[Tagging] Feature Proposal - RFC - boundary=forestry(_compartment) relations (Was "Feature Proposal - RFC - boundary=forest(_compartment) relations")

David Marchal penegal.fr at protonmail.com
Sun Feb 14 11:46:40 UTC 2021


‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
Le dimanche, 14. février 2021 11:13, Peter Elderson <pelderson at gmail.com> a écrit :

> Managed... You could add the tag by mechanical edit to the whole of Europe.

Not necessarily; for instance, many privately-owned wooded lands, once-managed woodlands, are left unmanaged in France, often because it was inherited by people that simply don't care or because the wooded land is too small to hope economical advantage through its management (for instance, too few wood for sale).

> Many mappers will not bother themselves with verifying actual, legal or whatever management. You will end up with a mixed situation, where nobody knows what it means, so data users will not bother and treat it all the same.

As long as they use natural=wood for mapping these areas, it is still an improvement to tell "this area is wooded, but I don't know/care about it being managed or not"; with current tagging customs, you cannot be sure if the mapper cared about management or simply maped a wooded land. With the new tagging scheme, expliciting its managed state with boundary=forestry is a bonus, but saying the land is wooded while unambiguously saying "there is no information about management, it can be managed or not" is still good, as it removes an ambiguity (is it unmanaged, or did the mapper not know/care?).

> Denotation urban/wild/decorative: Does not add much value. In an urban environment, it's urban and/or decorative, elsewhere it's non-urban.

If I understand correctly this tag, it is about isolated trees, and is out of proposal scope.

> I think natural=<main_landcover> is better than using the landuse key to tag the landcover.
> Natural then has the meaning of: growing or flowing itself, even if arranged or guided by man.

Totally agree with that; the proposal follows that.

> Renderers and other data users still have to deal with natural areas of all sizes enclosed in or overlapping landuses vice versa.

The proposal is only about forestry areas; such overlapped uses will still be relevant in, for instance, wooded residential areas.

> And I still can't imagine all occurrences of landuse=forest being replaced with appropriate new keys and values, ever, unless a major OSM-wide approved automatic edit is organised/performed by the DWG itself.

This change should not be automated, because we can't be sure that the mapper that used landuse=forest meant it is managed; maybe he/she simply wanted to say that the land is wooded, and did not know/care about management. The change will likely take months or years, but removing the current confusion between managed wooded lands/forestry areas is worth it, I think.

> In reality I think we are now adding yet another scheme to the ones already there.

Not really, as the new scheme explicitly separates management issues (forestry) from landcover (wooded lands), which had not been done in an unambiguous way before AFAIK.

Regards.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210214/9a3a47a6/attachment-0001.htm>


More information about the Tagging mailing list