-
Notifications
You must be signed in to change notification settings - Fork 1.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
"only_*" turn restrictions shouldn't affect turns in opposite direction of way #4849
Comments
Yes, this is something that came up in our internal testing and I took an approach that might not be perfect.. I was trying to follow the rule on the osm wiki:
We agree that that this wording is probably an oversimplification - when the FROM way is bidirectional. I ended up adding some code in here to say "only follow this rule and restrict other paths that are local to the intersection". The situation we ran into with testing was that the FROM way could trail off for miles and that I think your approach of " It would make me even happier if we could just say "
Yeah that's a consequence of the "local via" rule I mentioned above.. What is considered "local" depends on that slider. |
Should also add - If this is the case then we should definitely change iD to match what they do 😆 |
here's an example: relation/2227207 OSRM and Graphhopper happly route across this turn restriction in both directions: |
Yeah I thought about this more.. I'll definitely change it and push out an update in the next day or so. |
+1 |
I noticed the following regression related to the reworked turn restrictions editor (see #4768 / #2622): When a short road segment in a (moderately) complex intersection is used as a from member of a only-type turn restriction, not only are all "outgoing" turns marked as prohibited (i.e. turns one could otherwise take when driving on the road towards the via node or way(s)), but also all turns on the other end of this road, basically transforming it into a dead-end road. Here's an example (location):
Notice how the southern end of the selected way turns into an all-red situation after adding the only-restriction at the northern end of the road.
Another confusing follow-up effect of this is that some intersections are displayed differently on how far one is zoomed in, as seen here (same location as above):
//edit: a second side effect of this is that all previously mapped no restrictions on the opposite end of a way are removed after adding an only restriction on one end. In this example, the initial southern no_u_turn restricton is affected by this.
I think that only-type turn restrictions should only affect the direction of the way towards the respective via member(s) (even though the osm wiki page about turn restrictions doesn't state this explicitly). And as far as I can see, this is also how every routing software implements these.
The text was updated successfully, but these errors were encountered: