Opened 12 years ago
Last modified 3 years ago
#8622 new enhancement
Junction validator?
Reported by: | Debeet | Owned by: | team |
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | Core validator | Version: | |
Keywords: | roundabout | Cc: |
Description (last modified by )
Is is possible to make a junction validator to check for a right highway tag?
Attachments (2)
Change History (13)
comment:1 by , 12 years ago
Owner: | changed from | to
---|---|
Status: | new → assigned |
by , 12 years ago
Attachment: | junctions.osm added |
---|
follow-up: 3 comment:2 by , 12 years ago
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
In 5902/josm:
comment:3 by , 12 years ago
Replying to Don-vip:
In 5902/josm:
I am sorry, but do I understand it right that your code depends on only one closed way as roundabout ? If so, this is wrong. There are quite a lot of roundabounts which are cut in parts.
Weather it is a single way or several parts, we even do not have a check that checks that a roundabout is a closed way. We need this, too.
Just my thoughts maybe I am totally wrong.
comment:4 by , 12 years ago
Resolution: | fixed |
---|---|
Status: | closed → reopened |
comment:5 by , 12 years ago
Owner: | changed from | to
---|---|
Status: | reopened → new |
This is a lot more complicated to handle theses cases, I have no time for this right now.
comment:8 by , 12 years ago
Replying to #8751 anonymous:
On the type of highway on the roundabout.
The motoroad and trunk roads are circulating specific and often slow-moving vehicles are prohibited (pedestrian, animal, motorcycle, bike, ..)
Testing the type of highway JOSM encourages roundaboud put in trunk when a trunk is connected to the roundabout.
Hence the roundabout is in spite of the connection trunk road or segondary primary (or lower) and slow vehicle (see below) can not take the roundabout.
In most cases (in the field) signs indicate the end and the beginning of trunk excluding the roundabout.
I am not sure about what signs you are talking about. Do you mean motorroad=yes
?
See also:
- http://wiki.openstreetmap.org/wiki/Tag:highway=trunk
- http://wiki.openstreetmap.org/wiki/Highway:International_equivalence
It seems preferable not to submit trunk on roundabout if there is a lower type connection. to let the vehicle slow to use the roundabout.
So we are talking about "motorroad=yes/no"
Similar to motorway
Motorway
is by far different to all other highways as it implies quite a lot. (Should be cleaned up but different topic.)
Anyway, these definitions with not fit to many countries in Afrika for example.
If you have more than one trunk
connected to the roundabout the roundabout is at least a trunk_link
and if we talk about one trunk
leading through a roundabout it is a trunk
though with possible motorroad=no
.
Replying to #8751 comment:1 anonymous:
in the case of connection on a roundabout:
1 primary
1 primary_link
2 residential
residential proposal, while I think the primary would be desirable.
Should be used as the * _link highway without link.
I am not sure about highway=*_link
on roundabout but why not ?
The check needs some rework. It nicely already checks for oneway but it does not know how to handle *_link
and what to do if you have several different grades (e.g. primary*
plus secondary*
plus tertiary
).
Please have a look at attached example file.
follow-up: 10 comment:9 by , 12 years ago
I meant trunk, as in France motorroad is not advisable or rarely used (not documented).
After reading the documentation motorroad indeed if trunk (motorroad = yes / no) connect the roundabout, the roundabout must either motorroad = no (if of the lower connections, primary, ...).
For example, I would put the roundabout of the largest type (excluding motorway, trunk).
Seeing the example, I see that you have taken care to separate the connections on different nodes. I think a test verification of this state would be an improvement. Indeed, if two nodes share the same connections so we can consider that the vehicle passes an input to an output directly without entering the roundabout (and rules of conduct roundabout).
comment:10 by , 12 years ago
Replying to anonymous:
I meant trunk, as in France motorroad is not advisable or rarely used (not documented).
In Germany and Spain it is well used. Maybe discuss this in the Frensh Community.
After reading the documentation motorroad indeed if trunk (motorroad = yes / no) connect the roundabout, the roundabout must either motorroad = no (if of the lower connections, primary, ...).
Maybe it would help to add motorroad=no to the info of trunk*.
For example, I would put the roundabout of the largest type (excluding motorway, trunk).
This does not fit.
I think we need the highest one if at least two connected ways with the same riding direction exist. Otherwise use the second highest. This includes all *_link.
Seeing the example, I see that you have taken care to separate the connections on different nodes. I think a test verification of this state would be an improvement. Indeed, if two nodes share the same connections so we can consider that the vehicle passes an input to an output directly without entering the roundabout (and rules of conduct roundabout).
There might be bypasses and I think there exists a tag already but you have to map them on their own.
For improvements of or new checks for validator, please, open new tickets with the component "Core validator". Thanks
comment:11 by , 3 years ago
Description: | modified (diff) |
---|---|
Keywords: | roundabout added |
test case