Open jeisenbe opened 5 years ago
Good suggestion. Rendering pitch numbers at high zoom levels would be great. Here is an example of a scout camp site that already has numbers on its pitches, but it simply uses name
=50
on landuse
=grass
. Is using ref
better than using name
? I suppose pitches could have both a ref number and a name, which should be rendered in that case?
I'm not aware of many campsites with named pitches. Usually there is a number or letter and number for the pitch.
Such numbers or strings of numbers and letters should use the ref=* key, as suggested at the wiki page: https://wiki.openstreetmap.org/wiki/Tag:tourism%3Dcamp_pitch
I've reviewed the current uses of name=* with tourism=camp_pitch.
There are 618, compared to over 5000 with ref=, and it looks like
almost all of them are something like name=
I'd still see that as a rarely used tag. It is a form of micromapping, if you consider one campsite having potentially 100 or more pitches, you'd divide the tag count accordingly. Thus, only few campsites are really using it.
A camp pitch is usually the size of a house, so I wouldn't call this micromapping; perhaps mini-mapping? :-)
While there are large campsites, just like hotels the come in all sizes, from 1 pitch up to hundreds, and the smaller ones are much more common.
I found 42 nodes, 107 ways, and 8 relations tagged tourism=camp_site or tourism=caravan_site within 200 meters of a tourism=camp_pitch, so that's at least 157 campground and caravan sites using tourism=camp_pitch - https://overpass-turbo.eu/s/L7S (includes all types)
There are another 410 tourism=camp_site and tourism=caravan_site within 200 meters of a camp_site=camp_pitch (the deprecated, older tagging), so I suspect a few hundred more features will switch tags or be added in the next few months.
On 7/28/19, polarbearing notifications@github.com wrote:
I'd still see that as a rarely used tag. It is a form of micromapping, if you consider one campsite having potentially 100 or more pitches, you'd divide the tag count accordingly. Thus, only few campsites are really using it.
-- You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub: https://github.com/gravitystorm/openstreetmap-carto/issues/3827#issuecomment-515708305
"size of a house" is a wide range. Anyway I'd consider a camp pitch to be the size of a tent or a caravan.
Here is an example of a camping that has all pitches mapped. They are about the same size of neighbouring buildings, so it makes sense to show pitch numbers when also showing house numbers. However, I agree that this tag is not used very much. Not many camp sites have their pitches mapped.
"size of a house" is a wide range. Anyway I'd consider a camp pitch to be the size of a tent or a caravan.
I don't know how it is in other places, but there's a lot of individual tent camp sites in California where the sites are rather large due to including bbq's, picnic tables, and a parking space for a vehicle. It doesn't seem clear from the wiki if the area of those other things are included or if it's just confined to the small spot of the individual camping area where the tent goes.
On Mon, Jul 29, 2019 at 7:39 AM Adamant36 notifications@github.com wrote:
I don't know how it is in other places, but there's a lot of individual tent camp sites in California where the sites are rather large due to including bbq's, picnic tables, and a parking space for a vehicle. It doesn't seem clear from the wiki if the area of those other things are included or if it's just confined to the small spot of the individual camping area where the tent goes.
I haven't started trying to map individual pitches at any of the campgrounds I've visited, but some are quite large indeed because their areas include vegetation for visual screening from the neighbouring sites.
I agree that the Wiki isn't 100% clear, but mapping just the tent pad seems, uhm, narrow-minded, literally.
@polarbearing, for future reference, how many tourism=camp_site
and tourism=caravan_site
features should be mapped with pitches before we should consider rendering the pitch ref?
There are currently 100,000 tourism=camp_site features, but only 33,000 mapped as an area (the others may be single-pitch sites), and 26,000 caravan_site features (14,000 areas).
Perhaps 1000 would be sufficient? This might translate to 20 to 40k individual tourism=camp_pitch features.
It's been six months since the issue was opened and there are now 12,255 uses of tourism=camp_pitch (double the original amount) and 7,526 of them have ref=* tags. So, any chance of revisiting this now that the usage numbers have increased a lot?
The current usage map at taginfo shows there are several places where this has been mapped extensively (e.g. California, Washington, a couple other states in the US, central Europe, and some spots in Australia), though there are one or two locations in many other countries:
Any updates on this?
25k uses of tourism=camp_pitch
now (70% of which have ref=
), I think this is enough to consider it worth of rendering.
Thought I would also point to https://opencampingmap.org The render (for me) is with older data but it's a useful site anyways. OpenCampingMap Example Same site on OSM
It's impressive that he number of tourism=camp_pitch
features with a ref=
tag has increase from 7526 on 18-Jan-2020 (https://github.com/gravitystorm/openstreetmap-carto/issues/3827#issuecomment-575877146) to 18187 today, 5-Sep-2020 - more than 1000 have been added every month.
The map shows increased adoption of the tag in Japan, Taiwan, South Korea, Namibia, East Africa and Mexico, since I lasted checked (https://github.com/gravitystorm/openstreetmap-carto/issues/3827#issuecomment-575894855) - https://taginfo.openstreetmap.org/tags/tourism=camp_pitch#map - though usage continues to increase fastest in North America and Europe.
It was mentioned above that many of these might be found within one campsite (campground). While that is true, in my random check of sites around New England, only one site had over 100 pitches, and the others had 1 to 2 dozen nodes located within a tourism=camp_site area. Most of the rest in Vermont and adjacent areas were just 1 or 2 pitches by themselves, not in an area, which is not how the tag was expected to be used, but the numbers are small. If we limit the search to just those with a ref=* they are more likely to be in a campsite.
Looking at the whole USA, there appear to be about 200 different campsites with pitches mapped. From looking at the rest of the map, I would estimate that about ~500 different campsites have been mapped with pitches and refs, world-wide. This is perhaps not enough for rendering, but considering the relatively rapid growth in use of the tag it should be common enough in a few more months. So perhaps we could work on getting a PR ready.
What's wrong with rendering ones without a ref=
tag?
I mapped a camp/caravan site myself and the refs for them where only visible on some pitches, the others didn't have signs.
This would mean that only half of the pitches in that campsite would be rendered. (This is the site I used in my example above)
What's wrong with rendering ones without a ref= tag?
I assume only rendering ones with refrence numbers weeds out the ones not being mapped at camp-sites. Although, I think that would be faulty because like you say some just don't have refrence numbers, even if they are in camp sites, and only rendering ones that do might incourage people adding a refrence number in cases where there shouldn't be one.
We could attempt to render tourism=camp_site
areas with a fill color, but I do not enjoy the prospect of trying to find a color which would work in all cases. This would also only work for those which are mapped as areas (closed ways or multipolygon relations), not for nodes, while rendering the ref=
is equally possible for nodes.
Since tourism=camp_pitch
features are mostly unvegetated, they should not be green (which are are trying to reserve for types of vegetation, and which is used for leisure=pitch
). They could be similar to landuse=residential
or landuse=farmyard
semantically based on usage, or @bare_ground
areas based on typical landcover (though this varies widely). That would suggest using
Currently tourism=camp_site
and tourism=caravan_site
are rendered as @campsite:
#def6c0
Compare the other options:
@residential:
#e0dfdf
@farmyard:
#f5dcba
@bare_ground
: #eee5dc
- used for gravel, shingle, mud etc
@sand:
#f5e9c6
There has been a lot of discussion about pitches being in or not in a camp site. Not all pitches are located in a camp site. For example, primitive camp pitches are often found along long trails. Rendering these are incredibly useful for people planning multi-day hikes.
At high zoom levels if simple icons indicating tent, caravans, electricity, water, etc. could also prove useful as long as it doesn't make the map too busy.
Is there not a case to be made that "if you book it, they will come", i.e. if OSM users see individual sites rendered, they'd be more likely to add sites themselves in the future? I added a few and thought they'd been deleted, which was disappointing, until I just chatted to a helpful person on the IRC who suggested I turn on 'map data' and check for the issue here. There might be hundreds like me.
@jeisenbe is a fill colour a requirement for rendering? If not, and the site just displays the boundary over the underlying vegetation (campground), that would seem sensible in my opinion...
At high zoom levels if simple icons indicating tent, caravans, electricity, water, etc. could also prove useful as long as it doesn't make the map too busy.
I fully agree with this. When out of cell signal range, OSM really shines, and having these symbols would be great. I added multiple taps to a campground recently, but again was discouraged from doing more after they didn't show up.
As I understand it, there's no option to force rendering of such details in OSMand even when one knows about this technical issue, so those in the field are unable to benefit from others' inputs.
Is there not a case to be made that "if you book it, they will come"
Not everything can or should be rendered. Instead, there needs to be a balance between rendering things that have a likely chance of sticking around and not rendering things that don't. Which is mostly based on the current usage. 33,755 (22,888 with a ref number) is a lot though. But from what I remember and just read over the reason it hasn't been rendered yet is more due how exactly it should not being resolved yet then it is about the usage. I could be wrong about though since it's been a while.
I added multiple taps to a campground recently, but again was discouraged from doing more after they didn't show up.
If you think water taps are worth rendering no one is going to object to you opening an issue requesting it. That said, per the Wiki you should probably be adding amenity=drinking_water
to them instead of drinking_water.
I'm pretty sure they will render as water fountains if you do. Not that you should tag for the rendering though.
When out of cell signal range, OSM really shines, and having these symbols would be great.
The main style on the website, which is what this issue tracker is about, doesn't load when there's no signal. I guess you could technically load it in Docker without an internet connection, but I doubt anyone is doing that while they are camping.
As I understand it, there's no option to force rendering of such details in OSMand even when one knows about this technical issue, so those in the field are unable to benefit from others' inputs.
That's something that you should take up with on OsmAnd GitHub page. This style isn't connected to them.
At high zoom levels if simple icons indicating tent, caravans, electricity, water, etc. could also prove useful as long as it doesn't make the map too busy.
I don't think you could make a legible icon at higher (or any) zoom levels that would indicate all those things. The only way to do it is render each one individually. Which wouldn't work at higher zoom levels. Really, it probably wouldn't work at lower ones either.
One of the main problems with this style IMO is that it started out to focused on rendering things that exist in big European cities. Which came at the cost of rendering more rural, non-Eurocentric POIs. Unfortunately, for various reasons that can't be changed now. We're kind of stuck with what we have.
Stumbled across this issue while micromapping campgrounds. To recap in concise terms what is the blocker so consensus can be reached and at least some base level of rendering started? Deciding in what cases to render?:
There are some very notable consumers of OSM data. For example, the National Park Service in the US uses OSM data both for their browser map (ex: https://www.nps.gov/yose/planyourvisit/maps.htm) and mobile. While they technically use Mapbox for tiles, OSM leading the way in showing camp pitches could encourage other tiling services which would be notably beneficial.
There is currently no specific decision concerning this were we lack consensus. So far we lack a concrete proposal in the form of a concrete code change how this can be depicted in a way that takes into account the previous discussions and the constraints of this style.
The tag itself is not very widely but quite consistently used (with the constraints that have been mentioned) in many parts of the world so it is a fairly good candidate for adding rendering in principle - if a suitable way to do so can be found.
Now tourism=camp_pitch
has been used over 40,000 times and has doubled in the past 2 years, but that is still only 1/3rd as many times as the number of tourism=camp_site
objects: https://taginfo.openstreetmap.org/compare/tourism=camp_site/tourism=camp_pitch
67% have a ref=* tag (29,500 objects).
The tag is up to 86k occurrences now, which is 60% as many as tourism=camp_site
. Considering that this ratio has grown by 2x since @jeisenbe last commented, I'd advocate for rendering it.
Should we render tent pitch locations within a campsite (tourism=camp_pitch)?
Currently tourism=camp_site and tourism=caravan_site are rendered. Campsites often have numbered and signed pitch locations, where a tent can be pitched or a camper trailer can be parked. These are not usually considered addresses so they are not rendered.
This situation is somewhat like that with to allotments=plot, which is a reference number that is useful for routing and orientation, but not an official address. We could render the "ref=*" tag for tourism=camp_pitch at high zoom levels.
tourism=camp_pitch was recently approved. Although it had already been in use, camp_site=camp_pitch had been the more common option until recently. Now there are 6837 uses of tourism=camp_pitch, versus 4419 of camp_site=camp_pitch, and 5066 tourism=camp_pitch have a ref tag as well: https://overpass-turbo.eu/s/L6B