Re: [OSM-talk] Woods vs Forests

[JB]

Hello,
Someone hinted it earlier, but may I repeat?
Is @tagging (list) not working correctly?
JB.

Le 02/11/2017 à 13:34, Martin Koppenhoefer a écrit :


sent from a phone


On 2. Nov 2017, at 11:34, Warin <61sundow...@gmail.com> wrote:

With landuse=residential there are no sub tags to indicate the kind of houses 
there or apartment blocks, colours, height etc.


we might do this though. If we were all urbanists and architects we most likely 
would do it.



If you want that kind of detail then map the physical houses with that detail.

You could do the same with trees ..


We lack the manpower to map like this, and even more to maintain it. I’ll 
assume you forgot to put a smiley there


cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk



___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

sent from a phone

> On 2. Nov 2017, at 11:34, Warin <61sundow...@gmail.com> wrote:
> 
> With landuse=residential there are no sub tags to indicate the kind of houses 
> there or apartment blocks, colours, height etc.


we might do this though. If we were all urbanists and architects we most likely 
would do it.


> 
> If you want that kind of detail then map the physical houses with that detail.
> 
> You could do the same with trees ..


We lack the manpower to map like this, and even more to maintain it. I’ll 
assume you forgot to put a smiley there 


cheers,
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

sent from a phone

> On 2. Nov 2017, at 11:34, Warin <61sundow...@gmail.com> wrote:
> 
> What next - the difference between a large group of houses and a smaller 
> group?


we already do this. We don’t tag a village as tiny town or a city as huge 
village. Different size can lead to different qualities


cheers,
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

sent from a phone

> On 2. Nov 2017, at 11:24, Warin <61sundow...@gmail.com> wrote:
> 
> How is a small area of trees different from a larger one?


it offers habitat for animals, creates its own microclimate and develops 
different soil. A forest is different from a group of trees. Just go into a 
forest and you’ll see, feel and smell the difference.

cheers,
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Lester Caine]

On 02/11/17 10:34, Warin wrote:
> On 02-Nov-17 09:21 PM, Lester Caine wrote:
>> On 02/11/17 09:40, Martin Koppenhoefer wrote:
>>> ONE tag to say what? You are still owing an answer to this.
>> I think the problem is similar to the multiple areas problem. There are
>> several layers of complexity so should landuse=residential enclose the
>> whole area including the grass and wooded areas or should they all be
>> isolated areas? Adding leisure=park within landuse=residential area just
>> makes things even more difficult? Just what is a small area of trees?
>> within the leisure=park or landuse=forest because it's not a
>> natural=wood 'creation' ... it's something the planning authorities have
>> requested as perhaps a barrier or simply as an amenity ... or has been
>> preserved as it has been there for hundreds of years ...
>>
>> We need to build a proper hierarchy of LANDUSE into which more detail
>> can be added if required?
>>
> If you want to tag the presence of trees then it is a land cover you
> want, not a land use.
Rather than natural= ? ... My point was that there should be an agreed
non-overlaping set of landuse=tags ... landuse=agricultural for areas
between landuse=residential or landuse=industrial where appropriate with
the farm builds, yards, orchards, coppices, and other detail secondary
tags to the landuse one ... rather than having to define every field
with it's own landuse tag.

Actually ... how difficult would it be to identify areas that don't have
a boundary around them? I have always though it would be useful if one
could fill in the gaps between things like landuse= areas.

> The difference between a large group of trees compared to a smaller group?
Size of the area is not relevant ...
But adding landcover=trees inside landuse=residential is perhaps a
better solution? But should a large park outside a residential area
still be tagged leisure=park? To fill in the landuse coverage it should
perhaps be landuse=park ... with wooded areas tagged within it.

> What next - the difference between a large group of houses and a smaller
> group?
We can accurately every building cleanly that is not a problem! It's
adding the other details of the development which is ... somewhat hit
and miss?

> landuse=forest does not mean there are trees there all the time, they
> could be logged and later replanted.
> 
> With landuse=residential there are no sub tags to indicate the kind of
> houses there or apartment blocks, colours, height etc.
> 
> If you want that kind of detail then map the physical houses with that
> detail.
> 
> You could do the same with trees .. map each one with its height,
> species and genus .. I'll leave that to others...
Actually the plans for the current local developments HAVE all of that
detail. Not that I expect them to actually follow the signed off plans
on the ground ;) But it does define the areas of the developments that
are not covered with buildings, highway elements, private gardens and
fencing. But again for consistency should the whole area be re-tagged
residential from farmland, or the preserved wooded areas inside the
development be tagged differently?

One of the developments south of here has individual trees that have to
be protected but they are less of a problem since they are individual
objects.

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Marc Gemis]

On Thu, Nov 2, 2017 at 11:24 AM, Warin <61sundow...@gmail.com> wrote:
> On 02-Nov-17 08:58 PM, Tomas Straupis wrote:
>>
>> 2017-11-02 11:24 GMT+02:00 Marc Gemis wrote:
>>>
>>> The current situation is not helping in producing useful maps. Too
>>> often I find myself in a residential area with large gardens and trees
>>> when I expected to find a real forest based on what OSM is displaying.
>
>
> What is a 'real forest'? Compared to this 'residential area with large
> gardens and trees'?
>
> Would not the area give away what you would find there?
>
> How is a small area of trees different from a larger one?

What do you expect from

* http://osm.org/go/0Erceo7JJ-
* http://osm.org/go/0EjtkaxH
* http://osm.org/go/0ErRW49us-

?

In the first 2 the main use of the land is residential. Why do I see
landuse=forest ? IMHO it should be landuse=residential ;
leisure=garden;access=private ; landcover=trees

m.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Lester Caine]

On 02/11/17 10:17, Warin wrote:
> A botanic Garden contains lots of different plants, including grass for
> the ones I have been to.
> Mapping each individual plant with its species and genus ... no thanks.
> I did map one tree though, just to be inconsistent. :)

But there is nothing stopping the staff of that Botanic Garden adding
all the footpaths, private areas, beds, features and so on if they are
working to produce their own map of the site? In much the same way that
universities and collages are mapping campuses in more and more detail.
Some areas have considerably more detail than others depending on who is
generating the data.

It's Tomas's interpretation of landuse=forest and natural=wood which is
a little at odds with others who would tag large 'unmanaged' forests as
natural=wood ... we need perhaps two levels of tagging for macro and
micro, rather than implying different interpretations on existing tags
depending on where they are used?

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 02-Nov-17 09:21 PM, Lester Caine wrote:

On 02/11/17 09:40, Martin Koppenhoefer wrote:

ONE tag to say what? You are still owing an answer to this.

I think the problem is similar to the multiple areas problem. There are
several layers of complexity so should landuse=residential enclose the
whole area including the grass and wooded areas or should they all be
isolated areas? Adding leisure=park within landuse=residential area just
makes things even more difficult? Just what is a small area of trees?
within the leisure=park or landuse=forest because it's not a
natural=wood 'creation' ... it's something the planning authorities have
requested as perhaps a barrier or simply as an amenity ... or has been
preserved as it has been there for hundreds of years ...

We need to build a proper hierarchy of LANDUSE into which more detail
can be added if required?

If you want to tag the presence of trees then it is a land cover you 
want, not a land use.


The difference between a large group of trees compared to a smaller group?

What next - the difference between a large group of houses and a smaller 
group?




landuse=forest does not mean there are trees there all the time, they 
could be logged and later replanted.


With landuse=residential there are no sub tags to indicate the kind of 
houses there or apartment blocks, colours, height etc.


If you want that kind of detail then map the physical houses with that 
detail.


You could do the same with trees .. map each one with its height, 
species and genus .. I'll leave that to others...





___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 02-Nov-17 08:58 PM, Tomas Straupis wrote:

2017-11-02 11:24 GMT+02:00 Marc Gemis wrote:

The current situation is not helping in producing useful maps. Too
often I find myself in a residential area with large gardens and trees
when I expected to find a real forest based on what OSM is displaying.


What is a 'real forest'? Compared to this 'residential area with large gardens 
and trees'?

Would not the area give away what you would find there?

How is a small area of trees different from a larger one?



   This is exactly why I started the topology rules topic. What we're
doing in Lithuania is we have to separate types: general forest, and
forest inside residential, commercial, industrial zones. The later one
is usually just a small number of trees in an area which is marked as
say residential zone in official maps. The later one can easily be
skipped in a map and result would not have "holes".

   So even if we're using two tags in Lithuania, I'm fine with choosing
one tag for all forests/woods/trees/whatever and then if someone
needs/wants - they could add subtags for details.

   Introducing even new tags seems impossible (and impractical) because
absolute majority of mappers just want to tag "forest/wood". And they
don't care about the details, so they will not tag it. And I do not
know maps which would somehow use such data, especially when such
detailed data would only be filled by a few, so it will not be filled
in a large enough regions to do any reasonable analysis.



In that case used 'natural=wood' and be done, no sub tags.

That has no implication of 'managed' and according to parts of the 
OSMwiki includes 'unnatural' too.





___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Lester Caine]

On 02/11/17 09:40, Martin Koppenhoefer wrote:
> ONE tag to say what? You are still owing an answer to this.

I think the problem is similar to the multiple areas problem. There are
several layers of complexity so should landuse=residential enclose the
whole area including the grass and wooded areas or should they all be
isolated areas? Adding leisure=park within landuse=residential area just
makes things even more difficult? Just what is a small area of trees?
within the leisure=park or landuse=forest because it's not a
natural=wood 'creation' ... it's something the planning authorities have
requested as perhaps a barrier or simply as an amenity ... or has been
preserved as it has been there for hundreds of years ...

We need to build a proper hierarchy of LANDUSE into which more detail
can be added if required?

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 02-Nov-17 08:49 PM, Lester Caine wrote:

On 02/11/17 09:13, Tomas Straupis wrote:

IMHO there are semantic implications in the key, as has been said many
times, <...>

   And that is subjective -> nobody is wrong -> everybody is right ->
everybody thinks THEIR proposal is the right one -> this topic is not
settled for so many years -> I suggest doing a compromise and agreeing
on ONE tag.
   (Compromise is currently done on rendering/data extraction side.
Nobody cares there about natural/landuse/landcover whatever. It's one
forest and that is it)

   The only other way is to use de facto situation - natural=wood and
landuse=forest - and forget this discussion.

In terms of topology, the idea from some that 'landuse' only applies to
land that is 'used' for something implies that large areas of the planet
are 'unused'?


Or unmapped for human use.

Given that the tag landuse may not have a good value for what the mapper thinks 
the area is used for,

or they think a different tag ... such as used for a National Park is all that 
is needed to describe the land use so they leave the tag landuse off.
So I think rather than 'unused' they are 'unmapped' and/or the definition of 
'use' has not included all the possible tags that could be interpreted to be 
human use.


A single layer of areas defining the current 'landcover'
should be something that is managed even if that includes 'wood-managed'
and 'wood-unmanaged'. The current historic situation has never been
right but then so have other long-standing compromises in tagging.


P.S. And all I wanted was to talk about topology rules... BTW: here is
an example of topology rules in Lithuania:
https://wiki.openstreetmap.org/wiki/WikiProject_Lithuania/Topology_rules

In terms of the UK, Land Use and Land Cover is well defined with a set
of clear classifications
https://www.gov.uk/government/statistics/national-land-use-database-land-use-and-land-cover-classification
except they are not really THAT clear. In your rules #2 and #5 seem to
be at odds? and in the UK classifications, just how do you define the
wooded areas of a park ... which may or may not be 'managed' ... and are
combined with 'grassland' and other natural or managed landcover. We can
define landuse=park, but that park can have a lot of detail contained
within it ... We are looking to render blocks of trees, grass, and other
objects how ever created?


A botanic Garden contains lots of different plants, including grass for the 
ones I have been to.
Mapping each individual plant with its species and genus ... no thanks.
I did map one tree though, just to be inconsistent. :)


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

2017-11-02 11:49 GMT+02:00 Lester Caine wrote:
>> P.S. And all I wanted was to talk about topology rules... BTW: here is
>> an example of topology rules in Lithuania:
>> https://wiki.openstreetmap.org/wiki/WikiProject_Lithuania/Topology_rules
>
> <...> In your rules #2 and #5 seem to be at odds?

  Landuse=forest is a real forest. These polygons are used to say
calculate total percentage of forests in a region and they and only
they are used for small scale maps. And natural=wood is NOT used in
forest percentage calculation, they are skipped from small scale maps
together with other micromapping stuff like sidewalks alongside the
reoads, treerows etc. Rule #5 was introduced for the specific purpose
to help filtering out "residential forests" from the map and
calculations.

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

2017-11-02 11:24 GMT+02:00 Marc Gemis wrote:
> The current situation is not helping in producing useful maps. Too
> often I find myself in a residential area with large gardens and trees
> when I expected to find a real forest based on what OSM is displaying.

  This is exactly why I started the topology rules topic. What we're
doing in Lithuania is we have to separate types: general forest, and
forest inside residential, commercial, industrial zones. The later one
is usually just a small number of trees in an area which is marked as
say residential zone in official maps. The later one can easily be
skipped in a map and result would not have "holes".

  So even if we're using two tags in Lithuania, I'm fine with choosing
one tag for all forests/woods/trees/whatever and then if someone
needs/wants - they could add subtags for details.

  Introducing even new tags seems impossible (and impractical) because
absolute majority of mappers just want to tag "forest/wood". And they
don't care about the details, so they will not tag it. And I do not
know maps which would somehow use such data, especially when such
detailed data would only be filled by a few, so it will not be filled
in a large enough regions to do any reasonable analysis.

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Lester Caine]

On 02/11/17 09:13, Tomas Straupis wrote:
>> IMHO there are semantic implications in the key, as has been said many
>> times, <...>
> 
>   And that is subjective -> nobody is wrong -> everybody is right ->
> everybody thinks THEIR proposal is the right one -> this topic is not
> settled for so many years -> I suggest doing a compromise and agreeing
> on ONE tag.
>   (Compromise is currently done on rendering/data extraction side.
> Nobody cares there about natural/landuse/landcover whatever. It's one
> forest and that is it)
> 
>   The only other way is to use de facto situation - natural=wood and
> landuse=forest - and forget this discussion.

In terms of topology, the idea from some that 'landuse' only applies to
land that is 'used' for something implies that large areas of the planet
are 'unused'? A single layer of areas defining the current 'landcover'
should be something that is managed even if that includes 'wood-managed'
and 'wood-unmanaged'. The current historic situation has never been
right but then so have other long-standing compromises in tagging.

> P.S. And all I wanted was to talk about topology rules... BTW: here is
> an example of topology rules in Lithuania:
> https://wiki.openstreetmap.org/wiki/WikiProject_Lithuania/Topology_rules

In terms of the UK, Land Use and Land Cover is well defined with a set
of clear classifications
https://www.gov.uk/government/statistics/national-land-use-database-land-use-and-land-cover-classification
except they are not really THAT clear. In your rules #2 and #5 seem to
be at odds? and in the UK classifications, just how do you define the
wooded areas of a park ... which may or may not be 'managed' ... and are
combined with 'grassland' and other natural or managed landcover. We can
define landuse=park, but that park can have a lot of detail contained
within it ... We are looking to render blocks of trees, grass, and other
objects how ever created?

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

2017-11-02 10:24 GMT+01:00 Marc Gemis :

>
> I like the landcover=trees idea, but this does not describe the
> complete picture. I also want t be able to indicate what is between
> the trees: ground, grass, grassy plants, bushes.
> And when is something landcover=trees + bushes and when
> landcover=bushes + some trees ?



+1, some ideas:

1. this could be mapped as "surface" with the disadvantage that "surface"
values are typically detailed, so it would impose the burden of doing
detailed mapping on the mapper

2. a new tag (e.g. "ground_cover") to say how area near the ground is made.
This tag would have more generalized values than what we find in "surface".

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

2017-11-02 10:13 GMT+01:00 Tomas Straupis :

> this topic is not
> settled for so many years -> I suggest doing a compromise and agreeing
> on ONE tag.



ONE tag to say what? You are still owing an answer to this.

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Marc Gemis]

The current situation is not helping in producing useful maps. Too
often I find myself in a residential area with large gardens and trees
when I expected to find a real forest based on what OSM is displaying.

So there is room for improvement.

I like the landcover=trees idea, but this does not describe the
complete picture. I also want t be able to indicate what is between
the trees: ground, grass, grassy plants, bushes.
And when is something landcover=trees + bushes and when
landcover=bushes + some trees ?

m.

On Thu, Nov 2, 2017 at 10:13 AM, Tomas Straupis  wrote:
>> IMHO there are semantic implications in the key, as has been said many
>> times, <...>
>
>   And that is subjective -> nobody is wrong -> everybody is right ->
> everybody thinks THEIR proposal is the right one -> this topic is not
> settled for so many years -> I suggest doing a compromise and agreeing
> on ONE tag.
>   (Compromise is currently done on rendering/data extraction side.
> Nobody cares there about natural/landuse/landcover whatever. It's one
> forest and that is it)
>
>   The only other way is to use de facto situation - natural=wood and
> landuse=forest - and forget this discussion.
>
> P.S. And all I wanted was to talk about topology rules... BTW: here is
> an example of topology rules in Lithuania:
> https://wiki.openstreetmap.org/wiki/WikiProject_Lithuania/Topology_rules
>
> --
> Tomas
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

> IMHO there are semantic implications in the key, as has been said many
> times, <...>

  And that is subjective -> nobody is wrong -> everybody is right ->
everybody thinks THEIR proposal is the right one -> this topic is not
settled for so many years -> I suggest doing a compromise and agreeing
on ONE tag.
  (Compromise is currently done on rendering/data extraction side.
Nobody cares there about natural/landuse/landcover whatever. It's one
forest and that is it)

  The only other way is to use de facto situation - natural=wood and
landuse=forest - and forget this discussion.

P.S. And all I wanted was to talk about topology rules... BTW: here is
an example of topology rules in Lithuania:
https://wiki.openstreetmap.org/wiki/WikiProject_Lithuania/Topology_rules

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

2017-11-02 8:14 GMT+01:00 Tomas Straupis :

> Currently according to taginfo the most popular are:
> natural=wood 4,5M
> landuse=forest 3,5M
> others are way behind. for example landcover=trees - 11000 objects...
>
> So maybe there is a point to choose one of the two popular tags and be
> done with it?
>
>

IMHO there are semantic implications in the key, as has been said many
times, "landuse" is about the human use of land. "natural=wood" reads as
"is a wood". Many trees don't grow in a wood but along streets, in small
groups that aren't woods, in gardens, etc. The main reason natural=wood
4,5M and landuse=forest 3,5M are used much more than any other tag is
existing rendering rules in OSM carto and presets. As this topic is popping
up frequently since at least 10 years, I guess there is some problem with
the established tagging, hence looking at the naked numbers doesn't bring
us further. If you look at the actual objects that have these tags, you'll
find that many are neither "wood"s nor "forest"s according to any
definition. They are simply groups of trees or areas covered with trees. A
forest and also a wood are more than just trees, they are also implying an
ecosystem, microclima, etc.


> If anybody wants more detail - subtags could be used: wood=xxx or
> forest=xxx depending on which one is chosen? Editors would remove
> other tags from presets, changes will be done in the database and then
> rendering and data extraction could be simplified.
>


what is the "thing" you want to tag with the tag to be chosen (what are the
basic characteristics, that are implied by the main tag)? Trees growing
there? A forest?

There's also still the problem with names: typically any bigger forest with
a name has smaller parts with their own names, which again have smaller
parts with their own names, etc.
It doesn't (IMHO) make sense to have nested same value landuses. The
solution could be either "natural" objects or place objects (or yet another
new key).
In any bigger (named) forest you will also typically find areas which are
by common interpretation part of the forest (e.g. clearings, lakes) but
aren't actually tree covered (again an argument against landcover).

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

Currently according to taginfo the most popular are:
natural=wood 4,5M
landuse=forest 3,5M
others are way behind. for example landcover=trees - 11000 objects...

So maybe there is a point to choose one of the two popular tags and be
done with it?

If anybody wants more detail - subtags could be used: wood=xxx or
forest=xxx depending on which one is chosen? Editors would remove
other tags from presets, changes will be done in the database and then
rendering and data extraction could be simplified.

Because none of the discussed options is wrong, it leads nowhere (and
it has led nowhere for the last 10 years).

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 02-Nov-17 05:42 PM, Daniel Koć wrote:

W dniu 02.11.2017 o 07:14, Warin pisze:

On 02-Nov-17 02:31 PM, Stephane Goldstein wrote:


Do you have any other words that smean tree planting, growing
and then harvesting?? And don't mean anything else?
The closest I have is 'forestry'.


landuse=forestry is a good option as well.


I'd like one word that includes harvesting of the trees as well.


Currently there's quite popular tag for such activity:

https://wiki.openstreetmap.org/wiki/Tag:landuse%3Dlogging


That only applies when logging is taking place.. and it is taken for 
'shrubs'.
It would be temporary  - not something I would tag. It would be like 
tagging a farm at the time of harvest of a crop .. say landuse=harvest 
and then having to change it back a week or so later to landuse=farm. 
Not something I woulod advise as it is too temporary.



I'd like one plain word that means;

An area where trees are grown for harvesting of products for human use. 
It would encompass planting, growing, thinning, clear felling, selective 
felling (and other things I have not thought of to do with harvesting of 
trees).



___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Daniel Koć]

W dniu 02.11.2017 o 07:14, Warin pisze:

On 02-Nov-17 02:31 PM, Stephane Goldstein wrote:


Do you have any other words that smean tree planting, growing and
then harvesting?? And don't mean anything else?
The closest I have is 'forestry'.


landuse=forestry is a good option as well.


I'd like one word that includes harvesting of the trees as well.


Currently there's quite popular tag for such activity:

https://wiki.openstreetmap.org/wiki/Tag:landuse%3Dlogging

--
"My method is uncertain/ It's a mess but it's working" [F. Apple]

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 02-Nov-17 02:31 PM, Stephane Goldstein wrote:


If OSM were to have landuse=plantation there would need to be a
further tag
plantation=cotton/tobacco/coffee/sugar_cane/trees/banana///house/bushes/*

Do you have any other words that smean tree planting, growing and
then harvesting?? And don't mean anything else?
The closest I have is 'forestry'.


landuse=forestry is a good option as well.


I'd like one word that includes harvesting of the trees as well.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Stephane Goldstein]

>
> If OSM were to have landuse=plantation there would need to be a further
> tag plantation=cotton/tobacco/coffee/sugar_cane/trees/banana*/*
> house/bushes/*
>
> Do you have any other words that smean tree planting, growing and then
> harvesting?? And don't mean anything else?
> The closest I have is 'forestry'.
>

landuse=forestry is a good option as well.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 01-Nov-17 03:10 PM, Stephane Goldstein wrote:



On Wed, Nov 1, 2017 at 10:17 AM, Warin <61sundow...@gmail.com 
> wrote:


On 01-Nov-17 09:24 AM, Stephane Goldstein wrote:


B)
depreciate landuse=forest and introduce a clearly defined
landuse=forestry that only includes tree areas that produce
base material for human use.

Managed forest for product harvesting  are commonly identified as
Plantations,


Possibly in your area of the world.
But not in Australia.


I live in Tasmania, and work in forestry.
http://www.agriculture.gov.au/abares/forestsaustralia/profiles/industrial-plantations


Ha. Sorry .. one of the 'common' people here ..  'forests' is the term I 
come across regularly.
"pine plantations" might be a goer .. but I would not use that term for 
hardwood forests amongst the 'common' people.
I would take a 'plantation' to be a uniform row on row planting, quite a 
few hardwood forests are not configured this way. Thus they don't fit 
with my perception of a 'plantation'.


In Alabama plantations were cotton farms .. but are now houses?
https://en.wikipedia.org/wiki/List_of_plantations_in_Alabama
Looks to be similar now in adjacent states.

Plantation is used for more than timber .. and that is a problem in that 
people will use it for all the other things ... just like they have used 
landuse=forest inappropriately.


If OSM were to have landuse=plantation there would need to be a further 
tag 
plantation=cotton/tobacco/coffee/sugar_cane/trees/banana///house/bushes/*


Do you have any other words that smean tree planting, growing and then 
harvesting?? And don't mean anything else?

The closest I have is 'forestry'.



My Dictionary says plantation means;
1) a farm or estate, especially in a tropical or sub tropical
country, on which cotton, tabacco, coffee, sugar or the like is
cultivated, usually by resident labours.
2) a group of planted trees or plants
[late ME from planting]



which I think would be most appropriate term for landuse tag in
that case.
https://en.wikipedia.org/wiki/Plantation

http://wiki.openstreetmap.org/wiki/Tag:landuse%3Dplantation



Those too include things other than trees, and even with trees
they don't exclude trees not intended for production of material
for human use.



___
talk mailing list
talk@openstreetmap.org 
https://lists.openstreetmap.org/listinfo/talk





___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk



___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Stephane Goldstein]

On Wed, Nov 1, 2017 at 10:17 AM, Warin <61sundow...@gmail.com> wrote:

> On 01-Nov-17 09:24 AM, Stephane Goldstein wrote:
>
> B)
>> depreciate landuse=forest and introduce a clearly defined
>> landuse=forestry that only includes tree areas that produce base material
>> for human use.
>>
>
> Managed forest for product harvesting  are commonly identified as
> Plantations,
>
>
> Possibly in your area of the world.
> But not in Australia.
>

I live in Tasmania, and work in forestry.
http://www.agriculture.gov.au/abares/forestsaustralia/profiles/industrial-plantations


>
> My Dictionary says plantation means;
> 1) a farm or estate, especially in a tropical or sub tropical country, on
> which cotton, tabacco, coffee, sugar or the like is cultivated, usually by
> resident labours.
> 2) a group of planted trees or plants
> [late ME from planting]
>
>
> which I think would be most appropriate term for landuse tag in that case.
> https://en.wikipedia.org/wiki/Plantation
> http://wiki.openstreetmap.org/wiki/Tag:landuse%3Dplantation
>
>
> Those too include things other than trees, and even with trees they don't
> exclude trees not intended for production of material for human use.
>
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 01-Nov-17 09:24 AM, Stephane Goldstein wrote:


B)
depreciate landuse=forest and introduce a clearly defined
landuse=forestry that only includes tree areas that produce base
material for human use.

Managed forest for product harvesting  are commonly identified as 
Plantations,


Possibly in your area of the world.
But not in Australia.

My Dictionary says plantation means;
1) a farm or estate, especially in a tropical or sub tropical country, 
on which cotton, tabacco, coffee, sugar or the like is cultivated, 
usually by resident labours.

2) a group of planted trees or plants
[late ME from planting]



which I think would be most appropriate term for landuse tag in that case.
https://en.wikipedia.org/wiki/Plantation
http://wiki.openstreetmap.org/wiki/Tag:landuse%3Dplantation


Those too include things other than trees, and even with trees they 
don't exclude trees not intended for production of material for human use.



___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Stephane Goldstein]

>
> B)
> depreciate landuse=forest and introduce a clearly defined landuse=forestry
> that only includes tree areas that produce base material for human use.
>

Managed forest for product harvesting  are commonly identified as
Plantations, which I think would be most appropriate term for landuse tag
in that case.
https://en.wikipedia.org/wiki/Plantation
http://wiki.openstreetmap.org/wiki/Tag:landuse%3Dplantation
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 01-Nov-17 02:26 AM, Christian Rogel wrote:


"Landuse=forest" may remain for large forests off the inhabited places.



No!

Only if the tree area is to be used for the production of material for 
human use.


IF not then it is not a 'land use' but a 'land cover' and should not be 
tagged 'landuse'.


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Christian Rogel]

> Le 31 oct. 2017 à 10:18, Warin <61sundow...@gmail.com> a écrit :
> A problem comes from the name "natural" that to most people (including 
> mappers) carries a meaning that excludes some meanings accepted in the 
> OSMwiki definition.
> For this reason it would be better to use a tag that carries no distorted 
> meaning, say landcover=trees for instance.

I can see any problem with "natural=wood" as there is a "natural=tree_row" not 
natural in the reality. Same of "leisure=sport" for a, say, tennis training 
camp (not very much leisure there ;-) ).
Let us the mappers know there are tags that could not be as transparents they 
wish.
We don't need a "landcover" for a very limited purpose : keep one type of 
landuse off overlapping others.
Creating a new tag could be confusing and time-costing.
"Landuse=forest" may remain for large forests off the inhabited places.

Christian Rogel___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Eugene Alvin Villar]

On Tue, Oct 31, 2017 at 5:18 PM, Warin <61sundow...@gmail.com> wrote:

> A problem comes from the name "natural" that to most people (including
> mappers) carries a meaning that excludes some meanings accepted in the
> OSMwiki definition.
> For this reason it would be better to use a tag that carries no distorted
> meaning, say landcover=trees for instance.
>

I can definitely agree with this. landcover=trees carries no weird meanings
unlike natural=wood (natural? man-made?). I can accept using landuse=forest
for land that is used for forestry-related uses (like production of timber
or a national forest), but for everything that is covered by a dense group
of trees, landcover=trees seems the way to go.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Yves]

This discussion looks like a chance to write proposals that could be voted by 
more than 15 people. 
Yves ___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 31-Oct-17 07:54 PM, Tobias Knerr wrote:

On 31.10.2017 07:59, Martin Koppenhoefer wrote:> one tag for what? An
area with trees? A forest? How would you define> "forest"?
One tag that can be used for mapping both the things currently mapped as
landuse=forest, and the things currently mapped as natural=wood.

Whether that is a tag for "forest" or for "tree-covered area" is a
worthwhile discussion. But going beyond that and asking for a precise
definition of "forest" seems like an almost impossible requirement, and
I'd rather have a less than perfect definition than no change.


this is really a non-issue, just evaluate these 2 tags the same way and you’re 
done.

That's an easy way out for data consumers, but not for mappers. When
mapping forests, you are currently forced to make a distinction that you
may not care about and that you may not even be qualified to make: You
can't just map a forest without also including a statement regarding its
naturalness or use for forestry purposes.


Presently the tag natural=wood within OSM means any tree area, it can be 
'natural', 'unnatural', 'managed', 'unmanaged',
'used for forestry', 'not used for forestry' or anything else you can dream of!

A problem comes from the name "natural" that to most people (including mappers) 
carries a meaning that excludes some meanings accepted in the OSMwiki definition.
For this reason it would be better to use a tag that carries no distorted 
meaning, say landcover=trees for instance.

Evaluating both tags the same way removes data that some have correctly tagged,
if they were rendered differently these differences will motivate mappers to 
tag with more care.


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tobias Knerr]

On 31.10.2017 07:59, Martin Koppenhoefer wrote:> one tag for what? An
area with trees? A forest? How would you define> "forest"?
One tag that can be used for mapping both the things currently mapped as
landuse=forest, and the things currently mapped as natural=wood.

Whether that is a tag for "forest" or for "tree-covered area" is a
worthwhile discussion. But going beyond that and asking for a precise
definition of "forest" seems like an almost impossible requirement, and
I'd rather have a less than perfect definition than no change.

> this is really a non-issue, just evaluate these 2 tags the same way and 
> you’re done.

That's an easy way out for data consumers, but not for mappers. When
mapping forests, you are currently forced to make a distinction that you
may not care about and that you may not even be qualified to make: You
can't just map a forest without also including a statement regarding its
naturalness or use for forestry purposes.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 31-Oct-17 05:57 PM, Martin Koppenhoefer wrote:


sent from a phone


On 31. Oct 2017, at 00:17, Warin <61sundow...@gmail.com> wrote:

Depreciate natural=wood and introduce landcover=trees.


landcover=trees is already introduced, I’m using it for years and others do the 
same. Just use it.



Sorry .. depreciate natural=wood and recommend on its OSMwiki page 
landcover=trees

I too use it. But I dual tag for the renders.


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

sent from a phone

> On 31. Oct 2017, at 00:36, Dave F  wrote:
> 
> then I'd be happy to go with that as the /real/ problem is OSM is currently 
> using two different tags (both key & value!) to represent the same entity,


this is really a non-issue, just evaluate these 2 tags the same way and you’re 
done. A real problem is the contrary: using the same tag for different things.

Ciao, Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

sent from a phone

> On 31. Oct 2017, at 00:17, Warin <61sundow...@gmail.com> wrote:
> 
> Depreciate natural=wood and introduce landcover=trees.


landcover=trees is already introduced, I’m using it for years and others do the 
same. Just use it.


Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

sent from a phone

> On 30. Oct 2017, at 18:25, Daniel Koć  wrote:
> 
> "Although forest is a term of common parlance, there is no universally 
> recognised precise definition, with more than 800 definitions of forest used 
> around the world."
> 
> [ https://en.wikipedia.org/wiki/Forest#Definition ]
> 
> I think it's hopeless for us to coin good definition.


we just have to find one definition we can agree upon, if we don’t, everyone 
will use his own personal interpretation and that’s probably worse.

I’m sure, using any of these asserted 800 definitions, there are a lot of areas 
where trees grow that aren’t forests.


Cheers,
Martin ___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

On 30/10/2017 17:25, Daniel Koć wrote:


"Although /forest/ is a term of common parlance, there is no 
universally recognised precise definition, with more than 800 
definitions of forest used around the world."


This is part of the OSM problem - It's been misappropriated as a verb to 
indicate that work is performed on a group of trees & also merged to 
include a few of those 800 variations (size, density etc). Using 
sub-tags some of those numerous definitions can, where appropriate, be 
added.


Separating them into individual sub-tags makes it *much* easier & 
accurate to filter if required rendering or data analysis.




[ https://en.wikipedia.org/wiki/Forest#Definition ]

I think it's hopeless for us to coin good definition.

FWIW, there's already a compatible tag to say: "area with trees", and 
that is landcover=trees.


+1 - this one is clear for me.


I'd think natural=wood is adequate (all the main renderers are 
displaying it), but if there's a belief another tag is better, then I'd 
be happy to go with that as the /real/ problem is OSM is currently using 
two different tags (both key & value!) to represent the same entity,


Are any renderers currently rendering 'landcover'?

DaveF



---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 30-Oct-17 10:23 PM, Dave F wrote:


On 29/10/2017 21:42, Warin wrote:


And then when the trees are harvested in a forestry operation the tag 
natural=wood could be removed with the result that the land use would 
be lost..


Irrelevant, it could also be removed if it were landuse=forest.


But it should not be removed ... the ground truth remains as the landuse 
remains the same.


Where are the ground truth of the presence of trees changes.


until such time as the tress grow again then the natural=wood could 
be reintroduced, but then the land use would have to be rediscovered 
and then retagged.


Again, could be the same for landuse=forest.
Again ... the ground truth remains as the landuse remains the same, so 
it should not be removed.


Where are the ground truth of the presence of trees changes




At the moment landuse is a separate main tag and is not subservient 
to another tag. That should remain.


Why?

landuse=residential .. to be made subservient to??
landuse=commercial .. to be made subservient to??
landuse=industrial .. to be made subservient to??
landuse=agriculture .. to be made subservient to??





I see that some might see a necessity of tagging tree areas with both 
landuse=forest and natural=wood.


Why?! They're the /same/ thing.


No they are NOT!

One marks the USE of the land.
The other marks the COVER of the land.

They are different things entirely.




However the one does not imply the other, to the extend that I only 
tag the landuse=forest and leave off the natural=wood.


To repeat, they're the same entity. 

See above.





Then there may be others who see natural=wood and think that their 
area of trees are not natural by their definition so falsely use 
landuse=foresty under the impression that any tree are that is 
'managed' is suitable for landuse=forest.


Solutions?

For the landuse=forest problem?

A) ?
Change the definition of landuse=forest to exclude the word 'managed',


Forest does *not* mean 'managed'. Never has, never will.
Yet OSM mappers take it to be so .. possibly because of the OSMwiki 
words on https://wiki.openstreetmap.org/wiki/Forest ! 'Managed' all over 
the place.



Some will object to the change of meaning of such a 'frequently used 
tag', no mater how confusing it may be.


It's been "frequently" misused. Most have used it without any 
understanding it's implied meaning.
Implied meanings can be taken in any direction by those doing the 
implication. The 'implication' needs to be stated to be clear as to the 
meaning taken.


As I indicated before Approach 2 is most appropriate.


And that is (having been trimmed out)
B)
depreciate landuse=forest and introduce a clearly defined 
landuse=forestry that only includes tree areas that produce base 
material for human use.




Leaving the natural=wood problems of not being used for 'managed' and/or 
'not natural'  what solution do you prefer?


A) ?
Depreciate natural=wood and introduce landcover=trees.

B) ? something else?



https://wiki.openstreetmap.org/wiki/Forest


That page just perpetuates the misuse of landuse=forest.
It does not advise the novice of how to tag a tree area but leaves the 
decision up to them .. that is not good educational practice!
It should at the very beginning state how to tag a tree area ... managed 
or not, natural or not, used to produce something or not. Just a simple 
statement.






___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Daniel Koć]

W dniu 30.10.2017 o 18:13, Martin Koppenhoefer pisze:
one tag for what? An area with trees? A forest? How would you define 
"forest"?


"Although /forest/ is a term of common parlance, there is no universally 
recognised precise definition, with more than 800 definitions of forest 
used around the world."


[ https://en.wikipedia.org/wiki/Forest#Definition ]

I think it's hopeless for us to coin good definition.

FWIW, there's already a compatible tag to say: "area with trees", and 
that is landcover=trees.


+1 - this one is clear for me.

--
"My method is uncertain/ It's a mess but it's working" [F. Apple]

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

2017-10-29 15:16 GMT+01:00 Tobias Knerr :

> On 27.10.2017 00:49, Dave F wrote:
> > The woods/forest problem is one of the worst tagging cock-ups in OSM.
>
> Indeed. The current mess is especially disappointing because it hasn't
> always been that way: The status quo is the result of an attempt to
> "improve" the tagging years ago.
>
> From my point of view, it's plainly obvious that there should be only
> one main tag instead of two.




one tag for what? An area with trees? A forest? How would you define
"forest"?
FWIW, there's already a compatible tag to say: "area with trees", and that
is landcover=trees.

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

On 29/10/2017 21:42, Warin wrote:


And then when the trees are harvested in a forestry operation the tag 
natural=wood could be removed with the result that the land use would 
be lost..


Irrelevant, it could also be removed if it were landuse=forest.

until such time as the tress grow again then the natural=wood could be 
reintroduced, but then the land use would have to be rediscovered and 
then retagged.


Again, could be the same for landuse=forest.



At the moment landuse is a separate main tag and is not subservient to 
another tag. That should remain.


Why?




I see that some might see a necessity of tagging tree areas with both 
landuse=forest and natural=wood.


Why?! They're the /same/ thing.



However the one does not imply the other, to the extend that I only 
tag the landuse=forest and leave off the natural=wood.


To repeat, they're the same entity.



Then there may be others who see natural=wood and think that their 
area of trees are not natural by their definition so falsely use 
landuse=foresty under the impression that any tree are that is 
'managed' is suitable for landuse=forest.


Solutions?

For the landuse=forest problem?

A) ?
Change the definition of landuse=forest to exclude the word 'managed',


Forest does *not* mean 'managed'. Never has, never will.


Some will object to the change of meaning of such a 'frequently used 
tag', no mater how confusing it may be.


It's been "frequently" misused. Most have used it without any 
understanding it's implied meaning.


As I indicated before Approach 2 is most appropriate.
https://wiki.openstreetmap.org/wiki/Forest

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

On 29/10/2017 14:16, Tobias Knerr wrote:

On 27.10.2017 00:49, Dave F wrote:

The woods/forest problem is one of the worst tagging cock-ups in OSM.

Indeed. The current mess is especially disappointing because it hasn't
always been that way: The status quo is the result of an attempt to
"improve" the tagging years ago.


True.
The really silly thing is that due to 'landuse=forest' being added as a 
selectable default in most editors, 99.9% of them were added with no 
knowledge of them being 'managed'.


It really is pointless data.

DaveF

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 30-Oct-17 01:16 AM, Tobias Knerr wrote:

On 27.10.2017 00:49, Dave F wrote:

The woods/forest problem is one of the worst tagging cock-ups in OSM.

Indeed. The current mess is especially disappointing because it hasn't
always been that way: The status quo is the result of an attempt to
"improve" the tagging years ago.

 From my point of view, it's plainly obvious that there should be only
one main tag instead of two.

Details on whether the area is used for forestry, whether it is in a
"natural" state (whatever that means), or other such information can be
gathered in addition to that main tag. Gathering that secondary
information should not be a requirement for being able to map the
forest/wood in the first place.


all areas of trees should
be primarily tagged as natural=wood. As with other entities, any further
details which gives clarity should be provided in sub-tags.

That would work nicely as far as I'm concerned.



And then when the trees are harvested in a forestry operation the tag 
natural=wood could be removed with the result that the land use would be lost..
until such time as the tress grow again then the natural=wood could be 
reintroduced, but then the land use would have to be rediscovered and then 
retagged.

At the moment landuse is a separate main tag and is not subservient to another 
tag. That should remain.

I see that some might see a necessity of tagging tree areas with both 
landuse=forest and natural=wood.

However the one does not imply the other, to the extend that I only tag the 
landuse=forest and leave off the natural=wood.

Then there may be others who see natural=wood and think that their area of 
trees are not natural by their definition so falsely use landuse=foresty under 
the impression that any tree are that is 'managed' is suitable for 
landuse=forest.

Solutions?

For the landuse=forest problem?

A) ?
Change the definition of landuse=forest to exclude the word 'managed',
 emphasise the use for the production of goods for human use
e.g. the base material for the production of paper, guitars, floor 
boards,furniture, house frames etc

Some will object to the change of meaning of such a 'frequently used tag', no 
mater how confusing it may be.

B) ?
Alternative - depreciate landuse=forest and introduce a clearly defined 
landuse=forestry that only includes tree areas that produce base material for 
human use.

C) ? others

For the natural=wood problem?

A) ?
Depreciate natural=wood and introduce landcover=trees.

B) ?



___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tobias Knerr]

On 27.10.2017 00:49, Dave F wrote:
> The woods/forest problem is one of the worst tagging cock-ups in OSM.

Indeed. The current mess is especially disappointing because it hasn't
always been that way: The status quo is the result of an attempt to
"improve" the tagging years ago.

From my point of view, it's plainly obvious that there should be only
one main tag instead of two.

Details on whether the area is used for forestry, whether it is in a
"natural" state (whatever that means), or other such information can be
gathered in addition to that main tag. Gathering that secondary
information should not be a requirement for being able to map the
forest/wood in the first place.

> all areas of trees should
> be primarily tagged as natural=wood. As with other entities, any further
> details which gives clarity should be provided in sub-tags.

That would work nicely as far as I'm concerned.

Tobias

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

On 28/10/2017 22:02, Warin wrote:


Maybe 'purposes' was a bit confusing, I see landuse=residential as a 
primary tag. Sub-tags are 'descriptive', 'adjective'. The 'cuisine' 
of a restaurant, or 'managed' for woods. for example.


I see landuse as the primary tag, the values used with it are 
descriptive.

There are secondary or sub tags such as 'name' that add details.


I agree 'landuse' /is/ primary' but the 'forest' value is being 
*misused* to describe various attributes of a group of trees: How 
they're managed, their size & density.






The 'landuse' tag when combined with forest is a misuse of a primary 
tag as it's being used as an adjective.


The landuse=forest will not always have trees on it. From time to time 
they may be harvested and result in no trees.


That has nothing to do with my point.





In this case, whether it's managed or not. Actually, it's use is even 
more confusing with people using it to describe the size of the area 
& density of trees, which, again, should be described with sub-tags.


The density of trees in a forestry area may change over time, Usually 
these areas are, when first planted, fairly dense in the number of 
plants then they are thinned as the trees grow to select the better 
trees to reach maturity.


Again, irrelevant to the clear stated problem.

