I propose not to encode climbing ethics into OSM maps. I do not see myself adding tags such as “start=sitting” 
Exactly. That is exactly why I proposed that we go to Fontainebleau last week-end, in all honesty: so as to explore these issues practically.
1 Like
Maybe I have one to propose. It consists in acknowledging that the core concept is routability (by humans in the field, then by algorithms on the map) and that different object types involve different routing methods. This can apply to OSM ways as much as to OSM relations:
- ways with
highway=*involve the most usual routing method: straight lines between points; - relations with
route=*involve a derived routing method: routing through successive ways in the relation; - actual or hypothetical other relations can involve other routing methods, mostly based on nodes in the relation: a ferry route, a route made of cairns, a climbing circuit, etc;
- we could even apply it to new types of ways defined with very few points, e.g. a beach crossing, a scramble.
In that context, a climbing circuit would be a route made of only key points (the bouldering problems) and it is up to the data user to decide how to render it
Disclaimer, I’m a (bad 
) climber.
what is the exact meaning of
climbing=boulder? is it “this is a boulder”? or “this is a climbing route (a problem) on a boulder?”
For me, climbing=boulder, as it is a subkey of sport=climbing, means “a climbing route on a boulder, where you don’t need any specific climbing equipment, just climbing shoes and crashpad, as the max height is ~4-5m”. It is both used for artificial inddor climbing (see) or natural outdoor climbing. We have the same confusion in french between the stone and the climbing on the stone, we use the same word : bloc.
If it’s one path or a few paths along the numbered boulders, how about (existing tags) leisure=track & sport=climbing & climbing=boulder ?
At least in Fontainebleau, the boulder circuit is a succession of boulder (natural=stone) with one or more climbing=boulder on it. each climbing=boulder has a cotation, a name, and a ref prior to this circuit. A circuit has a name, a min/mean/max cotation, and an (ordered) list of climbing=boulder. But he path linking the climbing=boulder is either informal, or non-existant. Most of the time there are no paths.
Are there designated, maybe even waymarked / signposted paths between the boulders that tell you how to get from one problem to the next?
Never saw one. The only thinh you have, is at most a number/color at the start of the climbing=boulder.
How do people know what the next boulder is and how to get there?
Written on a (paper) topo, and soon in OSM. But most of the time, you just have to open your eyes ;(
Is there a particular order in which people have to clear the boulders?
Do they have e.g. maps showing where the boulders are and which difficulty these have?
Do people get a list somewhere with e.g. easy, challenging, expert boulder sets?
Yes, the topo describes the succession of the climbing=boulder, with each difficulty and name.
So finally, I would map a bouldering route as a relation type=route + route=climbing + climbing=boulder + eventually name= + `` + color= + climbing:grade:french:min/max=*. And the element of the relation beeing POI with sport=climbing + climbing=boulder + climbing:grade:french=* + name= + ref= + any external ref on topo.
Beware to not confuse with climbing=route which describe a (long) classic climbing route with a route_bottom, route_top, and (a lot of) bolts.
1 Like
This is exactly what happens if you just map the boulders as nodes.
That’s no ground truth at all for any route there. Only the boulders, which can of course be mapped as such.
This is exactly the same ground truth as for hiking routes : random paint signs that you must manage to follow using your eyes.
Not sure this is a useful functional definition in terms of height. I think El Capitan has been climbed solo without protection, and originally this was a multi-day climb. I believe some free soloers (Ron Fawcett comes to mind) may jump off climbs from a considerable height (20 m) if thwarted: a controlled jump is safer than an uncontrolled fall. Wikipedia separates out bouldering, highball bouldering and free solo climbing, with 4 m being roughly separation between first two, and 20 m between the last two.
Also in the case of Fontainebleau the boulders are just in the forest, so no-one has set up facilities. A climbing wall as in the first example is just that and although the climbing style is bouldering it is still a climbing wall. I would presume the wall does not provide a way of protecting the climb with a top rope: for artificial facilities this might be better expressed with an explicit tag.
I think the climbing wall example above illustrates the issue quite well: I think separating the two notions is less important for Fontainebleau, but significant for other bouldering locations, and especially the small unattended climbing wall type.
In the UK whether someone adopts a bouldering approach to a climb or uses traditional protection may largely depend on ability. Somewhere like Stanage may have a wide-range of abilities, climbing styles etc within a short span of rock. I think therefore climbing styles is actually a better approach.
I think a route of bouldering problems is very similar to a fitness course with the difference being that instead of a fitness station there is a bouldering problem at each station. I would therefore adapt the ideas and documentation of route=fitness_trail to something like route=bouldering_trail), and this would also favour nodes for the base of each bouldering route.
2 Likes
That’s no ground truth at all for any route there. Only the boulders, which can of course be mapped as such.
Disagree, there is one “ground truth”. Usually, each climbing=boulder of a given route will be painted at the begining of the boulder with a number, painted in a color than define the difficulty of the route. So, just with the ground, you can go from n°1 blue to n°x blue, and you’ll do the “blue circuit of the massif taratata”
1 Like
I think therefore climbing styles is actually a better approach
I do agree, but is there a difference between (climbing:boulder=yes + other value to no) and climbing=boulder ?
I think a route of bouldering problems is very similar to a fitness course with the difference being that instead of a fitness station there is a bouldering problem at each station. I would therefore adapt the ideas and documentation of
route=fitness_trailto something likeroute=bouldering_trail), and this would also favour nodes for the base of each bouldering route.
Totally agree; except fot a fitness course, you have proper path between fitness station no ?
Yes, there are things that we need to adapt. Not sure we need roles, for instance. I would not call it a trail either. But the general principe is the same.
1 Like
Not random: they are placed in view, especially at turns, in a specific manner to indicate the pre-destined route, and the symbols are route- (or network-)specific. Together they form a specific fixed route.
Usually yes, AFAIK. Sometimes the path is for general use and sometimes it may continue around a corner or past a crossing. In general, I map the fitness stations, not the route. Only if it’s a really dedicated circuit course with an entrance/exit, I may think of creating a route relation, but still I decide it’s not worth it. Same as for the Stations of the Cross, really. If a dedicated fitness station course also had a real proper name, I might be persuaded.
If it’s a bunch of fitness stations in one area without paths, I would map the stations and the area.
At the near edge of the overlap wall ( left most as you face the rock) is a short problem with an obvious water hole bucket.
I long for making my problems eternal by posting them to some portal!
Wikipedia has all the answers! What I observed, a boulder does not need a top, at least outside of competitions. The water hole bucket problem reinforces my observation.
1 Like
There’s no reason why a path needs to be mapped, you could just include the bouldering problems as an ordered list in the relation (from time-to-time this approach is suggested for bus routes where the driver is free to select the actual route driven as long as they visit the stops in the correct order: quite common in the UK). Data consumers can calculate a path based on available mapped paths/trails at the point of data consumption. One caveat I can think of is that in the UK routes to the base of crags and other climbing problems are often only suitable for people with some climbing experience (sure-footed, head for heights etc), and even then “easy ways down” are notorious for falls as people let their guard down. See the red line image at the bottom of this thread (not mapped on OSM).
2 Likes
definitely very different from Fontainebleau, where toddlers play around the blocks.
1 Like
After making an experiment to discuss it with the locals (the French outdoors group, mainly) I was surprised at how difficult it was to find a rendering that supports climbing. Any suggestions?
What is the main difficulty?
In not finding? well, it appears when you have looked at all the outdoors maps you can think of, and you have no more idea where to look 
What tags are you looking for, and how would you expect them to be displayed?