 18 November 2008 (UTC)
Are these routes marked with a sign? If so, then it's definitely designated. Note that "designated" basically means "marked with a sign" and has no relation to the word "designed". If the only way to know that they're good bike routes is by the map, it's a grey area; however, especially since the city publishes that map, I think it's fair to consider the map to be an official designation too. Hope this helps. --Hawke 18:14, 18 November 2008 (UTC)
I will follow your advice and tag them as designated. Sometimes they are marked by destination markers with a little bicycle on them. Other times the marker is not there but it's pretty obvious given the design of the cycle path system that these roads were intended as access between cycle paths and then I will mark them as designated too. Thanks! Norpan 23:29, 18 November 2008 (UTC)
It seems that this leads to weird routing for cars. It will not use the road as it is marked as designated for bicycles. But even without the tag, bicycle routing will use the road because it's shorter. It's simply a rendering/mapping issue to have the tag. Norpan 13:16, 26 November 2008 (UTC)

Index __NOTOC__

I have removed __NOTOC__ on the main page. I think this page has become so big that it needs a reference table on top.--Extremecarver 13:05, 15 December 2008 (UTC)

NCN / LCN conflicts

In Oxford we have a situation where part of the NCN uses a section of road that is (much) too busy to be recommended as part of the Local Cycle Network. We've discussed the matter with Sustrans locally, who are happy for the NCN to be suppressed for local presentations. We could go down the line of creating a special rendering, but would prefer this was dealt with by the main OSM Cycle Map rendering.

My best thought on dealing with this so far is to change the NCN markings to a much narrower shaded line at zoom levels 13-18 when the NCN coincides with a lcn=* or lcn_ref=* tag. This would be combined with _not_ rendering for lcn when the tag is lcn=no. This would leave the NCN as still marked, but much less dominant.--RichardMann 14:54, 12 February 2009 (UTC)

If you have special rendering requirements, why not just making them yourself? It's a bit awkward to ask for changing the rendering of all cycle routes in the world just because there's a stretch where the NCN passes a busy road... In 99% of the cases routes tagged as ncn are much more important than those tagged with lcn. --Eimai 15:11, 12 February 2009 (UTC)
Mine is a local example, but illustrative of a more general problem. Maybe a more general solution would be a better bet. At the moment, LCN overwrites NCN at zoom1-12, and NCN dominates LCN at zoom13-18 (brighter colour, wider shading). This is the wrong way round - NCN should dominate at small scale, LCN should be dominant at large scale. LCN is not a subsidiary network to NCN, it serves a different function at a much more local level. The more fundamental solution would be for NCN to show as narrow red shading at zoom13-18 (regardless of LCN tagging). Maybe the NCN shading should narrow in stages as you progress to higher zoom levels.--RichardMann 11:52, 13 February 2009 (UTC)
IMHO ncn should always look more important than rcn, and rcn should always look more important than lcn, and that should be at all zoom levels. It's the only logical way of keeping the hierarchy clear. It would just be confusing otherwise. --Eimai 13:39, 13 February 2009 (UTC)
I suspect you may be fairly unique in using lcn as a third level in the hierarchy in Antwerpen. Mostly people seem to tag lcn for routes that have a local destination as an objective, or are interesting, or signposted, and I feel that this is as much a legitimate use of the tag as yours. Andy's words: "At low zoom levels it is intended for overviews of the National Cycle Network; at higher zoom levels it should help with planning which streets to cycle on, where you can park your bike and so on." I stand by my suggestion for rendering; I think what I'm seeking to be rendered is a reflection how most people tag, which has an lcn independent of ncn/rcn, and not as a third layer in a hierarchy. It would appear that we must differ.--RichardMann 22:09, 13 February 2009 (UTC)
It's what lcn is used for in the Netherlands and in Belgium: local routes, usually around 30-50km, often with a certain "theme", making a loop, and only for touristic use. The routes have nothing to do with those classified as ncn or rcn. As far as I understood the UK situation, it's used there for routes like those from the London cycle network: non-touristic routes crossing the street that actually bring you somewhere. And ncn was touristic, wasn't it? So how can the two be related to each other and should their rendering match that non-existing connection between ncn and lcn? --Eimai 19:43, 14 February 2009 (UTC)
Further study of Antwerpen shows the LCN tag is being used there to show cycle tours rather than a network. These sometimes follow NCN/RCN and sometimes go their own way. They don't fit into the NCN/RCN hierarchy - they are an overlay. I think there might be a case for a fourth colour for signposted tours - perhaps green. But I also think that these macro-scale networks need to fade at larger scales (perhaps just become reference numbers at zoom15), to let the micro-scale LCN take the foreground. Otherwise opencyclemap just becomes dominated by touring info.--RichardMann 12:54, 15 February 2009 (UTC)
Actually they are a kind of network: the routes are (usually) signed with the same kind of signs in each province (and it's usually the provinces that manage the routes), and the routes are often connected to each other as well (so you could combine two or more loops). --Eimai 13:09, 15 February 2009 (UTC)
And the ncn/rcn/lcn hierarchy works as follows in Belgium: ncn is for the touristic routes crossing the country, so no discussion there. rcn (also for touristic use only, and better known as the cycle node network) can be used to cycle between places that are far apart (from region to region, as it were) and the lcn routes can only be used to stay in one area (i.e. local). --Eimai 13:16, 15 February 2009 (UTC)
Some Dutch/German towns seem in practice to use RCN for their main local network, others use LCN (and in the UK we use LCN, because RCN is used for secondary linear touristic routes). The Dutch RCN seems pretty local in scale, even if one of its purposes is regional touring. I'm thinking perhaps that we need 2 colours for tours (one for NCN and the other for routes like your local tours and UK RCN), and 2 colours for local-scale routes (Dutch RCN and LCN). I'd still like to see the tours fade as you zoom in to a more local scale, but I think getting a bit more harmonisation on how we show secondary touring routes might be a sensible first step, otherwise the rendering will always be a compromise.--RichardMann 22:30, 15 February 2009 (UTC)
RCN a local thing in the Netherlands and Belgium? RCN is exclusively used for the cycle node network, and that network almost completely covers both countries with one network (and there's even a bit in Germany). I'm sure that most cyclists on the network just do a local tour on a single day, but at least it's also possible to use it to make bigger trips between cities (or even cross both countries), and that's not something which can be said from the routes tagged with lcn. Take a look at the map here to get a better understanding of how things work in both countries. Select and deselect the right overlays to see only lcn (lokaal), rcn (regionaal) or ncn (nationaal) routes.
Why do you make a distinction between the routes in Belgium and the Netherlands btw? We both have exactly the same kind of cycle routes and our tagging rules are basically the same. --Eimai 12:13, 16 February 2009 (UTC)
On the original point, I'd be unhappy if the NCN was suppressed at high zoom levels on the main OSM cyclemap. NCN routes can be difficult to follow in built-up areas (more likelihood of damage to signs, etc.) - I now make it a rule to print out a map of each urban area before setting out on an NCN ride. If you're referring to NCN 57 along Cowley Road in this case, this is a good example - on the day I followed NCN 57 east from Oxford (mapping it as I went), the sign where it leaves the Cowley Road had been turned, meaning that I initially overshot by some way. Altering the cartography so that both sets of routes are visible would be a good solution, but not solely showing info for one set of users at the expense of another. cheers --Richard (Charlbury)

To attempt to summarise - Netherlands (and parts of Belgium and Germany too) use all three cycle network levels for "touring", with none in theory available for short-distance urban utility routes (though see Bonn area, and the two towns NE of Amsterdam - some people are mapping urban networks; some are using rcn and some using lcn). The UK (so far) uses two levels for touring (ncn/rcn) and one level for urban (lcn). My conclusion is we need a fourth network, say TCN for tours (ie what the Dutch/Belgians/Germans are officially using LCN for), to leave LCN for urban networks (I'd suggest changing the Dutch/Belgian/German tours because most are set up as relations and thus a lot easier to change). On the question of rendering, the Dutch perhaps show the lead - the cycle networks should be overlays, not part of the base. I still think there's a case for being clever at the base layer, putting all the routes in but not letting anything dominate. Then you leave it to the overlays to add the preferred emphasis. But just having NCN/RCN as an overlay that can be switched on/off would solve my original problem.--RichardMann 23:36, 16 February 2009 (UTC)

And if renderers implement a fourth network tag we may consider it. Right now these three layers work nicely, and also make the tagging the same as the walking networks here, where we have exactly the same kind of networks as for cycling (nwn for national routes crossing the country, rwn for the walking node networks spanning entire regions, and lwn for small loops, usually 10-15km). But indeed there's the problem of urban networks (in Brussels for example, currently using rcn but would probably be better to use lcn, "r"cn was chosen by the person mapping them since it's the Brussels-Capital "Region" that manages the routes).
But do we really need a fourth network tag? It should be sufficient to have another tag in there so we can distinguish two networks if they both are in the same "lcn" category. And I'm sure that we have places where there are more layers necessary (the ncn routes for example can be divided into two layers in Belgium and the Netherlands). The current three layers are sufficient to give an idea about the kind of network in the international maps. Better maps can then be generated by the countries or regions themselves taking the other tags into account. --Eimai 12:31, 17 February 2009 (UTC)
Further to the above, it isn't the case that UK NCN/RCN routes are purely touring routes. Many are also urban commuter routes: Birmingham (NCN 5 along the BCN Main Line towpath), Cardiff (NCN 8), Worcester (NCN 45) and Gloucester spring to mind. The situation in Oxford isn't necessarily the case elsewhere. --Richard 19:53, 18 February 2009 (UTC)
Many are also quiet urban commuter routes, it's just that some are not.--RichardMann 12:10, 22 February 2009 (UTC)
Yep, absolutely. In my experience Cowley Road is definitely the exception rather than the rule, but then Oxford is difficult in any case due to the limited number of river crossings. --Richard 12:13, 22 February 2009 (UTC)
As a solution, perhaps a "priority=(numerical value)" tag could be added to relations (of individual routes) and used to override the default rendering hierarchy. This would allow for differing hierarchies on a region-to-region basis if the default hierarchy is not appropriate locally.
In London, the 'highest priority' routes (Cycle Superhighways > Quietways) are now mapped as rcn; these routes extensively "clash" with the next highest priority routes (the semi-retired 'London Cycle Network') which are mapped as lcn; NCN routes would probably be the next level of priority, or possibly be higher than some of the 'unsigned LCN routes'. However there are also a number of 'proposed' rcn routes mapped, which (until they actually launch as 'highest priority' routes) would ideally have a lower priority than any existing lcn routes that they might be replacing in the future.
The default hierarchy just doesn't cater for regional/complicated route conflicts. However if a "priority=*" tag (or similar tag) was recognised, then at higher zoom cycle routes could instead be rendered according to local area requirements.--MacLondon (talk) 18:41, 16 October 2018 (UTC)

United Kingdom / CTC National Byways Network?

According to the CTC, "The CTC’s Byways Network is a unique resource available to members. Despite its name, the 13,000 miles of routes, covering the whole of the British Isles (including Eire) are mainly on-road. Routes have been designed to utilise, as far as practically possible, quiet scenic roads. Route Sheets explain the route verbally and give some information about places passed through. The route is described in both directions." info link. AFAIK none of the network is actually signposted, but they do sell maps of the routes. It would be nice to have an online digital version. Should it be mapped in OSM? ChrisB 18:38, 24 May 2009 (UTC)

Although I'm sure the CTC recommendations are a fine resource, we should only be mapping routes which are physically signed on the ground. It's a little confusing though how similar the name is to the National Byway, which is independent of the CTC and is signed on the ground. Monxton 23:20, 28 April 2011 (BST)

Bike race routes?

It might be interesting to many to have routes of bike races marked as relations in OSM. For example, you may want to have the route on your GPS for training, or you may want to ride it even after the race itself is over. But how should these be marked and tagged? We don't want them to appear as regular bike-routes. --mok0 10:25, 25 August 2009 (UTC)

problem 1: bike races don't follow the same route each year, so would you then have a relation for each year?
problem 2: how to map them in the first place? Is there a source you're allowed to use? They're not signed like the other routes we tag. One could probably try to follow the route when the race is taking place, but that's close to impossible.
IMHO this should be held somewhere outside OSM. Unsigned and temporary routes is just something we don't map. --Eimai 10:56, 25 August 2009 (UTC)
Sometimes they're marked in some way -- often chalk on the pavement (US sense). I agree otherwise with the problems you've raised (even problem 2, since they're not always marked.) --Hawke 11:54, 26 August 2009 (UTC)

U.S. ACA route network

What about the ACA route network? It seems to be distinct from the U.S. Bike Route System. ChrisB 13:32, 17 January 2010 (UTC)

I don't think it's marked on the roads; it's more like a driving route you'll get from the auto club. --NE2 02:59, 18 July 2010 (UTC)

NCN RCN LCN in United States

I been mapping the County Bike routes in Santa Clara County (based on the County bike plan). The Dilemma is that there are potential four levels of Cycle routes. National (NCN), State (??), County (??), City (LCN). While the State of California does not have may state cycle routes there are ones that are multiple counties in the San Francisco bay area. Metropolitan Transportation Commission Regional Bicycle Planning Santa Clara county has it own routes. Naturally the many of the cites have their own routes that should be LCN. This problem is partial because of California being 1.74 times the size of the United Kingdom. ~bmhr 11/12/2010

Feel free to invent a new one and document it. If you wanted to fit with the current naming scheme you could have something like network=pcn (for Provincial Cycling Network) for the State networks, and allocate County routes to network=rcn. --Richard 08:51, 13 November 2010 (UTC)
Are there any places where these county and city systems actually overlap? It seems that it would be more beneficial for them to work together - and even if they don't, county systems would probably cover only county roads outside city limits. --NE2 06:56, 14 November 2010 (UTC)
What I've been doing for my area, (Solano County, CA, USA) is consider routes that don't leave the county as local. Since the Bay Area does have the Bay/Ridge trails which go between counties, I suggest them regional. --WikiLeon 11:37, 29 July 2012 (BST)

Can we talk about this bit: "A cycleway built to move cyclists off the main roadway on urban interstates and US routes would also qualify for a NCN route relation"? This doesn't seem right; just because a bike route corresponds to a route in the national motor vehicle network, that doesn't mean it's part of the national cycle network. A route should only be NCN if it's part of the US Bike Route System or possibly another long-distance route. --NE2 06:56, 14 November 2010 (UTC)

Cycle lanes on motorways?

There seems to be a rendering problem with cycle lanes on motorways. e.g. highway=motorway, bicycle=yes, cycleway=lane. The cycle lane does not show. Is there some way to raise a rendering bug? --Ebenezer 23:29, 12 December 2010 (UTC)

Assuming you are referring to OpenCycleMap, you can add a bug report on Trac (choose the component "opencyclemap"): Preferably with a link to a location with an example of the problem. Though do you really have cycle lanes on a motorway? Sounds interesting, do you have any links to photos of it? --Vclaw 01:49, 13 December 2010 (UTC)
There's one in Miami (I guess it's debatable whether it's really a motorway, but it does meet standards west of the traffic signal):,-80.16486&spn=0.002034,0.00515&t=k&z=19 --NE2 03:37, 13 December 2010 (UTC)
Here's one for Sydney, Australia on the M5 Motorway (not that I'd want to ride on it).,151.087632&z=20&t=k&nmd=20100916 I'll add something to trac. --Ebenezer 05:29, 24 January 2011 (UTC)
Actually that one is missing bicycle=yes. I will update it and check it before submitting a ticket.
Is that a bike lane or a shoulder with sharrows? --NE2 11:00, 25 January 2011 (UTC)

All the Other Roads

Among all the other roads (those not part of a local or national cycle network) there are often good and bad choices. How can we tag roads to avoid and roads to prefer, if they are not part of an official network? Brycenesbitt 23:27, 12 July 2011 (BST)

What makes a good or bad choice? Tag that. Alv 08:11, 13 July 2011 (BST)

Alternative routes

In the UK, NCN routes are often improved (eg addition of a cycle path bypassing a hill etc), but the signage seems to often be left in place for all of the old route, except where it splits from the current route. I would like to suggest that these could be tagged as an alternative route (maybe ncn_ref_alt) and rendered with a dotted line (though preferebly distinct from and probably less stong than the proposed route dotted line. I would suggest they dissapear at the lowest zoom levels where a national overview is provided). I foresee several benefits to doing this:

  • When a mapper comes across an NCN sign on the old route, they can see that the route has already been mapped as an alternative, so that they don't map it as part of the current route, causing confusion.
  • The old routes were often perfectly good cycle routes, and they may happen to go near someone's starting point/end point/intermediate destination such as a shop or accomodation. It's helpful to keep them mapped.
  • If someone is planning a there-and-back journey along an NCN route, it can add variation to use an alternative route in one direction.

Daveemtb 11:39, 20 July 2011 (BST)

NCN links

Sustrans encourage rangers to sign links between the proper NCN and nearby points such as train stations, universities etc. These should be signed on the ground with the route number in brackets, although sometimes are not (sadly there is a lot of inconsistency in signing across the country). These could be tagged as 'network=ncn_link, ref=XXX', or even add the additional ways into the main relation with a 'link' role. My preference is mildly towards the former option, the latter seems more elegant but a bit obscure. Any opinions on this?

For an example, see this OpenCycleMap rendering. The route is actually signed on the ground as '8', but is meant to be a link route between 8 and 82. This is how I've tagged it now, but I think it could be improved. Spark 15:54, 23 December 2011 (UTC)

My opinion... Given that OSM is international, and that the idea of a 'link' route is rather unique not even to the UK but really just to Sustrans itself, I don't think these should be mapped... any more than we'd map the route described by equivalent signage on the road network. On the road this style of sign is clearly understood not to describe a route as such, but just the direction to be driven in to get to a numbered road. So it's not a 'link route' that anyone would map... just a set of signs indicating the direction to go to get to the route in question. In most respect this is how the Sustrans signage is used in most places - as an indication of the direction to travel in to reach the route, not as a route itself.
There are exceptions, but that's more a product of odd local circumstances (or Sustrans mapping) than anything else. Perhaps where (oddly) an actual route of some substance is actually signed with a number in brackets it should be recorded. There's one link here, but as to all intents and purposes it's signed as a bit of NCN route (with a red and white number symbol which happens to include brackets) it seems quite sensible that it's recorded as part of the NCN. The fact that Sustrans (and or the collaborating councils) has this odd additional designation shouldn't push us OSM mappers into inventing something equally unusual to account for it. Rostranimin (talk) 22:30, 13 June 2018 (UTC)
On UK road signage, "links" like this are given with the route number in brackets e.g. see [3], meaning "the route in brackets can be reached this way". That doesn't mean the route leading to it is shown on maps as *part* of the route it leads to, obviously. The same logic ought to apply to cycle routes signed in the same way. In some cases, though, I do this it's useful to create a relation for the "local cycle link" or whatever it is. Here's an example I added recently [4] It's in Shoreham-by-Sea in West Sussex. It's just a set of residential roads that West Sussex County Council have added cycle signs too (in one direction, pointing to the train station, in the other, saying "Downs Link"). There's no special infrastructure on the roads, but it is by far and away the best route to get to the start of the nice off-road "Downs Link" path, which just kinda arbitrarily stops without connecting to anywhere else. As such, I feel it's a useful think to show on any cycle map - but it doesn't need to be tagged as part of the NCN (and it isn't shown as such on Sustrans' map)--Feline1 (talk) 08:40, 14 June 2018 (UTC)

Bicycle Boulevards

Bicycle boulevards are intended to be different from generic "bike route" streets. Bicycle boulevards typically feature lower speed limits and traffic-calming features, plus signs and, often, special pavement markings (examples are from Albuquerque, other communities use different signs and markings as seen in [5], page 19). Depending on the community, generic "bike route" streets can be low-volume, low-speed streets, but they may also be higher-volume, higher-speed streets that are designated to connect a network of multi-use trails, bike lanes, or bicycle boulevards. The Key:cycleway page and Portland suggest cycleway=shared_lane or cycleway=sharrow for bicycle boulevards. The suggestion for lcn=yes seems more appropriate for generic "bike route" streets than for bicycle boulevards.