DaveF


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 28-Oct-17 10:54 PM, Dave F wrote:


On 27/10/2017 20:53, Warin wrote:

On 27-Oct-17 08:25 PM, Dave F wrote:
You appear to be differentiating based on size & location which, 
seeing OSM's output is visual & geospatial seems unnecessary.


*All* groups of trees are 'natural' so there should only be one 
primary tag. All "purposes" should be within sub-tags.





Your definition of 'natural' must be different for mine. :)


Quelle surprise :)



A tree that is grown in a nursery from grafted stock, planted and 
nurtured in a green house and then finally planted outside ... to me 
is not 'natural'.
A 'natural' tree grown from a seed that comes off a tree by natural 
means, falls to the ground and than grows without human interference 
to full size.


I don't really agree with this, but for the purpose of my main 
argument: 'Unifying the key tag for groups of trees'; 'natural' is 
interchangeable with your preferred 'landcover'. What key used is 
arguable, but, whichever, there should only be *one*.




--
? "All "purposes" should be within sub-tags. "
Umm  so you would remove landuse? landuse=residential would be a 
subtag .. under what?


Maybe 'purposes' was a bit confusing, I see landuse=residential as a 
primary tag. Sub-tags are 'descriptive', 'adjective'. The 'cuisine' of 
a restaurant, or 'managed' for woods. for example.


I see landuse as the primary tag, the values used with it are descriptive.
There are secondary or sub tags such as 'name' that add details.



The 'landuse' tag when combined with forest is a misuse of a primary 
tag as it's being used as an adjective.


The landuse=forest will not always have trees on it. From time to time 
they may be harvested and result in no trees.




In this case, whether it's managed or not. Actually, it's use is even 
more confusing with people using it to describe the size of the area & 
density of trees, which, again, should be described with sub-tags.


The density of trees in a forestry area may change over time, Usually 
these areas are, when first planted, fairly dense in the number of 
plants then they are thinned as the trees grow to select the better 
trees to reach maturity.


I don't bother  adding the landcover to landuse=forest as it changes and 
I'm not prepared to track that and map it.


To me the landcover is secondary to the landuse tag in forestry areas as 
one is fairly permanent while the other changes over time.







___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

On 27/10/2017 20:53, Warin wrote:

On 27-Oct-17 08:25 PM, Dave F wrote:
You appear to be differentiating based on size & location which, 
seeing OSM's output is visual & geospatial seems unnecessary.


*All* groups of trees are 'natural' so there should only be one 
primary tag. All "purposes" should be within sub-tags.





Your definition of 'natural' must be different for mine. :)


Quelle surprise :)



A tree that is grown in a nursery from grafted stock, planted and 
nurtured in a green house and then finally planted outside ... to me 
is not 'natural'.
A 'natural' tree grown from a seed that comes off a tree by natural 
means, falls to the ground and than grows without human interference 
to full size.


I don't really agree with this, but for the purpose of my main argument: 
'Unifying the key tag for groups of trees'; 'natural' is interchangeable 
with your preferred 'landcover'. What key used is arguable, but, 
whichever, there should only be *one*.




--
? "All "purposes" should be within sub-tags. "
Umm  so you would remove landuse? landuse=residential would be a 
subtag .. under what?


Maybe 'purposes' was a bit confusing, I see landuse=residential as a 
primary tag. Sub-tags are 'descriptive', 'adjective'. The 'cuisine' of a 
restaurant, or 'managed' for woods. for example.


The 'landuse' tag when combined with forest is a misuse of a primary tag 
as it's being used as an adjective.


In this case, whether it's managed or not. Actually, it's use is even 
more confusing with people using it to describe the size of the area & 
density of trees, which, again, should be described with sub-tags.


DaveF

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

On 27/10/2017 13:00, Tomas Straupis wrote:

Fine. Let's say in higher level there is only one "forest". Then my
topic moves one layer down and stays exactly the same otherwise.
What I'm talking is about virtual hierarchy.
OSM tagging comes AFTER that.

As I map & tag what I see in reality; could you expand on what you mean by
"virtual hierarchy"?

   When you create a map (not a GIS database), you start with hierarchy
of objects which you're going to display. After that you specify what
exactly each of those items in the hierarchy is in your datasets.


You've put the cart before the carthorse.

Entities are mapped & tagged into a database. Individual renderers 
*then* decide which objects to include & how they appear.


I believe your points are irrelevant to unifying the key tag used to 
describe groups of trees.


DaveF

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

2017-10-27 22:44 GMT+03:00 Warin wrote:
> What you are talking about looks to be the rendering into layers and which
> layer comes higher than the other.
>
> That is the choice of the render and what could be higher in one rendering
> could be the lower in another rendering.

  While I agree with you on a theoretical level, I cannot come up with
any example of different order depending on "renderer" when we talk
about BASEMAPS (as opposed to thematic maps). Can you help me and give
an example?

P.S. Roads (or other objects) depicted wider/larger than they really
are for visibility overlapping building polygons is not enough,
because the problem here is that something is exaggerated on purpose.

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 27-Oct-17 08:21 PM, Daniel Koć wrote:



I have no idea what "landform" can be, so I don't have an opinion on 
that.


Some land forms;
peak
cliff
saddle
ridge
valley



However "natural" key for trees ( 
https://taginfo.openstreetmap.org/tags/natural=trees maybe?) sounds 
perfectly valid for me.



And not to me.


Does it still sound valid for a flower bead that is replanted with fresh 
fully grown flower plants each year?


Or would that be better tagged as landcover?


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 27-Oct-17 09:17 PM, James wrote:

landuse= man made and maintained
natrual= it made itself(which is 99.9% of the time the case)


Two different things.

'landuse' does not imply man made, but the use of the land.

'natural' implies made by nature, and this is hard for a mapper to be 
certain of.




___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 27-Oct-17 08:25 PM, Dave F wrote:
You appear to be differentiating based on size & location which, 
seeing OSM's output is visual & geospatial seems unnecessary.


*All* groups of trees are 'natural' so there should only be one 
primary tag. All "purposes" should be within sub-tags.





Your definition of 'natural' must be different for mine. :)

A tree that is grown in a nursery from grafted stock, planted and 
nurtured in a green house and then finally planted outside ... to me is 
not 'natural'.
A 'natural' tree grown from a seed that comes off a tree by natural 
means, falls to the ground and than grows without human interference to 
full size.


--
? "All "purposes" should be within sub-tags. "
Umm  so you would remove landuse? landuse=residential would be a subtag 
.. under what?


I think landuse is a good classification and should remain.
Areas used for forestry should be able to be tagged under a landuse tag. 
If the present landuse=forest is confusing then change it .. to, say 
landuse=forestry.




___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 27-Oct-17 09:49 PM, Tomas Straupis wrote:

2017-10-27 12:25 GMT+03:00 Dave F wrote:

You appear to be differentiating based on size & location which, seeing
OSM's output is visual & geospatial seems unnecessary.

   If we make no such distinction, then in order to be topographically
correct, we would have to "cut out" (create multipolygons) for each
small wood areas with 10 trees inside say residential area.


*All* groups of trees are 'natural' so there should only be one primary tag.
All "purposes" should be within sub-tags.

   Fine. Let's say in higher level there is only one "forest". Then my
topic moves one layer down and stays exactly the same otherwise.
   What I'm talking is about virtual hierarchy.
   OSM tagging comes AFTER that.



What you are talking about looks to be the rendering into layers and 
which layer comes higher than the other.


That is the choice of the render and what could be higher in one 
rendering could be the lower in another rendering.



Within the data base of OSM the distinctions need to be clear between 
these classifications so there is no cross over, no confusion.


Which classification is 'higher' than another has no effect on how it is 
stored in the OSM data base.


And tagging is about the storage of things in the OSM data base - trying 
to make it clear, organised and usable for both tagger and render.




___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 27-Oct-17 09:20 PM, Martin Koppenhoefer wrote:


I can't ignore the landcover argument in this context, and still 
believe the natural= key should mean: "a geographic feature", not 
"something natural" (as opposed to artificial). I would tag a peak 
with natural=peak regardless of human intervention, it's a peak.  In 
this sense, natural=wood means a "wood", and as not all areas of trees 
are woods, I'd question this statement.


A peak, yes. But where the entire hill is made from the tailings of an 
open cut mine (very large - both the hill and the mine) I don't think it 
can be said to be 'natural'. It is man made.
And this particular peak is prominent in the surrounding landscape, it 
is used as a tourist view point for the district.


As with other entities, any further details which gives clarity
should be provided in sub-tags.



as always, all tags should make sense, subtags are for further 
details, not to adjust/relativise the meaning of the main tag.

+1
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

>>Fine. Let's say in higher level there is only one "forest". Then my
>> topic moves one layer down and stays exactly the same otherwise.
>>What I'm talking is about virtual hierarchy.
>>OSM tagging comes AFTER that.
>
> As I map & tag what I see in reality; could you expand on what you mean by
> "virtual hierarchy"?

  When you create a map (not a GIS database), you start with hierarchy
of objects which you're going to display. After that you specify what
exactly each of those items in the hierarchy is in your datasets. So
in the beginning you define an abstract reason/purpose, only then you
specify technical details (f.e. tags).
  I call this a virtual hierarchy of information.

  I do understand that a number of different approaches currently
exist in OSM. You could look around, see some objects which seem
interesting/important to you and then tag them by your best knowledge
using natural language concepts. And that is a problem, because even
with English speakers those "natural language concepts" are subjective
for a number of different reasons. So when we combine a set of
objects, tagged by different people with different subjective
understanding it is very hard to make a logical system which is
essential for quality use. And it also introduces too much arguing
later as we can see with this ten year long forest topic which is
nowhere close to agreement... even not closer than it was ten years
ago.

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

On 27/10/2017 11:49, Tomas Straupis wrote:


   If we make no such distinction, then in order to be topographically
correct, we would have to "cut out" (create multipolygons) for each
small wood areas with 10 trees inside say residential area.


Well, depending if it's a communal area or privately owned (& whether I 
can be bothered), I have done that. People don't, generally, live in trees.


Referring back to your previous comment about not overlapping; you 
appear to assume that 'landuse' is being used as the key.


The confusion of having two primary 'keys' for the same object is my 
main point.




   Fine. Let's say in higher level there is only one "forest". Then my
topic moves one layer down and stays exactly the same otherwise.
   What I'm talking is about virtual hierarchy.
   OSM tagging comes AFTER that.


As I map & tag what I see in reality; could you expand on what you mean 
by "virtual hierarchy"?


DaveF



---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

2017-10-27 12:25 GMT+03:00 Dave F wrote:
> You appear to be differentiating based on size & location which, seeing
> OSM's output is visual & geospatial seems unnecessary.

  If we make no such distinction, then in order to be topographically
correct, we would have to "cut out" (create multipolygons) for each
small wood areas with 10 trees inside say residential area.

> *All* groups of trees are 'natural' so there should only be one primary tag.
> All "purposes" should be within sub-tags.

  Fine. Let's say in higher level there is only one "forest". Then my
topic moves one layer down and stays exactly the same otherwise.
  What I'm talking is about virtual hierarchy.
  OSM tagging comes AFTER that.

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Oleksiy Muzalyev]

On 27.10.17 12:20, Martin Koppenhoefer wrote:

[...]

isn't there a difference between using the wood that grows naturally 
(without being planted) and growing wood for using it?


[...]


Now woods are being planted also for the renaturation. Here is, as an 
example, an information board of the project of the renaturation on the 
river l'Hermance between the bridges Pont Neuf and Pont des Golettes 
https://commons.wikimedia.org/wiki/Category:Pont_Neuf_(Hermance)?uselang=fr#/media/File:Pont_Neuf_GE_10.jpg 
.


The river was canalized several decades ago, and now it is being 
returned to its original state, including woods along the river. It is 
practically impossible to tell if it is a natural or man-made forest, 
unless one knows it is a result of the renaturation.


Best regards,

O.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

> In; F1 there are the words "general landuse polygons"
>
> F2 there are the words "residential, commercial, industrial zones" that
> clearly imply land use.
>
> So your discussion is clearly about land use? Fine - that is ok.

  No. It is about virtual layers, calculated from OSM data for
cartographic, statistical or maybe some other use cases.

> I have an area that is used for recreation - picnics, walks, etc. It is a
> designated "National Park".
> So human land use is 'recreation'. There are native animals in there .. but
> the plan of management is primarily for 'recreation' and has been for many
> decades.

  My understanding is that all parks (national/regional and local
ones) are on a "higher" level "different GIS layer". That is on the
ground you still have a forest, water, meadow, rock, whatever. And
then on TOP of that you have an area of a "park". If you're interested
in parks - you render them on top of forest, meadow etc. objects. But
if you're not interested in parks you skip park objects and should
still get a good result with forests, meadows etc.
  Also if you're calculating how much forests there is in a region,
you want forests. It is not important if that forest is IN the park,
or not. You can simply ignore park objects.

  Or in other words, park is something I must KNOW. It is not
something I can see from say the airplane.

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Martin Koppenhoefer]

2017-10-27 0:49 GMT+02:00 Dave F :

> I think you'd be hard pressed to find any area of trees which hasn't been
> managed in one way or another by humans; especially in the Western world.
> Even in the depths of the Amazonian rainforest or Borneo the locals use
> wood for tools/fire/building etc.
>


isn't there a difference between using the wood that grows naturally
(without being planted) and growing wood for using it?




>
> Ignoring the landcover argument for a moment, all areas of trees should be
> primarily tagged as natural=wood.




I can't ignore the landcover argument in this context, and still believe
the natural= key should mean: "a geographic feature", not "something
natural" (as opposed to artificial). I would tag a peak with natural=peak
regardless of human intervention, it's a peak.  In this sense, natural=wood
means a "wood", and as not all areas of trees are woods, I'd question this
statement.



> As with other entities, any further details which gives clarity should be
> provided in sub-tags.
>


as always, all tags should make sense, subtags are for further details, not
to adjust/relativise the meaning of the main tag.

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[James]

landuse= man made and maintained
natrual= it made itself(which is 99.9% of the time the case)

On Oct 27, 2017 5:27 AM, "Dave F"  wrote:

> You appear to be differentiating based on size & location which, seeing
> OSM's output is visual & geospatial seems unnecessary.
>
> *All* groups of trees are 'natural' so there should only be one primary
> tag. All "purposes" should be within sub-tags.
>
> DaveF
>
> On 27/10/2017 08:52, Tomas Straupis wrote:
>
>> Some info on how/why forest/wood tagging is used in Lithuania. I will
>> not give specific tags (forest vs wood, landuse vs natural etc),
>> because in my opinion that is a secondary issue. Let's say we have
>> tags F1 and F2.
>>
>> F1 is for general forests. Those are the ones depicted on small scale
>> maps (full country/region).
>>
>> F2 is for small wooded areas INSIDE other polygons, usually inside
>> residential, commercial, industrial zones.
>>
>> This approach ignores utility as such (managed, non managed, natural,
>> left for full nature cycles as mention in Oleksiy's post). This
>> information could be added as a sub-tag if needed for some thematic
>> maps or specific statistical calculations.
>>
>> What I'm saying is that maybe we should:
>> 1. first decide the PURPOSES of having "tree cluster" polygons tagged
>> separately.
>> 2. Then PRIORITISE the purposes (based on ACTUAL usage ignoring all
>> "it could theoretically be used to/for...")
>> 3. and then decide which info goes to primary tag, which goes to
>> secondary tag(s).
>> 4. And only THEN decide on actual tags (keys, values).
>> Doing it the other way round will take us back to this forest
>> discussion as it has been here for the last ten years like discussing
>> what the words "forest", "wood", "natural", "landuse", "landcover"
>> etc. actually mean.
>>
>>
>
> ---
> This email has been checked for viruses by Avast antivirus software.
> https://www.avast.com/antivirus
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Dave F]

You appear to be differentiating based on size & location which, seeing 
OSM's output is visual & geospatial seems unnecessary.


*All* groups of trees are 'natural' so there should only be one primary 
tag. All "purposes" should be within sub-tags.


DaveF

On 27/10/2017 08:52, Tomas Straupis wrote:

Some info on how/why forest/wood tagging is used in Lithuania. I will
not give specific tags (forest vs wood, landuse vs natural etc),
because in my opinion that is a secondary issue. Let's say we have
tags F1 and F2.

F1 is for general forests. Those are the ones depicted on small scale
maps (full country/region).

F2 is for small wooded areas INSIDE other polygons, usually inside
residential, commercial, industrial zones.

This approach ignores utility as such (managed, non managed, natural,
left for full nature cycles as mention in Oleksiy's post). This
information could be added as a sub-tag if needed for some thematic
maps or specific statistical calculations.

What I'm saying is that maybe we should:
1. first decide the PURPOSES of having "tree cluster" polygons tagged
separately.
2. Then PRIORITISE the purposes (based on ACTUAL usage ignoring all
"it could theoretically be used to/for...")
3. and then decide which info goes to primary tag, which goes to
secondary tag(s).
4. And only THEN decide on actual tags (keys, values).
Doing it the other way round will take us back to this forest
discussion as it has been here for the last ten years like discussing
what the words "forest", "wood", "natural", "landuse", "landcover"
etc. actually mean.




---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Daniel Koć]

W dniu 27.10.2017 o 08:36, Warin pisze:


The landuse key is clearly to tag the use of the land by humans.


It's also to indicate use of the water, hence it's not 100% clear and I 
understand why some people don't like it.


The natural key is unclear - it seams to be for both things made by 
nature and things made by man! To me this confused all and the key 
should be discouraged.
It should be replaced by the keys landcover and landform, these have 
no implication of human or nature but simply describe the type of 
feature.


Let's look at natural=tree - it doesn't matter if the tree was seeded by 
man or by natural means, the tree is natural object, which was not 
created by man (even GMO is about _modyfying_, not creating). There can 
be however man_made=tree - we have a popular artwork in Warsaw, which is 
a palm made of plastic (tagging has changed, but it's a nice example):


https://en.wikipedia.org/wiki/Greetings_from_Jerusalem_Avenues
https://www.openstreetmap.org/node/1795659661

Landcover is neutral (what one can see on the surface). I like it 
because it's closest to the "ground truth", and is very useful when we 
don't know more details. However we could promote "surface" tag as a 
primary and it would also make sense for me (currently it's defined as 
additional tag: "used to provide additional information about the 
physical surface of roads/footpaths and some other features").


I have no idea what "landform" can be, so I don't have an opinion on that.

However "natural" key for trees ( 
https://taginfo.openstreetmap.org/tags/natural=trees maybe?) sounds 
perfectly valid for me.


--
"My method is uncertain/ It's a mess but it's working" [F. Apple]


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 27-Oct-17 06:52 PM, Tomas Straupis wrote:

Some info on how/why forest/wood tagging is used in Lithuania. I will
not give specific tags (forest vs wood, landuse vs natural etc),
because in my opinion that is a secondary issue. Let's say we have
tags F1 and F2.

F1 is for general forests. Those are the ones depicted on small scale
maps (full country/region).
Topology: They cannot overlap with other general landuse polygons:
water, reservoirs, riverbanks, meadows, scrub, sand, residential,
commercial, industrial zones etc.
Usage:
cartography: when generating small scale map we get a topologically
correct mosaic - non overlapping polygons - we do not have to worry
about overlapping polygons, draw order.
statistics: used to calculate percentage of forest coverage for a region

F2 is for small wooded areas INSIDE other polygons, usually inside
residential, commercial, industrial zones.
Topology: They MUST be above (fully inside) residential, commercial or
industrial polygon. If the F2 forest area is too large to be included
in say residential area - change it to F1.
Usage:
cartography: ignored for small scale maps. for large scale maps
(detailed small area) they are drawn on top of residential, commercial
and industrial areas.
statistics: ignored when calculating percentage of forest coverage.

This approach ignores utility as such (managed, non managed, natural,
left for full nature cycles as mention in Oleksiy's post). This
information could be added as a sub-tag if needed for some thematic
maps or specific statistical calculations.

What I'm saying is that maybe we should:
1. first decide the PURPOSES of having "tree cluster" polygons tagged
separately.
2. Then PRIORITISE the purposes (based on ACTUAL usage ignoring all
"it could theoretically be used to/for...")
3. and then decide which info goes to primary tag, which goes to
secondary tag(s).
4. And only THEN decide on actual tags (keys, values).
Doing it the other way round will take us back to this forest
discussion as it has been here for the last ten years like discussing
what the words "forest", "wood", "natural", "landuse", "landcover"
etc. actually mean.



In; F1 there are the words "general landuse polygons"

F2 there are the words "residential, commercial, industrial zones" that clearly 
imply land use.

So your discussion is clearly about land use? Fine - that is ok.

I have an area that is used for recreation - picnics, walks, etc. It is a designated 
"National Park".
So human land use is 'recreation'. There are native animals in there .. but the 
plan of management is primarily for 'recreation' and has been for many decades.

Some of the area is grassed, some trees ... but the human use of the land is 
not 'grass' nor 'trees', but 'picnic', 'walks' and 'peace and quite'.

At the moment there is no declared landuse tag on this area, it does have a 
admin boundary for the National Park, which could also in this case be used for 
the land use. It does have separately tagged land covers of paved, unpaved, 
trees, grass and water. But the single land use should be recreation.

As the 'use' is separate from 'cover' these things should be considered 
separately.




___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Tomas Straupis]

Some info on how/why forest/wood tagging is used in Lithuania. I will
not give specific tags (forest vs wood, landuse vs natural etc),
because in my opinion that is a secondary issue. Let's say we have
tags F1 and F2.

F1 is for general forests. Those are the ones depicted on small scale
maps (full country/region).
Topology: They cannot overlap with other general landuse polygons:
water, reservoirs, riverbanks, meadows, scrub, sand, residential,
commercial, industrial zones etc.
Usage:
cartography: when generating small scale map we get a topologically
correct mosaic - non overlapping polygons - we do not have to worry
about overlapping polygons, draw order.
statistics: used to calculate percentage of forest coverage for a region

F2 is for small wooded areas INSIDE other polygons, usually inside
residential, commercial, industrial zones.
Topology: They MUST be above (fully inside) residential, commercial or
industrial polygon. If the F2 forest area is too large to be included
in say residential area - change it to F1.
Usage:
cartography: ignored for small scale maps. for large scale maps
(detailed small area) they are drawn on top of residential, commercial
and industrial areas.
statistics: ignored when calculating percentage of forest coverage.

This approach ignores utility as such (managed, non managed, natural,
left for full nature cycles as mention in Oleksiy's post). This
information could be added as a sub-tag if needed for some thematic
maps or specific statistical calculations.

What I'm saying is that maybe we should:
1. first decide the PURPOSES of having "tree cluster" polygons tagged
separately.
2. Then PRIORITISE the purposes (based on ACTUAL usage ignoring all
"it could theoretically be used to/for...")
3. and then decide which info goes to primary tag, which goes to
secondary tag(s).
4. And only THEN decide on actual tags (keys, values).
Doing it the other way round will take us back to this forest
discussion as it has been here for the last ten years like discussing
what the words "forest", "wood", "natural", "landuse", "landcover"
etc. actually mean.

-- 
Tomas

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Warin]

On 27-Oct-17 04:51 PM, Oleksiy Muzalyev wrote:

On 27.10.17 00:49, Dave F wrote:

[...]
I think you'd be hard pressed to find any area of trees which hasn't 
been managed in one way or another by humans; especially in the 
Western world. [...]


There is a theory nowadays that woods should be left alone to natural 
cycles which may last hundreds of years. At least that a forest is not 
a park where everything should be cleaned up and tidy. Dead wood in a 
forest is the food for numerous insects. These insects are the basis 
of a biodiversity pyramid. Here is some information on it: 
http://blogs.helsinki.fi/deadwoodmeeting/files/2016/02/Barbalat_dead-wood-insects_web.pdf


Best regards,

O.


For thousands of years the Australian Aborigines have used fire to 
manage their lands.
There is a view that current fire dangers in Australia are a result of 
the lack of regular fire burning practices.
There is also the view that these burning practices encourage native 
vegetation.
And yet another view that these burning practices would discourage 
introduced weeds.
There are many who want regular patterned fire burns conducted for the 
above reasons.


Having said that, there are at several areas that have not been managed 
by humans by fire for many, if not thousands of, years - one where the 
Wollemi Pine was found and a few where cycads remain in central and 
northern Australia.


--
My view;

The landuse key is clearly to tag the use of the land by humans.

The natural key is unclear - it seams to be for both things made by 
nature and things made by man! To me this confused all and the key 
should be discouraged.
It should be replaced by the keys landcover and landform, these have no 
implication of human or nature but simply describe the type of feature.



___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Woods vs Forests

[Oleksiy Muzalyev]

On 27.10.17 00:49, Dave F wrote:

[...]
I think you'd be hard pressed to find any area of trees which hasn't 
been managed in one way or another by humans; especially in the 
Western world. [...]


There is a theory nowadays that woods should be left alone to natural 
cycles which may last hundreds of years. At least that a forest is not a 
park where everything should be cleaned up and tidy. Dead wood in a 
forest is the food for numerous insects. These insects are the basis of 
a biodiversity pyramid. Here is some information on it: 
http://blogs.helsinki.fi/deadwoodmeeting/files/2016/02/Barbalat_dead-wood-insects_web.pdf


Best regards,

O.


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Woods vs Forests

[Dave F]

(Split to a separate thread)

The woods/forest problem is one of the worst tagging cock-ups in OSM. 
It's bad enough when alternate values are used to differentiate what is 
actually the same object, but in this case it's also the key!


I think you'd be hard pressed to find any area of trees which hasn't 
been managed in one way or another by humans; especially in the Western 
world. Even in the depths of the Amazonian rainforest or Borneo the 
locals use wood for tools/fire/building etc.


Ignoring the landcover argument for a moment, all areas of trees should 
be primarily tagged as natural=wood. As with other entities, any further 
details which gives clarity should be provided in sub-tags.


Approach 2 is the appropriate example: 
https://wiki.openstreetmap.org/wiki/Forest,


The four render options on the website render wood & forest primary tags 
the same


DaveF

--
On 26/10/2017 13:37, Janko Mihelić wrote:,> A problem i find is with 
landuse=forest. Formally, those are zones that are used for growing 
trees. But practically in OSM, that tag is used for any land that is 
covered with trees. So formally, landuse=forest shouldn't overlap with 
other zones, but practically, until a new tag (landcover=trees) is 
rendered, this rule isn't going to be followed.


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Woods Vs Forests (Was Topology rules)

[Dave F]

Started a new thread as it's gone of subject.

DaveF.


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk