[OSM-talk] Church building rendering bug?

Andy Allan gravitystorm at gmail.com
Thu Sep 9 13:29:01 BST 2010


On Wed, Sep 8, 2010 at 5:08 PM, M∡rtin Koppenhoefer
<dieterdreist at gmail.com> wrote:
> The buildings below highway-areas is IMHO a mapnik bug, but one by
> decision.

I'll nitpick and disagree with you on this one - it's supposed to be a
map, not a top-down-view of the world. So if the roads are, from a
cartographic perspective, more important than the buildings or
anything above the road, it's fine for the road to obscure them. For
example, if you have a cycle path and a road running parrallel and
close to one another, it's up to the cartographer to decide which one
is more important. For example, compare these:

http://www.openstreetmap.org/?mlat=51.4532&mlon=-0.2595&zoom=15&layers=C
http://www.openstreetmap.org/?mlat=51.4532&mlon=-0.2595&zoom=15&layers=M

You'll notice that the cycle map shows the cycle path over the road,
since that's more important. Anyway, a long way of saying that
highways obscuring buildings is legitimate in a map, since it's not
supposed to be an exact potrayal of the world.

> Anyway, it will render right in mapnik as well when mapped correctly
> (multipolygon). Move the tags from the area to the relation, apply
> role outer to the way and inner to the building, it's simple.

Totally agree with you on this one. The church isn't either on top of
or under the pedestrian precinct, they are not overlapping and so the
pedestrian precinct should have a church-shaped hole in it. If anyone
stuggles to see why, imagine you've been asked to paint the surface of
the square green, for a party. Would you be painting inside the
church, under it or anything else? No.

Any abuse of the layer tags would be just that.

Cheers,
Andy



More information about the talk mailing list