Re: [OSM-talk] Key:layer update

[Richard Z.]

On Wed, Mar 12, 2014 at 09:06:41PM -0400, Russ Nelson wrote:
> Pieren writes:
>  > On Wed, Mar 12, 2014 at 3:22 PM, Frank Little  wrote:
>  > > Richard Z wrote
>  > 
>  > > As mapped, the waterway=stream (Way #138911739) runs underground 
> (layer=-1),
>  > > probably through a culvert given the way the stream left and right are
>  > > separately outlined as waterway=riverbank (and without layer=*). The way
>  > > (stream) should be tagged as a culvert.
>  > >
>  > > Perhaps there is in reality a bridge not a culvert, in which case the 
> road
>  > > needs splitting and the appropriate new road segment tagged as 
> bridge=yes.
>  > >
>  > > In either case, a layer tag is not needed for rendering.
>  > +1
> 
> Nonetheless I add one out of habit. But I would be happy to stop,
> because as noted, the bridge or culvert carries an implicit layering.
 
With a new type of bridge we could do it. The current state is that if 
there is no layer tag the bridge has a layer=0 which is not what you 
want. The old definition can't be changed because it would affect many
existing crossings.

The implicit layering that you mention is a technical workaround that 
software does to avoid problems with OSM data. Those are not necessarily 
bugs in OSM data but very often it is missing information - someone was 
not sure is there a bridge/culvert or perhaps a ford.
Rendering and other software needs to make a guess in such cases. Although 
it might be more correct to paint a question mark there most renderers 
assume a culvert and implicit layering in such cases.

It isn't good to rely on that for technical reasons, many alternative 
renderers are much less fault tolerant than Mapnik and you will get 
strange results.

Also the validators need to improve their error checking to catch accidental
errors and this is nearly impossible until people agree on how to use layer
(and level and other similar tags).


Richard

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

Re: [OSM-talk] Key:layer update

[Tobias Knerr]

On 13.03.2014 00:28, Frank Little wrote:
> Tobias Knerr wrote:
>> http://wiki.openstreetmap.org/wiki/Proposed_features/default_layer_for_bridge_and_tunnel
>>
[...]
> This is however not the "implicit layer tags" argument which Richard Z.
> gives, which he suggests means:. "so every tunnel and bridge should now
> have one".

That's exactly what was discussed back then. Yes, a mapper who is
generally critical of mechanical edits additionally pointed out that
such edits might result from a successful vote, but that doesn't change
the proposal itself.

> We model bridges and tunnels in a specific way in OSM which
> means we do not need to add additional layer tags (but feel free if you
> do want to).

Well, according to the result of the vote, properly mapped
bridge/tunnels do need layer tags, and that's what the Key:layer page
should document today.

I don't see a reason to start the discussion again (especially in this
thread), but should you manage to gather broad support for changing the
definition, then the Key:layer page may eventually reflect that.

> I have always understood that we do not use a layer tag for
> road/waterway crossings, as osm help also suggests here:
> https://help.openstreetmap.org/questions/3049/how-to-correct-waterways-intersecting-roads-without-a-junction-node

That's not what is being suggested here. The option b) in the question
was to just add layer=-1 to large sections of a waterway without
surveying what actually happens in those places. Which is clearly not a
good solution.

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

Re: [OSM-talk] Key:layer update

[Martin Koppenhoefer]

> Am 13/mar/2014 um 00:28 schrieb "Frank Little" :
> 
> We model bridges and tunnels in a specific way in OSM which means we do not 
> need to add additional layer tags (but feel free if you do want to).


IMHO we do not yet model explicit bridges or tunnels at all, we only add 
attributes to roads or other objects that they are on a bridge or in a tunnel, 
but we omit the actual tunnel and bridge objects (besides some few relations 
that are modeled). 

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

Re: [OSM-talk] Key:layer update

[Russ Nelson]

Pieren writes:
 > On Wed, Mar 12, 2014 at 3:22 PM, Frank Little  wrote:
 > > Richard Z wrote
 > 
 > > As mapped, the waterway=stream (Way #138911739) runs underground 
 > > (layer=-1),
 > > probably through a culvert given the way the stream left and right are
 > > separately outlined as waterway=riverbank (and without layer=*). The way
 > > (stream) should be tagged as a culvert.
 > >
 > > Perhaps there is in reality a bridge not a culvert, in which case the road
 > > needs splitting and the appropriate new road segment tagged as bridge=yes.
 > >
 > > In either case, a layer tag is not needed for rendering.
 > +1

Nonetheless I add one out of habit. But I would be happy to stop,
because as noted, the bridge or culvert carries an implicit layering.

 > Bridge or culvert, one way has to be split (the road for bridge or the
 > stream for culvert)

My rule of thumb here is how it looks from above: if the people on the
top would perceive it as a bridge, I split that way and mark is as a
bridge. Otherwise I split the waterway and mark it as a culvert.

-- 
--my blog is athttp://blog.russnelson.com
Crynwr supports open source software
521 Pleasant Valley Rd. | +1 315-600-8815
Potsdam, NY 13676-3213  | Sheepdog   

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

Re: [OSM-talk] Key:layer update

[Frank Little]

Tobias Knerr wrote:

On 12.03.2014 16:53, Frank Little wrote:

But when did this happen as reported by Richard Z.:
QUOTE
The initiative to have implicit layer tags for those feature was voted
down so every tunnel and bridge should
now have one.
UNQUOTE


There was a proposal to set default layer=1 for bridge=yes and layer=-1
for tunnel=yes. It was rejected by a clear majority of those who voted:

http://wiki.openstreetmap.org/wiki/Proposed_features/default_layer_for_bridge_and_tunnel

My thanks. I'd agree with the comment which Frederik Ramm made (while 
abstaining) on that proposal: If the idea was to automatically add a layer tag 
to bridges and tunnels (or to remove it if present), I would oppose it (for 
reasons he gave).


This is however not the "implicit layer tags" argument which Richard Z. gives, 
which he suggests means:. "so every tunnel and bridge should now have one". We 
model bridges and tunnels in a specific way in OSM which means we do not need 
to add additional layer tags (but feel free if you do want to).


I have always understood that we do not use a layer tag for road/waterway 
crossings, as osm help also suggests here: 
https://help.openstreetmap.org/questions/3049/how-to-correct-waterways-intersecting-roads-without-a-junction-node



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

Re: [OSM-talk] Key:layer update

[Tobias Knerr]

On 12.03.2014 16:53, Frank Little wrote:
> But when did this happen as reported by Richard Z.:
> QUOTE
> The initiative to have implicit layer tags for those feature was voted
> down so every tunnel and bridge should
> now have one.
> UNQUOTE

There was a proposal to set default layer=1 for bridge=yes and layer=-1
for tunnel=yes. It was rejected by a clear majority of those who voted:

http://wiki.openstreetmap.org/wiki/Proposed_features/default_layer_for_bridge_and_tunnel

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

Re: [OSM-talk] Key:layer update

[Frank Little]

Pieren wrote:

On Wed, Mar 12, 2014 at 3:22 PM, Frank Little  wrote:

Richard Z wrote


As mapped, the waterway=stream (Way #138911739) runs underground 
(layer=-1),

probably through a culvert given the way the stream left and right are
separately outlined as waterway=riverbank (and without layer=*). The way
(stream) should be tagged as a culvert.

Perhaps there is in reality a bridge not a culvert, in which case the road
needs splitting and the appropriate new road segment tagged as bridge=yes.

In either case, a layer tag is not needed for rendering.

+1
Bridge or culvert, one way has to be split (the road for bridge or the
stream for culvert) to identify the structure and one of the ways
needs a tag layer. It's true that the shared node is wrong here
excepted if it's a ford. It needs a local survey.
Same mapping issues can be found between roads and railways, railways
and rivers, etc

Pieren
Unless it is a level crossing of road/rail, of course (in which case a tag is 
needed on the common node, just as with fords).


But when did this happen as reported by Richard Z.:
QUOTE
The initiative to have implicit layer tags for those feature was voted down so 
every tunnel and bridge should

now have one.
UNQUOTE 



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

Re: [OSM-talk] Key:layer update

[Pieren]

On Wed, Mar 12, 2014 at 3:22 PM, Frank Little  wrote:
> Richard Z wrote

> As mapped, the waterway=stream (Way #138911739) runs underground (layer=-1),
> probably through a culvert given the way the stream left and right are
> separately outlined as waterway=riverbank (and without layer=*). The way
> (stream) should be tagged as a culvert.
>
> Perhaps there is in reality a bridge not a culvert, in which case the road
> needs splitting and the appropriate new road segment tagged as bridge=yes.
>
> In either case, a layer tag is not needed for rendering.
+1
Bridge or culvert, one way has to be split (the road for bridge or the
stream for culvert) to identify the structure and one of the ways
needs a tag layer. It's true that the shared node is wrong here
excepted if it's a ford. It needs a local survey.
Same mapping issues can be found between roads and railways, railways
and rivers, etc

Pieren

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

Re: [OSM-talk] Key:layer update

[Frank Little]

Richard Z wrote



>In practice this rule is broken more often than you would think: Hamburg
>is full
>of waterways connected with roads on bridges through a tag obstacle.
>France is
>full of bridges sharing a node with the waterway bellow.


...

http://www.openstreetmap.org/node/1522876252



I would say this is an example of incorrect mapping and I would correct it if
I came across it on survey. It is either a culvert or a bridge, or a ford. A
layer tag should not be necessary in any of these three cases.

As mapped, the waterway=stream (Way #138911739) runs underground (layer=-1),
probably through a culvert given the way the stream left and right are
separately outlined as waterway=riverbank (and without layer=*). The way
(stream) should be tagged as a culvert.

Perhaps there is in reality a bridge not a culvert, in which case the road
needs splitting and the appropriate new road segment tagged as bridge=yes.

In either case, a layer tag is not needed for rendering.

Given the layer tag, it is unlikely that the mapper thought it was a ford. (I
don't know whether the cadastre data distinguishes between these features).
Fords do not need layer tag, by definition.

The solution will be to find and remap such examples.


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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Tue, Mar 11, 2014 at 02:52:02PM +, Dave F. wrote:
> On 09/03/2014 14:45, Martin Koppenhoefer wrote:

> >+1, but adding a layer=1 to a lake in a park isn't clearer or more
> >accurate, they are both on the same layer, the lake is in the
> >park, not above (usually).
> 
> Which confirms my point perfectly. You're are correct: The lake &
> park /are/ at the same level, which is why the layer tag is needed.
> It's used purely to let the renderer know which entity to put on top
> of the pile show it display properly.

if a layer tag is needed to display a lake in a park then you have some
other problem. Show us an example.

natural=water + layer has no meaning unless in combination with tunnel,
bridge or similar.

Richard

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

Re: [OSM-talk] Key:layer update

[moltonel 3x Combo]

On 11/03/2014, Dave F.  wrote:
> On 09/03/2014 14:45, Martin Koppenhoefer wrote:
>> For instance if you had an area tagged 'park' & another area
>> within it tagged 'lake' you could add a 'layer' tag to 'lake'
>> to ensure the render displayed it.
[...]
> Which confirms my point perfectly. You're are correct: The lake & park
> /are/ at the same level, which is why the layer tag is needed. It's used
> purely to let the renderer know which entity to put on top of the pile
> show it display properly.

Go fix your rendering stylesheet if you feel it should display lakes
on top of parks but doesn't. Stylesheets make this type of decision
for plenty of objects without needing a layer tag. The decision to
render a lake on top of a park isn't a universal one.

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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 10:26:59PM +, Matthijs Melissen wrote:
> On 9 March 2014 10:30, Richard Z.  wrote:
> > for some time now I have been working on the wiki page to state the rules
> > as clearly as possible.. hope that most of the improvements are fairly
> > uncontroversial.
> 
> Thank you for doing this, it's very useful to have this properly
> documented. I have been working on layering in the main CartoCSS
> stylesheet, and found that at the moment, indeed not all aspects of
> the layering model are defined precise enough.

just remembered, there is also a rather special rule for 
  tunnel=building_passage
"The layer has to be the same as the building, with the above mentioned
 exception when several tunnels are passing on different levels. "
https://wiki.openstreetmap.org/wiki/Key:tunnel#tunnel.3Dbuilding_passage

It makes sense but is so much different from normal tunnels that I am
wondering if it should not have been done with a different tag.

Richard


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

Re: [OSM-talk] Key:layer update

[Robert Kaiser]

Richard Z. schrieb:

On Sun, Mar 09, 2014 at 12:34:31PM +, Dave F. wrote:

On 09/03/2014 12:24, Richard Z. wrote:


it says "point", not "node" the difference probably needs to be emphasized
very strongly. There is a difference between mathematicaly precise and
intuitive formulations:((

https://www.google.co.uk/#q=node%20definition

"a point in a network or diagram at which lines or pathways
intersect or branch"


changed to "precise location" instead of point.. is that better?



+1 for "location", that's really hard to misinterpret but is clear on 
what it means.


Robert Kaiser


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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Tue, Mar 11, 2014 at 02:51:23PM +, Dave F. wrote:
> On 09/03/2014 12:21, Richard Z. wrote:

> >In practice this rule is broken more often than you would think: Hamburg is 
> >full
> >of waterways connected with roads on bridges through a tag obstacle. France 
> >is
> >full of bridges sharing a node with the waterway bellow.
> 
> Could you link to an example please?

http://www.openstreetmap.org/node/1522876252

stupid question - suppose I have this node selected in JOSM - what is
the quickest way of getting an URL like the above?

Richard

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

Re: [OSM-talk] Key:layer update

[Martin Koppenhoefer]

2014-03-11 15:52 GMT+01:00 Dave F. :

> I agree it does say which object is above another. For the benefit of the
> renderer. Who else needs that data? Certainly not routers.
>


well, everybody who tries to understand what this specific part of the map
represents will have to have this information in the case of objects that
overlap in 2D.




>
>
>
>>  For instance if you had an area tagged 'park' & another area within it
>>> tagged 'lake' you could add a 'layer' tag to 'lake' to ensure the render
>>> displayed it.
>>>
>> -1, you should not add any layer in this case (tagging for the renderer)
>>
>
>  The correct expression is 'don't tag incorrectly for the renderer'.
>> There's /nothing/ wrong in making OSM data clearer & more accurate.
>
>
>
>  +1, but adding a layer=1 to a lake in a park isn't clearer or more
> accurate, they are both on the same layer, the lake is in the park, not
> above (usually).
>
>
> Which confirms my point perfectly. You're are correct: The lake & park
> /are/ at the same level, which is why the layer tag is needed. It's used
> purely to let the renderer know which entity to put on top of the pile show
> it display properly.
>


no, it would be wrong to use the layer tag here, as it would move the lake
out of the park and above. OK, this sounds unprobable to a human, and he
might still understand what was the intention (by interpretation and common
sense), but the modelling remains "wrong" (IMHO).

Btw.: a lake is a physical object, while a park is an abstract object, so
they aren't on the same level anyway (but on the same layer) ;-)

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

Re: [OSM-talk] Key:layer update

[Dave F.]

On 09/03/2014 14:45, Martin Koppenhoefer wrote:


2014-03-09 13:17 GMT+01:00 Dave F. >:


it says which object is above and which below when they cross,
it is not only for rendering



I agree it does say which object is above another. For the benefit of 
the renderer. Who else needs that data? Certainly not routers.





For instance if you had an area tagged 'park' & another area
within it tagged 'lake' you could add a 'layer' tag to 'lake'
to ensure the render displayed it.

-1, you should not add any layer in this case (tagging for the
renderer)


The correct expression is 'don't tag incorrectly for the
renderer'. There's /nothing/ wrong in making OSM data clearer &
more accurate.



+1, but adding a layer=1 to a lake in a park isn't clearer or more 
accurate, they are both on the same layer, the lake is in the park, 
not above (usually).


Which confirms my point perfectly. You're are correct: The lake & park 
/are/ at the same level, which is why the layer tag is needed. It's used 
purely to let the renderer know which entity to put on top of the pile 
show it display properly.


Dave F.


---
This email is free from viruses and malware because avast! Antivirus protection 
is active.
http://www.avast.com
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Key:layer update

[Dave F.]

On 09/03/2014 12:21, Richard Z. wrote:


the same conceptual problem exists with pylons where they are shared by two 
bridges
or aerial tramways. Actualy every pylon breaks the rule by definition because it
connects "ground" with layer=0 with something else at a different level.
How do you want to model such cases better? Lifts in buildings?

In practice this rule is broken more often than you would think: Hamburg is full
of waterways connected with roads on bridges through a tag obstacle. France is
full of bridges sharing a node with the waterway bellow.


Could you link to an example please?

Dave F.

---
This email is free from viruses and malware because avast! Antivirus protection 
is active.
http://www.avast.com


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

Re: [OSM-talk] Key:layer update

[SomeoneElse]

Dave F. wrote:
Could you give some visual examples, maybe temporarily creating them 
in OSM (& deleting them afterwards)


... or on the dev server:

http://api06.dev.openstreetmap.org/

Cheers,

Andy


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

Re: [OSM-talk] Key:layer update

[Dave F.]

Could you give some visual examples, maybe temporarily creating them in 
OSM (& deleting them afterwards) to clarify what you mean?


Dave F.


On 09/03/2014 22:26, Matthijs Melissen wrote:

On 9 March 2014 10:30, Richard Z.  wrote:

for some time now I have been working on the wiki page to state the rules
as clearly as possible.. hope that most of the improvements are fairly
uncontroversial.

Thank you for doing this, it's very useful to have this properly
documented. I have been working on layering in the main CartoCSS
stylesheet, and found that at the moment, indeed not all aspects of
the layering model are defined precise enough.

A question: a single road can contain sections on multiple layers, so
there will be a point where the sections that are on different layers
meet. At that point, there might even be a side street. However, no
vertical ordering should be assumed at such a point. It is written
that "The vertical ordering established by the layer values is valid
exactly only in the point where the ways cross or objects overlap."
Perhaps 'crossing' should be interpreted here as crossing without
node, but that causes problems with bridge/waterway.

In other words, I am wondering for each of the following situations if
the roads should be interpreted as meeting on the same or different
levels:
- A node where two waterways on layer 1 and two roads on layer 2 meet;
- A node where two roads on layer 1 and two roads on layer 2 meet;
- A node where two roads on layer 1 and one road on layer 2 meet;
- A node where one road on layer 1 and one road on layer 2 meet.

Perhaps some text that answers questions like this should be added to
the Wiki-page.

-- Matthijs

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




---
This email is free from viruses and malware because avast! Antivirus protection 
is active.
http://www.avast.com


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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 10:26:59PM +, Matthijs Melissen wrote:
> On 9 March 2014 10:30, Richard Z.  wrote:
> > for some time now I have been working on the wiki page to state the rules
> > as clearly as possible.. hope that most of the improvements are fairly
> > uncontroversial.
> 
> Thank you for doing this, it's very useful to have this properly
> documented. I have been working on layering in the main CartoCSS
> stylesheet, and found that at the moment, indeed not all aspects of
> the layering model are defined precise enough.
> 
> A question: a single road can contain sections on multiple layers, so
> there will be a point where the sections that are on different layers
> meet. At that point, there might even be a side street. However, no
> vertical ordering should be assumed at such a point. It is written
> that "The vertical ordering established by the layer values is valid
> exactly only in the point where the ways cross or objects overlap."
> Perhaps 'crossing' should be interpreted here as crossing without
> node, but that causes problems with bridge/waterway.

it should be indeed crossing without a shared node, have already updated 
the text to clarify that. 

> In other words, I am wondering for each of the following situations if
> the roads should be interpreted as meeting on the same or different
> levels:
> - A node where two waterways on layer 1 and two roads on layer 2 meet;

the roads should join at the same physical level in this node. Except the
node is a pylon connecting two bridges on two levels and a waterway or 
a similarly weird exception which is not described in the wiki but happens 
in real life and probably somewhere in OSM data as well.
Generally, if the node does not have a special type (lift, pylon, part of 
buildings with level) the roads should join as expected.

Nothing is certain about the waterway unless the node is of type ford
or pylon, or the layering is otherwise obvious such as when the road is
on a dam or weir.
Exceptions and errors in data are currently very common where waterways 
and roads have shared nodes.
Once I have mapped a weir with a highway ford on top it and part of the 
water going through a pipe through the weir.

Conceptually I am thinking of it so that certain constructions such as
a dam establish a "connection" in the sense that both the road and the
river are passing over/through it and hence are connected to the dam 
without really meeting in this point - the dam establishes its own layering
rules.

> - A node where two roads on layer 1 and two roads on layer 2 meet;
> - A node where two roads on layer 1 and one road on layer 2 meet;
> - A node where one road on layer 1 and one road on layer 2 meet.

they should all join without steps and exceptions should be extremely
rare.(maybe lifts and such)
More precisely "layer" does not say anything in this situation so the 
default rule applies - roads are expected to join without steps.

It is important to understand that the meaning of "layer" is very limited:

- it applies exactly only in the point (without shared node) of the crossing
  and has absolutely no meaning anywhere else
- it has absolutely no defined meaning if not in combination with one of bridge,
  tunnel and the other tags listed in the wiki (I may have forgotten some but 
you
  get an idea)
- a number of other tags (covered, location, level, dam and probably some 
other) 
  define own layering concepts or modify layer in strange ways

Richard

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

Re: [OSM-talk] Key:layer update

[Matthijs Melissen]

On 9 March 2014 10:30, Richard Z.  wrote:
> for some time now I have been working on the wiki page to state the rules
> as clearly as possible.. hope that most of the improvements are fairly
> uncontroversial.

