Wikipedia talk:WikiProject UK Waterways/Coordinates + microformats

Coordinates + microformats

For examples of how I see coordinates being used, including Geo and hCard microformats, please see:

• Digbeth Branch Canal
• Dudley Tunnel (with additional OS grid refs.)
• Manchester Ship Canal
• Netherton Tunnel Branch Canal
• Norwood Tunnel

See WikiProject Microformats for more on microformats in Wikipedia. Andy Mabbett 19:27, 13 April 2007 (UTC)

I'd be grateful for people's comments on these. Andy Mabbett 10:01, 14 April 2007 (UTC)

Seems rather too detailed information to me, although it is easy for readers to ignore if they're not interested. Could be useful for features that are not shown on an OS map, for example, demolished structures. I don't think I'd want to see it added to route maps as it would make them unnecessarily wide.

EdJogg 21:50, 14 April 2007 (UTC)

I'd agree, this seems very cluttered. For further arguements and a test case, please see Tinsley Viaduct and talk:Tinsley Viaduct#A poll. L.J.Skinner^{wot|I did} 12:16, 17 April 2007 (UTC)

Discussion of a short viaduct is not relevant to discussion of lengthy canals. Andy Mabbett 12:24, 17 April 2007 (UTC)

I still feel it is cluttered and unnecessary. You have mine and EdJogg's reasons Pigsonthewing. If you must revert, at least leave the fixes in place. The zoom is dreadful, yet again! L.J.Skinner^{wot|I did} 12:33, 17 April 2007 (UTC)

PLEASE do not start arguing here over the provision of coordinates, as at talk:Tinsley Viaduct#A poll. Andy, you are obviously very passionate about including multiple coordinates in articles, but the evidence from the talk pages I've seen suggests that your views are very much in the minority. I can see some benefit in providing a discreet (and discrete!) table of coords, provided that they are kept in proportion with the size of the article and the article's subject matter. Furthermore, where something like a canal or railway line links to detailed articles about things large enough for their own coords, the parent article should not contain duplicates of the coordinates.

Your attention to coords is obviously 'good faith' (as is this message intended to be), even if others do not agree, and it would be a great shame if this issue were to drive you away from WP editing.

So, as this is so important to you, may I suggest that you take your ideas to the WP discussion areas / policy makers, so that a true concensus may be reached. And, while that discussion is taking place, you'll be able to channel your energies more productively towards this project.

EdJogg 13:45, 17 April 2007 (UTC)

Editing wars don't solve differences in opinion. A coordinate table may not be appropriate to Tinsley Viaduct but it is to a canal or a railway. You cannot give one coordinate for a canal and say that's where it is. A canal needs the coordinates of its beginning, its end and notable features in between. How this information is presented is a secondary issue. I, for one want to know where notable features on a canal are, coordinates are a sensible way of providing this information. Martin Cordon 15:07, 17 April 2007 (UTC)

I agree with Martin Cordon entirely. The Tinsley Viaduct case is completely different. Adambro 13:12, 18 April 2007 (UTC)

So do I. A coordinate table would be arguably a requirement for canals, railways, (long distance) footpaths (for which similar route diagramming facilities need to be developed as are being developed here), and also for some notable roads. It is the extended nature of these "extended-location entities" which is a major distinguishing factor between them and the "point-location entities" that makes a coordinate table so useful for them, and not for the other kind.  DDStretch  (talk) 14:14, 18 April 2007 (UTC)

Co-ordinate tables are not very human-readable. Encyclopedias, by contrast, are generally human-readable. Whilst I admire the aims of the microformats project, I am not at all persuaded that Wikipedia would benefit from the inclusion of large tables of numeric data into articles. I have still not seen a good explanation of why co-ordinate data cannot appear in the individual articles describing the point locations traversed by railways, canals, etc. AlexTiefling 14:35, 18 April 2007 (UTC)

I have no trouble reading tables of coordinates; nor, I'm sure , do people who use maps, GPS units, tor who studied geography. And by putting them in a table they're easily skipped by people who can't read them - just as I skip the, to me, incomprehensible musical notation in Symphony No. 5 (Beethoven). many of the points do not have, nor warrant, individual articles. Side by side comparison of coordinates enables direction and distance to be judged; and allows the automated downloading of sets of waypoints as a bonus. Andy Mabbett 14:43, 18 April 2007 (UTC)

