Talk:Proposed features/traffic jam warning
From OpenStreetMap Wiki
- Instead of just hh:mm-hh:mm I'd also include days and possibly include things like public holidays, you might want to look at opening_hours=* for some more ideas -- Delta foxtrot2 12:04, 26 May 2010 (UTC)
- Also jams=* might be a bit ambiguous, there might be jam shops, what about traffic_jam=yes/time instead? -- Delta foxtrot2 10:38, 28 May 2010 (UTC)
- jams might be ambiguous but I would at least stick to the plural form to make clear that there is no actual jam but the potential for jams. traffic_jams=yes/time would be okay for me. -- SnakeBDD 10:52, 28 May 2010 (UTC)
- It would be useful if you could specify which side of the road the traffic jams would be on. This tag would be especially useful for routing programs. I think that a tag such as traffic_jams=yes sounds like it is saying that there is currently a traffic jam, I also think that jams=* is ambiguous. Maybe a tag such a s queues_likely=yes would be better, I don't know about other countries but in Great Britain "Queues Likely" is what is written under traffic signs no. 124. -- Awardle 2010-05-28 17:16 (UTC)
- I also thought of this road side issue. Many major streets are mapped one track per direction. In that case, one-way jams can be tagged the proposed way. Things get complicated when you only have one track for both directions. You could add an extra tag like oneway_jams=yes/opposite explaining that the jams=* tag only applies to one direction. But most roads are jammed in one direction in the morning and in the other direction in the afternoon, making it difficult to tag jams=time. --SnakeBDD 22:24, 28 May 2010 (UTC)
- For other attributes a commonly used, even if not formally acknowledged, way has been to add :forward or :backward to the key, i.e. traffic_jams:forward=*, just like maxspeed:forward=*. That doesn't describe adequately the case where a lane for, say, turning left is jammed at times, but traffic turning right flows freely regardless of the time. Alv 13:12, 30 May 2010 (UTC)
For four reasons:
- it's not verifiable if there is no permanent traffic sign indicating the jams
- it's dynamic in time: rush hours from monday are not the same of wednesday or saturday.
- it's dynamic in importance. You may wait 5 minutes in a traffic jam up to hours. Are you going to introduce a waiting time in your proposal ? After how many minutes are you considering it as a jam ?
- it's dynamic in geography. Marking the main road as risked for jams is not enough. All adjacent roads are impacted and routing shall be aware of that as well if we continue your logic.
Conclusion : traffic jam impacting routing is a complex science and is dynamic. You try to copy existing commercial features with a workaround which finally will be useless. --Pieren 08:43, 30 May 2010 (UTC)
- I'm not convinced that the proposed criteria is sufficient. A jam is in a way statistical concept, and therefore the verifiability must and can rely on statistical values: if the criteria is required to be fulfilled on, say, at least four out of five (work)days, it's possible to verify any location by surveying it on more than five days. Frequently jammed locations don't disappear unless they, or some other location nearby, is heavily modified physically. Jam hazard traffic signs are apparently erected only where the slow queues can come as a surprise to faster traffic - the permanent ones must rely on a statistical criteria. Alv 10:17, 30 May 2010 (UTC)