Wikipedia talk:WikiProject Geographical coordinates/Linear

Let's draft some recommendations

First of all, good idea, long overdue (we had this a year or two ago on de.wp, but it was inconclusive). Maybe we can draft a few guidelines.

IMO several coordinates per feature are fine, as long as one is designated as the article coordinate.

feature	main coordinate	further recommended coordinates
river	estuary/mouth	source, major confluences
tunnel	center	both portals

etc. --Dschwen 21:54, 2 October 2008 (UTC)

Thank you. I've copied your table to the project page and added to it. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 22:08, 2 October 2008 (UTC)

Whatever rules are adopted (or even if none can be agreed), it ought always to be made clear to the reader what point is being referred to by a given set of coordinates. I once proposed adding a location parameter to the coordinates template that would display this information. For example, instead of just "Coordinates xyz xyz" we could display for example "Coordinates (mouth) xyzxyz" (for a river), "Coordinates (geographical center) xyxyz" (for a country), "Coordinates (town hall) xyzxyz", etc. This idea was rejected as others preferred to use deliberate vagueness in the numbers as a substitute for explicit specification of meaning, but I still think, as an encyclopedia, we should be about giving information rather than withholding it.--Kotniski (talk) 08:13, 3 October 2008 (UTC)

You have a good point; that's often best done in an infobox (some of your examples, though, do not apply to linear features; you may do better to raise them on WT:GEO). Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 10:38, 3 October 2008 (UTC)

For many rivers, neither the source nor the mouth are clearly defined points. When displaying a map for a river, it would be desirable that most of its course would be visible on a map centered on the main coordinate. Using the mouth (or estuary) as primary coordinate that would waste a lot of space. Therefore it has been proposed earlier to use a point somewhere "in the middle". As guidance could be given: the point on the river closest to the smallest circle containing all its branches. Although theoretically not unique, this would in practice be relatively easy to determine. So I propose to change the "main coordinate" in the table accordingly and move the estuary/mouth to "further coordinates". −Woodstone (talk) 14:03, 11 October 2008 (UTC)

The suggestions to use the river mouth came from the Rivers wikiproject; I suggest you seek consensus there, before changing anything here. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 14:23, 11 October 2008 (UTC)

Valleys

Valleys seem to be missing here. --Berland (talk) 12:53, 7 October 2008 (UTC)

I've added them; though you would have been welcome to do so. Also, note that the page says "Linear features include, but are not limited to...". Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 13:01, 7 October 2008 (UTC)

Keep it simple

I am interested that a newbie editor is not deterred from posting an article by the complexity of the process. So I see the instruction - the lowest end- to be good and -the geographical centre determined by taking the geometrical mean -to be with the fairies. If there is an esoteric definition accepted by notable bodies- bung it in the text with comment and references and tag something we can see on Google Maps.

Applying these principles to roads- choose the end that joins the major road. In the case of the D2 that connects the N86 with the N304- the easy rule is to choose the road with the lowest number 44.6103N 4.7193E.

That was in France- but in the UK, it would work. A cursory look at a US road atlas shows this rule working. The 247 in New Mexico,(Lincoln County) runs between the US 54 and the US 285 for example- giving an easy choice.

Canals are a problem- but if you look at the lock gates on Google maps you can see which way the water is flowing. Some canals, are fed from their midpoint and flow to each end- no matter, one end will be lower than the other.

In the UK- railways have an UP and a DOWN line- the UP normally points to London. It is this end that should be tagged- but this must be different in each country. However, in most cases there is a mainline.

OK- just a few thoughts. But we do need a list of simple recommendations that we can publish. ClemRutter (talk) 20:39, 15 October 2008 (UTC)

I also added a reminder that flexibility is needed. We don't need to mark the middle of an object with a precision of millimeters, but we should put our marker on something which represents the object (mark a road outside a tunnel, not put the dot on top of a mountain which happens to have the road inside it). There will be some disagreements over exactly where to mark, but at least there is guidance as to where to try to put the mark. -- SEWilco (talk) 19:31, 21 October 2008 (UTC)

Main coordinates

FWIW, I tend to think in every case we should be making a recommendation as to the main coordinate - the one we would wish to see elevated to the title line. — Preceding unsigned comment added by Tagishsimon (talk • contribs)

That's the main coordinate column in the table on the project page. That said, I'm not sure how sensible it is to give one set of "main" coordinates for a very long feature such as, say, the M6 motorway. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 22:55, 15 October 2008 (UTC)

I, too, am not sure whether it is sensible, but I do always hanker after a title coord. The draft recommendations are at the moment inconsistent; canals get a main whilst railways & roads do not. If we consider short railways, such as Alnwick branch line, a midpoint probably works well. I agree it is more tenuous on the ECML. From the point of view of making recommendations, I think our main column should give advice useful to those dealing with the short cases, even if that advice might not extend to the long cases. --Tagishsimon (talk) 23:12, 15 October 2008 (UTC)

I've made some revisions - please see what you think, now. I have also sugegsted elsewhere that, for very long, yet narrow features, {{kml}} could also generate a "map of all points" link in the title bar. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 23:22, 15 October 2008 (UTC)

Yup, I see them; they're good, thanks. --Tagishsimon (talk) 23:46, 15 October 2008 (UTC)

All features should have a geotag.

The term mid-point needs more work- it is ambiguous - it can be calculated in many ways but would be suitable for short features such as bridges, the centre point would be less ambiguous.

All the suggestions seem to tend to default to using a mid point rather than an endpoint. This is wrong, as it is ambiguous and on examination in the field it is less useful than in theory.

• The road, which may cross many thousand of miles, we should use the end that joins the major road, in the cases of roads that join roads of equal status, then we tag the end that joins the road with the lowest road number. (I've explained elsewhere that it works in US, UK France. For a long road all of these have limited use- but I can't think of any practical use, or a way to define a midpoint.
• The Railway, in the UK usually has a name Chatham Main Line so that town should be chosen. In other cases West coast mainline the principal railway works on that line- i.e Crewe. This is a tentative proposal- that needs more work.
• A river has tributaries, maybe a gorge, then a mouth/ or bifurcation as it joins another water body. It has no defined midpoint. A line can be drawn from source to mouth, measured and divided by 2 giving a midpoint that will not actually be on the river at all. Or one can lay a piece of string along the centre of the watercourse. This can be measured, and divided by 2. As the lower courses will meander, the midpoint will be far closer to one end than the other. As an alternative I propose that the principal coordinate should be at the mouth.
• If a canal has no obvious mouth that will usually be its lowest point that should be used, otherwise use the point in its highest section where it crosses the watershed. This I propose.
• Bridge- mathematical centre point
• The problem of using the centre or midpoint for a tunnel is you can't see it when look at Google Maps. A portal is visible.
• Valleys as for rivers. Though many upland valleys have a point- known as the head of the valley- not all do- and we are back to the problem of working out which is the principle head.
• Tracks and trails- as for roads. For circular walks choose the principal starting point. Needs more work.
• Borders- Many are circular not linear- border of Austria- state border of New Mexico. If it is a short line segment of a border- then principal crossing point. Open to any suggestion
• Mountain ranges, the highest point on the watershed line.

In all cases I am looking for something unambiguous that can be located with the bookmarklet, and something that a guy on the ground, could get out of his car and visit, something that can be photographed, that is not a product of own interesting original research. I maintain that the esoteric or theoretical can and should be explained in the text. ClemRutter (talk) 22:42, 16 October 2008 (UTC)

Many of your assertions are flawed, or false. For instance some canals have no locks, and are all at one level. Some railway lines have no single named town, and no works. Visibility on proprietary system like Google maps should not dictate how we geo-tag tunnels or other features. The use of "mouth" for rivers is already proposed. Clearly, the reference to "borders" in this context is to linear borders, not circular (sic) borders - note that the draft guidance specifically excludes polygons. Using a mid-, rather than an end-, point centres maps correctly. Can you imagine the partisan edit wars over, say, whether the English or French portal of the Channel Tunnel is the "main" one? Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 23:12, 16 October 2008 (UTC)

I have had a short break, and I would like you to justify some of the above. Firstly, canals- please name an example of a lockless canal that has no obvious mouth. Were this to be a significant problem it is solved by simple rewording.

Some railway lines have no single named town, and no works. Could this be why I wrote This is a tentative proposal- that needs more work., while I am perfectly happy to savaged, I do believe that opinion should be read and considered. You have the experience and skills to constructively add or ammend the tentative proposal.

Propriety systems- it ain't just Google. In this case, I expect to be able to find it using a map and compass- not a map and compass and a spade.

Rivers: does that mean you are accepting that there is a consensus here. There were at least two other proposals going the rounds.

You haven't commented on bridges, mountain ranges, valleys and trails and tracks

As a midpoint cannot be uniquely defined- it cannot be used in a proposal.

Elsewhere on this page, this discussion has moved to presenting a linear feature as a set of undefined midpoints/ points of undefined notability, which must tend to an infinite set. Have you any comments on how infinite sets can be accommodated.

All this obscures the issue that out there we have multiple templates pointing the reader to this page (via a redirection), telling them that they will be given clear instructions on how to add a single tag to a page- and we as experienced editors are failing them. So, can we all settle which of definitions we are going to use in October 2008- then those of us that want to refine it can do so at out leisure

A final comment on canals supporting using the portal rather than the highest point from a Railway Engineer of my acquaintance. If I want to take my narrow boat to a canal I find on wikipedia- I am going in at one of the ends- not dropping it in in the middle!

ClemRutter (talk) 21:38, 20 October 2008 (UTC)

To answer your first point: the Netherton Tunnel Branch Canal & Bumble Hole Branch Canal are just two examples local to me. As for your final point, your friend should sick to railways, and leave canal enthusiasts to enter the Dudley Canal at Windmill End junction (which is at neither end); or the BCN Main Line at Horseley Fields Junction (ditto), Factory Junction (ditto), Dudley Port Junction (ditto), Albion Junction (ditto), Pudding Green Junction (ditto), or Old Turn Junction (ditto). Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 22:08, 20 October 2008 (UTC)

So can we focus on the task in hand- of devising rule for the main coordinate for each of the categories above. I thank you for Netherton Tunnel Branch Canal & Bumble Hole Branch Canal but they are adequately covered by the clause If a canal has no obvious mouth. ClemRutter (talk) 09:04, 21 October 2008 (UTC)

Hmm. How about Texas Coastal Bend? What do we do for that one? For now, I just eyeballed a midpoint. Travisl (talk) 03:40, 17 October 2008 (UTC)

That's not really linear; I've added an exception. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 08:24, 17 October 2008 (UTC)

For linear features, the most important objective for a "main coordinate" is to show the whole feature on a properly scaled map centered on it. Using either end as main coordinate would waste a lot of space. Therefore a better solution is to use a point somewhere "in the middle". As guidance could be given: the point on the feature closest to the smallest circle containing all parts of the feature. Although theoretically not unique, this would in practice be relatively easy to determine. −Woodstone (talk) 18:13, 18 October 2008 (UTC)

Let's see how that would work in practice, taking the Loire River (one of the largest rivers of France) as an example. The smallest circle enclosing the whole river would have its center about 50 km north of Limoges. The nearest point on the river would be near Tours. To show the whole course of the river, the scale should be about 1:3,000,000, 47.393611°N 0.689167°E / 47.393611; 0.689167. On such a map in Google Maps, the river is only visible for a short stretch near its mouth, so you still don't know where it runs. The mouth of a river is usually well defined, and rivers are usually at their most significant (width, flow rate, population density) near their mouths, so I would prefer to use the mouth as the main location. Markussep ^Talk 19:00, 18 October 2008 (UTC)

File:SevernBoth.png

Which is most helpful - a map based one the coordinates of the river's mouth, or a map showing the coordinates of all crossings?

You're quite right in the first part; but a map centred on the mouth, and encompassing the whole river, would cover twice the width and four times the area, and make finding the river's outline even more difficult; all the more reason why giving a single point for such a feature is unhelpful to our users; see the section on an an alternative to "main" coordinates, below, for a proposed solution. Again, compare these maps of the River Severn centred on the mouth and showing showing all crossings. Which is most informative? Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 19:26, 18 October 2008 (UTC)

Of course a complete outline is more informative, but it's also a lot of work, and I don't see anyone creating them for all linear features in the near future. Anyway, what do people use the title coordinates for? I'd say to get a quick idea of where the location is. For the complete course of a feature, a drawn map migth be more useful. Markussep ^Talk 19:48, 18 October 2008 (UTC)

"I don't see anyone creating [complete outlines] for all linear features in the near future" - Perhaps they won't, but WP improves incrementally, one small step at a time, and we must provide the capabilities for people to make those improvements. I don't see anyone creating drawn maps for all linear features in the near future, either, but lists of coordinates are the more useful option, because they can be displayed on maps of the user's choosing, accessed individually, independently verified, downloaded as GPX/ XML, uploaded to GPS devices, built incrementally and collaboratively, etc. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 20:08, 18 October 2008 (UTC)

The possibility to create lists of coordinates is already there, as you've demonstrated. Very nice to have, but not essential. This discussion is only about what to use as the "main coordinates". If it has to be one point, I'd rather choose a meaningful point (mouth, source) than a random point about halfway the feature. Markussep ^Talk 20:30, 18 October 2008 (UTC)

My point is that the selection of a single set of coordinates, whether at one end or in the middle, is often not appropriate for linear features; and my proposal takes account of that. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 20:49, 18 October 2008 (UTC)

Valleys

I've added valley with opening suggestions. I shied away from tagging wharfedale, iirc, as I was in want of this advice. I note that valleys are not strictly linear, but then neither are rivers, though perhaps they deviate to a lessor extent. I note that valleys have linear features as options, such as rivers & watersheds; not sure how useful that observation is, but I have listed these as additional options. Happy to see valleys reverted, or amended, as you like. --Tagishsimon (talk) 00:04, 16 October 2008 (UTC)

It might be best to adopt the reverse procedure - tag the article, see if you can get consensus for the method you use, or an alternative, then document the results here. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 23:38, 16 October 2008 (UTC)

I'm not going to have time to play in valeys until Tuesday next: feel free to remove the valley line in the linear document if you wish. --Tagishsimon (talk) 08:31, 17 October 2008 (UTC)

No rush ;-) Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 17:54, 17 October 2008 (UTC)

An alternative to "main" coordinates

As mentioned above, one alternative to having a "main" set of coordinates for very long features (such as roads or rivers), where a single set of coordinates is not appropriate, is to have {{GeoGroupTemplate}}'s "map of all coordinates" link in the "title" position, usually occupied by an article's single set of coordinates. I've requested an experimental template edit at Template talk:GeoGroupTemplate#Title display, which would facilitate this. If that generates consensus, I'll speak to my contacts at Google to advise them that they can map such articles in Google Maps and Google Earth. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 23:36, 16 October 2008 (UTC)

The top right corner of geographically related articles is reserved for coordinate information about the article, not for a link to an external service only. A single point will in most cases still indicate where a feature is located in the world, even if it's linear or more complex. It's essential to keep Wikipedia article layout consistent, so that information can be expected to be found and found in the same position, even after the click on the link. Special additional features should be implemented in GeoHack and/or GeoTemplate, not at the top of articles. --Para (talk) 09:47, 17 October 2008 (UTC)

My proposal would keep layout consistent. Pages would either have a "title" display of Coordinates: 1.23 N 4.56.E or of Coordinates: see all (or somesuch). This is consistent and logical, and avoids "forcing" the use of coordinates for a single point in a feature which may be 100s of kilometres in length, but only a metre or so wide. If, as you claim "top right corner of geographically related articles is reserved for coordinate information" (cite, please); then that could be experesesed by something to the effect Coordinates: This linear feature has multiple coordinates (map). It would only be a single external link until such time as a multi-point equivalent of GeoTemplate is available. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 14:12, 17 October 2008 (UTC)

Is the fact that multiple coordinates can be given for a feature really the most important thing about it, so much that Wikipedia should mention it before anything else? Displaying something else in place of coordinates would be inconsistent with the rest of Wikipedia when the coordinates are known (re coord missing). The circumference of the earth is 40 thousand kilometres, so there are plenty of possible locations for such a 100 km long feature. For many readers the geographical description of a feature doesn't need to be so precise, and a single point along the line will be helpful for anyone who doesn't know where the feature is located. If a feature has multiple coordinates for some reason, that's no reason not to have one of them in the same place all other articles have coordinates. --Para (talk) 16:57, 17 October 2008 (UTC)

I suspect the most important feature of it is that there's no compelling main coordinate, but there is a desire to use the top right title space to display links to, in this case, multiple coordinates. For many readers a single arbitrary point on a very long line is just silly. Shoehorning the single coordinate into that place is also silly. (Just my view, of course). I support the idea of having the ability to provide a link in the normal geolink location to a set of coordinates. And as an example: the ECML. Providing a single coordinate at, say, Kings Cross, will not allow the user to trace the rest of the line and as such is of limited use. Providing a link to a set of ECML coordinates seems to me to be the natural thing to do. --Tagishsimon (talk) 17:03, 17 October 2008 (UTC)

...and doing so where people expect to find the default map link for an article is sensible. There are a number of straw men in Para's post; no-one has said that the existence of multiple coordinate pairs is the "most important" thing; I only suggested mentionig that at all, to satisfy the supposed rule that "top right corner of geographically related articles is reserved for coordinate information", for which he has not provided the requested citation. Who says that the thing mentioned first is the most important? Does anyone think that single sets of coordinates always the most important datum on a page? I'm unclear as to how a map link for a single coordinate pair is more useful to our readers than a link to maps of all the relevant coordinates for a linear feature. Compare these two links, for the "main coordinates" and all coordinates for the same article; and see which you find most useful and informative - and that's for a feature 4km, not 100km, in length. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 17:47, 17 October 2008 (UTC)

Most people in the world have no idea where Netherton Tunnel Branch Canal or ECML are, but if coordinates are where they are with all other articles, people can easily click on the WikiMiniAtlas to find out, or through the coordinates to see the locator map, or finally click on the same "All coordinates" Google link that you'd like to lead them to directly. Just because all sorts of information is available in many services does not mean that all readers are always interested in it, or that we should design all interfaces for the users who are after the most detailed information. --Para (talk) 23:44, 17 October 2008 (UTC)

Does WikiMiniAtlas handle multiple sets of coordinates? Or is your opposition to this proposal related to that facility - which is only available to users who have Javascript enabled - not doing so? This proposal would still cater for your second and third scenarios, as now. I doubt that simply reading coordinates will help most people to visualise the location of a feature; that's why the article's lede should include such information (those mentioned above say "Netherton Tunnel Branch Canal in the English West Midlands, is ..." and "The East Coast Main Line (ECML) is the electrified high-speed railway link between London and Edinburgh connecting Yorkshire, North East England and Scotland."); or a map diagram should be placed in an infobox, or elsewhere in the article. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 10:26, 18 October 2008 (UTC)

The Wikipedia community has chosen to always display coordinates as such and not hide them under link texts or icons, so I would suspect that the main purpose of having coordinates at such a prominent place is to make the information consistently visible. The linking effect of that information would then be secondary, like we have seen with many articles linking to external map services from the external links section despite the link from the coordinates. Wikipedia attempts to be a complete source of encyclopedic information and where information isn't suitable for inclusion, it can be linked following Wikipedia:External links. The silly shoehorning in this proposal is therefore the attempt to add a Google Maps link on the first row of articles, before any other content. Note that the guideline does not have any such recommendation for Really Important external links, and there's no reason to deviate from the rule just to be able to place a link there without any information in itself. How about just making maps ourselves and placing them at an equally prominent position? When that's started, we can work on coord, GeoHack, GeoTemplate and GeoGroupTemplate to make the external services for groups of coordinates more visible. --Para (talk) 23:44, 17 October 2008 (UTC)

The Wikipedia community has made that choice based on the use of single sets of coordinates to represent the subject of an article. The proposal in hand deals with cases where that is not appropriate, and is to develop a demonstrator in order that the community can make a more informed choice about articles on linear features, which are more sensibly represented by multiple coordinates pairs (though it might also apply to articles which list multiple places, such as lists of buildings; something else which the current position does not cater for). Placing a piece of arbitrary information in the top right-hand corner of an article, rather than conveying the correct information, that the article's subject cannot be represented by a dingle point, is unhelpful - indeed, misleading - to our readers. Please don't dismiss the proposal as "silly shoehorning"; it is no such thing and derogatory comments of that nature are far from helpful. The proosal is not to add a Google maps link, but to link to a GeoTemplate style page, as soon as one is available. (One interim solution might be to use the top-right corner to host text saying Coordinates for this linear feature are listed below, with an in-page link to the relevant section header; or to the page's instance of the "maps of all coordinates" template.). Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 10:10, 18 October 2008 (UTC)

