I don't know where the notion came from saying a greenhouse is not a
building. It is not a residence but it is most certainly a building - it
has doors, walls and a roof, Just because they're made of glass doesn't
disqualify it from the building category.
On Thu, May 19, 2016 at 3:29 PM, Martin
2016-05-19 20:50 GMT+02:00 Marc Gemis :
> On Thu, May 19, 2016 at 5:10 PM, Marc Zoutendijk
> wrote:
> > That's because everything that is a building should be tagged as a
> building”
>
> so I looked at the wikipedia definition of building [1], and
On Thu, May 19, 2016 at 5:10 PM, Marc Zoutendijk wrote:
> That's because everything that is a building should be tagged as a building”
so I looked at the wikipedia definition of building [1], and their
definition of non-buildings [2]
At this moment I think that
What's about
building=roof
layer=1
roof:material=plants
rendered on map, way may pass below, no new tags necessary
Kurt
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging
> Op 19 mei 2016, om 14:46 heeft Marc Gemis het volgende
> geschreven:
>
> On Wed, May 18, 2016 at 6:45 PM, Marc Zoutendijk
> wrote:
>> And as a side note I could say:
>> Combining man_made=* AND building=* on the same object (as is often done to
Thanks Marc, yes, shelter_type is the one I've used in the past. I was
using my phone while boarding a plane and didn't have time to do a proper
look-up. There have been discussions in the past about the use of the
"type" key and so shelter_type would be much preferred.
On Thu, May 19, 2016 at
On Wed, May 18, 2016 at 6:45 PM, Marc Zoutendijk wrote:
> And as a side note I could say:
> Combining man_made=* AND building=* on the same object (as is often done to
> make it appear on the map) is as wrong as using highway=* and waterway=* on
> the same object. It is
Rory McCann wrote:
> Both of the example maps of Russia/Ukraine and India/Pakistan
> require the use of another data set. Which is a shame. One should
> be able to generate that from OSM entirely.
Why?
OSM's selling point is not "all geodata, ever, in one place". OSM's selling
point is
I see two parallel subjects here:
1) how do we represent disputed borders and "different versions of the
truth" in OSM
2) how do we use that mechanism responsibly
Whatever criteria are used for 2), the chances are there is always going
to be a need for 1).
//colin
On 2016-05-19 12:08,
2016-05-19 11:55 GMT+02:00 Simon Poole :
> The current mapping of de-facto boundaries of effective control is
> easily defensible and has a certain logic that even the greatest
> nationalists typically will accept (that knowing who really controls an
> area is helpful in avoiding
As somebody who regularly has to respond to complaints from officials
and others on boundary matters, I fail to see how mapping additional
borders (and a lot of them, given that there are nearly no countries
without disputes) is going to help. Matter of fact it is guaranteed to
make things worse
2016-05-19 10:05 GMT+02:00 Rory McCann :
> I'm not suggesting mapping every little "someone in $COUNTRY thinks
> $AREA should be in their country", I'm suggesting mapping areas which
> governments claim. Imagine you had to make a map for the government of
> $COUNTRY, and
On 07/05/16 11:54, Andy Townsend wrote:
> The problem with answering Rory's original question directly is that in
> OSM we try and "map what's on the ground", and don't map stuff that's
> never going to happen (for example, if a village thinks that it'd be
> really nice if there was a bypass
Why not the established shelter_type [1], which is used over 69.000
times according to taginfo [2]
[1] http://wiki.openstreetmap.org/wiki/Key:shelter_type
[2] http://taginfo.openstreetmap.org/search?q=shelter_type
On Thu, May 19, 2016 at 6:09 AM, David Swarthout
wrote:
14 matches
Mail list logo