I think it reasonable to treat areas with mixed use according to dominant use of the area. So following your example, a retail building with residential space would just be considered retail especially if it is inside of a similar landuse=retail. As a result any of the surrounding roads would be considered unclassifiedor greater. The exception being that if a particular street had adjoining residential building.
The reverse could be true. If there was retail inside a primarily residential building. It would likely be included in any nearby landuse=residential.
In either case, the lack of or location of residential parking would have no effect on the building type or larger landuse classification.
Where is the threshold for that? Do two levels of flats in a building overrule one level with retail?
I guess dominance would not be equivalent to counting and comparing square meters, a building with 2-3 retail levels and 6 apartment levels would probably be dominated by retail use for example, e.g. when looking at the numbers of people who go there for shopping and for visiting people.
Any industrial use can easily dominate a whole quarter, but this isnât the kind of dominance this is about, as it should ultimately be looking at individual plots, agreed?
I think ideally land use classification would be based on a broader, more holistic view than scrutinizing a particular building, incorporating considerations such as whether the area is considered a âbusiness districtâ, the kind of landscaping, and other considerations that would be common sense to locals. If one were to scrutinize a particular building purely based on square footage, it could go either way. Traditionally in the U.S., and probably elsewhere, shopkeepers lived upstairs from their shops or rented out the upstairs to families as living spaces or to professionals as offices; conversely, these days, apartments lease the downstairs space for retail.
By the way, classifying land use appropriately could have some effects besides a colored fill on a rendered map. Local laws about implicit speed limits, U-turns, and street parking can depend on whether a given stretch of roadway is within a âbusiness districtâ, defined in a multitude of ways. It would be tempting for a tool like @westnordostâs osm-legal-default-speeds to use the abutters of a roadway or the surrounding landuse area as a proxy for the nuanced provisions of these laws.
maybe this can work around you, but here it is an orthogonal property. The law imposes a different speed limit according to context (âinsideâ and âoutsideâ a settlement), also here, but it has to be signposted (the beginning and end of a settlement in the sense of the traffic law). Just someone living there is not sufficient.
I always take as a rule of thump how most people use the land. In a shopping area with appartments above, the dominant landuse is residential.
interesting, because I would also take as a rule of thumb how most people use the land, and in a âshopping areaâ I cannot imagine more people living there than going for shopping, so I would definitely tag a âshopping areaâ with landuse=retail
it is rare that proposals get 62% opposing votes, but any tag you like, and landuse:secondary doesnât clash with common mapping practice in terms of key collisions, so it is probably not harmful at least, likely better than nothing, but I would rather be adding building usage by floor (or floor range).
Agreed. Even here where there arenât necessarily signs to that effect, the exact criteria under the law are so exacting (such as considering the percentage of residential abutters over a certain distance) that probably no one would know for sure until a driver has been hauled out to court over an alleged infraction. The point I was trying to make is that this speed limit evaluation tool already uses abutters as a proxy for the actual criteria, in jurisdictions where land use is relevant (not your jurisdiction).