[Tagging] [Talk-us] Heavily-wooded residential polygons

Greg Troxel gdt at lexort.com
Tue Jun 2 11:11:00 UTC 2020


stevea <steveaOSM at softworkers.com> writes:

> As I mentioned to Doug I exchanged a couple of emails with
> user:jeisenberg (a principal contributor to Carto) about what was
> going on with some examples of this, and Mr. Eisenberg explained to me
> (in short) that it is a complicated ordering (or re-ordering) of
> layers issue, both Doug and I continue to scratch our heads about what
> "best practice" might be here.  (For "heavily wooded residential"
> polygons, which are frequent in Northern California).  While Doug and
> I both tend towards the preference of the "superimposed look," it is
> not always simple to achieve, due to complexities in the renderer and
> data/tagging dependencies.  And, Doug and I are certainly aware of
> "don't code for the renderer."  However, given that Doug and I are
> fairly certain that others have noticed this, but aren't certain that
> others know what best to do (we don't, either), we ask the wider
> community "what do you think?" and "What are best practices here?"

Agreed this is really hard.

First, I'm going to assume that polygons for landuse=residential do or
are intended to align with property boundaries.  I'm also going to
assume that natural=wood aligns with the actual location of trees, which
is (in mass) almost always not aligned with property boundaries.  I have
thought it an error to have natural=wood tagged on a polygon that shows
conservation land, as the adjacent non-conservation land almost always
similarly has trees (around me).

I would suggest that perhaps a "this land has some trees" landcover tag
(cover != use, strongly agreed) may make sense.  I am not sure you are
talking about this, or not.   I find natural=wood to imply that the land
has none to very little built structures, mostly trees, and the usual
understory plants.   I would definitely not want to use this tag on an
landuse=residential area with houses, but I might use it on the rear
parts of a housing area that are basically trees.   I also would not
want to stop at the subdivision line.

The basic problem here is that it's pretty straightforward to render a
map that primarily shows landuse, and it's pretty straightforward to
render a map tha primarily shows landcover.  What carto does, and what a
lot of people want, is a way to show both of them.

I would suggest that if tagging heroics are needed there is something
suboptimal in the renderer.  I think renderers probably need fancier
code to choose which of landuse/landcover to emphasisize depending on
local scale.  Or a deconfliction of symbology.

To have a way forward, I think we need a coherent design for a style
(not code, but an articulation of how it ought to work, first) that uses
some kind of symbology for landuse and some other kind for landcover.
I naively lean to solid fill, tending to lighter shades, for landuse,
and stipple patterns for landcover.  I think this is what you are suggesting.

It is interesting to think about the 80s USGS topo maps, and surely also
interesting to look at other traditional maps for inspiration.  The USGS
ones were primarily land cover and very little landuse.   But they did
have a gray "house omission tint" in built-up areas, which I'd say is
"this area has many buildings" and is sort of landcover, even though
it's a proxy for landuse.



More information about the Tagging mailing list