[OSM-dev] Tagging of multipolygons with holes within holes

[Igor Brejc]

On Wed, Oct 17, 2012 at 12:58 PM, Martin Koppenhoefer <
dieterdreist at gmail.com> wrote:

> 2012/10/17 Igor Brejc <igor.brejc at gmail.com>:
> > Another thing that's odd is the duality of inner rings: they represent
> > holes, but they can also represent holes AND separate polygons at the
> same
> > time. Decision on which is which has to be made based on a vague notion
> of
> > "difference" in tagging.
>
>
> This is something I don't find odd. An inner ring is always a "hole"
> in the multipolygon, but it is clear that what is inside this ring is
> not "nothing" but something, and this something will often also be
> tagged. But it is tagged on the ring itself (or on a different
> multipolygon where it is an outer way), so I don't see where the
> "decision" comes into play. Or are there implementations where the
> tags of inner ways get also inherited to the multipolygon-area?
>
>
Well, reading this (from the Wiki):


>    - If the inner way represents something in itself (e.g., a forest with
>    a hole where the hole is a lake), then *the inner way may be tagged as
>    such*.
>
>
>    - Otherwise the inner way(s) should be left *untagged*.
>
>
What does "untagged" mean, exactly? That the way is completely without tags
or that it just doesn't have tags that are in conflict with the outer ring?
Example: if I put "source=Wikipedia" on the inner tag, that still
represents a hole. However, if I put landuse=residential, then it's no
longer just a hole, but a smaller polygon in itself. Jochen talked about
this in a previous reply.

But yes, you are right: in the end, you can always look at any hole as
being "something" and not just a hole. I guess it depends on what you want
to do with it.

Igor
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/dev/attachments/20121017/ac13b20b/attachment.html>