I studied geography, and can read and use maps, thank you. It's the simple appearance of blocks of numbers without apparent differentiation that's not very human-readable. Being able to find a location from an OS grid reference is not the same as being able to extract, with any useful speed, information about the shape and whereabouts of a feature from the co-ordinates of selected waypoints. You are right that putting such data in a table makes them easy to skip. I have no opposition to the inclusion of co-ordinate tables such as those you have already included - I would simply skip them if uninterested. However, for this reason I would ask you not to include them in rail and canal routemaps - because I am interested in those, and do not want to skip them! AlexTiefling 14:45, 18 April 2007 (UTC)

Thank you for the clarification. You can, surely, ignore a column (and I suggest they be kept to the right-hand side, for that reason) just as easily as you can skip a table. Andy Mabbett 14:54, 18 April 2007 (UTC)

Thank you, both for your words of encouragement (and the barnstar!) I definitely think we need an option to add coordinates column (and an OS grid ref column, for that matter) to route diagrams, so we can merge, for example, the two tables on Netherton Tunnel Branch Canal. I don't have sufficient knowledge of template coding, though :-( Andy Mabbett 14:40, 18 April 2007 (UTC)

Adding (visible) coordinates and OS refs will make the route maps very wide (they can have upto 5 icons side-by-side now), causing them to dominate the article. Also, the coords/references will occupy more space than the 'useful information' (ie canal feature icons, mileage, description) that is the reason for adding the route map in the first place. For a short map, such as Netherton Tunnel Branch Canal, you could get away with combining the two tables, and it would improve the article. However, for longer routes, where the route map will occupy one or more screens height-wise, the routemap will become the article -- anyone viewing on a narrow window will see nothing else. The obvious alternative, having a coordinate table towards the bottom of the article, ensures that all the references are kept neatly in one place, and allows people to read or ignore them as appropriate (they will ignore the entire table, not just the coords). The majority of readers, I would guess, are NOT interested in coordinates, and will find our routemaps too cluttered if the extra information is added.

I have no objection to coords/refs being added to the maps provided they are only visible on user request. They could be presented as a little icon that the interested reader can click or hover-over to find out more. However, the best solution would be to have them normally hidden, and provide a control that the user could click to select display of the coordinates – that way, both camps are satisfied, although the required template coding could be 'interesting'!

EdJogg 16:05, 18 April 2007 (UTC)

Your suggestions would breach accessibility guidelines - a simplrer and more accessible way to have them "available on request" would be on a separate, linked article. I would have no objections to this in principle, but we need to see how such tables and pages would work in practice. Andy Mabbett 16:12, 18 April 2007 (UTC)

EdJogg is on the right track, a column showing latitude and longitude in degrees, minutes and seconds is space hungry. Decimal degrees piped to a more accurate figure might be better. Some samples to look at might help this discussion along. Martin Cordon 16:16, 18 April 2007 (UTC)

Coord allows for input in either format; and output to the user's preference (defaulting to the format used for input). Andy Mabbett 17:20, 18 April 2007 (UTC)

(after several edit conflicts)I agree with Alex concerning the appearance of coordinates tables - they are usable by anyone that can read coordinates, but not readable in the sense that we read an encyclopedia. However, I disagree with him about routemaps and separate tables. Merging the routemap and coordinates table is definitely better than a separate table. If the coords cannot be included nicely in the text or something like a routemap (or something else that has a reason to exist without the coordinates), they probably shouldn't be there. As for having the "option" of an extra column, I really think it's better to first have at least one example where we have consensus that it would be useful, so that we're not just putting off the arguments until the option is actually used. JPD (talk) 16:18, 18 April 2007 (UTC)

It seems to me that route maps would be far more useful than coordinates. Look for example at Sheffield Outer Ring Road. The coordinates are not necessary for every intersection or notable crossing, but the routemap works well.

It would be possible to make a hidable table, indeed, I have been able to do so. L.J.Skinner^{wot|I did} 01:12, 19 April 2007 (UTC)

A map may be more useful to you than coordinates. That does not mean that it would be more useful to everyone, or that coordinate would be more useful to noone. It is possible to drink boiling water; that doesn't make that a good idea, either. Andy Mabbett 08:12, 19 April 2007 (UTC)

Obviously your statements are correct, Andy, but perhaps you could be more specific, and give an example of a way that plain text co-ordinates would be used by someone. It would be most helpful if the example is something expected of an encyclopedia, rather than some other source of information. JPD (talk) 10:26, 19 April 2007 (UTC)