Good luck with the Wikipedia:Self-reference! --Para (talk) 11:01, 18 October 2008 (UTC)

I look forward to your proposal to delete all TOCs. Alternatievly, we could just use Coordinates: This linear feature has multiple coordinates. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 11:16, 18 October 2008 (UTC)

WP:NOT

When does a list of coordinates for a feature cross over to WP:NOT? Coordinates may well be used as source data for making maps on Wikipedia, but how far is the textual presentation of information that detailed still encyclopedic? Will Wikipedia eventually have geographical information as detailed as on WikiMapia for example? I agree that it's cool being part of a free high profile project where information of our surroundings is recorded, but is it really notable in an encyclopedia article where the air vents of a tunnel are located? Is the point of such detail really to tell readers where the air vents are, or is it to outline the feature for external use? If the source data isn't notable itself and isn't intended for viewing, why show it? Where is the line, or is it not linear? --Para (talk) 00:11, 19 October 2008 (UTC)

Benefit for reuse

It was mentioned above that reusers of Wikipedia coordinate data should be informed of these possibly new linear features. What makes a feature linear for a machine reading the wikitext, and how are reusers expected to handle these features differently from other inline coordinates? Inline coordinates are in some applications already plotted on the map, but they may not always be wanted, as was noticed with a politically incorrect historical name recently. A connect-the-dots on a map for the points of interest perhaps, but is there enough detail for that, and how does the machine group together the points of a single feature from an article with coordinates possibly scattered all around the article? Many articles have coordinates in a table or a single section, but they're not always coordinates for the same feature, but just coordinates related to the topic of the article. What new features would these efforts amount to when Wikipedia data is viewed elsewhere? A different kind of icon or colour for the group, or perhaps a note that this point belongs to a group of coordinates thus indicated? --Para (talk) 12:56, 19 October 2008 (UTC)

Visibility

How about visibility as being a preferred characteristic for coordinates? The midpoint of a tunnel would be less preferred than an entrance, although either is better than no coordinate. If deciding between two points, the most southerly and then most easterly would be considered more visible (due to northern hemisphere aerial photography custom). Multiple coordinates are still encouraged, but visible ones preferred — while being aware that some markers of invisible things can help make hidden things more visible. For caves, the largest or most public entrance would be preferred over location of internal formations or mathematical centers. Locations which are legally or traditionally kept secret are not preferred, such as an archeological site (while a public museum or interpretive center for the site would be preferred). -- SEWilco (talk) 20:01, 21 October 2008 (UTC)

Mid-point, Midpoint, route

Good, more work is being done on the page. Does mid-point have the same meaning as midpoint, if not which one do we actually mean. What is a route? Isn´t Route 66 a road, or do we mean something like the Camina de Santiage de Compostella? Why the southwest and not the southeast? -ClemRutter (talk) 00:55, 11 December 2008 (UTC)

I don't think mid-point is proper English, at least not modern English. And yes, it has the same meaning: mathematically, the point equally far from the opposite ends. In practical terms, somewhere near the middle which is uniquely representative of the line.

Route is a series of paths, as in a highway route. Or a long distance hiking trail, a series of individual trails which (usually) connect. —EncMstr (talk) 01:39, 11 December 2008 (UTC)

Giving the southwest endpoint is more or less arbitrary. On a Cartesian coordinate plane, bottom left is "most negative"; all points up and right are more positive. Somebody at WT:USRD suggested it, perhaps for the same reason as the sequence of U.S. interstate highway numbering. —EncMstr (talk) 04:17, 11 December 2008 (UTC)

tunnel

The recommendation to give a tunnel's midpoint as its primary point is sometimes problematical, and often requires original research. For example, Vista Ridge Tunnels and long auto tunnels in Europe such as Mont Blanc Tunnel. In all cases, the tunnels curve substantially and there's no real way to know where the midpoint of the tunnel is. Even choosing any intermediate point on the tunnel involves guesswork. However, the ends of the tunnel are frequently visible directly by satellite, thus verifiable.

This Vista Ridge Tunnels satimage shows two entrances and two exits which are offset from each other. About half the tunnel curves. But is the west half of the tunnel circular, or is it elliptical? (Google map data shows one guess, or maybe it's from surveys and not a guess?)

Zooming in on—or visiting—a ten km tunnel's midpoint doesn't help actually find it, at least for 2-D thinking creatures. However, either endpoint is eminently useful in the same situation. —EncMstr (talk) 04:39, 11 December 2008 (UTC)

Yup. Same holds for pipeline, which is a special class of tunnel. --Tagishsimon (talk) 19:03, 27 February 2009 (UTC)

South West

If I have a route that runs from 52N 1E to 51N 2E which is the SW end? Surely it is better to choose the S end or the W end? So can I float this idea, if south west is good for routes, why can it not be applied to Tunnels? As in previous discussions, I generally prefer portals to mid points for irregular long items - but have difficulty in seeing how a defendable definition can be achieved. -ClemRutter (talk) 10:53, 11 December 2008 (UTC)

It is the convention of the U.S. Department of Transportation that E-W roads begin in the west and end in the east. N-S roads begin in the south and end in the north. Many other countries have adopted this convention, however I'm sure it's not universal. That is the origin of that proposal, I have no opinion on weather following this convention is appropriate for this project.

For the record, also per US DOT conventions, weather the route is NS or EW is defined by its number, not actual direction for travel. For example, U.S. Route 6 is a diagonal route from southern California to New England. Because it has an even number, it is an East-West route, even though in some states (i.e. California) its route would be more appropriately described as north-south. This convention is not-so-universal and is not even followed by all US states. However, it is surprising to me how many foreign countries I drive in that have adopted this convention. Dave (talk) 17:13, 15 December 2008 (UTC)

In that case, I've removed - for the time being - the proposal for routes to use the South-West end for coordinates, from this draft policy; since we shouldn't set policy for global articles based on the practice in one, or a few, countries. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 17:31, 15 December 2008 (UTC)

There is nothing wrong with the scheme just the way it expressed. If we can get the wording right then we have a policy.It must however be short, succinct and unambiguous. So, lets have a go.

• Southern entrance, or western entrance when more appropriate. (this is the first draft) ClemRutter (talk) 17:51, 15 December 2008 (UTC)

Andy, I am not sure that the heading of the table is worded correctly. There must be a better way to express the intention of the draft policy. When I have some time I may attempt a copy edit. ClemRutter (talk) 17:51, 15 December 2008 (UTC)

Query-to-map

Please take a look at: https://wiki.openstreetmap.org/wiki/Query-to-map. The project is not ready but it can be good to have this solution in mind. --Kolossos (talk) 20:27, 18 March 2009 (UTC)

Moving from draft to policy

I am just posting this message to revive interest in the project. To stimulate discussion here is an outrageous proposal to debate.

Delete all and replace with: All linear objects should be tagged with the geolocation of their most notable point, often this will be a portal but often it will be it's mid-point.

This would mean that the Ashton Canal would be geotagged at Asda in Ashton-under-Lyne, the Macclesfield Canal in Macclesfield but the Bridgewater Canal at the Barton Aquaduct, and the Manchester Ship Canal at its portal on the Mersey. The Edale Valley would be tagged at the railway station. Most valleys and rivers it would be at the mouth. Bridges -it would be at the point it crossed the midpoint of the valley's river. The Transpennine trail would be tagged where it crossed the watershed, at a mid-point but not the midpoint. The Sett Valley Trail would be at its lowest point in New Mills, which is the mouth of the River Sett- and consistency.

In most cases, the geotag would be placed by an editor who cared about the subject and his/her judgement as to notability is likely to be accurate.

Fixing a policy that delegates the responsibilty of choosing the appropriate point to a Wikiproject or individual removes the pressure from this team- removes the stress, and seems to be in line with the spirit of WP.

