subject:"Re\: \[Tagging\] Forest parcel with other landcover \(scrub, scree…\)\: how to map\?"

On Thu, 24 Jan 2019 at 16:52, Marc Gemis  wrote:

> I wonder how puzzled someone would be when they look at the original
> question "how to map scrub in forests" and then, without going through
> the whole thread, sees e.g the following mail in that thread:
>

Yes, way back up there ^ somewhere, someone mentioned size of things &
discussions sort of veered off that way.

Apologies, as I was one of the offenders :-(

Thanks

Graeme
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Marc Gemis

I wonder how puzzled someone would be when they look at the original
question "how to map scrub in forests" and then, without going through
the whole thread, sees e.g the following mail in that thread:

p.s. Joseph, please don't take this personally, I could have picked
another mail, but yours was the last.

On Thu, Jan 24, 2019 at 6:09 AM Joseph Eisenberg
 wrote:
>
> Chillagoe sounds like a village. It’s “only” 205 km (127 mi) W of the city of 
> Cairns
>
> Does it have a secondary school, eg for 14 to 18 year olds?
>
> The only town-level service is the “hospital”, but does it really offer full 
> services?
>
> > There are another half a dozen other small/er settlements within ~150k of 
> > Chillagoe for which it is "town".
>
> A town should usually serve more than a half dozen hamlets. That’s a 
> “village.”
>
> > “literally hundreds of similar towns.”
>
> That’s probably too many towns for a country with as low a population at 
> Australia.
>
> > “SW US”
>
> Americans in the rural West are used to driving for 2 to 3 hours to the 
> nearest town, 200 to 300km away. We used to drive 150km to town once or twice 
> a month. So most small settlements are losing their shops and businesses.
> On Thu, Jan 24, 2019 at 11:49 AM Graeme Fitzpatrick  
> wrote:
>>
>>
>>
>> On Thu, 24 Jan 2019 at 12:09, Warin <61sundow...@gmail.com> wrote:
>>>
>>> On 24/01/19 12:50, Joseph Eisenberg wrote:
>>>
>>> It’s not possible to have “town” level services with less than 1000 people. 
>>> A town has a major market (retail area) serving the surrounding area, as 
>>> well as basic educational, cultural and government facilities.
>>
>> That is the usual interpretation (& certainly matches the Western Europe / 
>> US definition) but here is an example of what I'm talking about: 
>> https://www.openstreetmap.org/#map=15/-17.1494/144.5257
>>
>> Chillagoe has a population, according to the 2016 census, of 251 people, so 
>> it "should" only be a =hamlet. However, it has a school, police station, 
>> general store, hardware store, post office, 2 pub's, 2 museums, hospital, 
>> bitumen airfield &, as it's a tourist destination in it's own right, about 6 
>> caravan parks / camping grounds!
>>
>> Dimbulah is ~100k to the east & is a bit bigger, but doesn't have a 
>> hospital, & only an unsealed airstrip.
>>
>> The next major town west is Normanton, which is ~700k away!
>>
>> There are another half a dozen other small/er settlements within ~150k of 
>> Chillagoe for which it is "town".
>>>
>>> In your area of the world I agree. You have a 'good' population density.
>>> In some places the nearest neighbour can be 400 km away. The population 
>>> 'centre' may have much less than 1,000 people in the local residential area 
>>> .. but may service 1,000s of square kilometres.
>>> Necessity makes this population centre very important for the few people 
>>> living in that area.
>>
>>
>> Yep! Have just started a discussion on the Australia list concerning this 
>> very point, after taking to somebody who lives in this area, & who made the 
>> comment that OSM is a bit useless, because when you open it, all you see is 
>> a massive blankness :-(
>>
>> & this, by no means, is an isolated example - Australia would have literally 
>> hundreds of similar towns. I would think the same would probably apply in 
>> places like Canada, Alaska, SW US, South Africa & a number of other 
>> civilised but empty countries?
>>
>> Thanks
>>
>> Graeme
>> ___
>> Tagging mailing list
>> Tagging@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/tagging
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

Chillagoe sounds like a village. It’s “only” 205 km (127 mi) W of the city
of Cairns

Does it have a secondary school, eg for 14 to 18 year olds?

The only town-level service is the “hospital”, but does it really offer
full services?

> There are another half a dozen other small/er settlements within ~150k of
Chillagoe for which it is "town".

A town should usually serve more than a half dozen hamlets. That’s a
“village.”

> “literally hundreds of similar towns.”

That’s probably too many towns for a country with as low a population at
Australia.

> “SW US”

Americans in the rural West are used to driving for 2 to 3 hours to the
nearest town, 200 to 300km away. We used to drive 150km to town once or
twice a month. So most small settlements are losing their shops and
businesses.
On Thu, Jan 24, 2019 at 11:49 AM Graeme Fitzpatrick 
wrote:

>
>
> On Thu, 24 Jan 2019 at 12:09, Warin <61sundow...@gmail.com> wrote:
>
>> On 24/01/19 12:50, Joseph Eisenberg wrote:
>>
>> It’s not possible to have “town” level services with less than 1000
>> people. A town has a major market (retail area) serving the surrounding
>> area, as well as basic educational, cultural and government facilities.
>>
>> That is the usual interpretation (& certainly matches the Western Europe
> / US definition) but here is an example of what I'm talking about:
> https://www.openstreetmap.org/#map=15/-17.1494/144.5257
>
> Chillagoe has a population, according to the 2016 census, of 251 people,
> so it "should" only be a =hamlet. However, it has a school, police station,
> general store, hardware store, post office, 2 pub's, 2 museums, hospital,
> bitumen airfield &, as it's a tourist destination in it's own right, about
> 6 caravan parks / camping grounds!
>
> Dimbulah is ~100k to the east & is a bit bigger, but doesn't have a
> hospital, & only an unsealed airstrip.
>
> The next major town west is Normanton, which is ~700k away!
>
> There are another half a dozen other small/er settlements within ~150k of
> Chillagoe for which it is "town".
>
>> In your area of the world I agree. You have a 'good' population density.
>> In some places the nearest neighbour can be 400 km away. The population
>> 'centre' may have much less than 1,000 people in the local residential area
>> .. but may service 1,000s of square kilometres.
>> Necessity makes this population centre very important for the few people
>> living in that area.
>>
>
> Yep! Have just started a discussion on the Australia list concerning this
> very point, after taking to somebody who lives in this area, & who made the
> comment that OSM is a bit useless, because when you open it, all you see is
> a massive blankness :-(
>
> & this, by no means, is an isolated example - Australia would have
> literally hundreds of similar towns. I would think the same would probably
> apply in places like Canada, Alaska, SW US, South Africa & a number of
> other civilised but empty countries?
>
> Thanks
>
> Graeme
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

On Thu, 24 Jan 2019 at 12:09, Warin <61sundow...@gmail.com> wrote:

> On 24/01/19 12:50, Joseph Eisenberg wrote:
>
> It’s not possible to have “town” level services with less than 1000
> people. A town has a major market (retail area) serving the surrounding
> area, as well as basic educational, cultural and government facilities.
>
> That is the usual interpretation (& certainly matches the Western Europe /
US definition) but here is an example of what I'm talking about:
https://www.openstreetmap.org/#map=15/-17.1494/144.5257

Chillagoe has a population, according to the 2016 census, of 251 people, so
it "should" only be a =hamlet. However, it has a school, police station,
general store, hardware store, post office, 2 pub's, 2 museums, hospital,
bitumen airfield &, as it's a tourist destination in it's own right, about
6 caravan parks / camping grounds!

Dimbulah is ~100k to the east & is a bit bigger, but doesn't have a
hospital, & only an unsealed airstrip.

The next major town west is Normanton, which is ~700k away!

There are another half a dozen other small/er settlements within ~150k of
Chillagoe for which it is "town".

> In your area of the world I agree. You have a 'good' population density.
> In some places the nearest neighbour can be 400 km away. The population
> 'centre' may have much less than 1,000 people in the local residential area
> .. but may service 1,000s of square kilometres.
> Necessity makes this population centre very important for the few people
> living in that area.
>

Yep! Have just started a discussion on the Australia list concerning this
very point, after taking to somebody who lives in this area, & who made the
comment that OSM is a bit useless, because when you open it, all you see is
a massive blankness :-(

& this, by no means, is an isolated example - Australia would have
literally hundreds of similar towns. I would think the same would probably
apply in places like Canada, Alaska, SW US, South Africa & a number of
other civilised but empty countries?

Thanks

Graeme
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

I’m in Papua, Indonesia. It’s quite different than Java or Bali.

I’ve tagged some places as “towns” with only 4000 or 5000 inhabitants in
the most sparsely populated parts around here, because they are the only
settlement larger than a few hundred people in an area of 100 x 100 km or
greater, and they have big markets, a variety of shops, a high school
(secondary education) and a few offices.

But even in Alaska or Siberia, I can’t see how a place with 500 inhabitants
could qualify as an OSM town. It might be the capital of an admin_level=4
and that could be a good measure of importance.

I’ve also been reluctant to tag my own town as a “city”. It’s the largest
settlement for 150 to 200km in all directions, by a factor of 4, but only
has 40,000 people, and many services require flying to the capital city of
the province. There are several institutes of higher education, but not a
university, and the airport only offers flights to one other city. But it’s
a borderline case. Probably in 10 years it will clearly qualify
On Thu, Jan 24, 2019 at 11:09 AM Warin <61sundow...@gmail.com> wrote:

> On 24/01/19 12:50, Joseph Eisenberg wrote:
>
> “an isolated "village" with only a few hundred people in it, but which is
> the main centre for this area will be a town, & maybe even a city?”
>
> It’s not possible to have “town” level services with less than 1000
> people. A town has a major market (retail area) serving the surrounding
> area, as well as basic educational, cultural and government facilities.
>
>
> In your area of the world I agree. You have a 'good' population density.
> In some places the nearest neighbour can be 400 km away. The population
> 'centre' may have much less than 1,000 people in the local residential area
> .. but may service 1,000s of square kilometres.
> Necessity makes this population centre very important for the few people
> living in that area.
>
>
> Similarly, a city must have more than a few thousand people, if it has a
> hospital, university, transportation hub (port / airport etc), government
> and business offices, and so on.
> On Thu, Jan 24, 2019 at 10:17 AM Graeme Fitzpatrick 
> wrote:
>
>>
>> On Thu, 24 Jan 2019 at 10:19, Martin Koppenhoefer 
>> wrote:
>>
>>>
>>> > On 23. Jan 2019, at 22:35, Paul Allen  wrote:
>>> >
>>> > According to the wiki, they're not
>>> > based on services either.
>>>
>>>
>>> if the wiki says it explicitly like this we should fix it.
>>>
>>
>> Just looking & it's a mixture, & also a bit of a mess
>>
>> =city: the largest settlement or settlements within a territory
>>
>> =town: an important urban centre ... normally have a good range of shops
>> and facilities which are used by people from nearby villages.
>>
>> =village: A settlement with between 1,000 and 10,000 inhabitants
>>
>> =hamlet: an isolated settlement, typically with less than 100-200
>> inhabitants
>>
>> =isolated_dwelling: the smallest kind of human settlement. They are
>> outside other settlements ... must not consist of more than 2 households
>>
>> =locality:  an unpopulated location (but the example photo is of an
>> inhabited village!)
>>
>> I would suggest we remove all reference to number of inhabitants, & base
>> the decision on each mappers own recognition of how "important" this place
>> is, so an isolated "village" with only a few hundred people in it, but
>> which is the main centre for this area will be a town, & maybe even a city?
>>
>> I'm happy to have a go at them if we're willing?
>>
>> Thanks
>>
>> Graeme
>> ___
>> Tagging mailing list
>> Tagging@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/tagging
>>
>
>
> ___
> Tagging mailing 
> listTagging@openstreetmap.orghttps://lists.openstreetmap.org/listinfo/tagging
>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

On 24/01/19 12:50, Joseph Eisenberg wrote:
“an isolated "village" with only a few hundred people in it, but which 
is the main centre for this area will be a town, & maybe even a city?”

