[Tagging] Feature Proposal - RFC - Shrubbery V2

[Peter Elderson]

For now, I would welcome an explicit consensus that
* *landuse* maps the use of the land,
* *natural* maps areas or objects that grow or flow of themselves, whether
the object or its growth or flow is [hu]man[aged|_made] or not.

I have the feeling that is the general or reigning opinion on forums,
mailing lists and in the documentation. The state of the database does not
yet fully reflect that. This results in groups of mappers still using
different methods and may oppose actual change, e.g.  because they want to
remain consistent within a territory. Others use it, but  most do not
correct old style tagging.

Landcover tagging could overcome this stand-still, providing improvement
opportunity without touching the gigantic backlog. Which would not
eliminate  the backlog for a long time, but would provide mapping
communities with a method to clean up an entire country and build on that.
Landcover tagging is still increasing despite not being rendered on OSM
Carto, so this is not idle thinking. But is it the best solution?

I have pondered all existing and possible useful values for landcover.
Turns out that, IMO, all these values are covered by and compatible with
the natural key, as defined above. New values for micromapping will fit
right in, and data users can decide to support or ignore, that's their area
of expertise. However, even if the entire voting community is in agreement,
the backlog problem is still the same, and without at least a commitment of
major data users, nothing significant will happen.

So now, closing in on the various shrubbery and scrub proposals, I think
natural=scrub, natural=shrub (node), natural=shrubs (area),
natural=shrubbery, natural=scrubland, are valid candidates within the
natural namespace. The exact word or spelling or plural/single is not that
important to me, though I have a preference for duck directness. If a
subtag is needed to know what is meant, I think a better value is
preferable. For attribute subtags: when you tag an area of shrubs,
density=* specifies the density of the shrubs for the area. No need to use
"reverse polish namespacing". I understand why it is proposed, but I prefer
simplicity, to encourage mapping and data use.

Peter Elderson


Op vr 16 jul. 2021 om 14:44 schreef Jeroen Hoek <mail at jeroenhoek.nl>:

> On 16-07-2021 13:32, Florian Lohoff wrote:
> > So i'd vote for extending the landcover hierarchy and work
> > on rendering those.
>
> Great! Except that has been tried, and has been, and will likely again,
> be voted against or otherwise blocked.
>
> So in effect this stance (not to single your response out, but to sketch
> the general problem) results in maintaining the status quo which is:
>
> natural=scrub just means any area of scrub, bushes, and shrubbery.
>
> So what we end up with is these standpoints (amongst others):
>
> * I want to maintain the status quo
>
> * I want landcover=* for such cultivated greenery
>
> * I want landcover=* for any land cover without a meaning (such as
> implied natural-ness or man-managed-ness) beyond that
>
> * I want cultivated bushes/shrubbery under a different natural=* key
>
> * I want cultivated bushes/shrubbery under natural=scrub with additional
> tags, but not those you are suggesting, and not those that other mapper
> is suggesting either
>
> * I want man_made=* for such cultivated greenery
>
> We could draft up a proposal with landcover=* and get your support, but
> probably be voted down by half the German mappers for breaking the
> existing established way of tagging (i.e., natural=scrub). We tried
> natural=shrubbery, and got voted down (in part by aforementioned German
> mappers for aforementioned reason). We could do man_made=bushes and get
> scolded (and rightly so) for even considering that namespace for a
> natural(-ish) feature, and get roasted by Australians for inappropriate
> use of the word 'bush'. And so on.
>
> Over the past few years any attempt to move beyond the status quo has
> been shot down and blocked, because many mappers active on the
> mailing-list hold one standpoint and will vote against any proposal for
> another, so that what we end up with is: the status quo. Which,
> implicitly just tells mappers to map cultivated greenery with
> natural=scrub, natural=wood (or landuse=forest), and landuse=grass, and
> effectively map for the renderer in this case, because we have no other
> way forward.
>
> Surely we can do better?
>
> _______________________________________________
> 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/20210716/dbfd929e/attachment.htm>