--ClemRutter (talk) 21:53, 11 June 2009 (UTC) --ClemRutter (talk) 11:56, 6 September 2009 (UTC)

You're right, let's get this thing finalized. I agree that we shouldn't define everything, but we can keep some of the recommendations as examples (e.g. mouth for rivers, midpoint for canals). Markussep ^Talk 17:30, 6 September 2009 (UTC)

Using coordinates for powerlines and cables

Currently, several articles about electric power transmission systems such as powerlines and cables have been overloaded with geographical coordinates. While geographical coordinates are useful for that kind of articles, there are some drastic examples, such as HVDC Italy–Greece where more than half of the article is just coordinates. Some other articles look very similar. It seems that there is no clear guidance for that kind of situation; therefore, I found like to start this discussion to have some consensus how to have all these coordinates in place and at the same time ensure readability of these articles. Beagel (talk) 17:28, 25 December 2010 (UTC)

Taking HVDC Italy–Greece as an example, I think the #Sites is ok, but #Waypoints of overhead line in Greece may be being overly detailed. But then again, such "waypoints" are quite cool and also quite encyclopedic. So I don't have a definitive answer. What I would do is simply select the clearest/most prominent part of the powerline/pipeline on the map, and add coords to that... Perhaps WP:COORDS could create a new guideline for such scenarios (or maybe it already has)? Hope this helps. Kind regards. Rehman 00:21, 26 December 2010 (UTC)

The solution to preserving readability and tables of unreadable data, is to provie the data in a collapsd format such that the user must expand the section to view it. I'm sure there'll be an example of this sort of thing somewhere, but as I'm on a steam-driven internet connection tonight, I'll not look for one now. --Tagishsimon (talk) 01:16, 26 December 2010 (UTC)

For an article primarily about more than the route of a linear feature—that is, it has content like history, geography, people, battles, etc.—a collapsible table of coordinates is realized by placing the templates {{Collapse top}} and {{Collapse bottom}} around the wikitable. Or, as someone has already done in HVDC IT-GR, use {{Collapse|1=<list of coordinates>}}. —EncMstr (talk) 19:42, 26 December 2010 (UTC)

Good. Neater yet if we could use GPX. Jim.henderson (talk) 11:46, 31 December 2010 (UTC)

Discussion occuring on the Roads project pages

FYI, A discussion about a style guideline used by several Roads projects has evolved into the perpetual debate on how to geo-tag road articles, discussion at: [1] Dave (talk) 23:35, 11 August 2011 (UTC)

Ring roads

Someone has tagged Manhattan Waterfront Greenway for needing coordinates. Is there a standard for geolocating such Ring roads? Jim.henderson (talk) 04:21, 6 August 2012 (UTC)

Marking

My ref is a railway line, I don't think you can drop a pin on it. But, any rail line would have a line map showing the towns it goes through and each of those towns has coords so why do it again? If someone needs to know where it is they aren't gonna look at an arbitrary pin, if they have the need, they'd have the nous to go look for the data.
Having ranted that! wouldn't any large item benefit from a wiki map link with a line or box showing neatly. This ref is to parish boundaries. A dropped pin tells you nothing about what it is inside of but a suburb style google map shows to scale what you're after. --Dave Rave (talk) 23:50, 25 July 2014 (UTC)

For many roads, we use {{Attached KML|disp=inline,title}} along with the appropriate KML file of the route of the roadway. That template calls the KML and adds links to view it on Google or Bing maps. (The other online mapping sites don't yet support KMLs.) The globe icon at the upper right of the article alsp pops up WikiMiniAtlas with the line showing the route from the KML. For an example, look at U.S. Route 31 in Michigan or Capitol Loop. (The latter article has the east and westbound directions shown individually.) It's just one option. Imzadi 1979 → 04:02, 26 July 2014 (UTC)

DUDE! almost first go Oberon railway line --Dave Rave (talk) 07:33, 27 July 2014 (UTC)

Main coordinates for rivers

The current recommendation is to set the main coordinates for rivers to the mouth of the river. In many cases, however, two rivers will merge into a larger river. In those cases, both of the rivers will have roughly the same coordinates, which can be confusing. Thoughts? Kaldari (talk) 20:48, 24 November 2018 (UTC)

Erroneous example

Severn Tunnel no longer works as this page claims as of Special:Diff/1101541494 in 2022. Uncle G (talk) 04:28, 27 December 2023 (UTC)