It’s not possible to have “town” level services with less than 1000 
people. A town has a major market (retail area) serving the 
surrounding area, as well as basic educational, cultural and 
government facilities.

In your area of the world I agree. You have a 'good' population density.
In some places the nearest neighbour can be 400 km away. The population 
'centre' may have much less than 1,000 people in the local residential 
area .. but may service 1,000s of square kilometres.
Necessity makes this population centre very important for the few people 
living in that area.

Similarly, a city must have more than a few thousand people, if it has 
a hospital, university, transportation hub (port / airport etc), 
government and business offices, and so on.
On Thu, Jan 24, 2019 at 10:17 AM Graeme Fitzpatrick 
mailto:graemefi...@gmail.com>> wrote:

On Thu, 24 Jan 2019 at 10:19, Martin Koppenhoefer
mailto:dieterdre...@gmail.com>> wrote:

> On 23. Jan 2019, at 22:35, Paul Allen mailto:pla16...@gmail.com>> wrote:
>
> According to the wiki, they're not
> based on services either.

if the wiki says it explicitly like this we should fix it.

Just looking & it's a mixture, & also a bit of a mess

=city: the largest settlement or settlements within a territory

=town: an important urban centre ... normally have a good range of
shops and facilities which are used by people from nearby villages.

=village: A settlement with between 1,000 and 10,000 inhabitants

=hamlet: an isolated settlement, typically with less than 100-200
inhabitants

=isolated_dwelling: the smallest kind of human settlement. They
are outside other settlements ... must not consist of more than 2
households

=locality: an unpopulated location (but the example photo is of an
inhabited village!)

I would suggest we remove all reference to number of inhabitants,
& base the decision on each mappers own recognition of how
"important" this place is, so an isolated "village" with only a
few hundred people in it, but which is the main centre for this
area will be a town, & maybe even a city?

I'm happy to have a go at them if we're willing?

Thanks

Graeme
___
Tagging mailing list
Tagging@openstreetmap.org 
https://lists.openstreetmap.org/listinfo/tagging

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

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

On 24/01/19 12:16, Graeme Fitzpatrick wrote:

On Thu, 24 Jan 2019 at 10:19, Martin Koppenhoefer 
mailto:dieterdre...@gmail.com>> wrote:

> On 23. Jan 2019, at 22:35, Paul Allen mailto:pla16...@gmail.com>> wrote:
>
> According to the wiki, they're not
> based on services either.

if the wiki says it explicitly like this we should fix it.

Just looking & it's a mixture, & also a bit of a mess

=city: the largest settlement or settlements within a territory

=town: an important urban centre ... normally have a good range of 
shops and facilities which are used by people from nearby villages.

=village: A settlement with between 1,000 and 10,000 inhabitants

=hamlet: an isolated settlement, typically with less than 100-200 
inhabitants

=isolated_dwelling: the smallest kind of human settlement. They are 
outside other settlements ... must not consist of more than 2 households

=locality: an unpopulated location (but the example photo is of an 
inhabited village!)

I would suggest we remove all reference to number of inhabitants, & 
base the decision on each mappers own recognition of how "important" 
this place is, so an isolated "village" with only a few hundred people 
in it, but which is the main centre for this area will be a town, & 
maybe even a city?

No, the population size is a handy guide and should be kept - as a 
guide. May add the word 'typically' where it is missing?

I'm happy to have a go at them if we're willing?

A problem is that the British have a system .. based on old definitions 
- using things like cathedrals to help define their definitions.

On 'one set of rules' is going to match world wide. The number of pubs 
may be a useful criteria in Australia, but not in Morocco.

The document needs to be a guide .. not a 'rule'.
One guide should be that surrounding places must be relative in level of 
important to the place that is being mapped. I don't think a place with 
no post office can be more important than a place with a post office in 
the same area.
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

“an isolated "village" with only a few hundred people in it, but which is
the main centre for this area will be a town, & maybe even a city?”

It’s not possible to have “town” level services with less than 1000 people.
A town has a major market (retail area) serving the surrounding area, as
well as basic educational, cultural and government facilities.

Similarly, a city must have more than a few thousand people, if it has a
hospital, university, transportation hub (port / airport etc), government
and business offices, and so on.
On Thu, Jan 24, 2019 at 10:17 AM Graeme Fitzpatrick 
wrote:

>
> On Thu, 24 Jan 2019 at 10:19, Martin Koppenhoefer 
> wrote:
>
>>
>> > On 23. Jan 2019, at 22:35, Paul Allen  wrote:
>> >
>> > According to the wiki, they're not
>> > based on services either.
>>
>>
>> if the wiki says it explicitly like this we should fix it.
>>
>
> Just looking & it's a mixture, & also a bit of a mess
>
> =city: the largest settlement or settlements within a territory
>
> =town: an important urban centre ... normally have a good range of shops
> and facilities which are used by people from nearby villages.
>
> =village: A settlement with between 1,000 and 10,000 inhabitants
>
> =hamlet: an isolated settlement, typically with less than 100-200
> inhabitants
>
> =isolated_dwelling: the smallest kind of human settlement. They are
> outside other settlements ... must not consist of more than 2 households
>
> =locality:  an unpopulated location (but the example photo is of an
> inhabited village!)
>
> I would suggest we remove all reference to number of inhabitants, & base
> the decision on each mappers own recognition of how "important" this place
> is, so an isolated "village" with only a few hundred people in it, but
> which is the main centre for this area will be a town, & maybe even a city?
>
> I'm happy to have a go at them if we're willing?
>
> Thanks
>
> Graeme
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

On Thu, 24 Jan 2019 at 10:19, Martin Koppenhoefer 
wrote:

>
> > On 23. Jan 2019, at 22:35, Paul Allen  wrote:
> >
> > According to the wiki, they're not
> > based on services either.
>
>
> if the wiki says it explicitly like this we should fix it.
>

Just looking & it's a mixture, & also a bit of a mess

=city: the largest settlement or settlements within a territory

=town: an important urban centre ... normally have a good range of shops
and facilities which are used by people from nearby villages.

=village: A settlement with between 1,000 and 10,000 inhabitants

=hamlet: an isolated settlement, typically with less than 100-200
inhabitants

=isolated_dwelling: the smallest kind of human settlement. They are outside
other settlements ... must not consist of more than 2 households

=locality:  an unpopulated location (but the example photo is of an
inhabited village!)

I would suggest we remove all reference to number of inhabitants, & base
the decision on each mappers own recognition of how "important" this place
is, so an isolated "village" with only a few hundred people in it, but
which is the main centre for this area will be a town, & maybe even a city?

I'm happy to have a go at them if we're willing?

Thanks

Graeme
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

> waterway=stream / river
> railway=rail/narrow_gauge
> highway=footway/pedestrian

These are narrow linear features. It’s unreasonable and less helpful to map
these as polygons when a simple linear way will do, so it’s good to have a
way to show the difference.

And there is also a big functional difference between =footway and
=pedestrian; the latter is wide enough for vehicles and usually allows
deliveries at some hours.

rail/narrow_gauge are not always functionally different, so that’s less
useful, but I suppose rail fanatics like to know these things

> shop=supermarket / convenience

Supermarkets here in Papua are often the same size as a big convenience
store in the USA. It’s the variety of products offered for sale that
determines the difference. Though it’s probanly true that a supermarket and
convenience store in the same town will always be different in size,
relatively.
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Martin Koppenhoefer



sent from a phone

> On 23. Jan 2019, at 22:35, Paul Allen  wrote:
> 
> According to the wiki, they're not
> based on services either. 


if the wiki says it explicitly like this we should fix it.


Cheers, Martin 

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Martin Koppenhoefer



sent from a phone

On 23. Jan 2019, at 21:44, Paul Allen  wrote:

>> place=island/islet what was a bad ideas and should not be repeated?
>> 
>> place=hamlet, place=village, place=town, place=city.
> 
> place=hamlet, place=village, place=town, place=city.


waterway=stream / river

railway=rail/narrow_gauge

highway=footway/pedestrian 

amenity=waste_basket / waste_disposal 

shop=supermarket / convenience 

and likely more. It is not completely a strange concept for osm to use 
different tags for different “sizes”.
It makes sense when there’s a different quality that comes along the different 
size

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

On Thu, 24 Jan 2019 at 09:03, Joseph Eisenberg 
wrote:

> > It would be usefull to know that a scrub area is so dense that it cannot
> be walked through,
> yet another scrub area is so sparse that a fire cannot propagate without
> wind.
>
> I agree! Those definitions are a pretty good start.
>
> Probably it should relate to the density of the “canopy” of the main type
> of vegetation. So a high-density woodland or scrubland or heath would have
> the branches each tree or shrub touching its neighbors, while sparse or low
> density would be widely spaced, with the “canopy” of the main vegetation
> type covering about 50% of the area only, with plenty of room in between.
>

The problem there is when you get a thick canopy, but the actual ground is
relatively clear.


> Or different fill colors could be used.
>

"Normal" mid-density is the current mid-green, open becomes light / pale
green, thick jungle becomes deep, dark green. Can't see that ever happening
though :-(

Thanks

Graeme
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

> It would be usefull to know that a scrub area is so dense that it cannot
be walked through,
yet another scrub area is so sparse that a fire cannot propagate without
wind.

I agree! Those definitions are a pretty good start.

Probably it should relate to the density of the “canopy” of the main type
of vegetation. So a high-density woodland or scrubland or heath would have
the branches each tree or shrub touching its neighbors, while sparse or low
density would be widely spaced, with the “canopy” of the main vegetation
type covering about 50% of the area only, with plenty of room in between.

This could be especially help info in desert and semi-arid climates.

> Even if those can be tagged, how would they be rendered?

If a renderer uses a pattern to show the type of tree or shrub, the icons
could be closer together or farther apart depending on the density, at
least at high zoom levels. Or different fill colors could be used.
On Thu, Jan 24, 2019 at 7:52 AM Warin <61sundow...@gmail.com> wrote:

> On 24/01/19 09:12, Kevin Kenny wrote:
>
> > On Wed, Jan 23, 2019 at 4:54 PM Warin <61sundow...@gmail.com> wrote:
> >> The main OSM map renders grass, trees as solid colours, and residential
> areas too .. that is not good to me.
> > Another example of the landuse/landcover confusion. We could make an
> > argument that we need a landcover value for 'densely developed'. The
> > US National Land Cover Dataset has several values for 'developed
> > land', just as it has several values for 'forested' (based on leaf
> > cycle) and several for 'farmland' (pasture, cropland, orchard,
> > vineyard, etc.)
> >
> > It's possible, though rare, to have a residential land USE with a
> > different land COVER - consider a small cluster of private houses on a
> > densely wooded parcel. I've been somewhat at a loss how to map the
> > combination, but it's just another example of the same problem: one is
> > physical and topographical, one is human and social.
> >
> Off the subject but:
>
> The density of things (houses, people, plants) is not something OSM does
> well or at all.
>
> It would be usefull to know that a scrub area is so dense that it cannot
> be walked through,
> yet another scrub area is so sparse that a fire cannot propagate without
> wind.
> Even if those can be tagged, how would they be rendered?
>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?


On 24/01/19 09:12, Kevin Kenny wrote:


On Wed, Jan 23, 2019 at 4:54 PM Warin <61sundow...@gmail.com> wrote:

The main OSM map renders grass, trees as solid colours, and residential areas 
too .. that is not good to me.

Another example of the landuse/landcover confusion. We could make an
argument that we need a landcover value for 'densely developed'. The
US National Land Cover Dataset has several values for 'developed
land', just as it has several values for 'forested' (based on leaf
cycle) and several for 'farmland' (pasture, cropland, orchard,
vineyard, etc.)

It's possible, though rare, to have a residential land USE with a
different land COVER - consider a small cluster of private houses on a
densely wooded parcel. I've been somewhat at a loss how to map the
combination, but it's just another example of the same problem: one is
physical and topographical, one is human and social.


Off the subject but:

The density of things (houses, people, plants) is not something OSM does well 
or at all.

It would be usefull to know that a scrub area is so dense that it cannot be 
walked through,
yet another scrub area is so sparse that a fire cannot propagate without wind.
Even if those can be tagged, how would they be rendered?


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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

On 24/01/19 09:06, Kevin Kenny wrote:

On Wed, Jan 23, 2019 at 4:36 PM Paul Allen wrote:

Way back in the thread you wrote:

OSM does not distinguish between the sizes of other thing other than by using
the area or a closed way, or dimensional tags.

That was Mateusz, not me!

Hey .. it was me!!! Can you not tell?:-P

In any case, if the sole determinant for whether something is a 'wood'
or a 'forest' is its land area, then the distinctive tag is redundant.
If it is an indicator of 'relative importance', then it may be
meaningful, but will likely run afoul of verifiability.

I am given to understand that in the UK, the 'hamlet', 'village',
'town', 'city' hierarchy is indeed loosely based on services: a
village has at least a church; a town has a market; a city has a
cathedral or a university. Other countries do it differently, and
you're right that at present the distinction is fairly subjective (and
amounts to tagging for the renderer: at what zoom level should a
municipality or settlement appear?) I'm OK with that because I can't
think of a better way to do it!

My point stands. OSM distinguishes between the sizes of localities (in order
to render them
differently at different zooms) by a means that is not an area or a dimension.
The choice of
hamlet/village/etc. is supposedly related to population size but only loosely,
especially when
some mappers take the number and type of available services into account as
well as
(or instead of) the population.

We appear to be in 'violent agreement', then. Making the distinction
based solely on a dimension is a mistake. If it's 'relative
importance' I can live with it, but need a better guideline about how
to make the distinction. (I don't insist on a quantitative one, just a
loose definition.)

Basing it 'importance' on the population is a 'guide'.
The actual population may not be known, keeps changing and the population could
be based on the area the place services,
not just its local residential area.

It has to be a subjective assessment. And it has to be locally relevant.
See
https://wiki.openstreetmap.org/wiki/Australian_Tagging_Guidelines#City.2C_Town_or_Village.3F
(Caution: I think I have had a hand in those words!)

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Kevin Kenny

On Wed, Jan 23, 2019 at 4:54 PM Warin <61sundow...@gmail.com> wrote:
> The main OSM map renders grass, trees as solid colours, and residential areas 
> too .. that is not good to me.

Another example of the landuse/landcover confusion. We could make an
argument that we need a landcover value for 'densely developed'. The
US National Land Cover Dataset has several values for 'developed
land', just as it has several values for 'forested' (based on leaf
cycle) and several for 'farmland' (pasture, cropland, orchard,
vineyard, etc.)

It's possible, though rare, to have a residential land USE with a
different land COVER - consider a small cluster of private houses on a
densely wooded parcel. I've been somewhat at a loss how to map the
combination, but it's just another example of the same problem: one is
physical and topographical, one is human and social.

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Kevin Kenny

On Wed, Jan 23, 2019 at 4:36 PM Paul Allen  wrote:
> Way back in the thread you wrote:
>>
>> >>> OSM does not distinguish between the sizes of other thing other than by 
>> >>> using the area or a closed way, or dimensional tags.

That was Mateusz, not me!

In any case, if the sole determinant for whether something is a 'wood'
or a 'forest' is its land area, then the distinctive tag is redundant.
If it is an indicator of 'relative importance', then it may be
meaningful, but will likely run afoul of verifiability.

I am given to understand that in the UK, the 'hamlet', 'village',
'town', 'city' hierarchy is indeed loosely based on services: a
village has at least a church; a town has a market; a city has a
cathedral or a university. Other countries do it differently, and
you're right that at present the distinction is fairly subjective (and
amounts to tagging for the renderer: at what zoom level should a
municipality or settlement appear?) I'm OK with that because I can't
think of a better way to do it!

> My point stands.  OSM distinguishes between the sizes of localities (in order 
> to render them
> differently at different zooms) by a means that is not an area or a 
> dimension.  The choice of
> hamlet/village/etc. is supposedly related to population size but only 
> loosely, especially when
> some mappers take the number and type of available services into account as 
> well as
> (or instead of) the population.

We appear to be in 'violent agreement', then. Making the distinction
based solely on a dimension is a mistake. If it's 'relative
importance' I can live with it, but need a better guideline about how
to make the distinction. (I don't insist on a quantitative one, just a
loose definition.)

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?


On 23/01/19 19:37, Marc Gemis wrote:

I think in many cases place=location would be fine.


Depends on what the name is seen as being.


A forest or wood ("bos" (small) or "woud" (for bigger area) in Dutch)
is typically an area with primarily trees, but also grass areas,
pools, cuttings for paths and tracks, etc.

So I was thinking that natural=wood (or landuse=forest) should only be
rendered as an boundary. It can have a name (which should be
rendered).
  For the describing the land-cover inside that area, one would use
landcover or natural=water. Those would be used to "fill the map with
colour". The landcovers areas can be multi-polygons, e.g. to cut out
water areas from landcover=tree


Rendering questions elsewhere :) However ...

To me the rendering question is firstly - chose to render what as a solid 
colour - land cover or land use.

The the other one can be rendered another way.

But choosing individual mixed land covers and land uses to the rendered as 
solid covers just leads to confusion.

Unfortunately many others disagree here ... :)
The main OSM map renders grass, trees as solid colours, and residential areas 
too .. that is not good to me.



m.

On Wed, Jan 23, 2019 at 9:08 AM Warin <61sundow...@gmail.com> wrote:

Nice question Marc.

What is this named area?
Does it have some (taggable) function?
Is it 'just' a location? place=location?


On 23/01/19 18:55, Marc Gemis wrote:

And where do you put the name of the forest/wood ? On the MP or on the
outer way ?
   I would think on the outer way, as the scrub is part of the named
area. But then I have an outer way with only a name tag. Is that
correct ?

m.

On Wed, Jan 23, 2019 at 8:51 AM Warin <61sundow...@gmail.com> wrote:

On 23/01/19 18:25, Peter Elderson wrote:

The rendering itself is a github issue of course, but it shoud be
based on consistent tagging, which is a tagging list concern.

I slipped up in the contradicting paragraphs... I meant, an area
landCOVER=grass within a landUSE=forest.

Main point is, let's recognise / support the growing use of the
landcover key for the three main values: trees, grass and scrub. Then,
bump the issues with the main renderers and editors. How to do that is
not for this list, you are absolutely right about that.

Only after that step, rediscuss the landuse key.

This developed from a simple question of how to map a tree area that has
holes in it of scrub etc.
A fairly simply question?

The simple answer is to map the tree area as a relation with;
natural=wood (even if not 'natural' as OSM acepts that the key 'natural'
encompass things that many regard as not 'natural'), type multipolygon,
The surrounding closed way with the role outer.
Then place simple closed way/s for the hole/s tagged natural=scrub as
appropriate .. and then place them in the relation with the role 'inner'.


The problem came the simple use of the word forest!

There is no need to wait for other steps to use the tag
landuse=forestry, it does not conflict with 'landcover' or 'natural'
tagging.
Many keys and values are developed in parallel.


___




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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Paul Allen

On Wed, 23 Jan 2019 at 21:05, Kevin Kenny  wrote:

Way back in the thread you wrote:

> >>> OSM does not distinguish between the sizes of other thing other than
> by using the area or a closed way, or dimensional tags.
>

To which I responded:

>> place=islet vs place=island
>

And I later responded:

> place=hamlet, place=village, place=town, place=city.
>
> Those are based on population and services, not on land area

Those aren't based upon population.  If they were, you could just specify
population=n
without having to use place=village (or whatever).  According to the wiki,
they're not
based on services either.  According to the wiki, the mapper has to choose
the one
appropriate based upon the population (if known) and I'm guessing most
mappers
do their own thing based on place name/local usage/whim/phase of the moon.

Whatever basis a mapper uses to decide whether to tag as hamlet or village,
etc. it is
NOT based upon area and only loosely based (if that) upon a population size
(which
the mapper may not know).

Which kind of refutes what you wrote much earlier (see above, but I'll
paste it here anyway)

>>> OSM does not distinguish between the sizes of other thing other than by
> using the area or a closed way, or dimensional tags.
>

My point stands.  OSM distinguishes between the sizes of localities (in
order to render them
differently at different zooms) by a means that is not an area or a
dimension.  The choice of
hamlet/village/etc. is supposedly related to population size but only
loosely, especially when
some mappers take the number and type of available services into account as
well as
(or instead of) the population.

-- 
Paul
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Kevin Kenny

>>> OSM does not distinguish between the sizes of other thing other than by 
>>> using the area or a closed way, or dimensional tags.
>>
>>
>> place=islet vs place=island
>>
>>
>> OSM does not distinguish between the sizes of other thing other than by 
>> using the area or a closed way,
>> or dimensional tags, except place=island/islet what was a bad ideas and 
>> should not be repeated?
>
>
> place=hamlet, place=village, place=town, place=city.

Those are based on population and services, not on land area (which
can easily be determined from a closed way or multipolygon).

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Mateusz Konieczny

Jan 23, 2019, 1:11 PM by pla16...@gmail.com:

> On Wed, 23 Jan 2019 at 03:15, Warin <> 61sundow...@gmail.com 
> > > wrote:
>
>
>> OSM does not distinguish between the sizes of other thing other thanby 
>> using the area or a closed way, or dimensional tags. 
>>
>
> place=islet vs place=island
>

OSM does not distinguish between the sizes of other thing other than by using 
the area or a closed way,
or dimensional tags, except place=island/islet what was a bad ideas and should 
not be repeated?

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Paul Allen

On Wed, 23 Jan 2019 at 03:15, Warin <61sundow...@gmail.com> wrote:

OSM does not distinguish between the sizes of other thing other than by
> using the area or a closed way, or dimensional tags.
>

place=islet vs place=island

-- 
Paul
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Marc Gemis

I'm not talking about large areas like the Black Forest, but smaller
forests such as Hondsbossen,
https://www.openstreetmap.org/search?query=sint-kathelijne-waver#map=17/51.07041/4.54562
Typically there is a map of the area near the entrances where the
exact area is shown. That map is similar to the one you see on
http://www.mechelsrivierengebied.be/index.php/natuurgebieden/natuurgebieden-sint-kat-waver/hondsbossen
For an ordinary mapper it might be difficult to map the exact extend,
but what about a government agency contributing to OSM ?

 In this case there is no lake in the middle, so no reason for a
multi-polygon. But in case there is a lake,  what do you do ?
Perhaps could map many of  them as protected areas, but that's
something we are investigating in the Belgian community.


On Wed, Jan 23, 2019 at 10:51 AM Joseph Eisenberg
 wrote:
>
> This idea of a wood or forest as a whole region, rather than the area that 
> actually contains trees, may be culturally limited.
>
> I grew up in a town that was surrounded by national forest on all side, but 
> the valley floor had some pasture and residential areas (300 people). Did we 
> live in the forest? Not sure. The official National Forest boundary and signs 
> were miles away, surrounding our hamlet, but the land was all privately 
> owned, unlike the actual forest land
>
> So it sounds like some people, especially in Europe, want to be able to tag a 
> whole region with the name of the forest.
>
> Would this mean that the whole, very large “Black Forest” region in Germany 
> would get tagged as landuse=forestry or whatever, even though it includes 
> towns and farms and many other types of landuse?
>
> I don’t see how that would be helpful.
>
> If you want region names, there was a proposal before. These have been used 
> to name mountain regions (eg sections of the Alps) for example. But here as 
> well as in the “Black Forest” example, the borders will be very hard to 
> define. Who’s to say where the Alps end and the foothills or valley begins?
>
> On Wed, Jan 23, 2019 at 6:12 PM Martin Koppenhoefer  
> wrote:
>>
>>
>>
>> sent from a phone
>>
>> > On 23. Jan 2019, at 08:55, Marc Gemis  wrote:
>> >
>> > And where do you put the name of the forest/wood ? On the MP or on the
>> > outer way ?
>> > I would think on the outer way, as the scrub is part of the named
>> > area. But then I have an outer way with only a name tag. Is that
>> > correct ?
>>
>>
>> we generally do not have a working concept for names of many kind of natural 
>> landscapes and features. These things typically have fuzzy borders and 
>> consist of different landcover and landuse. When I wrote the landcover 
>> proposal the idea was to separate named entities from both, landuse and from 
>> physical landcover, hence enabling more detailed mapping of landuse and 
>> landcover by not being constrained by the name question for the creation of 
>> objects .
>>
>> I had envisioned natural as key to define “natural features” with their 
>> names, because by then this was still a possible reading of the largest 
>> parts of the values (wood, beach, spring, cave, wetland , coastline, heath, 
>> grassland, cliff, peak,...) with few outliers but now we have so many things 
>> like “mud” , “sand”, “bare_rock”, “grass” that it became less probable it 
>> can be agreed on.
>> Maybe we could use “place” for it? In the end, place is about cultural 
>> objects, you could see named entities as result of a cultural process (they 
>> exist somehow in parallel to the “micro” reality, inside a forest you can 
>> find things that aren’t forest areas, but a person would still say they are 
>> inside that forest, e.g. a lake or a clearing or small fields. The locality 
>> nodes already are used like this.
>>
>>
>> Cheers, Martin
>> ___
>> Tagging mailing list
>> Tagging@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/tagging
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

This idea of a wood or forest as a whole region, rather than the area that
actually contains trees, may be culturally limited.

I grew up in a town that was surrounded by national forest on all side, but
the valley floor had some pasture and residential areas (300 people). Did
we live in the forest? Not sure. The official National Forest boundary and
signs were miles away, surrounding our hamlet, but the land was all
privately owned, unlike the actual forest land

So it sounds like some people, especially in Europe, want to be able to tag
a whole region with the name of the forest.

Would this mean that the whole, very large “Black Forest” region in Germany
would get tagged as landuse=forestry or whatever, even though it includes
towns and farms and many other types of landuse?

I don’t see how that would be helpful.

If you want region names, there was a proposal before. These have been used
to name mountain regions (eg sections of the Alps) for example. But here as
well as in the “Black Forest” example, the borders will be very hard to
define. Who’s to say where the Alps end and the foothills or valley begins?

On Wed, Jan 23, 2019 at 6:12 PM Martin Koppenhoefer 
wrote:

>
>
> sent from a phone
>
> > On 23. Jan 2019, at 08:55, Marc Gemis  wrote:
> >
> > And where do you put the name of the forest/wood ? On the MP or on the
> > outer way ?
> > I would think on the outer way, as the scrub is part of the named
> > area. But then I have an outer way with only a name tag. Is that
> > correct ?
>
>
> we generally do not have a working concept for names of many kind of
> natural landscapes and features. These things typically have fuzzy borders
> and consist of different landcover and landuse. When I wrote the landcover
> proposal the idea was to separate named entities from both, landuse and
> from physical landcover, hence enabling more detailed mapping of landuse
> and landcover by not being constrained by the name question for the
> creation of objects .
>
> I had envisioned natural as key to define “natural features” with their
> names, because by then this was still a possible reading of the largest
> parts of the values (wood, beach, spring, cave, wetland , coastline, heath,
> grassland, cliff, peak,...) with few outliers but now we have so many
> things like “mud” , “sand”, “bare_rock”, “grass” that it became less
> probable it can be agreed on.
> Maybe we could use “place” for it? In the end, place is about cultural
> objects, you could see named entities as result of a cultural process (they
> exist somehow in parallel to the “micro” reality, inside a forest you can
> find things that aren’t forest areas, but a person would still say they are
> inside that forest, e.g. a lake or a clearing or small fields. The locality
> nodes already are used like this.
>
>
> Cheers, Martin
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Martin Koppenhoefer

sent from a phone

> On 23. Jan 2019, at 08:55, Marc Gemis  wrote:
> 
> And where do you put the name of the forest/wood ? On the MP or on the
> outer way ?
> I would think on the outer way, as the scrub is part of the named
> area. But then I have an outer way with only a name tag. Is that
> correct ?

we generally do not have a working concept for names of many kind of natural 
landscapes and features. These things typically have fuzzy borders and consist 
of different landcover and landuse. When I wrote the landcover proposal the 
idea was to separate named entities from both, landuse and from physical 
landcover, hence enabling more detailed mapping of landuse and landcover by not 
being constrained by the name question for the creation of objects .

I had envisioned natural as key to define “natural features” with their names, 
because by then this was still a possible reading of the largest parts of the 
values (wood, beach, spring, cave, wetland , coastline, heath, grassland, 
cliff, peak,...) with few outliers but now we have so many things like “mud” , 
“sand”, “bare_rock”, “grass” that it became less probable it can be agreed on.
Maybe we could use “place” for it? In the end, place is about cultural objects, 
you could see named entities as result of a cultural process (they exist 
somehow in parallel to the “micro” reality, inside a forest you can find things 
that aren’t forest areas, but a person would still say they are inside that 
forest, e.g. a lake or a clearing or small fields. The locality nodes already 
are used like this.

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Mateusz Konieczny

I admit that I am putting it on MP, despite that it would be more correct to
put in an outer way with some tagging indicating that it refers to this entire 
forest.

Unfortunately, I have no good idea for tagging it properly. 

place=locality is rather for small places
landuse=logging, landuse=forestry would be also incorrect in nearly all cases
natural=forest, natural=wood - well, there is a reason why there is MP here

Maybe place=forest would fit well?

And in some cases like village or town inside forest second MP would be needed. 

Jan 23, 2019, 8:55 AM by marc.ge...@gmail.com:

> And where do you put the name of the forest/wood ? On the MP or on the
> outer way ?
>  I would think on the outer way, as the scrub is part of the named
> area. But then I have an outer way with only a name tag. Is that
> correct ?
>
> m.
>
> On Wed, Jan 23, 2019 at 8:51 AM Warin <> 61sundow...@gmail.com 
> > > wrote:
>
>>
>> On 23/01/19 18:25, Peter Elderson wrote:
>> > The rendering itself is a github issue of course, but it shoud be
>> > based on consistent tagging, which is a tagging list concern.
>> >
>> > I slipped up in the contradicting paragraphs... I meant, an area
>> > landCOVER=grass within a landUSE=forest.
>> >
>> > Main point is, let's recognise / support the growing use of the
>> > landcover key for the three main values: trees, grass and scrub. Then,
>> > bump the issues with the main renderers and editors. How to do that is
>> > not for this list, you are absolutely right about that.
>> >
>> > Only after that step, rediscuss the landuse key.
>>
>> This developed from a simple question of how to map a tree area that has
>> holes in it of scrub etc.
>> A fairly simply question?
>>
>> The simple answer is to map the tree area as a relation with;
>> natural=wood (even if not 'natural' as OSM acepts that the key 'natural'
>> encompass things that many regard as not 'natural'), type multipolygon,
>> The surrounding closed way with the role outer.
>> Then place simple closed way/s for the hole/s tagged natural=scrub as
>> appropriate .. and then place them in the relation with the role 'inner'.
>>
>> 
>> The problem came the simple use of the word forest!
>>
>> There is no need to wait for other steps to use the tag
>> landuse=forestry, it does not conflict with 'landcover' or 'natural'
>> tagging.
>> Many keys and values are developed in parallel.
>>
>>
>> ___
>> Tagging mailing list
>> Tagging@openstreetmap.org 
>> https://lists.openstreetmap.org/listinfo/tagging 
>> 
>>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org 
> https://lists.openstreetmap.org/listinfo/tagging 
> 
>

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-23 Thread Marc Gemis

I think in many cases place=location would be fine.
A forest or wood ("bos" (small) or "woud" (for bigger area) in Dutch)
is typically an area with primarily trees, but also grass areas,
pools, cuttings for paths and tracks, etc.

So I was thinking that natural=wood (or landuse=forest) should only be
rendered as an boundary. It can have a name (which should be
rendered).
 For the describing the land-cover inside that area, one would use
landcover or natural=water. Those would be used to "fill the map with
colour". The landcovers areas can be multi-polygons, e.g. to cut out
water areas from landcover=tree

m.

On Wed, Jan 23, 2019 at 9:08 AM Warin <61sundow...@gmail.com> wrote:
>
> Nice question Marc.
>
> What is this named area?
> Does it have some (taggable) function?
> Is it 'just' a location? place=location?
>
>
> On 23/01/19 18:55, Marc Gemis wrote:
> > And where do you put the name of the forest/wood ? On the MP or on the
> > outer way ?
> >   I would think on the outer way, as the scrub is part of the named
> > area. But then I have an outer way with only a name tag. Is that
> > correct ?
> >
> > m.
> >
> > On Wed, Jan 23, 2019 at 8:51 AM Warin <61sundow...@gmail.com> wrote:
> >> On 23/01/19 18:25, Peter Elderson wrote:
> >>> The rendering itself is a github issue of course, but it shoud be
> >>> based on consistent tagging, which is a tagging list concern.
> >>>
> >>> I slipped up in the contradicting paragraphs... I meant, an area
> >>> landCOVER=grass within a landUSE=forest.
> >>>
> >>> Main point is, let's recognise / support the growing use of the
> >>> landcover key for the three main values: trees, grass and scrub. Then,
> >>> bump the issues with the main renderers and editors. How to do that is
> >>> not for this list, you are absolutely right about that.
> >>>
> >>> Only after that step, rediscuss the landuse key.
> >> This developed from a simple question of how to map a tree area that has
> >> holes in it of scrub etc.
> >> A fairly simply question?
> >>
> >> The simple answer is to map the tree area as a relation with;
> >> natural=wood (even if not 'natural' as OSM acepts that the key 'natural'
> >> encompass things that many regard as not 'natural'), type multipolygon,
> >> The surrounding closed way with the role outer.
> >> Then place simple closed way/s for the hole/s tagged natural=scrub as
> >> appropriate .. and then place them in the relation with the role 'inner'.
> >>
> >> 
> >> The problem came the simple use of the word forest!
> >>
> >> There is no need to wait for other steps to use the tag
> >> landuse=forestry, it does not conflict with 'landcover' or 'natural'
> >> tagging.
> >> Many keys and values are developed in parallel.
> >>
> >>
> >> ___
> >> Tagging mailing list
> >> Tagging@openstreetmap.org
> >> https://lists.openstreetmap.org/listinfo/tagging
> > ___
> > Tagging mailing list
> > Tagging@openstreetmap.org
> > https://lists.openstreetmap.org/listinfo/tagging
>
>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?


Nice question Marc.

What is this named area?
Does it have some (taggable) function?
Is it 'just' a location? place=location?


On 23/01/19 18:55, Marc Gemis wrote:

And where do you put the name of the forest/wood ? On the MP or on the
outer way ?
  I would think on the outer way, as the scrub is part of the named
area. But then I have an outer way with only a name tag. Is that
correct ?

m.

On Wed, Jan 23, 2019 at 8:51 AM Warin <61sundow...@gmail.com> wrote:

On 23/01/19 18:25, Peter Elderson wrote:

The rendering itself is a github issue of course, but it shoud be
based on consistent tagging, which is a tagging list concern.

I slipped up in the contradicting paragraphs... I meant, an area
landCOVER=grass within a landUSE=forest.

Main point is, let's recognise / support the growing use of the
landcover key for the three main values: trees, grass and scrub. Then,
bump the issues with the main renderers and editors. How to do that is
not for this list, you are absolutely right about that.

Only after that step, rediscuss the landuse key.

This developed from a simple question of how to map a tree area that has
holes in it of scrub etc.
A fairly simply question?

The simple answer is to map the tree area as a relation with;
natural=wood (even if not 'natural' as OSM acepts that the key 'natural'
encompass things that many regard as not 'natural'), type multipolygon,
The surrounding closed way with the role outer.
Then place simple closed way/s for the hole/s tagged natural=scrub as
appropriate .. and then place them in the relation with the role 'inner'.


The problem came the simple use of the word forest!

There is no need to wait for other steps to use the tag
landuse=forestry, it does not conflict with 'landcover' or 'natural'
tagging.
Many keys and values are developed in parallel.


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

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




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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Marc Gemis

And where do you put the name of the forest/wood ? On the MP or on the
outer way ?
 I would think on the outer way, as the scrub is part of the named
area. But then I have an outer way with only a name tag. Is that
correct ?

m.

On Wed, Jan 23, 2019 at 8:51 AM Warin <61sundow...@gmail.com> wrote:
>
> On 23/01/19 18:25, Peter Elderson wrote:
> > The rendering itself is a github issue of course, but it shoud be
> > based on consistent tagging, which is a tagging list concern.
> >
> > I slipped up in the contradicting paragraphs... I meant, an area
> > landCOVER=grass within a landUSE=forest.
> >
> > Main point is, let's recognise / support the growing use of the
> > landcover key for the three main values: trees, grass and scrub. Then,
> > bump the issues with the main renderers and editors. How to do that is
> > not for this list, you are absolutely right about that.
> >
> > Only after that step, rediscuss the landuse key.
>
> This developed from a simple question of how to map a tree area that has
> holes in it of scrub etc.
> A fairly simply question?
>
> The simple answer is to map the tree area as a relation with;
> natural=wood (even if not 'natural' as OSM acepts that the key 'natural'
> encompass things that many regard as not 'natural'), type multipolygon,
> The surrounding closed way with the role outer.
> Then place simple closed way/s for the hole/s tagged natural=scrub as
> appropriate .. and then place them in the relation with the role 'inner'.
>
> 
> The problem came the simple use of the word forest!
>
> There is no need to wait for other steps to use the tag
> landuse=forestry, it does not conflict with 'landcover' or 'natural'
> tagging.
> Many keys and values are developed in parallel.
>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?


On 23/01/19 18:25, Peter Elderson wrote:
The rendering itself is a github issue of course, but it shoud be 
based on consistent tagging, which is a tagging list concern.


I slipped up in the contradicting paragraphs... I meant, an area 
landCOVER=grass within a landUSE=forest.


Main point is, let's recognise / support the growing use of the 
landcover key for the three main values: trees, grass and scrub. Then, 
bump the issues with the main renderers and editors. How to do that is 
not for this list, you are absolutely right about that.


Only after that step, rediscuss the landuse key.


This developed from a simple question of how to map a tree area that has 
holes in it of scrub etc.

A fairly simply question?

The simple answer is to map the tree area as a relation with;
natural=wood (even if not 'natural' as OSM acepts that the key 'natural' 
encompass things that many regard as not 'natural'), type multipolygon,

The surrounding closed way with the role outer.
Then place simple closed way/s for the hole/s tagged natural=scrub as 
appropriate .. and then place them in the relation with the role 'inner'.



The problem came the simple use of the word forest!

There is no need to wait for other steps to use the tag 
landuse=forestry, it does not conflict with 'landcover' or 'natural' 
tagging.

Many keys and values are developed in parallel.


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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Peter Elderson

The rendering itself is a github issue of course, but it shoud be based on
consistent tagging, which is a tagging list concern.

I slipped up in the contradicting paragraphs... I meant, an area
landCOVER=grass within a landUSE=forest.

Main point is, let's recognise / support the growing use of the landcover
key for the three main values: trees, grass and scrub. Then, bump the
issues with the main renderers and editors. How to do that is not for this
list, you are absolutely right about that.

Only after that step, rediscuss the landuse key.

Vr gr Peter Elderson


Op wo 23 jan. 2019 om 08:11 schreef Mateusz Konieczny <
matkoni...@tutanota.com>:

>
>
>
> Jan 23, 2019, 1:00 AM by pelder...@gmail.com:
>
>
> Vr gr Peter Elderson
>
> Landcover tag now approaches 100 000 occurrences. Still growing despite
> not being rendered.  I would think rendering the top three landcover values
> is not out of place. The github issues are still there. Initially:
>
> And please use that GitHub issues for discussing this. Github issues of
> (even important) projects
> are not acceptable place to discuss tagging, discussing how specific
> projects should render
> things is utterly offtopic here.
>
> And I admit that continued "landcover must be rendered in this one
> specific project that
> I will not mention by name making my complaint not only offtopic but also
> utterly useless"
> mails on TAGGING mailing list make me think badly about this tagging
> concept.
>
> Hopefully we have not reached stage that I am falling for false flag.
>
> After that, further steps could be discussed. Until this is done, in my
> opinion every discussion about the usage of landuse tags is doomed to fail.
> There simply is no way forward if there is no rendering alternative.
>
> Anyone may make their own rendering. And if you are stuck then you can
> reach out to people
> who may help.
>
>
> If landcover=grass is rendered, a clearing in a landuse=forest could
> simply be tagged as a polygon with landuse=grass, without cutting up the
> forest.
>
> The main objection would be: existing base. But the landcover rendering
> does not harm the existing base; it's fully backwards compatible.
>
> This two paragraphs contradict each other.
>
>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?




Jan 23, 2019, 1:00 AM by pelder...@gmail.com:

>
> Vr gr Peter Elderson
>
> Landcover tag now approaches 100 000 occurrences. Still growing despite not 
> being rendered.  I would think rendering the top three landcover values is 
> not out of place. The github issues are still there. Initially:
>
And please use that GitHub issues for discussing this. Github issues of (even 
important) projects
are not acceptable place to discuss tagging, discussing how specific projects 
should render
things is utterly offtopic here.

And I admit that continued "landcover must be rendered in this one specific 
project that
I will not mention by name making my complaint not only offtopic but also 
utterly useless"
mails on TAGGING mailing list make me think badly about this tagging concept.

Hopefully we have not reached stage that I am falling for false flag.


> After that, further steps could be discussed. Until this is done, in my 
> opinion every discussion about the usage of landuse tags is doomed to fail. 
> There simply is no way forward if there is no rendering alternative. 
>
Anyone may make their own rendering. And if you are stuck then you can reach 
out to people
who may help.



> If landcover=grass is rendered, a clearing in a landuse=forest could simply 
> be tagged as a polygon with landuse=grass, without cutting up the forest.
>
> The main objection would be: existing base. But the landcover rendering does 
> not harm the existing base; it's fully backwards compatible.
>
This two paragraphs contradict each other. 



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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Peter Elderson

I think there can be no outcome of any discussion about landuse unless the
landcover is separated from the landuse first.

I think the only way forward is to actually do that first. The growth of
the landcover key shows that many mappers think that it's a good idea to
map landcover separately.

I would gladly see the landcover key recognised and rendered for the three
main values.
Only then, discuss the main landuse key values and modifyers again.

Else, we're just repeating the same discussion over and over.

Vr gr Peter Elderson


Op wo 23 jan. 2019 om 04:50 schreef Warin <61sundow...@gmail.com>:

> On 23/01/19 07:37, Mateusz Konieczny wrote:
>
> Jan 21, 2019, 12:03 AM by 61sundow...@gmail.com:
>
> The end to this madness is for renders to recognise that the
> landuse=forest needs to be rendered differently from natural=wood.
> The essential difference between the two is that landuse must have some
> human benefit, a produce, and a clear way of doing that is to add the
> rendering of a axe to the tree.
>
>
> (1) in a typical rendering this distinction is completely unimportant
> or at least not worth different rendering
>
> (2) other people have different mismatching ideas what is the
> "real" difference between natural=wood and landuse=forest
>
> (3) there is no consistent difference in how natural=wood and
> landuse=forest are used
> by mappers
>
>
> If the is no difference between the two then there will be no problem
> depreciating landuse=forest.
>
> There are some who do see a distinction of land use, and want to use that
> distinction.
> If some landuse=forest were to be re tagged landuse=forestry as it matches
> a definition of 'landuse' will those using landuse=forest be happy with
> that?
>
> Will they then be happy that landuse=forest becomes depreciated as it is
> seen as the same as natural=wood?
>
> --
> A 'managed' tree area does not necessarily match the land use definition.
> What is the purpose of this land management? Is there a produce that is
> derived from the trees?
> If there is no produce than it is not landuse=forestry.
>
> A national park is 'managed' .. In Australia no produce comes out of it.
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?




Jan 23, 2019, 4:49 AM by 61sundow...@gmail.com:

> On 23/01/19 07:37, Mateusz Konieczny  wrote:
>
>> Jan 21, 2019, 12:03 AM by >> 61sundow...@gmail.com 
>> >> :
>>
>>> The end to this madness is  for renders to recognise that the 
>>> landuse=forest needs to be  rendered differently from natural=wood.
>>> The essential difference  between the two is that landuse must have 
>>> some human benefit,  a produce, and a clear way of doing that is to 
>>> add the  rendering of a axe to the tree.
>>>
>>
>> (1) in a typical rendering this distinction is completely unimportant
>> or at least not worth different rendering
>>
>> (2) other people have different mismatchingideas what is the
>> "real" difference between natural=wood andlanduse=forest
>>
>> (3) there is no consistent difference in hownatural=wood and 
>> landuse=forest are used
>> by mappers
>>
>
> If the is no difference between the two then there will be noproblem 
> depreciating landuse=forest.
>

First of all: "there many, many opinions how natural=wood and landuse=forest 
differ and
some people think that his makes distinction between this tags useless" is not 
the same as
"there is no difference".

And landuse=forest is used more than three million times
https://taginfo.openstreetmap.org/tags/landuse=forest 



> If there is no produce than it is not landuse=forestry. 
>  
>
Note that many are not using "forestry" to mean "using  forest  toproduce wood".
See  https://en.wikipedia.org/wiki/Forestry 
 for an example.



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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

On 2019-01-23 04:14, Warin wrote:

> Temperate and Tropical moist/dry are climates... if those are to be mapped 
> them go right ahead .. but they are not confined to forests, so should be 
> mapped separately.

Those are simply the names biologists give to that forest types.  E.g. 
"Tropical moist" is a very specific type of forest. No mapping of the climate 
is implied (/but yes, //climate have influence on the forests of a particulare 
region, of course/).

And I missed the fact that we are (/or planning to/) map climates in OSM. Are 
we?  :-/

smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?


On 23/01/19 07:37, Mateusz Konieczny wrote:

Jan 21, 2019, 12:03 AM by 61sundow...@gmail.com:

The end to this madness is for renders to recognise that the
landuse=forest needs to be rendered differently from natural=wood.
The essential difference between the two is that landuse must have
some human benefit, a produce, and a clear way of doing that is to
add the rendering of a axe to the tree.


(1) in a typical rendering this distinction is completely unimportant
or at least not worth different rendering

(2) other people have different mismatching ideas what is the
"real" difference between natural=wood and landuse=forest

(3) there is no consistent difference in how natural=wood and 
landuse=forest are used

by mappers


If the is no difference between the two then there will be no problem 
depreciating landuse=forest.


There are some who do see a distinction of land use, and want to use 
that distinction.
If some landuse=forest were to be re tagged landuse=forestry as it 
matches a definition of 'landuse' will those using landuse=forest be 
happy with that?


Will they then be happy that landuse=forest becomes depreciated as it is 
seen as the same as natural=wood?


--
A 'managed' tree area does not necessarily match the land use definition.
What is the purpose of this land management? Is there a produce that is 
derived from the trees?

If there is no produce than it is not landuse=forestry.

A national park is 'managed' .. In Australia no produce comes out of it.

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?


On 23/01/19 07:31, Mateusz Konieczny wrote:




Jan 21, 2019, 9:44 PM by pla16...@gmail.com:

On Mon, 21 Jan 2019 at 20:21, Warin <61sundow...@gmail.com
> wrote:


My problem with going to landuse=forestry with natural=wood...

what happens to the remaining landuse=forest?
Will that finally be recognised as the same as natural=wood
and be migrated to natural=wood???


Ideally, if we get landuse=forestry and it eventually renders

Note that it is not obvious to me that a typical map would want to 
display landuse=forestry.




In Australia forestry areas are rendered differently. They are seen as a 
land use that is worth showing, like residential areas are worth showing.


The UK Landranger 1:50,000 show forestry commission land ...

I think you'll find a great many maps identify forestry.

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?


On 23/01/19 11:52, Sergio Manzi wrote:


Only about the cited point (/tagging natural forests as 
natural=wood/), I think a natural forest should be tagged as 
natural=forest (/quite logically, I would say.../), while natural=wood 
should be reserved for "small forests" (/which is one of the possible 
meaning of "wood" in English, if I'm not mistaken/).




The area (size) is given by the area of the closed way. No need to 
differentiate between 'small', 'medium' or 'large' (and each of them 
would have to be defined). Avoid this!
OSM does not distinguish between the sizes of other thing other than by 
using the area or a closed way, or dimensional tags.
I am firmly against having tags that distinguish between features based 
on their size.


 Also please consider that "forest" is very generic, and we could be 
willing to more exactly define which kind of forest we are tagging 
(/and eventually have it rendered accordingly/).



Trees and building are also 'generic' ...


To this extent the only way I see in the Wiki is to use the 
leaf_type=* tag (/and only"broadleaved", "needleleaved" and "mixed" 
are defined as possible values/).Wikipedia defines 6 types of forest:


But to me a forest is not just "a lot of trees", but an entire 
ecosystem dominated by trees. And those ecosystems can be very 
different between different continents and latitudes.


  * Temperate needleleaf
  * Temperate broadleaf and mixed
  * Tropical moist
  * Tropical dry
  * Sparse trees and parkland
  * Forest plantations



Temperate and Tropical moist/dry are climates... if those are to be 
mapped them go right ahead .. but they are not confined to forests, so 
should be mapped separately.



Whether something is 'natural' or not is up to your definition and 
knowledge. For most mappers they simply see trees/plants.

If you that 'natural' is something that is verifiable then tell us how.
How do you tell if a tree has been planted by humans and one that is 
self seeded? I cannot see a difference at all.


The density of plants (trees, shrubs, heath) is something to consider .. 
but it needs to be a separate thing and be able to be used with various 
vegetation types. I believe it has been raised before.
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Joseph Eisenberg

I would be very happy if more mappers start tagging leaf_type and
leaf_cycle. Certainly it is possible to add further detail. In temperate
regions one could tag the dominant species of tree in woodlands where the
majority of trees are one species, eg Ponderosa Pine or Douglas Fir in the
western USA.

For more complex forests, a new key might be needed, but it could be
difficult for mappers to recognize some of the finer distinctions. I grew
up in the Klamath National Forest, and I have some interest in Botany, but
I’m not sure I would know where to draw the line between different types of
woodland in my home area, beyond simple things like leaf_type, leaf_cycle,
and dominant species, and maybe tree or shrub density.

Maybe year-round rain vs wet season / dry season?

Here in my adopted homeland of Indonesia, it’s all just “hutan” (forest /
woodland of any type)

On Wed, Jan 23, 2019 at 10:24 AM Sergio Manzi  wrote:

> Hi!
> On 2019-01-23 02:10, Joseph Eisenberg wrote:
>
> The one thing that’s missing is a tag for the density of the main
> vegetation type; is it a dense canopy of trees, or dense scrubland, verses
> more widely spaced.
>
> Not only that (and the "leaf cycle" thing): again, a forest is not a bunch
> of trees (*maybe all of the same kind*), and there are very good
> definitions for every kind of forest.
>
> How would you tag a South Indian "Shola" or eastern Guadeloupe's tropical
> forest?
>
> We are not interested in this kind of detail when it comes to nature?
>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

Hi!

On 2019-01-23 02:10, Joseph Eisenberg wrote:
> The one thing that’s missing is a tag for the density of the main vegetation 
> type; is it a dense canopy of trees, or dense scrubland, verses more widely 
> spaced.

Not only that (and the "leaf cycle" thing): again, a forest is not a bunch of 
trees (/maybe all of the same kind/), and there are very good definitions for 
every kind of forest.

How would you tag a South Indian "Shola" or eastern Guadeloupe's tropical 
forest?

We are not interested in this kind of detail when it comes to nature?

smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Joseph Eisenberg

Oh, and we shouldn’t forget leaf_type=leafless


This is used for cactus and other succulents, and it’s currently rendered
by the Openstreetmap-Carto style, for wood and forest.

Leaf_cycle is rendered by the Alternative-Colors style made by Christoph,
if you want to see an example of how this can look:

http://blog.imagico.de/differentiated-rendering-of-woodland-in-maps/

And

http://blog.imagico.de/more-on-vegetation-rendering-in-openstreetmap-maps/


On Wed, Jan 23, 2019 at 10:10 AM Joseph Eisenberg <
joseph.eisenb...@gmail.com> wrote:

> > the only way I see in the Wiki is to use the leaf_type=* tag
>
> You can also use leaf_cycle= to tag deciduous vs evergreen, and also
> semi-deciduous, semi-evergreen and mixed:
> https://wiki.openstreetmap.org/wiki/Key:leaf_cycle
>
> This accounts for most types of woodland, along with leaf type. The rest
> can be inferred from latitude (eg forest/wood in the tropics is clearly
> tropical) and elevation (montane vs lowland rainforest) which is readily
> available information.
>
> Probably there is no need for mappers to tag elevation and
> latitude-related distinctions.
>
> The one thing that’s missing is a tag for the density of the main
> vegetation type; is it a dense canopy of trees, or dense scrubland, verses
> more widely spaced.
>
> And there isn’t a way to tag a grassland with scattered shrubs or trees,
> probably because this is not common in Northern Europe.
> On Wed, Jan 23, 2019 at 9:58 AM Sergio Manzi  wrote:
>
>> Well, sorry, obviously I did an editing mistake and the "*Wikipedia
>> defines 6 types of forest*" phrase jumped up in the wrong place: it
>> should be just above the dotted list of forest types...
>>
>> Sorry about the confusion...
>>
>> Sergio
>>
>>
>> On 2019-01-23 01:52, Sergio Manzi wrote:
>>
>> Only about the cited point (*tagging natural forests as natural=wood*),
>> I think a natural forest should be tagged as natural=forest (*quite
>> logically, I would say...*), while natural=wood should be reserved for
>> "small forests" (*which is one of the possible meaning of "wood" in
>> English, if I'm not mistaken*).
>>
>> Also please consider that "forest" is very generic, and we could be
>> willing to more exactly define which kind of forest we are tagging (*and
>> eventually have it rendered accordingly*).
>>
>> To this extent the only way I see in the Wiki is to use the leaf_type=*
>> tag (*and only"broadleaved", "needleleaved" and "mixed" are defined as
>> possible values*).Wikipedia defines 6 types of forest:
>>
>> But to me a forest is not just "a lot of trees", but an entire ecosystem
>> dominated by trees. And those ecosystems can be very different between
>> different continents and latitudes.
>>
>>- Temperate needleleaf
>>- Temperate broadleaf and mixed
>>- Tropical moist
>>- Tropical dry
>>- Sparse trees and parkland
>>- Forest plantations
>>
>> I think we should be (somehow) able to tag those types.
>>
>> Please also check the sixth forest type described in Wikipedia (*Forest
>> plantations*): I think it coincide with the "forestry" concept we are
>> talking about in the thread (*landuse=forestry, or whatever...*).
>>
>> Cheers,
>>
>> Sergio
>>
>>
>> On 2019-01-23 01:00, Peter Elderson wrote:
>>
>> Natural forests could be preferably tagged as natural=wood ...
>>
>> ___
>> Tagging mailing list
>> Tagging@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/tagging
>>
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Joseph Eisenberg

> the only way I see in the Wiki is to use the leaf_type=* tag

You can also use leaf_cycle= to tag deciduous vs evergreen, and also
semi-deciduous, semi-evergreen and mixed:
https://wiki.openstreetmap.org/wiki/Key:leaf_cycle

This accounts for most types of woodland, along with leaf type. The rest
can be inferred from latitude (eg forest/wood in the tropics is clearly
tropical) and elevation (montane vs lowland rainforest) which is readily
available information.

Probably there is no need for mappers to tag elevation and latitude-related
distinctions.

The one thing that’s missing is a tag for the density of the main
vegetation type; is it a dense canopy of trees, or dense scrubland, verses
more widely spaced.

And there isn’t a way to tag a grassland with scattered shrubs or trees,
probably because this is not common in Northern Europe.
On Wed, Jan 23, 2019 at 9:58 AM Sergio Manzi  wrote:

> Well, sorry, obviously I did an editing mistake and the "*Wikipedia
> defines 6 types of forest*" phrase jumped up in the wrong place: it
> should be just above the dotted list of forest types...
>
> Sorry about the confusion...
>
> Sergio
>
>
> On 2019-01-23 01:52, Sergio Manzi wrote:
>
> Only about the cited point (*tagging natural forests as natural=wood*), I
> think a natural forest should be tagged as natural=forest (*quite
> logically, I would say...*), while natural=wood should be reserved for
> "small forests" (*which is one of the possible meaning of "wood" in
> English, if I'm not mistaken*).
>
> Also please consider that "forest" is very generic, and we could be
> willing to more exactly define which kind of forest we are tagging (*and
> eventually have it rendered accordingly*).
>
> To this extent the only way I see in the Wiki is to use the leaf_type=*
> tag (*and only"broadleaved", "needleleaved" and "mixed" are defined as
> possible values*).Wikipedia defines 6 types of forest:
>
> But to me a forest is not just "a lot of trees", but an entire ecosystem
> dominated by trees. And those ecosystems can be very different between
> different continents and latitudes.
>
>- Temperate needleleaf
>- Temperate broadleaf and mixed
>- Tropical moist
>- Tropical dry
>- Sparse trees and parkland
>- Forest plantations
>
> I think we should be (somehow) able to tag those types.
>
> Please also check the sixth forest type described in Wikipedia (*Forest
> plantations*): I think it coincide with the "forestry" concept we are
> talking about in the thread (*landuse=forestry, or whatever...*).
>
> Cheers,
>
> Sergio
>
>
> On 2019-01-23 01:00, Peter Elderson wrote:
>
> Natural forests could be preferably tagged as natural=wood ...
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

Well, sorry, obviously I did an editing mistake and the "/Wikipedia defines 6 
types of forest/" phrase jumped up in the wrong place: it should be just above 
the dotted list of forest types...

Sorry about the confusion...

Sergio


On 2019-01-23 01:52, Sergio Manzi wrote:
>
> Only about the cited point (/tagging natural forests as natural=wood/), I 
> think a natural forest should be tagged as natural=forest (/quite logically, 
> I would say.../), while natural=wood should be reserved for "small forests" 
> (/which is one of the possible meaning of "wood" in English, if I'm not 
> mistaken/).
>
> Also please consider that "forest" is very generic, and we could be willing 
> to more exactly define which kind of forest we are tagging (/and eventually 
> have it rendered accordingly/).
>
> To this extent the only way I see in the Wiki is to use the leaf_type=* tag 
> (/and only"broadleaved", "needleleaved" and "mixed" are defined as possible 
> values/).Wikipedia defines 6 types of forest:
>
> But to me a forest is not just "a lot of trees", but an entire ecosystem 
> dominated by trees. And those ecosystems can be very different between 
> different continents and latitudes.
>
>   * Temperate needleleaf
>   * Temperate broadleaf and mixed
>   * Tropical moist
>   * Tropical dry
>   * Sparse trees and parkland
>   * Forest plantations
>
> I think we should be (somehow) able to tag those types.
>
> Please also check the sixth forest type described in Wikipedia (/Forest 
> plantations/): I think it coincide with the "forestry" concept we are talking 
> about in the thread (/landuse=forestry, or whatever.../).
>
> Cheers,
>
> Sergio
>
>
> On 2019-01-23 01:00, Peter Elderson wrote:
>> Natural forests could be preferably tagged as natural=wood ...


smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

Only about the cited point (/tagging natural forests as natural=wood/), I think 
a natural forest should be tagged as natural=forest (/quite logically, I would 
say.../), while natural=wood should be reserved for "small forests" (/which is 
one of the possible meaning of "wood" in English, if I'm not mistaken/).

Also please consider that "forest" is very generic, and we could be willing to 
more exactly define which kind of forest we are tagging (/and eventually have 
it rendered accordingly/).

To this extent the only way I see in the Wiki is to use the leaf_type=* tag 
(/and only"broadleaved", "needleleaved" and "mixed" are defined as possible 
values/).Wikipedia defines 6 types of forest:

But to me a forest is not just "a lot of trees", but an entire ecosystem 
dominated by trees. And those ecosystems can be very different between 
different continents and latitudes.

  * Temperate needleleaf
  * Temperate broadleaf and mixed
  * Tropical moist
  * Tropical dry
  * Sparse trees and parkland
  * Forest plantations

I think we should be (somehow) able to tag those types.

Please also check the sixth forest type described in Wikipedia (/Forest 
plantations/): I think it coincide with the "forestry" concept we are talking 
about in the thread (/landuse=forestry, or whatever.../).

Cheers,

Sergio


On 2019-01-23 01:00, Peter Elderson wrote:
> Natural forests could be preferably tagged as natural=wood ...


smime.p7s
Description: S/MIME Cryptographic Signature
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Peter Elderson

Vr gr Peter Elderson

Landcover tag now approaches 100 000 occurrences. Still growing despite not
being rendered.  I would think rendering the top three landcover values is
not out of place. The github issues are still there. Initially:
landcover=trees same rendering as natural=wood and landuse=forest
landcover=grass same rendering as landuse=grass
landcover=scrub same rendering as natural=scrub

After that, further steps could be discussed. Until this is done, in my
opinion every discussion about the usage of landuse tags is doomed to fail.
There simply is no way forward if there is no rendering alternative.

E.g. If landcover=trees is rendered, it is safe to restrict landuse=forest
to mean cultivated / exploited forest, used for a. forestry, b.
purposefully maintained forest for multipurpose usage such as all the
forests in the Netherlands.
Natural forests could be preferably tagged as natural=wood, and areas of
trees within a landuse such as a residential area could be tagged with
landcover=trees without cutting up the landuse.

landuse=forest would have a default landcover of trees.
landuse=forestry not. But, by using the landcover tags, you could still use
landcover=forestry is that is really what the use of the land is, and cover
it nicely with the appropriate landcovers.

If landcover=grass is rendered, a clearing in a landuse=forest could simply
be tagged as a polygon with landuse=grass, without cutting up the forest.

The main objection would be: existing base. But the landcover rendering
does not harm the existing base; it's fully backwards compatible. It will
not solve the completely mixed-up use of landuse=forest and natural=wood in
one blow, but it can provide more logical en clearer distinction (and
presets).
Over time, micromappers will simplify their own area (all those cut-up
residentials) just because they can and it's simpler and more logical. The
big forest area's often have dedicated mappers who will seize the
opportunity. New mappers will pick it up naturrally without even knowing it
was controversial.

I have spoken. Ugh.
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Graeme Fitzpatrick

On Wed, 23 Jan 2019 at 07:10, Tod Fitch  wrote:

>
> Perhaps the way forward would be to change the wiki to indicate that
> landuse=forest is deprecated due to its confused usage. Add some text to
> the page directing mappers to either landcover=trees if they are simply
> mapping the presence of trees or landuse=forestry if they are mapping an
> area used for the production of wood products (lumber, paper, etc.) that
> may or may not have trees on it a the moment.
>

That may be the bet way to go, as landcover=trees can cover anything from 3
trees in the middle of a roundabout, to a jungle covering 1000's of sq km,
& also natural forests as well as plantations, orchards, managed forests
etc.

>
> Not rendering landuse=forestry on the default OSM map to reduce “tagging
> for the renderer” is an interesting idea. I’ll have to think about that but
> it does have some appeal.
>

No, I think that =forestry should still be rendered, possibly by Warin's
idea of an axe in a tree trunk, to show that there are, or could be, trees
in that area. If it wasn't rendered, you'd finish up with some very
extensive areas of  blank map, frequently surrounded by trees

Thanks

Graeme
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Paul Allen

On Tue, 22 Jan 2019 at 21:10, Tod Fitch  wrote:

Not rendering landuse=forestry on the default OSM map to reduce “tagging
> for the renderer” is an interesting idea. I’ll have to think about that but
> it does have some appeal.
>

It doesn't appeal to me.  I'd prefer it to render but in a way that differs
from landcover=trees or
natural=wood (or whatever).

1) It's time-consuming and tedious to map large areas of land used for
forestry.  Many people
won't bother if it doesn't render and either use natural=wood for the
entire extent or use only
natural=wood where there are currently trees and ignore the areas of stumps
or saplings.

2) Somebody may be happy to put the effort into outlining the area used for
forestry but may
not want to put in the additional effort to use landcover=trees (or
whatever) to show where the
trees are today (actually, where they were when the aerial imagery was
generated) because it's
changeable.  So to permit minimal mapping we need forestry to show up.

3) Terry Pratchett's *Hogfather* points out that only drawings by children
leave a white gap between
the ground and the sky.  In an ideal world (where we had finally mapped
everything) there would
be no white areas of the map.  Not rendering land used for forestry would
prevent us ever
reaching that goal.  OK, that one's a bit of a stretch. :)

-- 
Paul
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Tod Fitch


> On Jan 22, 2019, at 12:52 PM, Adam Franco  wrote:
> 
> As someone who has mapped a lot of landcover & landuse 
>  in my local area, I 
> welcome sorting out the confusion that is the current state of 
> natural=wood/landuse=forest. Many parcels around me are managed for forestry 
> purposes but don't have trees currently while others had been cleared at one 
> point, but have returned to forest due to neglect and are not managed for 
> timber production.  My current practice is to map areas covered in trees as 
> landcover=trees + natural=wood, but I'd love to drop the natural=wood if 
> landcover=trees was rendered. Generally, I don't imagine that I'd map much 
> landuse=forestry, which is probably a good thing as I don't often know which 
> land is managed for productive forestry and which is more negligent forest 
> succession. In cases where the management is known and is important to be 
> known, then landuse=forestry becomes a useful tag as it is unambiguous as to 
> what it means.
> 
> I hope that a shift toward landuse=forestry would also include a shift toward 
> landcover=*, in particular landcover=trees as the rightful clear designation 
> that "there are trees here". Here is an old landcover=* proposal 
>  that might 
> be resurrected and updated.
> 
> I'm not sure if I would want landuse=forestry to be rendered by default or if 
> so, how I would like it to be styled. Generally in my region, areas managed 
> for forestry are more parcel boundaries than anything equating the land-cover 
> on the ground, so renderings that include iconography like trees are 
> problematic if those icons overlap and conflict with other land covers. I see 
> landuse=forestry as something more useful for custom maps or maybe something 
> that would be rendered as a subtle modifier to more-visible land-cover 
> renderings which are more directly visible and impactful when traversing the 
> landscape.
> 
> Best,
> Adam

+10 for this!

I also dual tag areas with trees as natural=wood and landcover=trees with the 
hope being that landcover=trees becomes the norm.

Perhaps the way forward would be to change the wiki to indicate that 
landuse=forest is deprecated due to its confused usage. Add some text to the 
page directing mappers to either landcover=trees if they are simply mapping the 
presence of trees or landuse=forestry if they are mapping an area used for the 
production of wood products (lumber, paper, etc.) that may or may not have 
trees on it a the moment.

Not rendering landuse=forestry on the default OSM map to reduce “tagging for 
the renderer” is an interesting idea. I’ll have to think about that but it does 
have some appeal.



signature.asc
Description: Message signed with OpenPGP using GPGMail
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Adam Franco

As someone who has mapped a lot of landcover & landuse
 in my local area,
I welcome sorting out the confusion that is the current state of
natural=wood/landuse=forest. Many parcels around me are managed for
forestry purposes but don't have trees currently while others had been
cleared at one point, but have returned to forest due to neglect and are
not managed for timber production.  My current practice is to map areas
covered in trees as landcover=trees + natural=wood, but I'd love to drop
the natural=wood if landcover=trees was rendered. Generally, I don't
imagine that I'd map much landuse=forestry, which is probably a good thing
as I don't often know which land is managed for productive forestry and
which is more negligent forest succession. In cases where the management is
known and is important to be known, then landuse=forestry becomes a useful
tag as it is unambiguous as to what it means.

I hope that a shift toward landuse=forestry would also include a shift
toward landcover=*, in particular landcover=trees as the rightful clear
designation that "there are trees here". Here is an old landcover=* proposal
 that
might be resurrected and updated.

I'm not sure if I would want landuse=forestry to be rendered by default or
if so, how I would like it to be styled. Generally in my region, areas
managed for forestry are more parcel boundaries than anything equating the
land-cover on the ground, so renderings that include iconography like trees
are problematic if those icons overlap and conflict with other land covers.
I see landuse=forestry as something more useful for custom maps or maybe
something that would be rendered as a subtle modifier to more-visible
land-cover renderings which are more directly visible and impactful when
traversing the landscape.

Best,
Adam

On Tue, Jan 22, 2019 at 2:39 PM Paul Allen  wrote:

> On Tue, 22 Jan 2019 at 18:14, David Marchal  wrote:
>
> Your landuse=forestry proposal seems good to me: it is clear enough, and
>> the transition process you describe here seems consistent with what I know
>> about such transitions which already happened. If I understand you, the
>> main problem for landuse=forestry is to include it in the standard style to
>> not discourage its use,
>
>
> Yup.  If it rendered, people who read this list would use it.  If enough
> people used it, editors
> would offer it as a preset (for iD somebody would have to raise the issue
> on github since
> Bryan Housel recently announced he was no longer following this list).  A
> couple of vicious
> circles there.
>
> but style devs rejected adding its rendering before its use spread a bit.
>
>
> I don't know if they have rejected this specific idea, or even if they
> were asked.  It's just
> that they often require that a tag has been used sufficiently in the wild
> before they consider
> adding it.
>
> Some sort of vicious circle, in fact?
>>
>
> As I said, two of them.  It won't be widely used until editors offer it as
> a preset and it
> renders.  So we're at an impasse.  A proposal to introduce it that
> suggests dual-tagging
> until it takes off enough for editors and carto to support it seems the
> only way forward -
> not guaranteed to succeed but it might.
>
> I might even write the proposal myself.  But only after I get a feel for
> the mood here.  So
> far nobody has heaped scorn on the idea, which is a good sign, but I'd
> like to see a little
> more support first because if people here don't see it as sensible then
> neither will
> most ordinary mappers.
>
> --
> Paul
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

Jan 21, 2019, 12:03 AM by 61sundow...@gmail.com:

> The end to this madness is for renders to recognise that the landuse=forest 
> needs to be rendered differently from natural=wood.
> The essential difference between the two is that landuse must have some human 
> benefit, a produce, and a clear way of doing that is to add the rendering of 
> a axe to the tree.
>

(1) in a typical rendering this distinction is completely unimportant
or at least not worth different rendering

(2) other people have different mismatching ideas what is the
"real" difference between natural=wood and landuse=forest

(3) there is no consistent difference in how natural=wood and landuse=forest 
are used
by mappers


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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?

2019-01-22 Thread Kevin Kenny

On Tue, Jan 22, 2019 at 2:39 PM Paul Allen  wrote:
> I don't know if they have rejected this specific idea, or even if they were 
> asked.  It's just
> that they often require that a tag has been used sufficiently in the wild 
> before they consider
> adding it.


A commoner objection is that it's redundant with landuse=forest. That
would even be true if anyone tagged landuse=forest to mean what it
says on the box.

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?




Jan 21, 2019, 9:44 PM by pla16...@gmail.com:

> On Mon, 21 Jan 2019 at 20:21, Warin <> 61sundow...@gmail.com 
> > > wrote:
>
>>
>> My problem with going to landuse=forestry with natural=wood...
>>  
>>  what happens to the remaining landuse=forest?
>>  Will that finally be recognised as the same as natural=wood and be migrated 
>> to natural=wood???
>>
>
> Ideally, if we get landuse=forestry and it eventually renders
>
Note that it is not obvious to me that a typical map would want to display 
landuse=forestry.

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

Re: [Tagging] Forest parcel with other landcover (scrub, scree…): how to map?