[Tagging] The showstoppers for mapping Scandinavian nature.

[Andrew Harvey]

On Mon, 28 Dec 2020 at 15:55, stevea <steveaOSM at softworkers.com> wrote:

> On Dec 27, 2020, at 6:01 PM, Andrew Harvey <andrew.harvey4 at gmail.com>
> wrote:
> > I don't see the issue with just mapping these areas as polygons, we have
> the 3,500km long Great Dividing Range mapped as fuzzy=50000
> https://www.openstreetmap.org/way/211234843. Like anything in OSM, it
> starts with a rough outline and over time can be better mapped in more
> detail. Would be silly to try and expect this to be mapped as a point and
> not a polygon.
>
> Andrew, while I appreciate that this seems like a slam-dunk easy answer to
> a non-specific question (Anders did start out with some wetlands which were
> confusing, I agreed), I click on "fuzzy" as a tag and find nothing.  While
> I appreciate that a tag of fuzzy=50000 "feels like to me" a method to say
> "rough in a significantly large way" I don't know if that its value is one
> of size, precision or what, as "fuzzy" has no wiki or otherwise docs for me
> to see what is meant (I want to say "exactly," but especially in this case
> — fuzzy! — that seems difficult).
>
> I get it that this is something like the fifth largest mountain range on
> Earth and that its shape and edges aren't precise nor widely agreed to.
> And that sets off my OSM spidey-sense that something isn't verifiable and
> we're back to that.  So, these remain open questions, especially as "fuzzy"
> isn't (well) documented (at all?  as a tag, I mean).
>

Yes we should work towards a tagging proposal for fuzzy to give it some
documentation, but to me it just means that this isn't an exact border but
within a tolerance of X meters satisfies most references and sources. We
might not be able to say exactly where the border is, but usually you can
say it's somewhere between A and B,


> Something doesn't add up:  naming natural things is a task in exactly
> defining the amorphous.  How DO we solve that?
>

A fuzzy or tolerance tag does exactly that, it says the actual extent falls
somewhere within the tolerance but we can't say exactly where.

Everything has a tolerance in OSM, nothing can be measured precisely, even
the location of a man_made=survey_point has some error in it. Just because
we can't map it exactly, doesn't mean we should not try to map it at all.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20201229/93651be3/attachment.htm>