[Talk-us] Why?

[stevea]

Paul Norman quotes my previous post in this thread and writes:
>  This is describing the actual landuse, not the legally permitted landuse.
>
>  An example of describing the zoning instead of the actual landuse is
>  marking areas of the desert with no development as landuse=residential
>  because the government has at some point in the past zoned them as
>  residential.

Paul, I agree; I understand this distinction.  For example, there was 
a tiff in Scotts Valley, California (not far from me) circa 2009 
where one OSM user entered landuse polygons directly from the 
published zoning map from the Scotts Valley City Council.  I began to 
correct these where actual on-the-ground data disagreed with the 
zoning map.  For example, many areas listed as zoned commercial are 
more like "intended to become commercial someday" but are truly 
residential in real life/on-the-ground, so I corrected them to be 
landuse=residential.

I take it as widely accepted that on-the-ground landuse is much 
preferred to be entered into OSM than is "zoned by the government" 
landuse.  The former is correct, the latter is not and should be 
removed or corrected.  Especially when the zoning represents an 
intention rather than reality.

What I understand Martin Koppenhoefer to say are essentially the same 
things, but I'm not sure if he understands (or agrees) with Escondido 
having large areas marked as landuse=residential.  These are not 
simply zoned residential (they are), they ARE (on-the-ground 
verifiable) residential.  So it is OK for them to be tagged as they 
are.  They might also receive more detailed tagging in addition to 
this simple landuse polygon, a highway=residential street running 
through them, and not much else.  These "skeletal" data are largely 
what are in OSM now across much of the USA, yes, I and many others 
know.  However, buildings, address data, and other micro-mapping 
detail are being added.  BOTH flavors of data are correct.  While 
skeletal data aren't exactly preferred to "largely complete" data, 
they are not incorrect, they are just not as complete as they might 
be.

Landuse data should show what actually IS, not "simply" what is zoned 
and especially not what is intended.  Yes, zoning data are a bit raw, 
and may be considered "early" or "a first step" for OSM.  They need 
updating, they change over time.  They may be "too broad" as where 40 
acres are tagged as landuse=farmland where only 39 of them actually 
are landuse=farmland, but one acre is a house (landuse=residential) 
and perhaps landuse=farmyard where the barn and tractor and 
irrigation supplies are.  Would I rather see this perfectly mapped in 
OSM, exactly as I describe such micro-mapped details?  Yes, 
absolutely.  Will I say that tagging all 40 acres as landuse=farmland 
is "totally incorrect?"  No, though if I or somebody else has the 
time to tag with those better details, OSM sure will appreciate it. 
Should OSM show landuse=commercial because the County Supervisors 
just approved a shopping center be built on this farmland in the 
future?  Absolutely not, especially if it is still a working farm and 
no construction has yet started.

Are we all agreed?  Thanks for good, productive discussion.

SteveA
California