Talk:Proposed features/Causeway

From OpenStreetMap Wiki
Jump to: navigation, search

Discussion related to the old (2007) proposal (mixed with discussion related to the new (Dec. 2009) proposal)

To confirm, would this also be used on causeways through marshes? Any number of roads in the Fens might use this. Morwen 12:14, 29 March 2007 (BST)
If they were known as causeways, yes I suppose so. There is a railway line over Rannoch Moor, I wonder if it is technically a causeway. Bruce89 17:58, 29 March 2007 (BST)
Causeways are flooded and non-passible at certain times, such as high tide. Smsm1 14:26, 5 May 2007 (BST)
This is not the general case in the US. Here, causeways are built up to prevent flooding except in exceptional situations. That's the whole purpose of the causeway.--turbodog 19:44, 1 December 2009 (UTC)
Instead of using the ambiguous tag man_made=causeway, why don't we use bridge=causeway for causeways that cross a body of water. klightspeed 21:55, 12 February 2008 (AEST)

What about embankment=yes (already in Map Features)? Another possibility is embankment=causeway. --Jttt 09:30, 8 May 2008 (UTC)

embankment=yes could describe the structure that the causeway is built on, but not the causeway itself, which might be built on concrete piles rather than an embankment.--turbodog 19:44, 1 December 2009 (UTC)
nice idea to have embankment=causeway I would use this for ways that are sometimes below water (tides) -- Dieterdreist 11:04, 1 December 2009 (UTC)
Again, flooding is possible, but not necessary. A causeway is built up to prevent flooding, for the most part. If the road is at ground level then it is not likely a causeway. "Water cover" more adequately defines a way that is periodically flooded. --turbodog 19:44, 1 December 2009 (UTC)
I should say "At least in the US." Possibly a causeway has a different definition in other countries (Australia?). That needs to be brought out to avoid confusion.--turbodog 19:57, 1 December 2009 (UTC)
In Australia "causeway" and "ford" are often used interchangeably to mean ford as described at Key:ford and at Wikipedia Ford(crossing). In that case "ford" should be used. --JSmart 21:41, 26 February 2012 (UTC)

A more appropriate approach for this tag would be to make it a property of a highway/railway/waterway, since a causeway is indeed a byway. A causeway can be any of these. "man-made=causeway" is inconsistent with other byway tags such as "bridge=yes". I believe "highway|railway|waterway=*, causeway=yes" is the more appropriate approach, in my opinion, or possibly "highway|railway|waterway=*, causeway=yes|embankment|bridge", since those appear to be the only construction bases for a causeway.--turbodog 19:57, 1 December 2009 (UTC)

coastline, bridge ?

I see two issues on this proposal :

  • distinction from a bridge is not clear ("the mapper will have to use personal judgment." is not enough).
    • Feel free to improve on it.--turbodog 20:28, 2 December 2009 (UTC)
  • what about modifying the coastline as suggested in the mailing list ?
    • I don't think it would be appropriate to modify the coastline for a causeway on pilings.--turbodog 20:28, 2 December 2009 (UTC)
    • I'm no expert on coastline mapping, but I suspect that brings a different set of issues to the table. Such as connecting two islands that were previously distinct, as in the mailing list. I'm sure there are ways to handle the fact that the islands may have distinct names, etc., but would that not be opening up a new can of mapping worms if you make their coasts (as well as that of the causeway) a contiguous coastline. It seems to me that making the coastlines contiguous is getting into micromapping. For those of us who are more interested in mapping navigable ways (at least this particular person), being able to use the causeway designation seems more appropriate, at least in the near term. --turbodog 20:40, 2 December 2009 (UTC)

Plus some questions/remarks: we need some links to real photos, not some roads crossing the see on OSM map (doesn't really help). What is the difference with the already existing tag embankment ? -- Pieren 08:50, 2 December 2009 (UTC)

  • Feel free to add photos. The wikipedia links included do have photos. More variety would be welcome. The precise point is that a causeway is not an embankment. A causeway is a highway or railway. An embankment is a man-made structure, usually of sand, dirt, gravel and/or rock. An embankment may or may not have a causeway atop it. A causeway may or may not be on an embankment. It could also be on pilings.--turbodog 20:28, 2 December 2009 (UTC)

What's the value?

This proposal describes two use cases:

  1. Solid-earth embankments (currently "embankment=yes")
  2. Embankments on pilings (currently "bridge=")
  • No, I have never seen an embankment on pilings. What would be the use of putting a pile of dirt on pilings? I think you have a misconception of the definition of embankment. --turbodog 20:28, 2 December 2009 (UTC)

The net value of this proposal therefore appears to be that the one tag can support both piling and embankment causeways, and that renderers might be able to make the transition between them look seamless.

The huge downside is that it creates two very large grey areas:

  1. What is the difference between an embankment, and a causeway that's an embankment? Is it just that the ground nearby is marshy or underwater? I recently tagged a tramway near me as embankment=yes. Now I have to wonder if it was built as a causeway. For what value?
  2. What is the difference between a bridge, and a causeway on pilings? The proposal admits there's not much difference. Will it have any value to anyone? Will it render differently?

I see lots of semantic hairsplitting, and almost no value. Stevage 10:09, 2 December 2009 (UTC)

  • A causeway is not an embankment. A causeway may be built on an embankment. An embankment can be many other things. This may be semantic, but it isn't hairsplitting. From the perspective of someone on the road or a renderer, there may be little difference between a causeway on pilings and a bridge. If, however, OSM is ever to be used for anything nautical, then someone on a large craft might want to know that the causeway, which is very low to the water, is there, and a bridge, which may be high enough to clear a large boat is a 1/2 mile to the north. Likewise (not that you made this an issue) someone in a small boat would like to know that the causeway is on pilings vs an embankment, and he can probably get under it with his rowboat/bassboat/johnboat/canoe/kayak--turbodog 20:28, 2 December 2009 (UTC)
    • In this last example, a 'person on a large craft' would probably need to know the clearance between the wtaer and the bottom of the causeway anyway. So a clearance = xx tag would be needed. This clearance tag could easily be added to an existing bridge = yes, so I don't see how having a causeway tag adds anything except confusion over whether something is a bridge or a causeway. Dmgroom 23:18, 14 December 2009 (UTC)
      • I was under the impression that the "clearance" tag was limited to the way that had the clearance issue, not necessarily the structure causing the clearance problem. For example, I would interpret that a "highway=*, bridge=yes, clearance=*" tagset would indicate that the highway had a clearance problem with the overhead bridge structure. Talk about confusion...--turbodog 22:11, 15 December 2009 (UTC)
      • I thought Maxheight was used for that. Maybe a 'clearance_under' tag is needed?Dmgroom 14:36, 16 December 2009 (UTC)
      • Yes, you are correct re Maxheight. I'm not sure how well adding a tag to something that is not on the route works, but it does solve the issue of adding a restriction to an area (as in a body of water) where the potential route is not restricted to fixed ways. Definitely worth consideration, independent of whether this particular proposal passes or fails. --turbodog 06:24, 17 December 2009 (UTC)

Causeway and other *way

According with the remarks made above (use of bridge, embankment...). I think also that the term itself could produce mistake.

We have a lot of *way in the OSM semantic: aeroway, highway, railway... Generaly it means a kind of network/equipment/transport. Causeway (I hope I understand the exact meaning that is chaussée in French) would introduce confusion.

--FrViPofm 09:07, 4 February 2010 (UTC)

example of use as an area

See http://www.openstreetmap.org/browse/way/75896966 . I've used it in the context of man_made=causeway and needed to embellish this with natural=land. --Ceyockey 17:37, 4 September 2010 (BST)

Have decided to use man_made=breakwater in this case instead. --Ceyockey 19:46, 4 September 2010 (BST)

tidal, sometimes covered, etc

I come to this conversation seeking answers to an equivalent question elsewhere, but I'm struck by how difficult it is as an issue to solve. For the record, coming to this afresh the system 'causeway=yes' seems by far the simplest (even given the confusion over what one is as it's clearly an inference to the presence of significant water rather than simply an embankment over land). Indeed I've been distracted into adding a section on the Wikipedia article on just this problem. One thing which is clear to me is the need for something indicating occasional water cover. Suggest everyone also follow up conversations happening (if you haven't already) about tidal roads/paths, fords. Strongly suggest people look at what I think is the most valuable proposal about water coverage: Proposed_features/Water_cover (and its associated discussion). Rostranimin 12:08, 2 March 2012 (UTC)