Actually, I think we need to see a range of possible displays, varying both the content and the presentation of information relevant to this particular discussion. That way we can all better see what the various proposals would look like on the page and thus become more informed in our opinions. It would be a more rational way to proceed, I suggest. And, what's more, perhaps by seeing them, a new alternative and better way (either content, presentation, or both) will occur to some or all of us that would be better than any which have been suggested so far.  DDStretch  (talk) 10:33, 19 April 2007 (UTC)

I'm very happy to proceed on that basis. Andy Mabbett 10:51, 19 April 2007 (UTC)

What? I just said I had created one of those possibilities - "It would be possible to make a hidable table, indeed, I have been able to do so.", which would keep all the coordinates as they are, in a format that does not overpower the article, but that could be expanded with one click. And you rejected it outright with a snide remark - "It is possible to drink boiling water; that doesn't make that a good idea, either". But when your buddy suggests it, it's a great idea! Why are you so unwlling to work with me that you will not accept my help in trying to keep your eighteen coordinates in the MSC article? I am looking for a compromise here. You seem to be trying to push an agenda. L.J.Skinner^{wot|I did} 13:59, 19 April 2007 (UTC)

"What?" - DDStretch referred to "possible displays". You're sugegsting not displaying content. Your "buddy" insinuation is unwarranted and unacceptable, as are your other personal attacks. Andy Mabbett 14:36, 19 April 2007 (UTC)

No Pigsonthewing, I was suggesting this. Try asking before you making assumptions. I feel the buddy comment was fine. It is known that you two are friends. I can tell, because you don't resort to personal attacks when he makes a reasonable comment. I ask you again, why are you so unwilling to work with me? L.J.Skinner^{wot|I did} 16:48, 19 April 2007 (UTC)

Further to my above comment, time-stamped "14:43, 18 April 2007", plain-text coordinates can be manually copied or cut-n-pasted into any of a myriad of available websites and software utilities. They can be manually entered into GPS devices. They are included on printed pages. They enable the user to form a mental picture of the relative positions and distances between sites, or to perform more precise calculations of such. They enable manual look-up on paper maps. I don't understand your latter criteria; not least because WP is reinventing the concept of an encyclopaedia. Andy Mabbett 10:51, 19 April 2007 (UTC)

That's a great idea, but I'd still like to be informed about the reasons for adding so many coordinates to an encyclopedia article. It's a lot easier to rationally consider different options when you have a better idea of what the aim is. JPD (talk) 11:25, 19 April 2007 (UTC)

(after edit conflicts)Wikipedia is reinventing the concept of an encyclopedia to a certain extent, but there are still limits. We are not at all including just any sort of information, in any format. For example, I don't think WP is meant to be a source of directions, but there are probably many opinions on this. More strongly, I would label inappropriate anything that needs to be copied into some other utility or have calculations performed on it to be useful. An encyclopedia is not a place for raw data. If we can display the data in a meaningful way, that is another matter. JPD (talk) 11:25, 19 April 2007 (UTC)

"I would label inappropriate anything that needs to be copied into some other utility or have calculations performed on it to be useful" - As I explained in my preceding comment, they don;t need that at all. Andy Mabbett 14:36, 19 April 2007 (UTC)

Actually, all of your examples (other than "included on printed pages", which isn't an example of being used at all) involved something that I would consider either copying or calculating, but even if they didn't, my point stands. The point is that the "myriad of available websites and software utilities", "GPS devices", and to a lesser extent the mental pictures formed, are all irrelevant or at most of secondary importance when deciding whether the material is appropriate for an encyclopedia. JPD (talk) 15:31, 19 April 2007 (UTC)

Cutting and pasting coordinates is not necessary in Wikipedia. Click on the coordinate and it takes you to a new page where a selection of web based map and aerial photo resources is presented. You can then choose the most appropriate for automatic display of a map or aerial photo of the object concerned. The coordinate link seamlessly connects Wikipedia to many other useful online resources. When one is dealing with geographic content such as canals and waterways coordinates are the easiest way of seeing how these objects fit into their environment. Have a go. Martin Cordon 11:42, 19 April 2007 (UTC)

Please don't be condescending. I have probably been making use of coordinate links on WP longer than you have. I fully support the inclusion of at least one set of co-ordinates on articles about a geographical feature, as a useful external link. However, it doesn't automatically follow from this viewpoint that tables of coordinates are a good idea. External links to maps may indeed be useful, but they are still on a level similar to raw data. The article should still be a an article, not a guide to a map. JPD (talk) 12:09, 19 April 2007 (UTC)

External links

Incidentally, would these tables fall under [[WP:EL]? as the page says, "This page in a nutshell: Adding external links can be a service to our readers, but they should be kept to a minimum of those that are meritable, accessible and appropriate to the article." Is 18 links, all to coordinates a contravention of this? L.J.Skinner^{wot|I did} 13:59, 19 April 2007 (UTC)

"would these tables fall under [[WP:EL]?" No. Andy Mabbett 14:36, 19 April 2007 (UTC)

Of course the external links fall under the External links policy. The co-ordinate links are a very special case, and the guideline was not written with them in mind, so the policy shouldn't be applied hastily, but it should be kept in mind. JPD (talk) 14:52, 19 April 2007 (UTC)

None of the links created by {{coord}} is external. Andy Mabbett 14:55, 19 April 2007 (UTC)

All the links are external on several different levels. Firstly, the direct link is to http://tools.wikimedia.de/~magnus/geo/geohack.php which is on wikimedia servers, but not in any sense an internal part of the en-Wikipedia project. Secondly, that page itself is simple a collection of external links, showing not only that such a page wouldn't be accepted as a Wikipedia page, but also that the intent of the original link is purely to provide access to external websites. The use of these links is quite contrary to Wikipedia practice in dealing with nearly all other external websites, and while I think it is worthwhile, it needs to be kept in perspective. To think of them as internal links is pushing the boundary way too far. JPD (talk) 15:19, 19 April 2007 (UTC)

Uh huh: http://en.wikipedia.org/w/index.php?title=Template:GeoTemplate Andy Mabbett 15:23, 19 April 2007 (UTC)

Is this relevant? It is a template, not used anywhere in the main namespace, which to some extent duplicates the external link used by {{coord}}. It's contents, as I already said, are the sort of thing that would be immediately rejected in just about any other context on Wikipedia. I think it is good that it is treated as a special case, and am not using slippery slope arguments to speak against it, but as I said, once you treat it as a normal internal link, you are hitting the bottom of the slope, and I disagree. It is a facility enabling users to access external, non-encyclopedic sources; a facility that is great as long as it is treated accordingly. JPD (talk) 16:26, 19 April 2007 (UTC)

Sub-project Required?

This discussion on coordinates is in danger of swamping the project, as people are discussing little else at present. My views are adequately described among the wildly differing viewpoints above, so may I suggest some ideas to progress this discussion towards a conclusion:

1. Move the coord discussion from here to a sub-page -- project members may then watch the main project talk page without having to see the coord discussions, if they wish :o).
2. Provide appropriate examples of how the coords will actually be displayed and used.
3. Advertise the suggestions more widely at Wikipedia:Village pump (proposals) (or elsewhere, if more appropriate) to gauge the views of non-project members

EdJogg 12:01, 19 April 2007 (UTC)

• Moved to here. Andy Mabbett 15:25, 19 April 2007 (UTC)

Adding coord to map template

It is a simple matter to include a template on the end of the BS rows, just type:

{{existing BS template content|{{additional template content}}}}

EdJogg 12:01, 19 April 2007 (UTC)

Netherton Tunnel Branch Canal (distances in miles)
0 BCN - Dudley Port Junction 52.52204°N 2.04331°W / 52.52204; -2.04331 0.5 Northern tunnel portal 52.51534°N 2.04948°W / 52.51534; -2.04948 Aston Road air vent 52.50850°N 2.05518°W / 52.50850; -2.05518 Regent Road air vent 52.50697°N 2.05708°W / 52.50697; -2.05708 99 Approx. tunnel mid-point 52.50435°N 2.05932°W / 52.50435; -2.05932 Southern tunnel portal 52.49335°N 2.06916°W / 52.49335; -2.06916 2.4 Dudley Canal - Windmill End Junction    52.49157°N 2.07019°W / 52.49157; -2.07019

Thank you. As can be seen from this short example, which uses the format {{BS|utKMW|99|Approx. tunnel mid-point}}|{{coord|52.50435|-2.05932|region:GB_type:landmark}}, there's no padding nor does it seem possible to add the necessary HTML class names for hCards (and thereby label each Geo with the respective feature's name). Andy Mabbett 15:28, 19 April 2007 (UTC)

I wouldn't worry too much about the hCards while we are just discussing possibilities. Presumably if we can come up with a visual output that everyone is happy with, it can be modified later to allow for hCards. JPD (talk) 16:18, 19 April 2007 (UTC)

And yet the numeric data, which does not suggest anything to me as text, is as wide as the descriptive text beside it. I know how to read co-ordinates, but I still don't have the slightest idea - without leaving the page and following the link - where in the world the feature is. AlexTiefling 15:52, 19 April 2007 (UTC)

<sigh> Full table added. By looking at those coordinates, I can see, without reference to a map, that the route diagram is arranged with North at the top, and that the canal runs with a slight NE-SW bias. Andy Mabbett 16:03, 19 April 2007 (UTC)

You are obviously very used to working with lat/long coords to be able to see that. To me they are 'just a string of numbers'.(I'm being honest, not provocative – my understanding is limited to knowing that Greenwich is at 0 :o)). BTW -- I've added some non-breaking spcaes to the last line of each table, so that you can gauge the effect. EdJogg 23:38, 19 April 2007 (UTC)

Yes, it does convery information, but not in a particularly efficient way. As for layout, the coords do take up an exceedingly large amount of space, although not as bad as it could be. JPD (talk) 16:18, 19 April 2007 (UTC)

Using parameter 5

Netherton Tunnel Branch Canal (distances in miles)
0 BCN - Dudley Port Junction 52.52204°N 2.04331°W / 52.52204; -2.04331 0.5 Northern tunnel portal 52.51534°N 2.04948°W / 52.51534; -2.04948 Southern tunnel portal 52.49335°N 2.06916°W / 52.49335; -2.06916 2.4 Dudley Canal - Windmill End Junction    52.49157°N 2.07019°W / 52.49157; -2.07019

Not sure what paramter 5 is normally used for, but just to give an idea of that it would look like with right justification and smaller text, I present another version. JPD (talk) 16:34, 19 April 2007 (UTC)

That's not so bad. My absolute favourite (if co-ordinates are called for) would be OS grid references, rather than Lat/Long co-ordinates, in field 5. That's the right-aligned note field, yes? AlexTiefling 16:37, 19 April 2007 (UTC)

I prefer decimal lat/long; but I've been involved in separate conversations with editors who wanted to see both included. Andy Mabbett 16:39, 19 April 2007 (UTC)

Thank you. The layout is better, and param 5 at least shows what I've been talking about is possible; but I find the smaller text hard to read. Generally, it's bad practise (in terms of accessibility) to reduce text below 90% of the user's default setting. Andy Mabbett 16:39, 19 April 2007 (UTC)

From Wikipedia:Railway_line_template:

• Parameter5: Note
• Parameter6: Right-aligned note

Andy Mabbett 16:44, 19 April 2007 (UTC)

That's BS2. By analogy, Param5 of BS is presumably the right-aligned note. It certainly looks that way. AlexTiefling 16:48, 19 April 2007 (UTC)

Yes, it's the right-aligned note. What sort of things are usually included as a right-aligned note? JPD (talk) 16:50, 19 April 2007 (UTC)

You're right. Struck out. Andy Mabbett 16:55, 19 April 2007 (UTC)

"Not sure what paramter 5 is normally used for".

Firstly, to point out that it is the last parameter: parameter '5' for 'BS', parameter '6' for BS2, etc, upto BS5 (and beyond?).

Secondly, AFAIK there is no agreement about the use of the right-hand parameter. In one or two places, I have used it to indicate that a branchline (junction) is closed.

EdJogg 23:38, 19 April 2007 (UTC)

In the both the examples so far, it has just occurred to me that there is a potential perceptual issue that might crop up. When trying to skip horizontal space "by eye", the accuracy can be greatly affected by the denseness and presence of other objects nearby on the page, thus causing slippage onto other lines. In long columns of figures, such as might happen here, with large white space between the end of the text and the beginning of the coordinates, it will potentially happen and thereby affect accuracy of readability. If need be, I can possibly dig out some psychological research that would back this up, but now that I am retired and don't have access to a university library it might be difficult for me to do so. The work of James Hartley who was at from Keele University on reading would be the stuff to look out for if one was moved to track these down. Very light "guide lines" or dots may help, but can make the display look ugly. Just something to be aware of.  DDStretch  (talk) 16:57, 19 April 2007 (UTC)

I believe that CSS3 has a method of applying two different background colours to alternate table rows, which can, of course, be pastel in shade. Andy Mabbett 17:01, 19 April 2007 (UTC)

Tables in own name-space

The proposal to give tables their own name-space may be relevant; and may allay concerns that table mark-up in article wiki-code is off-putting to novice editors. Andy Mabbett 09:57, 21 April 2007 (UTC)

Another reason to use coordinates

FYI: pages marked with 'coord' can be exported as KML or seen on Google Maps. Andy Mabbett | Talk to Andy Mabbett 13:48, 3 August 2007 (UTC)