There are problems for OSM with the surface types fine_gravel, gravel and compacted. There are subtle shifts in the way the word gravel is understood by various speakers of English. Nearly all roads are compacted, a mix of particle sizes is typically graded and rolled. The larger particles leave gaps which are filled by smaller particles, the gaps between them are filled by sand and between them are dust. Later, as the road degrades, the dust and fine particles are lost to water and wind and the larger particles dominate the top layer. The road user sees loose gravel but underneath a thin layer of loose gravel is a compacted road.
Both the photos at fine_gravel are probably compacted, File:Finegravel0.jpg and File:DSC05537a-Feinschotterweg.jpg
I think there is little difference between loose gravel and compacted, maybe only a year’s wind and rain.
Some final notes. Gravel can refer to river gravel or crushed rock. Both compact with a range of particle sizes in the mix but angular crushed rock compacts better. Compacted requires the user to make a judgement below the surface where they cannot see, we are encouraged to tag with what can be ground truthed. Railway ballast, 40mm or bigger is a rare surface in my experience, it seems counter productive to use “gravel” for it. Users will continue to tag compacted surfaces as “gravel” regardless what is written here, shouldn’t the wiki document actual tagging rather than try to impose definitions.
Yes, I think this debate is really fruitless tbh. The definitions (past & present) include surface texture as well as subsurface composition (eg constructed road bases). All combinations are possible. Photos only show the surface, not what’s beneath. Contributors emphasise factors such as “looseness” which is yet another feature again, and one which quickly changes after rain and time since grading. The definitions of the existing categories (compacted, fine gravel and gravel) are all a bit of a mess and, not surprisingly, their application across the world is even messier. The more often that people change the wiki, the worse things get. Adding selective photos to override the textual descriptions doesn’t help either. I’d suggest that we accept that it’s all pretty vague and move onto other, more productive initiatives.
I wouldn’t call it fruitless at all. It’s important if you want to do bike routing with OSM data, for example. There’s a big difference between a lightly gravel-topped firm towpath and an unimproved trackbed of railway ballast, but both are sometimes described as “gravel”.
Yes, maybe just move on. But if we are going to make any progress, I think that good quality photos are essential. I may be guilty of “Adding selective photos to override the textual descriptions” but if so, it is unintentional, I have tried to illustrate what is in the text and any shifts in meaning are unintentional. More photos please.
In Sweden I have been using surface=fine_gravel as a less technical/less precise term of a gravel road with relatively smooth surface, while I’ve seen surface=compacted as a more technical and actually describing more then just the surface, but how the road is constructed. I generally use just “unpaved” though for roads I don’t have more precise local knowledge of. Other Swedish mappers often use “gravel” as the generic term for unpaved road, as in daily Swedish speak we say “gravel road” about basically anything that isn’t paved.
Anyway, lately I’m starting to regret some of my fine_gravel uses and would probably have used compacted instead. The surface of a maintained compacted road varies over the year, here during the spring new coarse gravel is put on, which then gradually is worked into the road to provide a harder more even surface.
Yes, I ride a gravel bike. I fully agree that creating a separate tag for rail ballast is paramount. It’s what remains after that’s gone that’s a mess, given the poorly-differentiated mix of gravel, fine gravel and compacted in both the wikis and on the ground.
If you set an unusual surface=* value, then there is a high risk that navigators will not understand at all and skip this tag. It is also very important how the surface behaves in wet conditions. For similar cases, I usually set surface=unpaved.
I am curious to know where the wiki definition of gravel came from. “Broken/crushed rock with sharp edges, known as ballast on railways. Usually loosely arranged. Typical size for road and surface construction range between 4 and 8 cm”.
Does this match how roads are constructed and described in some countries? Which ones? Who is the gravel tag intuitive for? Are there a lot of roads in our database with ballast of 4 to 8 cm that are tagged as gravel?
As far as I can remember, I have come across exactly one road with 4-8 cm chunks of rock in my 51-year life — this one: Way: 1073325423 | OpenStreetMap . Since there is a photo in the wiki, and someone wrote the description in the first place, I suppose there must be more of them somewhere…
There’s a good description of the history of the expanded gravel tag on the Australian forum here. Short extract, “ TL;DR: Looks like a non-native English speaker individual created a continuum of of sand - fine_gravel - gravel in 2008 when sand - gravel - rough_gravel or just sand - gravel would have fitted the vernacular better. Never discussed or challenged…”. Lots more follows. I can’t comment on the veracity of this summation but my interpretation of it is that there may never have been a specific intention to include ballast specifically, instead the definition may have been expanded somewhat inadvertently. Who knows. Anyhow, the link is well worth a read.
I use “compacted” a lot in for Australian and Swedish roads. To me, compacted is a method going back to (European) roman times of having a layer of mud with embedded fine gravel fragments for cohesion and traction. Unless it has been very recently maintained, there is little or no LOOSE gravel. A hard surface is created. So for cycling or trying to drive a motor vehicle at speed, it is an entirely different experience to a gravel road of any kind. So definitely “fine_gravel” is for small loose gravel and does NOT equate to “compacted”.