There’s no specific definition of a “bicycle boulevard”, and so I don’t think there is a specific way to map it. It is the sum of its identity as a route and its bicycle facilities and traffic-calming features. If all or part of it has shared lanes or sharrows, or speed bumps or mini-roundabouts, then these should be mapped. If it is a named and designated route, then it makes sense to map it as an lcn. Michael Z. 2013-10-15 20:22 z

Signed and unsigned

Is a cycle route something which is signed?

I think that there's an unresolved issue in some areas - which could be resolved through either some attempt at a clearer definition of a cycle route, or some simple new tags.

Some examples to illustrate the issue - Edinburgh includes a European route, which only exists on paper, but which is currently mapped and is rendered prominently as you'd expect. Edinburgh also contains well recognised set of off road paths, which are clearly seen locally as part of the cycle network, but which aren't signed with any reference number or (well) recognised name.

A simple starting point would be to say that anything which is defined by a relation (route) should have signs on the ground confirming this - so the fact that a route is defined by some authority or organisation becomes meaningless (within OSM) until its signed. An slightly different approach to this would be to say instead that it should only be given a ref tag (in the relation) once signed. Are there other equivalent approaches?

An alternative view might be that well defined routes (e.g, by a local authority, council, government, etc) could be recorded, but that we need an additional tag - something like "signed=yes" or even "signs=yes/frequent/occasional/rare/no"

Any thoughts? Rostranimin (talk) 23:19, 14 October 2013 (UTC)

Don’t forget that bicycle=designated can also indicate that a way is officially for bicycles, whether part of a network or not. source:bicycle=sign might record that signs indicate a bicycle way or route (see source=*).
Here in Winnipeg we have a similar mapping problem. There is a series of decades-old recreational bicycle routes on quiet streets, indicated by signs. In recent years the city has started developing a network suitable for bike commuting, and have published a map. Many of the map routes are only planned bicycle boulevards with no bike-specific facility nor signage yet, and some of the recreational routes with signs are not marked on the new map. Do we map the old routes or the new, or both?
If there is community consensus that a route is good for bicycling, then it is probably worth mapping. The presence of facilities (cycle tracks and lanes), signage, official and unofficial maps, and usage patterns might all constitute evidence of bicycle routes, but I don’t know if there can be a universal rule. It’s probably a good idea to agree on some conventions with any other local and regional mappers, so you aren’t working at cross-purposes (I have unfortunately been guilty of this).
Do tourists from Europe arrive with plans to follow the European route? Then it should probably be mapped and noted as a route. Personally, I don’t think an informal network constitutes “routes,” in the same sense as numbered highway routes. But any long path with a conventional name might be a route, even if it isn’t designated or maintained by any authorities. Michael Z. 2013-10-15 22:04 z

I'm revisiting this because I think it's becoming an increasing problem that people add routes to OSM that they like, but which aren't signed. I'd like to begin using a tagging scheme as follows:

  • signed=yes (route overall can be followed without a map)
  • signed=frequently (route overall can mostly be followed without a map, but there are some problems)
  • signed=rarely (route signs exist, but can't be followed without a map)
  • signed=no (this is an unsigned route)

Does anyone else agree? Rostranimin (talk) 10:26, 23 August 2018 (UTC)

I think your suggestion has merit! A further complication is where the route has some kind of designated number as part of a network of routes (eg the London Cycle Network) but the route number isn't used on signage - instead there's just generic 'cycle' route signage and signs state destinations. This can make it hard to follow a particular numbered route without a map. But the route is still 'signed' (just not with its number) Feline1(talk) 23 August 2018
Nice point - we should specify that 'signed' means signs that include the route identifier (number or name or symbol). I've seen a few of these in my time too. --Rostranimin (talk) 21:10, 23 August 2018 (UTC)
If the route number doesn't appear on the signs, could its identifier be moved to admin_ref? I've done that on ways to stop their highways department ID appearing on the map, but keep it in the database. MJ Ray (talk) 23:22, 18 August 2020 (UTC)
I think that completely unsigned routes should not appear in OSM. It would remove "my favorite route between A and B". Are there "real" routes without any physical presence? Mateusz Konieczny (talk) 15:42, 3 September 2018 (UTC)
Well, I would agree with this in principle, Meteusz. Certainly I don' think routes simply devised by private individuals should have relations on OSM. In practice in the UK, however, the charity behind our 'National Cycle Network' quite often creates numbered routes that are 'on road, unsigned'... similarly bits of the 'London Cycle Network Plus' are in a similar situation. OpenStreetMap's mapping of these official routes is often better quality that these bodies' own maps, so in that respect, OSM serves a worthwhile purpose in making these poorly-signed routes usable for cyclists. --Feline1 (talk) 16:25, 3 September 2018 (UTC)
If a bit of NCN is "on road, unsigned", isn't that a fault that should be reported to the highways department and network operator Sustrans? MJ Ray (talk) 23:22, 18 August 2020 (UTC)

Actually I should be clear in saying that I agree that unsigned routes don't belong on OSM. However what I'm after here is a practical way forward. It is clearly the case that there are situations where signage, even on formalised routes, deteriorates to such an extent that a route or section of route is unsigned, but the route is still well known and to some extent 'official'. Then there are also situations where a route is clearly designated somewhere relatively official, but not really signed at all. And there are further situations where some enthusiastic group puts lots of effort into a half-signed mostly-unofficial route. We could be strict about this and say that a particular amount/quality of signage is the minimum necessary before a route is included on OSM, but I can't imagine this being effective or practical. Defining the amount/quality of signage required is going to be almost impossible. A practical way to deal with these situations is to allow their mapping, but to provide a way that the unsigned (or partially signed) status of the route can be recorded. Nobody can argue with this signage status being recorded. Nobody's going to think that me/us marking an unsigned route as being unsigned is being fussy or malicious. But the result (given time) should be that most standard maps won't show these unsigned routes with any significance... both giving an indication of the facts on the ground, and at the same time probably discouraging the mapping of more unsigned/unofficial routes by newcomers. --Rostranimin (talk) 22:33, 3 September 2018 (UTC)

For the routes you might have in mind, it is possible to have a cycle route relation without tagging with "network=*", or maybe better to tag with something like "not:network=lcn" to discourage any later addition of a network tag.
Any cycle route with this tagging would not get rendered on OpenCycleMap unless e.g. lcn_ref or "lcn=yes" is also tagged (either on the relation itself or on the ways in the relation.)
In addition, e.g. combining "ncn_ref=* + ncn=no" would hide the ncn route line but apply the ncn's ref as red text along the route. This can be used e.g. for labelling future proposed ncn upgrades along existing lcn routes. Similarly, this is useful in the UK where 'links' to ncn routes are signed as red "(x)" but are not themselves part of the ncn network; e.g. I've tagged a lcn 'link' between NCN16 and RCN50 as "network=lcn + ncn_ref=(16) + ncn=no + rcn_ref=(50) + rcn=no" to render a blue line with red "(16)" text and pink "(50)" text - see [6]
On OsmAnd app (if set up to display cycle routes), routes without 'network defining tagging' render as a brown colour in contrast to the red/pink/blue network routes; this brown colour 'gives way' to any lcn/rcn/ncn route(s) that might run along any part of it.
I've used this "not:network" tagging in London to hide some long term "state=proposed" routes (further using "proposed:route=bicycle" to completely hide these, e.g. [7]), and also for adding the ref for upcoming rcn routes (=future upgrades) as pink text along the already existing blue lcn network (=current).
In addition, this tagging has been useful to deal with clashes between lcn routes and the rcn/ncn network. Wherever lcn routes pass either along ncn routes or along completed sections of the newer (upgraded) rcn network in London, I've tagged the lcn section with "not:network=lcn" to avoid the blue colour rendering on top of the preferred pink/red colour; then by combining "lcn_ref=* + lcn=no", I can get the the lcn's ref to continue being rendered as blue text along the pink/red coloured network, e.g. [8]. --MacLondon (talk) 18:15, 6 March 2019 (UTC)

Copyright/Freeness status of the Cycle routes layer

What's the copyright status of this rendered (!) map layer? Also, how free is it? (Is the code which generates the tiles publicly known? Is it free software?) Which parties provide this map layer? (Do we depend on a single party or are there other sources to get this map from, too?)

--Wuzzy (talk) 15:18, 11 April 2015 (UTC)

Use of ncn_ref, rcn_ref and lcn_ref tags in the UK

Just wanted to revist the established convention for the use of these tags in the UK, which is for the value to just be a number, not any kind of letter prefix. To my mind, this works OK if the only routes being tagged are from a single operator (Sustrans), but of course doesn't work well when these routes have to be shown on the same map as routes from a different operator that has their own numbering scheme. The main place this is a problem is London, where TfL have Cycle Superhighway and Quietway routes, which we've decided to tag as rcn's... These have all been (sensibly, in my view) consistently tagged using letter prefixes in their 'rcn_ref' tags - e.g. 'rcn_ref=CS8' or 'rcn_ref=Q22'. To my mind it would also make more sense to have, for instance, 'ncn_ref=NCN1' or 'rcn_ref=RR90' or even lcn_ref='LCN25'. If we compare these to how road routes are tagged in the UK, it's analogous to the definitions of motorways, primary and secondary routes, all of which get tagged with 'M', 'A' and 'B' prefixes to their route numbers, respectively. I'm not sure what the reasoning was to omit the letter prefix on these tags for cycle routes, does anyone remember?--Feline1 (talk) 14:32, 2 August 2018 (UTC)

I don't know the original reasoning, but I suspect it would increase the risk of the useful end of the number being hidden on most map renderers (similar to how some truncate route names). Also, we still have no consensus on whether the Sustrans routes are N1, NCN1, NCR1 or Red 1 and any rarely appear on signs, unlike the London CS and Q identifiers which are always on signs. I see the usage as consistent with UK road routes which almost always have their M/A/B on the signs, sometimes even C. -- MJ Ray (talk) 17:58, 5 March 2019 (UTC)
Any invented values for '*ref' that are meant to improve rendering are IMHO frowned upon as mapping for the renderer. Either there is an official reference number with or w/o a prefix, suffix, code or whatever - or there isn't. 'ref' is not meant as a field where we happily invent names to make working with them easier or rendering in certain schemes better but document the existing and usually visible (on signposts, official maps etc.) IDs. Cyclists expect these IDs to be found "in the wild". I am not familiar with the peculiarities of these in the UK but I made this more clear in the general part of the page since the previous wording here and elsewhere in the wiki apparently led to quite some misuse of 'ref' to the point of it being almost useless in some regions. :( --Stefanct (talk) 02:24, 4 July 2021 (UTC)

Signage destinations

In my local town (King's Lynn, England), some key destinations (Hospital, Town Centre, certain suburbs and estates and so on) are consistently signed from neighbouring destinations along multiple routes all pointing to the same point on the network. At the moment, I've mentioned it in the "note" tag on the route relation. Could I tag these destinations as a cycle node network? Should I? -- MJ Ray (talk) 18:06, 5 March 2019 (UTC)

I'm not sure that a 'node network' is what you're looking for here. That would imply wanting to map all intersections of cycle routes to define a network. There's nothing to stop you doing that (even though you've already got the network defined as a route network), but it sounds like what you might actually be wanting is to highlight certain destinations... rather than use all the intersections of cycle routes as a network, which might provide little/no practical benefit.
If I'm interpreting what you really want correctly, then you could initially experiment with tagging a single destination node or an end node of a route to see how it renders before considering applying more widely, e.g. you could highlight the end point of the K4 route at the hospital by tagging the node as "lcn_ref=K4 + name=K4" which would appear similar to how this rcn end node is currently displayed, just with a blue lcn colour instead of a pink rcn colour.
Or you could add "lcn_ref=Hospital + name=Hospital + lcn=no" to add 'Hospital' as a destination label. In this instance, using "lcn=no" would remove OpenCycleMap's node 'bubble' which doesn't look good when longer text is used, as the bubble size doesn't expand with text length. As far as I'm aware, using the same value for 'lcn_ref' and 'name' seems to avoid issues with how these nodes are displayed by different renderers. --MacLondon (talk) 20:03, 6 March 2019 (UTC)