Re: [OSM-talk] Tagging continuous flow intersections
(I suppose your reply was meant to t...@? If not, sorry for posting it) On Mon, 24 Aug 2009 13:32:06 +0100, Matt Williams wrote: 2009/8/24 David Paleino d.pale...@gmail.com: On Sun, 23 Aug 2009 18:11:59 -0500, David Lynch wrote: I suspect that I'd end up creating a set of ways that look something like this, plus a whole bunch of oneway tags and turn restrictions: http://dl1050.dyndns.org:/images/osm/cfi.png Thanks for the suggestion -- but I'd avoid drawing different ways for different lanes in a single carriageway. In some cases it's acceptable to do that. As long as there's either a physical barrier or a hard no-changing-lanes restriction between the lanes in question. There's no physical barrier, and the lanes are divided by continuous lines -- that would be a no-changing-lanes restriction, but I'd still be uncomfortable with drawing two separate ways -- that doesn't reflect real world. Satellite image: http://maps.google.com/?ie=UTF8ll=38.12436,13.355808spn=0.001091,0.002411t=kz=19 (the street from NW is the one from S in my drawing -- http://imagebin.ca/view/SjGkG4.html ) If you zoom in, you can clearly see the horizontal signals (at least at the NW street, there's some shadow hiding those in the SW one). If there are no other suggestions, I'll try to think at something :/ Kindly, David -- . ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino : :' : Linuxer #334216 --|-- http://www.hanskalabs.net/ `. `'` GPG: 1392B174 | http://snipr.com/qa_page `- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174 signature.asc Description: PGP signature ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Tagging continuous flow intersections
On Mon, Aug 24, 2009 at 12:06 PM, David Paleinod.pale...@gmail.com wrote: There's no physical barrier, and the lanes are divided by continuous lines -- that would be a no-changing-lanes restriction, but I'd still be uncomfortable with drawing two separate ways -- that doesn't reflect real world. Satellite image: http://maps.google.com/?ie=UTF8ll=38.12436,13.355808spn=0.001091,0.002411t=kz=19 (the street from NW is the one from S in my drawing -- http://imagebin.ca/view/SjGkG4.html ) If you zoom in, you can clearly see the horizontal signals (at least at the NW street, there's some shadow hiding those in the SW one). If there are no other suggestions, I'll try to think at something :/ You also referred to http://en.wikipedia.org/wiki/Continuous_flow_intersection I don't remember ever seeing one of these in the wild. I don't think that the intersection that you are looking at is a continuous flow intersection as described in the wikipedia article. It appears to be a simple cross junction of two one way roads. Is that correct? If so I would map it with a single node at two crossing, one-way, ways. Here, in Ontario Canada, this junction would have no special signage, other than the one-way arrows (and do not enter, for the opposite). We have Right turn allowed on red light. http://en.wikipedia.org/wiki/Right_turn_on_red This is abstracted locally to include left turn allowed on red when the junction is one-way to one-way. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Tagging continuous flow intersections
On Mon, 24 Aug 2009 12:42:53 -0400, Richard Weait wrote: On Mon, Aug 24, 2009 at 12:06 PM, David Paleinod.pale...@gmail.com wrote: There's no physical barrier, and the lanes are divided by continuous lines -- that would be a no-changing-lanes restriction, but I'd still be uncomfortable with drawing two separate ways -- that doesn't reflect real world. Satellite image: http://maps.google.com/?ie=UTF8ll=38.12436,13.355808spn=0.001091,0.002411t=kz=19 (the street from NW is the one from S in my drawing -- http://imagebin.ca/view/SjGkG4.html ) If you zoom in, you can clearly see the horizontal signals (at least at the NW street, there's some shadow hiding those in the SW one). If there are no other suggestions, I'll try to think at something :/ You also referred to http://en.wikipedia.org/wiki/Continuous_flow_intersection I don't remember ever seeing one of these in the wild. I don't think that the intersection that you are looking at is a continuous flow intersection as described in the wikipedia article. It appears to be a simple cross junction of two one way roads. Is that correct? If so I would map it with a single node at two crossing, one-way, ways. You can't obviously see the vertical signals from the aerial imagery. In the street coming from NW, there is a continuous flow to left -- that means you can go left even with a Red-light-signal. In the street coming from SW, there is a continuous flow to right -- you can go right regardless of the traffic signal once again. They *are* CFIs. Kindly, David -- . ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino : :' : Linuxer #334216 --|-- http://www.hanskalabs.net/ `. `'` GPG: 1392B174 | http://snipr.com/qa_page `- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174 signature.asc Description: PGP signature ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Tagging continuous flow intersections
On Sun, 23 Aug 2009 18:11:59 -0500, David Lynch wrote: I suspect that I'd end up creating a set of ways that look something like this, plus a whole bunch of oneway tags and turn restrictions: http://dl1050.dyndns.org:/images/osm/cfi.png Wouldn't that make routers say turn left and at the traffic signals continue stright ahead, and then at the next traffix signals turn slightly left and finally turn left when it should be turn left at the traffic signals? Konrad ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Tagging continuous flow intersections
On Mon, Aug 24, 2009 at 12:25, Konrad Skerikon...@skeri.com wrote: On Sun, 23 Aug 2009 18:11:59 -0500, David Lynch wrote: I suspect that I'd end up creating a set of ways that look something like this, plus a whole bunch of oneway tags and turn restrictions: http://dl1050.dyndns.org:/images/osm/cfi.png Wouldn't that make routers say turn left and at the traffic signals continue stright ahead, and then at the next traffix signals turn slightly left and finally turn left when it should be turn left at the traffic signals? You pass through one set of signals when you cross the oncoming traffic lanes, and another set when completing the turn onto the cross street, so I would probably describe it as turn slightly left at the first traffic signal, go (distance) and turn left onto (street) at the second signal. I tend to err on the side of tagging every node where traffic might possibly stop for a red light, and there isn't a good way at the moment to indicate that there are two or more nodes tagged with highway=traffic_signal but only one set of signals. -- David J. Lynch djly...@gmail.com ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
[OSM-talk] Tagging continuous flow intersections
Hello, I'm having some issues with the tagging of the so-called CFI -- continuous flow intersections. These are junctions, often (if not always!) regulated by traffic signals, where vehicles in a particular lane can freely go into one specified direction, disregarding the indication of the traffic signals. http://en.wikipedia.org/wiki/Continuous_flow_intersection http://imagebin.ca/view/SjGkG4.html (sorry for the second link, don't blame me for my talent :-P) Surely I'd put an highway=traffic_signals on the junction node, but how to tag continuous flows? Sorry for not being clearer, please ask if you want/need more details. Kindly, David -- . ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino : :' : Linuxer #334216 --|-- http://www.hanskalabs.net/ `. `'` GPG: 1392B174 | http://snipr.com/qa_page `- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174 signature.asc Description: PGP signature ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Tagging continuous flow intersections
I suspect that I'd end up creating a set of ways that look something like this, plus a whole bunch of oneway tags and turn restrictions: http://dl1050.dyndns.org:/images/osm/cfi.png On Sun, Aug 23, 2009 at 17:33, David Paleinod.pale...@gmail.com wrote: Hello, I'm having some issues with the tagging of the so-called CFI -- continuous flow intersections. These are junctions, often (if not always!) regulated by traffic signals, where vehicles in a particular lane can freely go into one specified direction, disregarding the indication of the traffic signals. http://en.wikipedia.org/wiki/Continuous_flow_intersection http://imagebin.ca/view/SjGkG4.html (sorry for the second link, don't blame me for my talent :-P) Surely I'd put an highway=traffic_signals on the junction node, but how to tag continuous flows? Sorry for not being clearer, please ask if you want/need more details. Kindly, David -- . ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino : :' : Linuxer #334216 --|-- http://www.hanskalabs.net/ `. `'` GPG: 1392B174 | http://snipr.com/qa_page `- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174 ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk -- David J. Lynch djly...@gmail.com ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Tagging continuous flow intersections
On Sun, 23 Aug 2009 18:11:59 -0500, David Lynch wrote: I suspect that I'd end up creating a set of ways that look something like this, plus a whole bunch of oneway tags and turn restrictions: http://dl1050.dyndns.org:/images/osm/cfi.png Thanks for the suggestion -- but I'd avoid drawing different ways for different lanes in a single carriageway. David -- . ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino : :' : Linuxer #334216 --|-- http://www.hanskalabs.net/ `. `'` GPG: 1392B174 | http://snipr.com/qa_page `- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174 signature.asc Description: PGP signature ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