Thank you for doing this, it's very useful to have this properly
documented. I have been working on layering in the main CartoCSS
stylesheet, and found that at the moment, indeed not all aspects of
the layering model are defined precise enough.

A question: a single road can contain sections on multiple layers, so
there will be a point where the sections that are on different layers
meet. At that point, there might even be a side street. However, no
vertical ordering should be assumed at such a point. It is written
that "The vertical ordering established by the layer values is valid
exactly only in the point where the ways cross or objects overlap."
Perhaps 'crossing' should be interpreted here as crossing without
node, but that causes problems with bridge/waterway.

In other words, I am wondering for each of the following situations if
the roads should be interpreted as meeting on the same or different
levels:
- A node where two waterways on layer 1 and two roads on layer 2 meet;
- A node where two roads on layer 1 and two roads on layer 2 meet;
- A node where two roads on layer 1 and one road on layer 2 meet;
- A node where one road on layer 1 and one road on layer 2 meet.

Perhaps some text that answers questions like this should be added to
the Wiki-page.

-- Matthijs

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

Re: [OSM-talk] Key:layer update

[moltonel 3x Combo]

On 09/03/2014, Martin Koppenhoefer  wrote:
  For instance if you had an area tagged 'park' & another area within it
 tagged 'lake' you could add a 'layer' tag to 'lake' to ensure the render
 displayed it.
>>>
>>> -1, you should not add any layer in this case (tagging for the renderer)
>>
>> The correct expression is 'don't tag incorrectly for the renderer'. There's
>> /nothing/ wrong in making OSM data clearer & more accurate.
>
> +1, but adding a layer=1 to a lake in a park isn't clearer or more
> accurate, they are both on the same layer, the lake is in the park, not
> above (usually).

On top of tagging different layers for a park and its lake not being
"clearer and more accurate", it really is a renderer decision wether
to render a park on top of a lake or vice-versa. For example if your
rendering of "leisure=park" is a very transparent green area, then you
do want to render it on top of lakes and all other features. So please
do not use the layer tag for that purpose.

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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 04:55:51PM +0100, Tobias Knerr wrote:
> On 09.03.2014 14:18, Richard Z. wrote:
> > Pylons must share a node with the waterway bellow
> > in my opinion. They are a pretty relevant part of it.
> 
> Pylons will often be somewhere within the riverbank area - based on
> their exact positions in reality -, but I would not insert them into the
> waterway way.

somehow they ought to be connected to the river though, just beeing in the 
area is not enough. As they are relevant for navigation there can be situations 
where inserting them into the waterway way would be indeed the most logical
solution.

> What do you do if one pylon is left of the center of the waterway and
> one is right of it?

interesting question.. I will look again at the examples and ask the
author.

> > then there is the alternative of having a list of exceptions.
> 
> That's more reasonable, but exceptions should only be made where it is
> really necessary. I haven't encountered such an example yet.

Having seen a few examples of vertical lifts I am sure there will be 
more exceptions.


Richard

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

Re: [OSM-talk] Key:layer update

[Tobias Knerr]

On 09.03.2014 14:18, Richard Z. wrote:
> Pylons must share a node with the waterway bellow
> in my opinion. They are a pretty relevant part of it.

Pylons will often be somewhere within the riverbank area - based on
their exact positions in reality -, but I would not insert them into the
waterway way.

What do you do if one pylon is left of the center of the waterway and
one is right of it?

> then there is the alternative of having a list of exceptions.

That's more reasonable, but exceptions should only be made where it is
really necessary. I haven't encountered such an example yet.

Tobias

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

Re: [OSM-talk] Key:layer update

[Martin Koppenhoefer]

2014-03-09 13:17 GMT+01:00 Dave F. :

> it has implications on real world topology: it says which object is above
>> and which below when they cross, it is not only for rendering
>>
>
> Disagree. A bridge at the top of Everest would be tagged 'layer=1' exactly
> as it would be if in Death Valley (86m below sea level)



to me it seems your example does agree with my statement



>
>  For instance if you had an area tagged 'park' & another area within it
>> tagged 'lake' you could add a 'layer' tag to 'lake' to ensure the render
>> displayed it.
>>
> -1, you should not add any layer in this case (tagging for the renderer)
>

The correct expression is 'don't tag incorrectly for the renderer'. There's
> /nothing/ wrong in making OSM data clearer & more accurate.



+1, but adding a layer=1 to a lake in a park isn't clearer or more
accurate, they are both on the same layer, the lake is in the park, not
above (usually).

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

Re: [OSM-talk] Key:layer update

[Colin Smale]

 

Not all OSM nodes are also network/diagram nodes, which are points with
(AFAIK) three or more lines in common. Intermediate OSM nodes in the
middle of a way are not topologically significant. 

On 2014-03-09 14:00, Richard Z. wrote: 

> On Sun, Mar 09, 2014 at 12:34:31PM +, Dave F. wrote:
> On 09/03/2014 12:24, Richard Z. wrote: it says "point", not "node" the 
> difference probably needs to be emphasized very strongly. There is a 
> difference between mathematicaly precise and intuitive formulations:(( 
> https://www.google.co.uk/#q=node%20definition [1] "a point in a network or 
> diagram at which lines or pathways intersect or branch"

in OSM this is called "node". Better suggestions instead of "point"?

Richard

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk [2]

 

Links:
--
[1] https://www.google.co.uk/#q=node%20definition
[2] https://lists.openstreetmap.org/listinfo/talk
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 02:00:36PM +0100, Tobias Knerr wrote:
> On 09.03.2014 13:21, Richard Z. wrote:
> > the same conceptual problem exists with pylons where they are shared by two 
> > bridges
> > or aerial tramways. Actualy every pylon breaks the rule by definition 
> > because it 
> > connects "ground" with layer=0 with something else at a different level.
> > How do you want to model such cases better? Lifts in buildings?
> 
> Typical pylons aren't a problem because the "ground" is not an OSM
> element that they could share a node with. Pylons shared between more
> than one bridge are indeed an interesting problem for 3D mapping, but
> I'm not aware that this is commonly mapped or used by applications yet,
> so there is still some room for establishing good standard practice.
> 
> Lifts in buildings don't use layer, they use level. That tag follows
> different rules than layer.

I would be in favor of using level more widely but the rules are not so
much different because you can also have all kinds of highways and railways
on levels.

> > In practice this rule is broken more often than you would think: Hamburg is 
> > full
> > of waterways connected with roads on bridges through a tag obstacle. France 
> > is 
> > full of bridges sharing a node with the waterway bellow.
> 
> I would prefer correcting these errors instead of changing the rule they
> break.

are those really errors? Pylons must share a node with the waterway bellow
in my opinion. They are a pretty relevant part of it.

> > It may be worth to tag have such a rule restricted for "ways of the same 
> > type"
> > and a short well defined list of exceptions.
> 
> The rule is also needed for ways of different types, e.g. for ordering a
> stack of road, railway, and waterway bridges.

then there is the alternative of having a list of exceptions.

Richard


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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 12:34:31PM +, Dave F. wrote:
> On 09/03/2014 12:24, Richard Z. wrote:
> >
> >it says "point", not "node" the difference probably needs to be emphasized
> >very strongly. There is a difference between mathematicaly precise and
> >intuitive formulations:((
> https://www.google.co.uk/#q=node%20definition
> 
> "a point in a network or diagram at which lines or pathways
> intersect or branch"

changed to "precise location" instead of point.. is that better?

Also listed teh pylon as exception.

Richard

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

Re: [OSM-talk] Key:layer update

[Tobias Knerr]

On 09.03.2014 13:21, Richard Z. wrote:
> the same conceptual problem exists with pylons where they are shared by two 
> bridges
> or aerial tramways. Actualy every pylon breaks the rule by definition because 
> it 
> connects "ground" with layer=0 with something else at a different level.
> How do you want to model such cases better? Lifts in buildings?

Typical pylons aren't a problem because the "ground" is not an OSM
element that they could share a node with. Pylons shared between more
than one bridge are indeed an interesting problem for 3D mapping, but
I'm not aware that this is commonly mapped or used by applications yet,
so there is still some room for establishing good standard practice.

Lifts in buildings don't use layer, they use level. That tag follows
different rules than layer.

> In practice this rule is broken more often than you would think: Hamburg is 
> full
> of waterways connected with roads on bridges through a tag obstacle. France 
> is 
> full of bridges sharing a node with the waterway bellow.

I would prefer correcting these errors instead of changing the rule they
break.

> It may be worth to tag have such a rule restricted for "ways of the same type"
> and a short well defined list of exceptions.

The rule is also needed for ways of different types, e.g. for ordering a
stack of road, railway, and waterway bridges.

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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 12:34:31PM +, Dave F. wrote:
> On 09/03/2014 12:24, Richard Z. wrote:
> >
> >it says "point", not "node" the difference probably needs to be emphasized
> >very strongly. There is a difference between mathematicaly precise and
> >intuitive formulations:((
> https://www.google.co.uk/#q=node%20definition
> 
> "a point in a network or diagram at which lines or pathways
> intersect or branch"

in OSM this is called "node". Better suggestions instead of "point"?

Richard

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

Re: [OSM-talk] Key:layer update

[Tobias Knerr]

On 09.03.2014 13:17, Dave F. wrote:
> On 09/03/2014 12:03, Martin Koppenhoefer wrote:
>> it has implications on real world topology: it says which object is
>> above and which below when they cross, it is not only for rendering
> 
> Disagree. A bridge at the top of Everest would be tagged 'layer=1'
> exactly as it would be if in Death Valley (86m below sea level)

True, but that because layer models *relative* real-world elevation
*locally* (between vertically stacked objects), rather than globally.

Your statement points out that the layer relationship is neither global
nor representative of absolute elevation, but that was never contested.

Tobias

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

Re: [OSM-talk] Key:layer update

[Dave F.]

On 09/03/2014 12:24, Richard Z. wrote:


it says "point", not "node" the difference probably needs to be emphasized
very strongly. There is a difference between mathematicaly precise and
intuitive formulations:((

https://www.google.co.uk/#q=node%20definition

"a point in a network or diagram at which lines or pathways intersect or 
branch"


Dave F.

---
This email is free from viruses and malware because avast! Antivirus protection 
is active.
http://www.avast.com


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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 11:35:20AM +, Dave F. wrote:
> On 09/03/2014 10:30, Richard Z. wrote:
> >Hi,
> >
> >for some time now I have been working on the wiki page to state the rules
> >as clearly as possible.. hope that most of the improvements are fairly
> >uncontroversial. Some of the changes:
> >
> >* the vertical ordering established by the layer values is valid exactly only
> >   in the point where the ways cross or objects overlap
> 
> If you take that literally users will join rivers flowing under a
> bridge with a node & add the layer tag to it, which is incorrect:
> those ways should not join.


changed the intro text to say
<<
The layer=* tag is used to mark vertical relationships between crossing or 
overlapping features. The vertical ordering established by the layer values is 
valid exactly only in the point (not node!!!) where the ways cross or objects 
overlap. Joining the ways with a common node at the point where they cross 
would destroy the vertical order established by layer. The layer=* is not 
suitable to define vertical relationships of adjoining or nearby elements or 
areas. 
>>


Richard

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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 11:35:20AM +, Dave F. wrote:
> On 09/03/2014 10:30, Richard Z. wrote:
> >Hi,
> >
> >for some time now I have been working on the wiki page to state the rules
> >as clearly as possible.. hope that most of the improvements are fairly
> >uncontroversial. Some of the changes:
> >
> >* the vertical ordering established by the layer values is valid exactly only
> >   in the point where the ways cross or objects overlap
> 
> If you take that literally users will join rivers flowing under a
> bridge with a node & add the layer tag to it, which is incorrect:
> those ways should not join.

it says "point", not "node" the difference probably needs to be emphasized 
very strongly. There is a difference between mathematicaly precise and
intuitive formulations:((


Richard

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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 01:05:18PM +0100, Martin Koppenhoefer wrote:
> 
> 
> > Am 09/mar/2014 um 12:43 schrieb "Richard Z." :
> > 
> > It is broken by definition in at least one case: waterways ar supposed to 
> > share a node with the dam they are crossing, which means the highway 
> > passing 
> > across the dam will also share a node with the river passing thorugh a 
> > tunnel
> > or pipeline bellow it.
> 
> 
> -1, it is a modeling problem/error, the highway should not have a common node 
> with the waterway, if it has, it is wrong or should be tagged ford=yes ;-)

the same conceptual problem exists with pylons where they are shared by two 
bridges
or aerial tramways. Actualy every pylon breaks the rule by definition because 
it 
connects "ground" with layer=0 with something else at a different level.
How do you want to model such cases better? Lifts in buildings?

In practice this rule is broken more often than you would think: Hamburg is full
of waterways connected with roads on bridges through a tag obstacle. France is 
full of bridges sharing a node with the waterway bellow.

It may be worth to tag have such a rule restricted for "ways of the same type"
and a short well defined list of exceptions.

Richard

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

Re: [OSM-talk] Key:layer update

[Dave F.]

On 09/03/2014 12:03, Martin Koppenhoefer wrote:



Am 09/mar/2014 um 12:35 schrieb "Dave F." :

This is not my understanding of the layer tag. It is a tool to help renderers place 
objects on top of each other & has no real world implication in differences of 
height.


it has implications on real world topology: it says which object is above and 
which below when they cross, it is not only for rendering


Disagree. A bridge at the top of Everest would be tagged 'layer=1' 
exactly as it would be if in Death Valley (86m below sea level)





For instance if you had an area tagged 'park' & another area within it tagged 
'lake' you could add a 'layer' tag to 'lake' to ensure the render displayed it.

-1, you should not add any layer in this case (tagging for the renderer)


The correct expression is 'don't tag incorrectly for the renderer'. 
There's /nothing/ wrong in making OSM data clearer & more accurate.


Dave F.


---
This email is free from viruses and malware because avast! Antivirus protection 
is active.
http://www.avast.com


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

Re: [OSM-talk] Key:layer update

[Martin Koppenhoefer]

> Am 09/mar/2014 um 12:43 schrieb "Richard Z." :
> 
> It is broken by definition in at least one case: waterways ar supposed to 
> share a node with the dam they are crossing, which means the highway passing 
> across the dam will also share a node with the river passing thorugh a tunnel
> or pipeline bellow it.


-1, it is a modeling problem/error, the highway should not have a common node 
with the waterway, if it has, it is wrong or should be tagged ford=yes ;-)

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

Re: [OSM-talk] Key:layer update

[Martin Koppenhoefer]

> Am 09/mar/2014 um 12:35 schrieb "Dave F." :
> 
> This is not my understanding of the layer tag. It is a tool to help renderers 
> place objects on top of each other & has no real world implication in 
> differences of height.
> 



it has implications on real world topology: it says which object is above and 
which below when they cross, it is not only for rendering


> For instance if you had an area tagged 'park' & another area within it tagged 
> 'lake' you could add a 'layer' tag to 'lake' to ensure the render displayed 
> it.


-1, you should not add any layer in this case (tagging for the renderer)

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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 11:47:32AM +0100, Martin Koppenhoefer wrote:
> 
> 
> > Am 09/mar/2014 um 11:30 schrieb "Richard Z." :
> > 
> > * the vertical ordering established by the layer values is valid exactly 
> > only 
> >  in the point where the ways cross or objects overlap
> 
> 
> actually at the Point where a layer Way Connects to an layer=0 way both are 
> at the same height (e.g. where a bridge starts)


currently the assumption that everything meeting in a node is physically at the 
same 
elevation in this point is not valid in OSM.
It is broken by definition in at least one case: waterways ar supposed to 
share a node with the dam they are crossing, which means the highway passing 
across the dam will also share a node with the river passing thorugh a tunnel
or pipeline bellow it.

Some objects (such as dam, buildings) have the property to define their own 
physical level relations.


Richard

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

Re: [OSM-talk] Key:layer update

[Dave F.]

On 09/03/2014 10:30, Richard Z. wrote:

Hi,

for some time now I have been working on the wiki page to state the rules
as clearly as possible.. hope that most of the improvements are fairly
uncontroversial. Some of the changes:

* the vertical ordering established by the layer values is valid exactly only
   in the point where the ways cross or objects overlap


If you take that literally users will join rivers flowing under a bridge 
with a node & add the layer tag to it, which is incorrect: those ways 
should not join.



* define layer as higher value means above, lower value means bellow. Avoid
   the complicated layer=0 definition as "the natural ground level as it
   would be shown by contour lines on a topographic map". Explicit layer=0
   seems to be deprecated now.

* layer on ways should be used only in combination with one of tunnel=*, 
bridge=*,
   highway=steps, highway=elevator, covered=* or indoor=yes. For areas, it could
   be used in combination with tags such as man_made=bridge, building=* and 
similar.
   The motivation for this is to make it easy for validators to spot errors such
   as when the wrong segment is accidentaly tagged, bridge/tunnel forgotten, or
   someone tags excessively long ways for no good reason - common problem with
   waterways and elevated roads/railroads.
   I have validated this rule for ways in large parts of the world, there are
   exceptions which currently I do not know hot to tag better but those are 
rare.


This is not my understanding of the layer tag. It is a tool to help 
renderers place objects on top of each other & has no real world 
implication in differences of height.


For instance if you had an area tagged 'park' & another area within it 
tagged 'lake' you could add a 'layer' tag to 'lake' to ensure the render 
displayed it. Using multi-polygons is not the solution as it would take 
the lake /outside/ of the park, so if the renderer didn't want to render 
it's internal details (playground, wood buildings etc) it would end up 
with more holes in it than Swiss cheese.


Dave F.

---
This email is free from viruses and malware because avast! Antivirus protection 
is active.
http://www.avast.com


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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 10:58:35AM +, Dave F. wrote:
> Are you aware there's a separate wiki page for 'Layer'? No wonder
> users get confused!

yes, this is a problem. There are other meanings of layer so this
page probably has to stay at least as a disambiguator.

But yes, I am very much in favor of avoiding as much of the duplication
as possible.

Richard

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

Re: [OSM-talk] Key:layer update

[Richard Z.]

On Sun, Mar 09, 2014 at 11:49:17AM +0100, JB wrote:
> Hello,
> I globally agree, except here:
> >Explicit layer=0 seems to be deprecated now.
> Would you deprecate oneway=no, cycleway=no, horse=no for the same reason?
> We had a similar discussion on talk-fr a few weeks ago. layer=0 may
> be useless for the processing systems, but remains usefull for the
> contributor, disambiguating situations with multiple and complex
> crossings.

it was not my idea to deprecate layer=0 and I have no strong opinion on
this. So far the page says "should not be used explicitly" so we can add
some text that it may be appropriate in very complex situations.

I think this is somewhat different from oneway=no and similar.. which 
are a bit of a problem on its own. Not long ago I found a 3 mile long 
desert track somewhere on the Arab Peninsula with the attribute bridge=no.

Richard


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

Re: [OSM-talk] Key:layer update

[Dave F.]

Are you aware there's a separate wiki page for 'Layer'? No wonder users 
get confused!


I suggest these pages are combined into a page called 'Layer' as most 
users wouldn't search for 'Key:*'


On 09/03/2014 10:30, Richard Z. wrote:

Hi,

for some time now I have been working on the wiki page to state the rules
as clearly as possible.. hope that most of the improvements are fairly
uncontroversial. Some of the changes:

* the vertical ordering established by the layer values is valid exactly only
   in the point where the ways cross or objects overlap

* define layer as higher value means above, lower value means bellow. Avoid
   the complicated layer=0 definition as "the natural ground level as it
   would be shown by contour lines on a topographic map". Explicit layer=0
   seems to be deprecated now.

* layer on ways should be used only in combination with one of tunnel=*, 
bridge=*,
   highway=steps, highway=elevator, covered=* or indoor=yes. For areas, it could
   be used in combination with tags such as man_made=bridge, building=* and 
similar.
   The motivation for this is to make it easy for validators to spot errors such
   as when the wrong segment is accidentaly tagged, bridge/tunnel forgotten, or
   someone tags excessively long ways for no good reason - common problem with
   waterways and elevated roads/railroads.
   I have validated this rule for ways in large parts of the world, there are
   exceptions which currently I do not know hot to tag better but those are 
rare.

* in some cases "level" may be more appropriate than "layer"

https://wiki.openstreetmap.org/w/index.php?title=Key%3Alayer&diff=999107&oldid=935491


Richard

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




---
This email is free from viruses and malware because avast! Antivirus protection 
is active.
http://www.avast.com


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

Re: [OSM-talk] Key:layer update

[JB]

Hello,
I globally agree, except here:

Explicit layer=0 seems to be deprecated now.

Would you deprecate oneway=no, cycleway=no, horse=no for the same reason?
We had a similar discussion on talk-fr a few weeks ago. layer=0 may be 
useless for the processing systems, but remains usefull for the 
contributor, disambiguating situations with multiple and complex crossings.

JB

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

Re: [OSM-talk] Key:layer update

[Martin Koppenhoefer]

> Am 09/mar/2014 um 11:30 schrieb "Richard Z." :
> 
> * the vertical ordering established by the layer values is valid exactly only 
>  in the point where the ways cross or objects overlap


actually at the Point where a layer Way Connects to an layer=0 way both are at 
the same height (e.g. where a bridge starts)

Common nodes are always on the same layer regardless of layer values


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