This mainly affects clay and fine soils. For gravel, rock and most engineered surfaces, firmness is stable across weather conditions, which covers the majority of useful cases.
For weather-sensitive materials, the subtag structure handles it:
firmness=firm + firmness:wet=soft for clay roads
firmness=firm alone where weather doesn’t change behaviour meaningfully
The wiki can flag which surfaces have high weather-dependence and recommend subtags in those cases.
This is the core tension. firmness=* is most useful precisely where the same material produces different behaviour depending on compaction, maintenance or weather. That’s what surface=* can’t capture alone.
Probably true to some extent, but tracktype has the same problem and is still widely used and useful. Anchoring firmness to observable deformation under load at least gives mappers something more directly verifiable than the current mix of dimensions in tracktype.
On the naming deadlock: the poll results suggest firm / intermediate / soft as the path of least resistance, but the real work is in the descriptions. What actually helps a mapper decide is something like “no visible impression under a boot or tyre” vs “clear deformation but still passable” vs “significant sinking, risk of getting stuck”, not the label itself.
I think we agree on this, see #2 and #6 above. I think we best create a set of surface= values with well defined and described implied firmness values (incl. their dependence on humidity content), but we also need an independent firmness key, first of all to define what firmness is, and as a backup for the many existing unpaved surface values that don’t have a well-defined firmness and for cases where the firmness deviates from the default one derived from the surface value.
That’s another way of tagging it which is more explicit, but also more complicated with the risk that most mappers won’t tag it correctly, or won’t try to tag it at all.
We could agree that firmness should describe the worst-case scenario: for most (rocky) surfaces this is constant, but for the most interesting ones (clay and sand) this varies. We could assume by default that firmness=soft is for wet conditions (the worst case for most soft surfaces), except when surface=sand when dry conditions are worse.
What would be the use of that? One might as well tag fimness=hard on tracktype <= 2 and firmness=soft on tracktype>2. Or, like, skip introducing a new tag.
To me, the new key would only makes sense if it the key(s) invented have separate values for wet and dry conditions.
In fact, reading the discussion, I’ve become quite convinced that firmness by itself would be a bad tag name, as it would instantly be misunderstood by many mappers depending on the time of the year they map (some would map “dry firmness” and some “wet firmness”).
Thus I would propose two tags: wet_firmness=* and dry_firmness=* instead.
That way:
new information could be added to the map which wasn’t available before (firmness as a dependence on the weather conditions)
name itself is not confusing so there is much less potential for misunderstanding
people who only care about one but not the other can map just that one
even without reading the wiki (which is the real risk as we see) it is quite likely that most mappers would get the meaning mostly right
it can define both materials which get softer when wet (like dirt) and those who get softer when dry (like sand). And of course those which remain the same (like asphalt)
it allows for partial mapping instead of mappers guessing (i.e. if it is “firm” in the sun, you don’t have to guess whether it would become “soft” - much less how soft - in the rain, or not). You’d map just what you have verified.
editors (both in sense “apps” and in sense “humans”) would decide where it makes sense to tag (which could also be suggested by wiki). E.g. hypothetical StreetComplete (or SCEE) Quest could simply skip asking the quest for surface=asphalt.
it makes it clear when only partial information has been entered (i.e. opposed to firmness + firmness:wet when single firmness=* might be misinterpreted to mean “the same firmness in both conditions”, as :wet would be by definition “lower caste citizen” )
it does not require that mappers go map information only when it is raining (or soon thereafter) as e.g. suggested “only worst case scenario firmness=*” would)
In parts of the world with cold winters, the ground also freezes. This turns soggy, soft ground into firm ground. Would frozen be considered “dry” for the purposes of this tagging scheme, or would that also be confusing?
I’ve intentionally left that one out, even if I myself mentioned it before as one of the factors (temperature and pressure are factors in addition to composition of the soil which is often affected by presence of liquids - usually water)
Are there surfaces which are firm “when wet and warm” which become soft(er) when frozen?
If they are no such surfaces in practice, can it be assumed that anything which is soft when wet (and hard when dry), will become hard(er) when frozen? If so, there IMHO no much need to map that attribute separately, as it can be inferred from the existing wet_firmness and dry_firmness (and actual temperature).
but if there is a need for that, yes, frozen_firmness could be invented too by people interested in that. But I’ve not seen such need in discussions so far.
in any case, it could certainly be documented on the wiki to reduce possible confusion.
That was my initial idea, yes. But I’ve learnt that tracktype is multidimensional for most mappers, so that won’t work unless it’s known that that particular mapper is a “pure firmness” mapper (like me )
Interesting idea! The main disadvantage I see is that it would be 2 keys instead of one.
Well Key:winter_road - OpenStreetMap Wiki is often used through soggy terrain that is soft and impassible much of the year, but firms up in the winter. So there is precedent for mapping this kind of situation. @Matija_Nalis reasoning for why firmness when frozen doesn’t need to be included in this new tagging scheme makes sense to me though.
Not that I can think of. At least not any material that would be used on a road or path surface.
2 Likes
Mammi71
(One feature, Six mappers and still More ways to map it)
92
)