Re: [Tagging] RFC: new key Landcover
some good illustrative photos at http://en.wikipedia.org/wiki/Shrubland On Tue, Nov 16, 2010 at 10:45 AM, Richard Welty rwe...@averillpark.netwrote: On 11/16/10 10:29 AM, M∡rtin Koppenhoefer wrote: 2010/11/16 Richard Weltyrwe...@averillpark.net: On 11/16/10 10:11 AM, M∡rtin Koppenhoefer wrote: i like it. i would definitely move scrub from natural to landcover. there are large areas which are properly called that, with sporadic trees and the rest of the cover being of a bushier form. but wouldn't that be landcover=bushes (where they are)? Is scrub something like heath, beach, bay, wetland, fell? Then it would be OK in natural IMHO. Or something like tree, grass, rock? Then it would be OK for landcover IMHO. scrub is generally a mixture of trees and bush sized objects, frequently on sandy soil. the interior of Florida features extensive pine scrub, with a scattering of very large pine trees and lots of palmettos (bush sized relatives of the palm tree) filling the space. here in Albany (upstate NY) we have the Pine Bush, which again is pine trees interspersed with bushes. it's a very well defined ecological niche. richard ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
2010/11/16 Tobias Knerr o...@tobias-knerr.de: Maybe a few examples can help clarify this: What key would I use to describe the surface of a race track in a stadium - landcover or surface? surface Would the same surface material on the basketball court next to it (clearly an area, not linear) be tagged differently? no, that would be surface as well. I'd say the distinction is between the surface and the coverage (which comprises the surface). surface=bush or tree would not make any sense IMHO. surface=asphalt is fine for the surface, the landcover would be the street which is not only the surface of the street. cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
On 11/16/10 12:43 PM, M∡rtin Koppenhoefer wrote: no, that would be surface as well. I'd say the distinction is between the surface and the coverage (which comprises the surface). surface=bush or tree would not make any sense IMHO. surface=asphalt is fine for the surface, the landcover would be the street which is not only the surface of the street. my attempt at clarification: surface is used where the mapped entity is man-made (or modified, e.g. dirt roads.) richard ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
[Tagging] FW :Re: RFC: new key Landcover
If you wanted to describe both the soil and what is growing on the soil, would you use both a surface tag and also a landcover tag? For example, if you had a sand dune stabilized by beach grass, would you use surface=sand and landcover=beach_grass? ---Original Email--- Subject :Re: [Tagging] RFC: new key Landcover From :mailto:rwe...@averillpark.net Date :Tue Nov 16 11:48:51 America/Chicago 2010 On 11/16/10 12:43 PM, M∡rtin Koppenhoefer wrote: no, that would be surface as well. I'd say the distinction is between the surface and the coverage (which comprises the surface). surface=bush or tree would not make any sense IMHO. surface=asphalt is fine for the surface, the landcover would be the street which is not only the surface of the street. my attempt at clarification: surface is used where the mapped entity is man-made (or modified, e.g. dirt roads.) richard ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging Sent from my Verizon Wireless BlackBerry ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
2010/11/16 Brad Neuhauser brad.neuhau...@gmail.com: some good illustrative photos at http://en.wikipedia.org/wiki/Shrubland I think that it is clear what kind of areas are shrublands, the question is if it qualifies for landcover (e.g. desert or beach or wetland do not qualify for landcover). cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] FW :Re: RFC: new key Landcover
2010/11/16 j...@jfeldredge.com: If you wanted to describe both the soil and what is growing on the soil, would you use both a surface tag and also a landcover tag? For example, if you had a sand dune stabilized by beach grass, would you use surface=sand and landcover=beach_grass? I think that this is a good question. IMHO surface doesn't qualify for more then actually surface, that's why I don't think of it as a good tag for soil in general, because soil is comprising a whole layer (can be up to some metres thick), which is not what surface is about IMHO. cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
On 16/11/2010, at 16.11, M∡rtin Koppenhoefer wrote: I set up a proposal for a new key landcover. Thank you, Martin. This is an excellent proposal, that formalises in OSM's tagging scheme the classical distinction between physical ( landcover) and human (landuse) geography. This will enable users to generate a reference map [0] as well as a thematic map [1] from OSM data. It will facilitate more rigorous scientific uses of OSM in the fields of biology, geography and geology, and it will enable scientists to contribute their data to OSM, as well as the use of OSM as a scientific tool. Cheers, Morten [0] http://www.physicalgeography.net/physgeoglos/r.html#reference_map [1] http://www.physicalgeography.net/physgeoglos/t.html#thematic_map ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
Am 16.11.2010 18:48, schrieb Richard Welty: On 11/16/10 12:43 PM, M∡rtin Koppenhoefer wrote: no, that would be surface as well. I'd say the distinction is between the surface and the coverage (which comprises the surface). surface=bush or tree would not make any sense IMHO. surface=asphalt is fine for the surface, the landcover would be the street which is not only the surface of the street. my attempt at clarification: surface is used where the mapped entity is man-made (or modified, e.g. dirt roads.) While I understand, what you mean, there is a weakness in that logic: A path in the wood made by humans is man-made - so you would tag it e.g. as surface=dirt; but if it's made by animals on their way to the water, it's landcover=dirt? On the other hand the Lüneburger Heide in Germany is man-made some 100 years ago by exploitation of the woods (kept as it is by extensive sheep pasturing). So here it's surface? Just for further thinking about. regards Peter ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
On 16/11/2010, at 20.45, Peter Wendorff wrote: Am 16.11.2010 18:48, schrieb Richard Welty: On 11/16/10 12:43 PM, M∡rtin Koppenhoefer wrote: no, that would be surface as well. I'd say the distinction is between the surface and the coverage (which comprises the surface). surface=bush or tree would not make any sense IMHO. surface=asphalt is fine for the surface, the landcover would be the street which is not only the surface of the street. my attempt at clarification: surface is used where the mapped entity is man-made (or modified, e.g. dirt roads.) While I understand, what you mean, there is a weakness in that logic: A path in the wood made by humans is man-made - so you would tag it e.g. as surface=dirt; but if it's made by animals on their way to the water, it's landcover=dirt? On the other hand the Lüneburger Heide in Germany is man-made some 100 years ago by exploitation of the woods (kept as it is by extensive sheep pasturing). So here it's surface? surface is a property of something man-made, i.e. it's in the human geography domain. It makes perfect sense for roads and traffic infrastructure. As a matter of lingual preference, I also prefer surface to describe something 2-dimensional, i.e. I do not like e.g. surface=building. Generally, surface is the property of something else. -- Morten ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] geology taggin?
Am 16.11.2010 13:51, schrieb M∡rtin Koppenhoefer: 2010/11/16 Ulf Lampingulf.lamp...@googlemail.com: So what is the *exact* problem with surface? it extents the usage of surface as attribute for routable entities to all kind of entities, therefore reducing simplicity for the data consumers with no benefit at all. No, surface was meant (and is in fact used widely) to describe the surface material of something, being it a highway, beach or whatever. There is e.g. *no* problem to describe the surface of e.g. natural=beach with that tag. A router wouldn't try to search for surface, but for highway or alike. It might want to analyze surface in addition to another tag. I doesn't make sense to me, if people use surface as a standalone tag, because it should always be an addition to other tags. But that's not a problem with that tag, but how people using it in clear breach of the definitions. Another advantage of specialized tag landcover is that in contrast with surface it by itself implies area=yes. So what is the *exact* advantage of landcover? well, one you cited yourself. Did I? Another one was written above: trees, which are not representable with surface. I've never argued to use surface for trees, but the well established natural=wood / landuse=forest. Sorry, this kind of vague we might want to have xy because someone might want to ... is pretty much pointless. this is not vague at all, and people are frequently popping up with the landcover proposal, as there seems to be a desire for it. Reading your new proposal page, I only see a vague definition that is in direct conflict with landuse and natural and therefore will confuse mappers how to tag things. It remains unclear under which circumstances someone should use landcover, landuse and/or natural. Regards, ULFL ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] geology taggin?
2010/11/16 Ulf Lamping ulf.lamp...@googlemail.com: No, surface was meant (and is in fact used widely) to describe the surface material of something, being it a highway, beach or whatever. There is e.g. *no* problem to describe the surface of e.g. natural=beach with that tag. do you see the difference between surface and landcover as described later in this thread? I've never argued to use surface for trees, but the well established natural=wood / landuse=forest. well established for what? landuse=forest is for managed forests, natural=wood (strange enough it is not for woodland) is for unmanaged natural forests. For all other trees there are currently no tags, besides mapping them one by one. Reading your new proposal page, I only see a vague definition that is in direct conflict with landuse and natural and therefore will confuse mappers how to tag things. How can physical landcover be in conflict with landuse? Did you read the proposal? Natural is IMHO an ideal example of a tag to diffuse clarity and create confusion, because it is a mix of all sorts of features. It remains unclear under which circumstances someone should use landcover, landuse and/or natural. I guess you didn't read the proposal. It states that you are encouraged to combine them, because they are orthogonal. cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
2010/11/16 Morten Kjeldgaard m...@bioxray.dk: surface is a property of something man-made, i.e. it's in the human geography domain. It makes perfect sense for roads and traffic infrastructure. As a matter of lingual preference, I also prefer surface to describe something 2-dimensional, i.e. I do not like e.g. surface=building. Generally, surface is the property of something else. I partly agree, but surface is not limited to man made. There is also the surface of a water body (lake, ...) for instance. Surface is about the surface. I agree with your statement in your other post: surface is not suited for soil, geology or anything else that is above or below the surface. Surface is 2-dimensional, a very thin layer. cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] geology taggin?
Ulf Lamping napsal(a): Am 16.11.2010 13:51, schrieb M∡rtin Koppenhoefer: 2010/11/16 Ulf Lampingulf.lamp...@googlemail.com: So what is the *exact* problem with surface? it extents the usage of surface as attribute for routable entities to all kind of entities, therefore reducing simplicity for the data consumers with no benefit at all. No, surface was meant (and is in fact used widely) to describe the surface material of something, being it a highway, beach or whatever. There is e.g. *no* problem to describe the surface of e.g. natural=beach with that tag. And that's the main problem with using surface tag instead of specialized landcover. Currently if I want to simply describe what's on the ground of certain area I should: 1) search for a suitable tag value in landuse (although I have no idea what the land is used for, I just want the tag to say here you can see [that] covering the ground), 2) search for a suitable tag value in natural, 3) use surface='that'+area='yes'. I really fail to see the benefit of keeping this mess (and inevitably extending it with growing number of surface values), both data consumers and osm editors would imho benefit from introduction of better tagging scheme for this (e.g. landcover). Introducing landcover tag and moving some values of other tags under it would be good for several reasons: 1) Resolving the mess in tagging scheme described above. 2) The meaning of landuse tag could be finally truely limited to describe the usage of the land. 3) The meaning of surface tag could get back to its original - describing a property of a certain object. 4) Landcover tagging scheme is more foolproof: - You can easily check for errors like intersection of areas with landcover tag. - User does not need to remember to add area=yes (in contrast with extended usage of surface tag). 5) Landcover and landuse/natural are orthogonal - e.g. part of your beach can be covered by pebblestones, another by sand and the sand can continue to cover the ground outside the beach - now what? a) You could input the beach as 2 entities instead of one, add one of them surface=sand, another surface=pebblestone and then add third area with surface=sand+area=yes. b) Or you draw one polygon with natural=beach, and then two areas with landcover tag. I really think (b) is easier for both data consumers and osm editors. Petr signature.asc Description: OpenPGP digital signature ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
To expand upon my earlier example, imagine a sandy beach with some exposed rock outcroppings, and sand dunes with beach grass growing on them. The entire area would be tagged with natural=beach. Exposed rocks would be tagged with surface=rock, and the rest with surface=sand. In addition, areas with beach grass would be tagged with landcover=beach_grass. The reason for tagging areas with beach grass or other vegetation is that you are generally forbidden from driving on or otherwise damaging the vegetation, as this would allow the sand dunes to shift position when the wind blows. ---Original Email--- Subject :Re: [Tagging] RFC: new key Landcover From :mailto:dieterdre...@gmail.com Date :Tue Nov 16 15:22:44 America/Chicago 2010 2010/11/16 Morten Kjeldgaard m...@bioxray.dk: surface is a property of something man-made, i.e. it's in the human geography domain. It makes perfect sense for roads and traffic infrastructure. As a matter of lingual preference, I also prefer surface to describe something 2-dimensional, i.e. I do not like e.g. surface=building. Generally, surface is the property of something else. I partly agree, but surface is not limited to man made. There is also the surface of a water body (lake, ...) for instance. Surface is about the surface. I agree with your statement in your other post: surface is not suited for soil, geology or anything else that is above or below the surface. Surface is 2-dimensional, a very thin layer. cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging Sent from my Verizon Wireless BlackBerry ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] RFC: new key Landcover
M∡rtin Koppenhoefer napsal(a): I set up a proposal for a new key landcover. http://wiki.openstreetmap.org/wiki/Proposed_features/landcover It deprecates very few old values (mud and sand from natural, grass from landuse). Thank you very much for writing this down, this is exactly what OSM needs... Personally, I would include the suggested landcover=water, it makes perfect sense. Petr signature.asc Description: OpenPGP digital signature ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging