[Tagging] Feature Proposal - RFC - landuse bush

Bert -Araali- Van Opstal bert.araali.afritastic at gmail.com
Mon Feb 15 21:58:22 UTC 2021


Martin wrong subject ? Does your answer not belong in "[Tagging] Feature
Proposal - RFC - boundary=forestry(_compartment) relations (Was "Feature
Proposal - RFC - boundary=forest(_compartment) relations")"? I can't
align your comments with the ongoing discussion in "[Tagging] Feature
Proposal - RFC - landuse bush".

Greetings, Bert Araali

On 15/02/2021 23:29, Martin Søndergaard wrote:
> This whole discussion is reminding me of the previous discussions on
> fuzzy features and the man_made=clear_cut tag from the forestry
> compartments discussion. 
>
> On Fri, 5 Feb 2021 at 10:39, Mateusz Konieczny via Tagging
> <tagging at openstreetmap.org <mailto:tagging at openstreetmap.org>> wrote:
>
>     And if something is tree covered then it is taggable as either from
>     landuse=forest / natural=wood / landcover=trees
>
>
> I mostly agree with what you are saying and I am just using your reply
> as a jumping off point. 
>
> There is a significant difference between these three tags that I
> haven't seen anyone mention:
>
>   * *landcover=trees* is an objective piece of information.
>     Doesn't matter if the area is 100 m2 or 100 million m2. By default
>     it doesn't state anything about origin (man made or natural) or
>     about usage (forestry or park etc).
>   * *forest *and*wood* are both human constructs.
>     When an area of landcover=trees fulfills some set of criteria (of
>     which there are hundreds if not thousands different definitions)
>     then we as humans designate it as a forest or a wood. I would
>     never call the small patch of trees in my backyard a forest or a
>     wood, but I still have to tag it as such. 
>
> The fact is that a forest or wood can include many types of
> landcover because it is itself not a physical thing, but a human
> construct. I have personally had problems with this when tagging named
> forests. There might be a small lake or pond within the forest, a
> small clearing of grass, or (as mentioned in the forestry compartments
> discussion) a clear cut area. These areas are all still part of the
> forest, but you can't tag it as such without just leaving them
> overlapping (which is why many lakes now have trees "growing" in them
> on carto).
>
> It would be great if I could distinguish between the actual
> physical tree cover (having "physical" tags such as leaf_type and
> species) and the concept of the forest (having "non-physical" tags
> such as name, operator, opening_hours and access) since these two
> features often do not overlap exactly.
>
> A perfect tagging scheme should allow the mapper to add information
> progressively and leave out information which isn't known. I think
> this is why the landuse=forest vs natural=wood distinction was doomed
> to fail. These two tags forced the mapper to specify the origin of the
> tree cover (is it man made or natural) before they could even map it.
> landcover=trees (and similar landcover=grass, landcover=bushes, etc)
> fixes this problem.
>
> \Martin Søndergaard
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210216/68d732cb/attachment-0001.htm>


More information about the Tagging mailing list