[Talk-us] script for adding layer=1 to bridges
dale.puch at gmail.com
Sat Jan 30 07:00:35 GMT 2010
On Fri, Jan 29, 2010 at 5:31 PM, Alan Mintz
<Alan_Mintz+OSM at earthlink.net<Alan_Mintz%2BOSM at earthlink.net>
> At 2010-01-29 14:00, Dale Puch wrote:
> >Your talk of "absolute levels of elevation" got me thinking.
> >The tag has nothing to do with a measurement, only order of the layers.
> What I'm saying, though, is that, if you tag with an eye on making the
> layers reflect reality, it allows you to tag new features in the future
> without too much collision with existing features.
I think this is what leads ti it being so confusing. Your trying to set
layers based on possible future construction rather than the current layers
of the features.
Keep it simple!
> >Some examples might be an elevated (built up earth) interstate in the US
> >next to a city road. Both are layer 0 but not at the same elevation.
> I'd say it depends on how high it is in relation to the street. The
> question is, what do the streets that cross the freeway do? If the street
> underpasses go under the freeway without either the freeway or the roads
> changing much in elevation, it makes sense to make the freeway layer 1 and
> roads layer 0. If, OTOH, the roads have to drop substantially in elevation,
> then I would make the freeway and the road segments away from the freeway
> layer 0, and make the underpasses layer -1.
> As I said before, I don't think this has much to do with tagging something
> as a bridge or a tunnel - that is a matter of adherence to official
> categorization and/or rendering.
> I consider the "ground" (layer 0) level for a given area to be the position
> of the majority of the streets and street-level features.
> > A road on a cliff 1000 feet next to a river at the bottom are both layer
> I disagree. As far as the exact values to use, I'd want to look at the
> surrounding area, but I'd say the cliff has to be at least one layer higher
> than the river regardless. What happens when they build a bridge over the
> river to the cliff on the opposite side? You then have to raise the layer
> of the bridge by one, even if the ground truth doesn't reflect this (i.e.
> bridge at-grade with the cliff). If there was a cable car already at layer
> 1, you'd have to raise it too to avoid collision with the bridge, etc.
But my example is EXACTLY how it works everywhere else. Why are you making
special cases for a gorge vs. a simple stream?
Expand the example with a bridge also and compare it to the same setup, but
on flat ground. Why make them different just because the stream example
elevations are only a few feet vs ten times the height?
Your suggesting a road in the mountains should change layers if it at the
bottom of a valley vs on the valley wall? Is it layer 1 half way up? What
is it has a switchback or 10, what levels do you apply... It gets so
confusing if you start down that road. And again layers is just that, NOT
> >An elevated train track above city streets would be layer 1 (and probably
> >not tagged at bridge)
> Agreed, but why is this different from the freeway case, which will also
> have streets under it at some points?
> >Personally I would tag the top roads in Chicago as layer 0 and the ones
> >underneath as layer -1 and tunnel.
> There are, of course, places like this where a consistent "ground level"
> doesn't really exist, as traffic, natural features, and creative use of
> limited space have created more of a dual-layer structure, but these are
> more the exception than the rule I think.
> Alan Mintz <Alan_Mintz+OSM at Earthlink.net>
> Talk-us mailing list
> Talk-us at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Talk-us