[Tagging] RFC: Deprecating leisure=summer_camp in favour of tourism=camp_site

[stevea]

On Mar 20, 2022, at 5:48 PM, James Crawford via Tagging <tagging at openstreetmap.org> wrote:
> In my state, summer camps are often organized by religious organizations, such as this one: https://campmcdowell.org/
> 
> They often provide more services than just the summer camp, such as religious confrences, or other organized events. The summer camp portion can be signed up for, and has an organized program with employees and a fixed schedule. It is nothing like a normal campsite, since you can't really just make a reservation and bring a tent or anything like that. In this context I think that that tourism=camp_site is not the appropriate tag for a service like this. 

This was a point in my earlier post in this thread:  there are MANY different flavors of what are called "camps" around the world.  The concept of a "camp" is a bit like how "park" has a very extensive realm of definition in US English compared to OSM's definition of leisure=park as a smallish urban manicured landscape and how at the other end of a spectrum of "park" in OSM, boundary=national_park is "something else," but in the USA there are many thousands of "county parks" and "state parks" and even more (private parks) which are also called "park" (and are in-between a leisure=park and a boundary=national_park), but which OSM would prefer not be called park (except in their name=*).  But aside from that I won't get into these complexities.  It is indeed complex and it has taken a number of years to untangle and still isn't fully resolved, at least in the USA.

Similarly, we can see how the semantics of a wide and encompassing term like "camp" (and how that differs when translated into other languages and cultures) makes for a very rich set of semantics OSM wishes to tag to capture all the subtle nuances.  There should be some effort to "create an ontology" (a possibly large exercise of a worldwide categorization of related objects).  There WILL be some simultaneous "messy tagging" along the way while completing that, which should both give rise to the many myriad differences as well as to help align a future goal of how to "properly" categorize all of those things (in a sense of "OSM has discussed this and largely agrees amongst ourselves").  Tagging doesn't happen in a vacuum, and even though we shouldn't tag for renderers, we should certainly consider good tagging schemas them as we develop both tagging and their potential future use cases (like renderers).

That being done, a renderer (or other use case) like OpenCampingMap or similar endeavors can select from a wide (and appropriate) universe of tags, tagging schemes and related tagging schemes (having to do with "camp") and choose which to apply, render, allow selections from, display in a given context, et cetera.  Sure, these things can (and will) happen simultaneously, but they should be aware of one another and "chase each other" to a finish line of harmony.  There will be conflicts, awareness of this process can reduce this.  But without a full, global awareness of "all things camp," this is almost destined to either failure, or an endless re-hash of "that's not what WE call a camp" (in our language and/or culture).

This is a good example of how OSM might consider any given tagging topic as it relates to an end-use case like a specifically purposed renderer.  The chicken (real-world object to map), egg (tag, tagging scheme) and "cluck!" (successful use case) all happen at the same time, while who can say what really came first?  But we do need a clear understanding of a set of tags (scheme, whether formalized or ad hoc) if we are going to, say, build a distinctly purposed renderer.  As one chases the other and vice versa, conflicts will be noticed and must be addressed, or at the very least this sort of discussion must happen.  Results must include understanding that "the greater context" be more fully encompassed, thus yielding methodologies to minimize such conflicts.

This seems obvious once it's all said and done, but I've read this chapter before (with "park"), so I'd like to see a smoother, more learned trajectory towards "good camp mapping" (and we don't even seem to fully know what we mean by that — yet).

Thanks for reading.