ricoz.osm at gmail.com
Sun Oct 25 15:06:33 UTC 2015
On Sun, Oct 25, 2015 at 02:39:47PM +0100, Florian Lohoff wrote:
> On Sun, Oct 25, 2015 at 10:20:10AM +0000, Gerd Petermann wrote:
> > I don't think that this is a strong point.
> > Thinking about my own edits I'd say that the
> > length could be +/- 4m because typically I just try to place
> > the nodes somewhere neer the road, if I find an existing
> > node that looks good enough I use that.
> > On the other hand, I think we make
> > assumptions about the width of the road based on its
> > type (primary, secondary ,etc),
> > so this assumption would also apply on the culvert.
> So we implicitly assume a width of a road now we propagate
> this fault to all attached objects?
> We should reduce errors by implicitly assuming something not
> increase them.
agreed. But this is open-STREET-map so perhaps the streets should
be fixed first. Does not make much sense to map culverts with
sub-meter precission while freeways are still linear ways.
More information about the Tagging