You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jul 24, 2021. It is now read-only.
Reporter: stanton [Submitted to the original trac issue database at 10.52pm, Friday, 21st May 2010]
When different highway types share the same junction, rendering at the crossing point (which road is drawn on top) seems somewhat random. My suggestion for improvement is to introduce a priority based on the tags that affect rendering (most notably, highway and access):
highway=*_link "loses" over everything else (as a link by definition goes only from one road of the corresponding type to another road of the same or a lower type).
The only road of its kind to arrive at the junction "loses" over all others except *_link.
For roads that continue across the junction, the highest type wins - that is, first motorway, then trunk, then primary, then secondary, then tertiary, then unclassified, then residential and finally service.
A special case are any kinds of restrictions, such as denoted by the access tag and certain kinds of highway tags (e,g, highway=pedestrian). Restrictions denoted by the access tag are most likely interrupted when they continue beyond a junction. However, the case of an unclassified crossing a pedestrian is trickier: does the unclassified interrupt the pedestrian zone or vice versa? This is a semantic question and, where in doubt, is probably best denoted by tagging the node at which the ways cross accordingly.
Suggestion for this case: tag the node with the highway=* tag of the road that should continue across it (if necessary, same for the access=* tag) to make it clear. Some tags used exclusively on nodes (e.g. highway=crossing) may make this impossible but may by themselves contain the required information: in the case of the crossing between an unclassified and a pedestrian, if the node is tagged with highway=crossing or highway=traffic_signals, this implies that the unclassified interrupts the pedestrian way.
The text was updated successfully, but these errors were encountered:
Author: Ldp [Added to the original trac issue at 11.51pm, Thursday, 3rd June 2010]
Rendering priority ordering could indeed be better. CM has something like this, but the current stylesheet is not quite suited yet to that method.
Your other suggestions for junction handling (only road of its kind loses; which one continues across the intersection) are hard to handle, since roads are drawn serially, and they don't know about junctions and nearness of other roads.
Author: dareus[at]telefoninux.org [Added to the original trac issue at 8.50am, Thursday, 23rd May 2013]
After 3 years this bug (which I suppose is a tricky one) is still here.
It is quite bad in Italy where roundabout are almost everywhere and quite often there are many different highway tags connecting; e.g. [http://www.openstreetmap.org/?lat=45.596188&lon=8.860594&zoom=18&layers=M here] (I know it looks like a bean, but sometimes we like to be fanciful :) )
Reporter: stanton
[Submitted to the original trac issue database at 10.52pm, Friday, 21st May 2010]
When different highway types share the same junction, rendering at the crossing point (which road is drawn on top) seems somewhat random. My suggestion for improvement is to introduce a priority based on the tags that affect rendering (most notably, highway and access):
highway=*_link
"loses" over everything else (as a link by definition goes only from one road of the corresponding type to another road of the same or a lower type).*_link
.A special case are any kinds of restrictions, such as denoted by the access tag and certain kinds of highway tags (e,g,
highway=pedestrian
). Restrictions denoted by the access tag are most likely interrupted when they continue beyond a junction. However, the case of an unclassified crossing a pedestrian is trickier: does the unclassified interrupt the pedestrian zone or vice versa? This is a semantic question and, where in doubt, is probably best denoted by tagging the node at which the ways cross accordingly.Suggestion for this case: tag the node with the
highway=* tag
of the road that should continue across it (if necessary, same for theaccess=* tag
) to make it clear. Some tags used exclusively on nodes (e.g. highway=crossing) may make this impossible but may by themselves contain the required information: in the case of the crossing between an unclassified and a pedestrian, if the node is tagged withhighway=crossing
orhighway=traffic_signals
, this implies that the unclassified interrupts the pedestrian way.The text was updated successfully, but these errors were encountered: