traffic_sign:forward=*

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

traffic_sign:forward=*

Gerd Petermann

Hi all, 


it seems that the wiki 

http://wiki.openstreetmap.org/wiki/Key:traffic_sign

is too confusing. The overpass query

  node[ "traffic_sign"="maxspeed"]["source:maxspeed"="sign"];
shows 831 nodes. My understanding is that one tag is for the
node, the other for the way.
Also funny:
http://taginfo.openstreetmap.org/keys/traffic_sign%3Aforward#values
shows 1050 entries for the tag traffic_sign:forward=*
(yes, "*" as value)

My understanding so far is that I should see e.g. these tags on a node
which is part of a highway:
traffic_sign:forward=maxspeed
maxspeed=60

The corresponding way would start at this node and have
tags
maxspeed=60
source:maxspeed=sign

Am I right?

Gerd


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: traffic_sign:forward=*

dieterdreist


sent from a phone

Am 02.11.2015 um 19:18 schrieb Gerd Petermann <[hidden email]>:

The overpass query

  node[ "traffic_sign"="maxspeed"]["source:maxspeed"="sign"];
shows 831 nodes. My understanding is that one tag is for the
node, the other for the way.


yes, these nodes are likely mistagged, I usually tag nodes aside the road with
traffic_sign=maxspeed 
maxspeed=50 (etc.)

and on the highway
maxspeed =50
source:maxspeed =sign


Also funny:
shows 1050 entries for the tag traffic_sign:forward=*
(yes, "*" as value)

My understanding so far is that I should see e.g. these tags on a node
which is part of a highway:
traffic_sign:forward=maxspeed
maxspeed=60


tagging traffic signs on the road (unless they are road markings or maybe overhead signs), is strange as well, because you typically don't find signs on the road. In case of overhead signs they should get a higher layer and not be part of the highway.



The corresponding way would start at this node and have
tags
maxspeed=60
source:maxspeed=sign


it would not have to start there (because of repeating signs)


cheers 
Martin 

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: traffic_sign:forward=*

Gerd Petermann
dieterdreist wrote
> Also funny:
> http://taginfo.openstreetmap.org/keys/traffic_sign%3Aforward#values
> shows 1050 entries for the tag traffic_sign:forward=*
> (yes, "*" as value)
>
> My understanding so far is that I should see e.g. these tags on a node
> which is part of a highway:
> traffic_sign:forward=maxspeed
> maxspeed=60


tagging traffic signs on the road (unless they are road markings or maybe overhead signs), is strange as well, because you typically don't find signs on the road. In case of overhead signs they should get a higher layer and not be part of the highway.
I totally agree. I stumbled over some mistagged nodes (highway=maxspeed)
and started to remove them because I thought that they just copy the values
of the way, but then I understood that they are probably traffic_signs.
I read the wiki a few times and still did not fully understand how this
traffic_sign:forward idea should work.
dieterdreist wrote
> The corresponding way would start at this node and have
> tags
> maxspeed=60
> source:maxspeed=sign

it would not have to start there (because of repeating signs)
good point, that's of course right.

Gerd
Reply | Threaded
Open this post in threaded view
|

Re: traffic_sign:forward=*

Georg Feddern-2
Am 03.11.2015 um 17:43 schrieb GerdP:
> I read the wiki a few times and still did not fully understand how
> this traffic_sign:forward idea should work.

If you read (and use) _only_ traffic_sign:forward - I suppose you read
only the german wiki page and then I understand your problem, because
that can not work in all cases.
If you read the english wiki page, you may understand the intention
better - as there is also a traffic_sign:backward variant mentioned.

It shall work as in reality and as "computered" in the human mind:
Transform the information from the sign (node) to the way in the
relevant direction - with all possibilities, but even all obstacles also ...

I do not support this "node on way" strategy - I use only the node
beside way as tagging for the sign itself - and "always on the right
side" ;-) - at least here in Germany.

Georg

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: traffic_sign:forward=*

Gerd Petermann
Well, I typically look at the english page, if I don't understand
it, I check the german one as well. You are right, the german one
doesn't even mention the :backward suffix. Still,
there is no example that shows in detail how to use this tagging,
and that's probably the reason for the mistakes.
Also, JOSM doesn't seem to understand this tagging scheme.

Gerd
________________________________________
Von: Georg Feddern <[hidden email]>
Gesendet: Mittwoch, 4. November 2015 07:17
An: Tag discussion, strategy and related tools
Betreff: Re: [Tagging] traffic_sign:forward=*

Am 03.11.2015 um 17:43 schrieb GerdP:
> I read the wiki a few times and still did not fully understand how
> this traffic_sign:forward idea should work.

If you read (and use) _only_ traffic_sign:forward - I suppose you read
only the german wiki page and then I understand your problem, because
that can not work in all cases.
If you read the english wiki page, you may understand the intention
better - as there is also a traffic_sign:backward variant mentioned.

It shall work as in reality and as "computered" in the human mind:
Transform the information from the sign (node) to the way in the
relevant direction - with all possibilities, but even all obstacles also ...

I do not support this "node on way" strategy - I use only the node
beside way as tagging for the sign itself - and "always on the right
side" ;-) - at least here in Germany.

Georg

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: traffic_sign:forward=*

dieterdreist
In reply to this post by Georg Feddern-2

2015-11-04 7:17 GMT+01:00 Georg Feddern <[hidden email]>:
If you read (and use) _only_ traffic_sign:forward - I suppose you read only the german wiki page and then I understand your problem, because that can not work in all cases.
If you read the english wiki page, you may understand the intention better - as there is also a traffic_sign:backward variant mentioned.

It shall work as in reality and as "computered" in the human mind:
Transform the information from the sign (node) to the way in the relevant direction - with all possibilities, but even all obstacles also ...

I do not support this "node on way" strategy - I use only the node beside way as tagging for the sign itself - and "always on the right side" ;-) - at least here in Germany.



Looks as if we agree that traffic signs are point objects at the side of a road, not linear stuff on a road. It doesn't make sense to have a traffic sign on a long part of a road, it is a point but can have effect for a linear piece of road (but then, it is not the traffic sign but the effect you want to map).

Browsing the history of the page, I have found out that the idea of adding the key to a way is to collect a list of traffic signs that are valid for this way. IMHO this doesn't make a lot of sense, because the idea of storing the traffic signs was that of being able to verify actual tagging on the way (e.g. see from where to where a maxspeed is valid and where it changes for sure), but this idea is put from the top to the bottom if you repeat the actual effects (which normally do have their own tags, e.g. maxspeed, overtaking, access-tags, etc.) with the traffic sign tag.

I propose to at least discourage the use of the traffic_sign key on ways, if not deprecate.

Cheers,
Martin

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: traffic_sign:forward=*

Gerd Petermann


Hi Martin,


I think you are mixing two things now. I talked about the "As part of a way" part, not the "On a way or area" part,

which looks even more weird to me.


Besides that: Yes, I also think that we should map a traffic_sign as a node with the position of the sign.


Gerd



Von: Martin Koppenhoefer <[hidden email]>
Gesendet: Mittwoch, 4. November 2015 11:11
An: Tag discussion, strategy and related tools
Betreff: Re: [Tagging] traffic_sign:forward=*
 

2015-11-04 7:17 GMT+01:00 Georg Feddern <[hidden email]>:
If you read (and use) _only_ traffic_sign:forward - I suppose you read only the german wiki page and then I understand your problem, because that can not work in all cases.
If you read the english wiki page, you may understand the intention better - as there is also a traffic_sign:backward variant mentioned.

It shall work as in reality and as "computered" in the human mind:
Transform the information from the sign (node) to the way in the relevant direction - with all possibilities, but even all obstacles also ...

I do not support this "node on way" strategy - I use only the node beside way as tagging for the sign itself - and "always on the right side" ;-) - at least here in Germany.



Looks as if we agree that traffic signs are point objects at the side of a road, not linear stuff on a road. It doesn't make sense to have a traffic sign on a long part of a road, it is a point but can have effect for a linear piece of road (but then, it is not the traffic sign but the effect you want to map).

Browsing the history of the page, I have found out that the idea of adding the key to a way is to collect a list of traffic signs that are valid for this way. IMHO this doesn't make a lot of sense, because the idea of storing the traffic signs was that of being able to verify actual tagging on the way (e.g. see from where to where a maxspeed is valid and where it changes for sure), but this idea is put from the top to the bottom if you repeat the actual effects (which normally do have their own tags, e.g. maxspeed, overtaking, access-tags, etc.) with the traffic sign tag.

I propose to at least discourage the use of the traffic_sign key on ways, if not deprecate.

Cheers,
Martin

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: traffic_sign:forward=*

Pee Wee

Just want to let you know that in NL we have 3 kind of cycle ways. It is quite common to tag these ways with a traffic_sign in order to differentiate. This is not only unambiguous but it also makes things easy to change in the OSM database in case legislation changes. For example.. if a max speed is introduced in 1 of the 3 types it is easyly added in OSM. Something that would but a lot more work if the signs were only nodes.

Cheers

PeeWee32




2015-11-04 11:25 GMT+01:00 Gerd Petermann <[hidden email]>:


Hi Martin,


I think you are mixing two things now. I talked about the "As part of a way" part, not the "On a way or area" part,

which looks even more weird to me.


Besides that: Yes, I also think that we should map a traffic_sign as a node with the position of the sign.


Gerd



Von: Martin Koppenhoefer <[hidden email]>
Gesendet: Mittwoch, 4. November 2015 11:11
An: Tag discussion, strategy and related tools
Betreff: Re: [Tagging] traffic_sign:forward=*
 

2015-11-04 7:17 GMT+01:00 Georg Feddern <[hidden email]>:
If you read (and use) _only_ traffic_sign:forward - I suppose you read only the german wiki page and then I understand your problem, because that can not work in all cases.
If you read the english wiki page, you may understand the intention better - as there is also a traffic_sign:backward variant mentioned.

It shall work as in reality and as "computered" in the human mind:
Transform the information from the sign (node) to the way in the relevant direction - with all possibilities, but even all obstacles also ...

I do not support this "node on way" strategy - I use only the node beside way as tagging for the sign itself - and "always on the right side" ;-) - at least here in Germany.



Looks as if we agree that traffic signs are point objects at the side of a road, not linear stuff on a road. It doesn't make sense to have a traffic sign on a long part of a road, it is a point but can have effect for a linear piece of road (but then, it is not the traffic sign but the effect you want to map).

Browsing the history of the page, I have found out that the idea of adding the key to a way is to collect a list of traffic signs that are valid for this way. IMHO this doesn't make a lot of sense, because the idea of storing the traffic signs was that of being able to verify actual tagging on the way (e.g. see from where to where a maxspeed is valid and where it changes for sure), but this idea is put from the top to the bottom if you repeat the actual effects (which normally do have their own tags, e.g. maxspeed, overtaking, access-tags, etc.) with the traffic sign tag.

I propose to at least discourage the use of the traffic_sign key on ways, if not deprecate.

Cheers,
Martin

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging




--
Verbeter de wereld. Word mapper voor Openstreetmap.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging