[Tagging] Intermodal transport

[sabas88]

Hello,
I would like to start a discussion related to intermodal transport and its
facilities, and I drafted a page[0] with some links in it.

A tagging system should be proposed also considering the maritime tagging
(OpenSeaMap), or extending it to deal with inland facilities.
As an example, I saw harbour:LOCODE[1] key, but I preferred to use a (imho
more general) ref:LOCODE[2], as the older discussion didn't came to
results. See [3]

Regards,
Stefano

[0] http://wiki.openstreetmap.org/wiki/Proposed_features/Intermodal_Terminal
[1] http://wiki.openstreetmap.org/wiki/Key:harbour:LOCODE
[2] http://taginfo.openstreetmap.org/keys/ref%3ALOCODE
[3] http://comments.gmane.org/gmane.comp.gis.openseamap.maps/261
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] bicycle-only crossings

[Martin Koppenhoefer]

Am 09/apr/2013 um 20:39 schrieb "Alberto" :

> I've tagged some cycle crossing drawn as a single node with
> highway=crossing, foot=no and bicycle=yes. I would not use access=no because
> it can be misunderstood and routing software could prevent access to the
> main street.


>From your description it seems that the node might be forbidden to be crossed 
>by pedestrians (assuming that this crossing is between a cycle way and a 
>street it would prevent pedestrians to use the street as well)

cheers,
Martin 
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Mismatched Level of Detail in highways vs. other elements

[Eckhart Wörner]

Hi Martin,

Am Sonntag, 7. April 2013, 13:35:23 schrieb Martin Atkins:
> I wonder if the root problem is that we've conflated the idea of the 
> physical construct of a street with its parallel in the routing network.

IMHO the underlying main problem is that OSM is steadily moving towards some 
kind of OpenGeoDump™:

• Something has coordinates? – Put it into the OSM database, because OSM is 
about coordinates!

• Don't know how to set up a GIS system? – Put it into the OSM database, 
because nothing beats an existing GIS setup!

• Don't know how to keep data up-to-date? – Put it into the OSM database, 
because it's the duty of OSM people to maintain your data!

(You might notice some kind of pattern.)

Eckhart

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

Re: [Tagging] bicycle-only crossings

[Alberto]

I've tagged some cycle crossing drawn as a single node with
highway=crossing, foot=no and bicycle=yes. I would not use access=no because
it can be misunderstood and routing software could prevent access to the
main street.
The best solution is to tag intersection node with highway=crossing and draw
the cycle crossing with a way tagged with highway=cycleway and
cicleway=crossing.
Bye
Viking81


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

Re: [Tagging] Power generation refinement: power plant model evolution

[Alberto]

> This is where I still don't understand you: why do I need to specify 
> that
>> a feature XXX has the role XXX? Why do I need to specify, that a 
>> generator is a generator? A substation a substation? A dam a dam? A 
>> valve a valve? A weir a weir? And so on.
>>
>
> This is just because a role must be specified.
>

> Why do I have to specify a role?

Yes, roles can be omitted. In some existing types of relations members
haven't a role.
In this case we can omit them.
Viking81



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

Re: [Tagging] Power generation refinement: power plant model evolution

[François Lacombe]

Hi.


2013/4/9 Martin Vonwald (imagic) 

> A generator is different. It simply is a generator and its role within the
> plant is to be a generator. So there is no need to explicitly specify it.
> Furthermore think of the consumers. Example: the process one feature of
> the plant. If the plant is described by a relation the consumers have two
> sources about the type of that feature: the tags on the feature and the
> relation. As the relation is not always present they will trust the tags on
> the feature and ignore the role in the relation.
>

Ok. I agree & understand.



> Which association would they formalise? If a generator is part of a plant
> its function (aka role) is to work as generator. The association is created
> by adding the generator to the plant.
>

I've a relational approach more than a spatial one.
To me, associations must be formalized through hard structured description
in a DB.
I'll have to adapt myself to that kind of data processing.



> I'm just curious about the perimeter: what would we do if a plant has
> multiple perimeters? How would we map them?
>

That situation is under the dispersed power plant configuration.
A relation will gather all features, including fenced perimeterS.

I know you would recommend to only add multiple perimeters to the relation
and get features inside from closed areas but it would introduce one more
case and increase proposal complexity.

According to what is recommended to do for choosing the best power plant
perimeter, it's only when the same operator owns multiple and distinct
perimeters for the same power plant infrastructure.
Instead we would have many different enclosed power plants with different
operators.
http://wiki.openstreetmap.org/wiki/Proposed_features/Power_generation_refinement#Plant_area.2Fperimeter


I've mapped a French hydro pumped-storage power plant as an example of
dispersed power plants.
http://www.openstreetmap.org/browse/changeset/15642946
There are actually 2 power plants, have a look to relations inside that
changeset.



> Does it matter? Besides some special roles defined by the proposal, what
> would a consumer do with an unknown role?
>

Absolutely nothing.

So we can drop all roles?


Cheers.


-- 
*François Lacombe*

francois dot lacombe At telecom-bretagne dot eu
http://www.infos-reseaux.com
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] [OSM-talk] Crossroad names

[Alberto]

>I think we need some clean-up in the traffic signals tags/relations anyway.

+1 for single traffic signals tags and for a relations to manage them. Who
wants to make a proposal?
Viking81


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

Re: [Tagging] Power generation refinement: power plant model evolution

[Martin Vonwald (imagic)]

Hi!

Am 09.04.2013 um 17:22 schrieb François Lacombe 
:

> In my mind, define a role in a relation is mandatory but you say it's 
> definitely not right.

Roles can make sense. For example ways in a route relation may have the role 
forward or backward, if this specific way is only used in one direction. This 
is a property of the route and not of the way, therefore it is tagged as role 
in the relation.

A generator is different. It simply is a generator and its role within the 
plant is to be a generator. So there is no need to explicitly specify it.
Furthermore think of the consumers. Example: the process one feature of the 
plant. If the plant is described by a relation the consumers have two sources 
about the type of that feature: the tags on the feature and the relation. As 
the relation is not always present they will trust the tags on the feature and 
ignore the role in the relation. 


> I agree to say generator / substation and other roles are not relevant and 
> maybe redundant but they formalize association between features and power 
> plant, which features' tags (or plants' tags) won't do ever since they're 
> attached to objects, not to the link between objects.

Which association would they formalise? If a generator is part of a plant its 
function (aka role) is to work as generator. The association is created by 
adding the generator to the plant.


> To make proposal lighter I would consent to remove generator, substation 
> roles and most of the specific roles from non-enclosed power plants relations 
> (perimeter has already left the proposal as you maybe noticed).

I'm just curious about the perimeter: what would we do if a plant has multiple 
perimeters? How would we map them? 


> But that will allow mappers to use any values they want to use, not convince 
> them into letting the "role" field blank.

Does it matter? Besides some special roles defined by the proposal, what would 
a consumer do with an unknown role?


Best regards,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Power generation refinement: power plant model evolution

[François Lacombe]

No LM_1, nothing else than a power=generator could ever produce the right
sort of power (accordingly to what was defined at the plant scale with
plant:output=*) and be considered as this.


In my mind, define a role in a relation is mandatory but you say it's
definitely not right.
I agree to say generator / substation and other roles are not relevant and
maybe redundant but they formalize association between features and power
plant, which features' tags (or plants' tags) won't do ever since they're
attached to objects, not to the link between objects.

To make proposal lighter I would consent to remove generator, substation
roles and most of the specific roles from non-enclosed power plants
relations (perimeter has already left the proposal as you maybe noticed).
But that will allow mappers to use any values they want to use, not
convince them into letting the "role" field blank.

Any objections?

Cheers.



2013/4/9 LM_1 

> Could there not be something else than a generator=* in a role of a
> generator?
>
> LM_1
>
>
> 2013/4/9 Martin Vonwald 
>
>> Hi again :-)
>>
>> 2013/4/9 François Lacombe 
>>
>>> This is where I still don't understand you: why do I need to specify
 that a feature XXX has the role XXX? Why do I need to specify, that a
 generator is a generator? A substation a substation? A dam a dam? A valve a
 valve? A weir a weir? And so on.

>>>
>>> This is just because a role must be specified.
>>>
>>
>> Why do I have to specify a role?
>>
>>
>>
>>> When all features must be member of power=plant relation (because of
>>> lack of perimeter), what role can I associate to a generator except...
>>> generator?
>>>
>>
>> None at all. Because if a generator is a member of a plant, its role is
>> generator. So no need to specify it. Even more: if you force roles to be
>> specified, someone comes along and specifies a generator as dam. This is of
>> course can be very easily and automatically detected as all generators must
>> be generators. But wait! If it can be automatically detected, why should it
>> be specified?
>>
>> Imagine a basket ("relation") full of fruits ("members"). Everyone knows
>> the fruits and their names("tags"). Now someone puts notes ("roles") on the
>> fruits. On all apples now is a note saying "apple", on all pears it says
>> "pear", and so on. Would you think of those notes as helpful in any way?
>> You are a professional fruit merchant ("mapper"). Now your government ("the
>> one which writes the proposal" ;-) ) decides that in order to sell your
>> fruits you have to put one of those notes on each of them. The neighbours
>> boy comes along and switches all the notes. What do you think of the notes?
>> Are they worth the effort or would you consider changing your job (to
>> "google map maker")?
>> (Obviously I like illustrative comparisons)
>>
>>
>> Best regards!
>> Martin
>>
>>
>>
>> ___
>> Tagging mailing list
>> Tagging@openstreetmap.org
>> http://lists.openstreetmap.org/listinfo/tagging
>>
>>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/tagging
>
>


-- 
*François Lacombe*

francois dot lacombe At telecom-bretagne dot eu
http://www.infos-reseaux.com
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] bicycle-only crossings

[Richard Welty]

On 4/9/13 9:32 AM, Martin Vonwald wrote:

Good question. But in this context it would by obvious in my opinion - the
cycleway.


i would recommend against the approach of tagging the single node. better,
i think, to break the cycleway on either side of the shared node and tag
the way. this eliminates all ambiguity and doesn't require any special 
handling

in code for data consumers.

richard


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

Re: [Tagging] bicycle-only crossings

[Martin Koppenhoefer]

2013/4/9 Richard Fairhurst 

> Martin Vonwald (Imagic) wrote:
> > A node with highway=crossing (or level_crossing for railways) only
> > refers to the crossing of the highway, i.e. its access tags do not
> > refer to the highway itself.
>
> Genuine question - if you have a highway=cycleway crossing a
> highway=unclassified at a shared crossing node, how do you know which one
> is
> "the highway itself"?
>
>


+1, as the width of the street is rarely 0.0 it seems more practical to tag
a short piece of way with access-tags (and crossing).

cheers,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Power generation refinement: power plant model evolution

[LM_1]

Could there not be something else than a generator=* in a role of a
generator?

LM_1


2013/4/9 Martin Vonwald 

> Hi again :-)
>
> 2013/4/9 François Lacombe 
>
>> This is where I still don't understand you: why do I need to specify that
>>> a feature XXX has the role XXX? Why do I need to specify, that a generator
>>> is a generator? A substation a substation? A dam a dam? A valve a valve? A
>>> weir a weir? And so on.
>>>
>>
>> This is just because a role must be specified.
>>
>
> Why do I have to specify a role?
>
>
>
>> When all features must be member of power=plant relation (because of lack
>> of perimeter), what role can I associate to a generator except... generator?
>>
>
> None at all. Because if a generator is a member of a plant, its role is
> generator. So no need to specify it. Even more: if you force roles to be
> specified, someone comes along and specifies a generator as dam. This is of
> course can be very easily and automatically detected as all generators must
> be generators. But wait! If it can be automatically detected, why should it
> be specified?
>
> Imagine a basket ("relation") full of fruits ("members"). Everyone knows
> the fruits and their names("tags"). Now someone puts notes ("roles") on the
> fruits. On all apples now is a note saying "apple", on all pears it says
> "pear", and so on. Would you think of those notes as helpful in any way?
> You are a professional fruit merchant ("mapper"). Now your government ("the
> one which writes the proposal" ;-) ) decides that in order to sell your
> fruits you have to put one of those notes on each of them. The neighbours
> boy comes along and switches all the notes. What do you think of the notes?
> Are they worth the effort or would you consider changing your job (to
> "google map maker")?
> (Obviously I like illustrative comparisons)
>
>
> Best regards!
> Martin
>
>
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/tagging
>
>
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] bicycle-only crossings

[Martin Vonwald]

Good question. But in this context it would by obvious in my opinion - the
cycleway.

2013/4/9 Richard Fairhurst 

> Martin Vonwald (Imagic) wrote:
> > A node with highway=crossing (or level_crossing for railways) only
> > refers to the crossing of the highway, i.e. its access tags do not
> > refer to the highway itself.
>
> Genuine question - if you have a highway=cycleway crossing a
> highway=unclassified at a shared crossing node, how do you know which one
> is
> "the highway itself"?
>
> cheers
> Richard
>
>
>
>
>
> --
> View this message in context:
> http://gis.19327.n5.nabble.com/Re-bicycle-only-crossings-tp5756338p5756359.html
> Sent from the Tagging mailing list archive at Nabble.com.
>
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/tagging
>
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] bicycle-only crossings

[Pieren]

On Tue, Apr 9, 2013 at 3:21 PM, Richard Fairhurst wrote:

> Genuine question - if you have a highway=cycleway crossing a
> highway=unclassified at a shared crossing node, how do you know which one
> is
> "the highway itself"?
>
>
And even more genuine : why do you have to tag the shared crossing node
between an unclassified and cycleway ?

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

Re: [Tagging] bicycle-only crossings

[Richard Fairhurst]

Martin Vonwald (Imagic) wrote:
> A node with highway=crossing (or level_crossing for railways) only 
> refers to the crossing of the highway, i.e. its access tags do not 
> refer to the highway itself.

Genuine question - if you have a highway=cycleway crossing a
highway=unclassified at a shared crossing node, how do you know which one is
"the highway itself"?

cheers
Richard





--
View this message in context: 
http://gis.19327.n5.nabble.com/Re-bicycle-only-crossings-tp5756338p5756359.html
Sent from the Tagging mailing list archive at Nabble.com.

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

Re: [Tagging] bicycle-only crossings

[Martin Vonwald]

Hi!

2013/4/9 Richard Welty 

> but what is it crossing? if you put access=no on a node that is shared by a
> cycleway and a public highway, you'll deny access to cars.
>

A node with highway=crossing (or level_crossing for railways) only refers
to the crossing of the highway, i.e. its access tags do not refer to the
highway itself.

regards,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] bicycle-only crossings

[Richard Welty]

On 4/9/13 8:35 AM, Martin Vonwald wrote:

Hi!

2013/4/9 Volker Schmidt 


How to tag a bicycle-only crossing.
We have here occasionally road crossings that are separate for pedestrians
and for bicycles. They can be even 20meters apart.
Is it
highway=crossing
bicycle=yes
foot=no



Sounds reasonable. Although if it is really bicycle-only I would even use
access=no instead of foot=no, to prevent anyone on skies trying to use it
;-)


but what is it crossing? if you put access=no on a node that is shared by a
cycleway and a public highway, you'll deny access to cars.

richard


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

Re: [Tagging] bicycle-only crossings

[Martin Vonwald]

Hi!

2013/4/9 Volker Schmidt 

> How to tag a bicycle-only crossing.
> We have here occasionally road crossings that are separate for pedestrians
> and for bicycles. They can be even 20meters apart.
> Is it
> highway=crossing
> bicycle=yes
> foot=no
>
>
Sounds reasonable. Although if it is really bicycle-only I would even use
access=no instead of foot=no, to prevent anyone on skies trying to use it
;-)

regards,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Power generation refinement: power plant model evolution

[Martin Vonwald]

Hi again :-)

2013/4/9 François Lacombe 

> This is where I still don't understand you: why do I need to specify that
>> a feature XXX has the role XXX? Why do I need to specify, that a generator
>> is a generator? A substation a substation? A dam a dam? A valve a valve? A
>> weir a weir? And so on.
>>
>
> This is just because a role must be specified.
>

Why do I have to specify a role?



> When all features must be member of power=plant relation (because of lack
> of perimeter), what role can I associate to a generator except... generator?
>

None at all. Because if a generator is a member of a plant, its role is
generator. So no need to specify it. Even more: if you force roles to be
specified, someone comes along and specifies a generator as dam. This is of
course can be very easily and automatically detected as all generators must
be generators. But wait! If it can be automatically detected, why should it
be specified?

Imagine a basket ("relation") full of fruits ("members"). Everyone knows
the fruits and their names("tags"). Now someone puts notes ("roles") on the
fruits. On all apples now is a note saying "apple", on all pears it says
"pear", and so on. Would you think of those notes as helpful in any way?
You are a professional fruit merchant ("mapper"). Now your government ("the
one which writes the proposal" ;-) ) decides that in order to sell your
fruits you have to put one of those notes on each of them. The neighbours
boy comes along and switches all the notes. What do you think of the notes?
Are they worth the effort or would you consider changing your job (to
"google map maker")?
(Obviously I like illustrative comparisons)


Best regards!
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] bicycle-only crossings

[Volker Schmidt]

How to tag a bicycle-only crossing.
We have here occasionally road crossings that are separate for pedestrians
and for bicycles. They can be even 20meters apart.
Is it
highway=crossing
bicycle=yes
foot=no
?
Volker,
Padova, Italy
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Power generation refinement: power plant model evolution

[François Lacombe]

Hi Martin!

2013/4/9 Martin Vonwald 

> In my opinion this looks pretty nice. You still need to update the two
> examples Fukushima and Thémis - they are referring to relations and also to
> roles which are nowhere defined (perimeter was dropped, plant is not
> documented as role).
>

You're right.
I've updated accordingly the proposal's page.



>
> This is where I still don't understand you: why do I need to specify that
> a feature XXX has the role XXX? Why do I need to specify, that a generator
> is a generator? A substation a substation? A dam a dam? A valve a valve? A
> weir a weir? And so on.
>

This is just because a role must be specified.
When all features must be member of power=plant relation (because of lack
of perimeter), what role can I associate to a generator except... generator?



>
> I'm not sure if this is the best key to describe it, but I also don't have
> a better one for you, so lets say: it's fine ;-)
>

Update is still possible until the vote begin anyway.



> Finally I want to thank you for all your efforts on this topic!
>

Thank you for yours too.
It's a real pleasure to me.


Cheers.

-- 
*François Lacombe*

francois dot lacombe At telecom-bretagne dot eu
http://www.infos-reseaux.com
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] Mismatched Level of Detail in highways vs. other elements

[Martin Koppenhoefer]

2013/4/9 Martin Atkins 

>
> Right. It seems like the schematic vs. detail tagging situation is pretty
> good for streets if you accept the area:highway proposal:
> 
> http://wiki.openstreetmap.org/**wiki/Proposed_features/area:**highway
>
> Under this proposal you have area:highway as the detail element, and the
> existing highway ways as the routing network element, so the two tagging
> schemes can easily coexist without trampling one another.



there is also the area-relation proposal which aims at mapping (beside
other) implicit highway polygons:
http://wiki.openstreetmap.org/wiki/Relations/Proposed/Area

The idea is that you only map "parallels" to the highway way (at the outer
border of the sidewalk and eventually the kerbs) and then connect the two
sides via a relation. This way you don't need the traversing ways (usually
used to close the polygon on the small sides) which really helps a lot to
avoid misconnections of the routing graph and the highway-areas. Until now
this type of relation is not supported, but it is not really complicated to
do it. To help the renderers I guess that the direction of the two parallel
ways should be the same (otherwise you risk to get X-like connections).

cheers,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] [OSM-talk] Crossroad names

[Guillaume Allegre]

Le mar. 09 avril 2013 à 10:56 +0200, Pieren a écrit :

> Note that we have intersections with names and intersections with traffic
> signals. It does not have to be on the same relation types because it is
> imho two different features.

+1; with the increasing complexity in traffic signals to come (rather for
advanced contributors), there is no point in having additional complexity
for naming crossroads too.



-- 
 ° /\Guillaume AllègreOpenStreetMap France
  /~~\/\   allegre.guilla...@free.fr  Cartographie libre et collaborative
 /   /~~\tél. 04.76.63.26.99  http://www.openstreetmap.fr


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

Re: [Tagging] [OSM-talk] Crossroad names

[Pieren]

On Tue, Apr 9, 2013 at 9:36 AM, Vladimir Vyskocil <
vladimir.vysko...@gmail.com> wrote:

>
> Relation feature essentially means, "group up some objects to represent as
> a feature".
> So I think it could suite for our traffic_signal & junction naming,
> routing.
> (relation could express "label" membership)
>
> I also agree
>

I think we need some clean-up in the traffic signals tags/relations anyway.

>From its beginning, the tag "highway=traffic_signals" has been designated
to say "this intersection is controlled by traffic signals". It did not say
which lanes were controlled on which direction.  It was not intended to
identify each individual traffic signal.

Unfortunately, since we have hi-res images available, some contributors are
using the same tag to locate exactly each traffic signal next to the
intersection. Why not. We are not against improving data quality.

But the problem is that they use the same tag, most probably because it is
rendered and they just "draw a map". They don't care about the difference
"intersection controlled by traffic signals" vs "single traffic light".
They don't care if the same intersection is sometimes modeled with one
"highway=traffic_signals" or four or eight "highway=traffic_signals" nodes.
They don't care if it makes a difference for routing penalties or rendering
choices (e.g. zoom levels).

Once we decide to accept such level of details, we need to distinguish the
two methods of modeling traffic signals. We need a new tag to identify
single traffic signals . And we need of course a relation to say which
traffic signal belongs to which intersection (like for stops or giveway).
Once we have a correct modeling, it will be possible to use the data
correctly in applications like routing or rendering at different zoom
levels.
Note that we have intersections with names and intersections with traffic
signals. It does not have to be on the same relation types because it is
imho two different features.

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

Re: [Tagging] Power generation refinement: power plant model evolution

[Martin Vonwald]

Hi!

In my opinion this looks pretty nice. You still need to update the two
examples Fukushima and Thémis - they are referring to relations and also to
roles which are nowhere defined (perimeter was dropped, plant is not
documented as role).

2013/4/8 François Lacombe 

> Hi,
>
> I finally agree with you.
> I've began to update the proposal to remove relations in all cases except
> when power plant doesn't have any physical permimeter.
>
> We must keep role=generator for all generators (no need to distinguish
> them there) in such relation since other features may be added too.
>

This is where I still don't understand you: why do I need to specify that a
feature XXX has the role XXX? Why do I need to specify, that a generator is
a generator? A substation a substation? A dam a dam? A valve a valve? A
weir a weir? And so on.



> Concerning output/intermediate generators, I've introduced
> generator:plant=output or generator:plant=intermediate to follow your
> readability suggestion.
>

I'm not sure if this is the best key to describe it, but I also don't have
a better one for you, so lets say: it's fine ;-)


Finally I want to thank you for all your efforts on this topic!

Best regards,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging

Re: [Tagging] [OSM-talk] Crossroad names

[Vladimir Vyskocil]

On 2 avr. 2013, at 18:54, Satoshi IIDA  wrote:

> 
> Why don't use our old "relation = tarffic_lights"? (or junction?)
> http://wiki.openstreetmap.org/wiki/Relations/Proposed/Traffic_Lights
> (yes, it need more advanced!)

Yes, this is also my favorite solution to the problem. It only need some 
cleanup and specification : adding the (optional) name of the crossroads, 
naming roles,... ?

> 
> I would like to against write them as area.

+1

> 
> Relation feature essentially means, "group up some objects to represent as a 
> feature".
> So I think it could suite for our traffic_signal & junction naming, routing.
> (relation could express "label" membership)
> 

I also agree 

Vlad.

> 
> 
> 
> 2013/3/29 Pieren 
> On Thu, Mar 28, 2013 at 2:22 PM, Janko Mihelić  wrote:
> 
> > I saw this proposal before and I liked it. Now I like it even more because
> > it solves the junction name problem.
> 
> As area, only used 3 times by 2 different users in one year (date of
> the proposal):
> 
> 2 by the proposal writer "imagic":
> http://www.openstreetmap.org/browse/way/210448306
> http://www.openstreetmap.org/browse/way/202280167
> 
> 1 by "Soldier Boy":
> http://www.openstreetmap.org/browse/way/190569949
> 
> Another one is a mistake (roundabout):
> http://www.openstreetmap.org/browse/way/43114115
> 
> Not something I would call "a success".
> 
> But 1132 relations of type "junction":
> http://taginfo.openstreetmap.org/tags/type=junction
> http://taginfo.openstreetmap.org/relations/junction#overview
> http://wiki.openstreetmap.org/wiki/Relations/Proposed/Junctions
> 
> Pieren
> 
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/tagging
> 
> 
> 
> -- 
> Satoshi IIDA
> mail: nyamp...@gmail.com
> twitter: @nyampire
> ___
> Tagging mailing list
> Tagging@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/tagging

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