Jump to content

Wikipedia talk:Only make links that are relevant to the context/Archive 6

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1Archive 4Archive 5Archive 6Archive 7

Example

I thought: "Whats the problem of linking to whatever people want?" Then i saw this example taken of war elephant, an article i started. Notice especially mounted and ancient world. Obviously guidelines are needed here! I support the guidelines. Muriel Victoria 13:12, 27 Nov 2003 (UTC)

I guess I am confused about about what the object is to linking. I find it fascinating that someone would go through all the trouble of linking every word, so that the reader, especially if they are not an native language speaker can click on any word in the article and see what it means is the strength of the Wiki concept. Why remove links that people have put in?

Is it the fact that multiple links slow down the loading of a page? Is it the color of the links?

What is the rational? —Preceding unsigned comment added by 68.174.96.5 (talk) 13:38, 10 December 2007 (UTC)

The ultimate example of over-linking

See this and tell me that we dont need strict guidelines on the matter. Muriel 16:53, 17 Mar 2004 (UTC)

I changed movie theater because it had way too many links, including room! I'm hoping that everyone knows what a room is and, if they don't, are eager enough to type it into the search box. violet/riga (t) 08:46, 5 Sep 2004 (UTC)

a great example

[This version] of an article is, imo, a great example of linking gone wild. Unfortunately, whenever I try to clean it up, I seem to get reverted (at least, on the linking part). ONUnicorn 14:18, 2 August 2006 (UTC)

Yikes. It's pretty hideous. Not only is it overlinked, but the format is terrible as well, as in bad like a headache, a hangover, a motherinlaw, a flat tire... :) MotherFunctor 22:20, 23 August 2006 (UTC)

That's a good "bad example". Maybe someone wants to translate it. 62.116.68.236 14:51, 5 May 2006 (UTC) (de:User:TZM / commons:User:TZM).

Current example

Those are pretty good, but I just love the one that's currently on the page. I just about busted a gut, laughing! Mdotley 17:27, 13 September 2006 (UTC)

Linking of dates

Anyway, just passing by to notify those active on this talk page that something is about to change in the Dates & Numbers MoS, regarding linking of dates, please see Wikipedia_talk:Manual_of_Style_(dates_and_numbers)#P8c, which I suppose is the proposal that's going to make it to the guideline, and might lead to an update of the first bullet of Wikipedia:Only_make_links_that_are_relevant_to_the_context#What_should_be_linked. --Francis Schonken 09:46, 3 April 2006 (UTC)

I'm restoring this text. I think it's premature to say P8c "is the proposal that's going to make it to the guideline". Angr (talkcontribs) 10:09, 15 April 2006 (UTC)
Note that I had created Wikipedia:Only make links that are relevant to the context#Specific issues linking (under the caption "Dates") to the dates & numbers MoS. --Francis Schonken 12:29, 15 April 2006 (UTC)
Francis, you're deleting what the MoS says. Please leave it in. SlimVirgin (talk) 12:41, 15 April 2006 (UTC)

Oh SlimVirgin, always prone to whip up situations as close to edit-warring as you can make them. Sorry, the present formulation is not compatible with the MoS. Apart that the present MoS formulation is partly "suspended" (also by you, supporting BobbleWik's block over applying the MoS - prior to successfully getting the MoS adapted in that sense), and discussed at Wikipedia talk:Manual of Style (dates and numbers) in a way that makes clear that the present unammended version of the MoS is the least popular of all proposed options (P0 at Wikipedia talk:Manual of Style (dates and numbers)#A reworked version of Wikipedia:Manual of Style (dates and numbers)#Avoid overlinking dates), there are also these incompatibilities:

  • "special relevance" and "strong reason", the wording used in the MoS, are here narrowed down to "clearly help the reader to understand the topic". Afaik "special relevance" may include providing useful background info, not *strictly* needed for understanding a topic.
  • "Editors are not required" to link full dates according to this guideline; in the MoS the instruction to always link full dates is clear.

So, I'm going to find me a nice template warning against disruptive behavior, and post it on your talk page. --Francis Schonken 13:19, 15 April 2006 (UTC)

Stephen Turner's (ammended) proposal for the rewrite of the date linking issue has been inserted, by consensus, in the dates & numbers MoS earlier today, so I update WP:CONTEXT accordingly. That is: not by making a summary (which would be quite impossible with the amount of nuance now in the MoS), but, as I did before, by providing a link to the related section in the MoS. --Francis Schonken 07:53, 7 May 2006 (UTC)
I want to report a practical effect of the statement ""Dates when they contain a day, month, and year — 25 March 2004 — or day and month — February 10 — should be linked for date preference formatting". The "Cite web" template followed this advice in developing their template, and now every time we use the Cite-web template parameter for accessdate we get an automated wikilink to the year 2006 (as if that were useful) and to whatever date is put into the parameter for last-accessed-on X/X/XX (as if anyone has reason to know everything that happened on, say, March 25, or June 30, or August 12). If there are fifteen footnotes with cite-web-formatted templates in an article, that means the footnotes are now cluttered with 30 irrelevant wikilinks. The code-writers of this template then refer to Context#date as evidence of consensus for this practice. Maybe it's time to eliminate the wiki-wide assertion that "Dates when they contain a day, month, and year — 25 March 2004 — or day and month — February 10 — should be linked for date preference formatting". Shouldn't date-linking instead be fully dependent on a showing of relevance to the particular topic at hand? I can't imagine any genuine need to automate this kind of auto-linking for dates across the entire wiki. ... Kenosis 20:38, 25 September 2006 (UTC)
Apparently some Americans are disconcerted by seeing Christmas (for example) identified as "25 December", while some Brits don't like "December 25". These people can set their preferences so that any linked date will appear in their preferred format. You're right that the tradeoff for this convenience is some link clutter afflicting the rest of us. JamesMLane t c 18:26, 27 September 2006 (UTC)

I removed this from the project page:

Dates
see Wikipedia:Manual of Style (dates and numbers)#Dates

The linked page shows how to link dates but provides little guidance on when to link dates. Since this project page is about the when and not the how, the link added no useful information. Nova SS 19:43, 27 July 2006 (UTC)

From Wikipedia:Manual of Style (dates and numbers)#Dates:
  • "If a date includes both a month and a day, then the date should normally be linked..."
  • "If the date does not contain both a month and a day ...":
    • "There is consensus among editors that month and day names should not be linked unless there is a specific reason that the link will help the reader to understand the article."
    • ...
--Francis Schonken 16:59, 28 July 2006 (UTC)

Some Thoughts

Some thoughts: I think the bottom line throughout the encyclopedia should be consistency. Create a base and stay with it. I realize that is what a Manual of Style is supposed to create and maintain. What I see are a group of people, each with their own personal style to promote. There can and will be “chaos” in the substance of the material in the encyclopedia since it covers such a wide area of information. However, if the structure, the basic framework that is supposed to contain that information is chaotic, and changes from article to article, you have an unreadable disaster. One of the primary functions of Wikipedia (as in any encyclopedia) should be to organize this chaos.

Our brain depends a great deal on a certain amount of consistency and familiarity in our day-to-day surroundings. If the very basic elements of your life changed day to day you would soon be unable to function. But if the basic elements remain constant, you are able to deal with those elements that are different day to day. Each article in Wikipedia is a day; that basic structure should remain constant as the information changes.

Without this consistency it's like a group of architects arguing about what the basic structure of a building should be; each having their own creative concept; without considering that real people are going to have to navigate it every day. If that basic structure changed from day to day, the effect on the person trying to navigate it would be total disorientation.

Specific suggestions: Constants article to article - Presentation of Birth & Death Dates should be consistent, easily read and located in the same place (that's all the reader may be looking for); Location of Birth should be presented as close to the beginning of the body of the Article as possible (that's all the reader may be looking for); Links to other Articles should be relevant to the specific article it's keyed in. Imagine you're sitting in a library with unlimited resources; you come upon a term, name or other piece of information in the Article you're reading; would you get up, cross the room and pull out another text to look-up that reference? (I know that Pittsburgh is a city; I know it's located in Pennsylvania; is knowing more about the city of Pittsburgh going to enhance my knowledge of the specific subject I am researching at the time, or is it going to be an unnecessary sidetrack? The trunk of the tree is the primary subject you are researching; how many branches can there be before you no longer see that trunk?

I’m fairly new to Wikipedia, and am fascinated by its concept and execution. I plan to contribute as much as I can. More later… Michael David 13:12, 10 February 2006 (UTC)

A link is analogous to a footnote in a print medium.

It seems to me a wiki link is more analogous to a cross-reference than a footnote. Dforest 07:41, 11 February 2006 (UTC)

This is true. Deco 00:20, 13 February 2006 (UTC)
Yes, but really, it's the point that the wikilink is subtle. You only need to click on it if you want to. Personally, I don't find red links annoying, but I think we as editors need to make sure if we have one, there was no where else to put it. --ZachPruckowski 00:14, 4 April 2006 (UTC)
I've been noticing lately that when I'm reading for the purpose of finding information—when I have the "editor" part of my brain turned off—I'm most likely to only click on the most relevant links. But when I'm "exploring" (a.k.a., wasting time, just reading the encyclopedia for recreation) I'm most likely to click on less relevant links.
For instance, just now Pound sterling linked me to Australian pound, which linked me to Australian dollar, then to Australian coins and to Coat of Arms of Australia—nothing irrelevant about those links. But then I clicked Golden Wattle, which sent me to Order of Australia], and to British honours, then Bill Gates, National bank, Alliance (New Zealand political party), Social Credit Party (New Zealand), Social credit, Robert A. Heinlein, USS Roper (DD-147), Novorossiysk, Hero City and the Order of Lenin. I learned a lot of fascinating stuff—but at least half of those links weren't particularly relevant. I hope we don't go too far down the road of eliminating these kinds of links, they're part of the charm of Wikipedia for me. --TreyHarris 05:52, 4 April 2006 (UTC)

Well the kind of aimless "encyclopedia surfing" you just described would never be found in a regular encyclopedia because an encyclopedia's mission is to provide you with the information you came for, and any other relevant information that they know would be helpful to most. With Wikipedia's liberal standards as to when wikilinking is appropriate, it seems Wikipedia's mission is not to provide the reader with relevant information, but simply to provide the reader with as much information as possible. While it is true that nobody is forcing users to follow the wikilinks, they do nurture the very short attention span of today's modern user. The temptation to read a different article than the one they "should" be reading is often too great for many users. Irrelevant wikilinks almost seem to shout, "Why read this article, when you could be reading this one. You could say that everything is relevant to everything else in some way, no matter how obscure, but this thinking reeks of an encyclopedia written by teenagers. How hard is it to determine a standard of relevance for appropriateness of wikilinks. You know the highlighted blue terms in commercial websites that when moused over, turn into advertisements? See the connection? If Wikipedia is a non-profit, why does it advertise itself? —Preceding unsigned comment added by 67.167.120.36 (talkcontribs)

Rename it again, Sam!

Anyone up for renaming this guideline Wikipedia:Keep links relevant? Short, snappy, and actually more accurate, since the guideline is really encouraging the removal, rather than just the non-creation, of irrelevant links. Stevage 16:55, 6 April 2006 (UTC)

Support, for the reasons you mention. PizzaMargherita 17:27, 6 April 2006 (UTC)
Sounds good to me. I hate the length of the current title. But I've gotten used to shortening it to WP:CONTEXT, which would seem a little odd if the new name lost the word context. And "relevance" is usually a synonym for "notability", so people will get confused if we change the shortcut to something with relevance. But if folks don't mind the shortcut/long-title mismatch, that's great. If they do, maybe Wikipedia:Keep links contextually relevant or something... --TreyHarris 18:46, 6 April 2006 (UTC)
Agreed! I suggest a title which not only describes the content perfectly, but it also fits the current shortcut (WP:CONTEXT):
  • New title: "Contextual hyperlinking"
This is a recognized concept on the web, as this Google search reveals.
What is the best way to do the renaming? Since the shortcut remains unchanged, the redirect is the only problem, (or is it?) -- Fyslee 15:18, 16 August 2006 (UTC)
  • oppose. The current title says exactly what I mean when I remove some irrelevant links. — Reinyday, 16:11, 16 August 2006 (UTC)
  • We've changed the title of this page before. While the current title is a bit clumsy, it is very clear. The goal of our policy and guideline pages is to name them in ways that make them immediately understandable to a new reader/editor of Wikipedia. Be bold, Assume good faith, Don't include copies of primary sources, Ignore all rules, Use common sense, Process is important, Don't disrupt Wikipedia to illustrate a point, Sign your posts on talk pages, No original research, Build the web and "Only make links relevant to the context" are clear and simple descriptions that can be easily understood by the 40-something IT expert and by the 12 year old contributor to Pokemon articles. You might have to follow the link to understand all the nuances and specific details of the page but you get an immediate understanding just from looking at the link. They also tend to be action-oriented. In my opinion, having a verb in the title is generally a good thing. "Contextual hyperlinking", on the other hand, strikes me as a bit too abstract for our target audience.
    User:Stevage's suggestion to "Keep links relevant" works for me, though. I'm not worried about the shortcut mismatch. If we change it, I imagine that someone would make a new shortcut too and that the uses of the old shortcut would gradually get moved to the archives.Rossami (talk) 16:09, 16 August 2006 (UTC)

Two new suggestions

I would suggest to add somewhere the following consideration: "It may be advisable to wikilink a term when a reasonable interest is expected in the particular place and the term is not linked in the nearby text. For example (from "pie" article): "Some of these pies are pies in name only, such as the Boston cream pie, which is a cake." (Here a reader might be interested what exactly is the difference between a pie and a cake.)

Any thoughts? `'mikka (t) 18:09, 17 July 2006 (UTC)

I have always found it ugly when in lists some items are linked, while others are not: "Alabama, Atlanta, California, Delaware, Oregon". The main arguments against overlinking is that it distracts from reading and makes links not prominent. IMO in my example both arguments are invalid. On the contrary, irregular linking in a homogeneous list actually distracts the brain: broken typographical monotony unduly catches the eye. I would suggest something like this "in homogeneous enumerations it is reasonable to link all terms in cases when otherwise only several items would be unlinked"

Any comments? `'mikka (t) 18:18, 17 July 2006 (UTC)

Seconded. We should have a different standard for lists. --Ghirla -трёп- 07:19, 18 July 2006 (UTC)
I like this idea. Haukur 17:33, 27 July 2006 (UTC)


How about assigning relevance weight to links?

The debate over whether or not to include a link, based on its relevance to the context, potential to distract some readers, etc., seems to me like an artifact of the limitations of the Wiki software, which only allows a link to exist or not. If the software allowed editors to assign a relevancy weight to each link, readers could then decide which links to display based on relevance (or how to display the links, for example using colors, shading, mouseover text, etc., to indicate relevancy weight). A link on a date (which usually has very little relevance to the context) could have a low relevance weight, and a link on a jargon term essential to understanding the current article would have a high weight. Readers who find date links bothersome could easily filter most of them out, by setting their link-display relevancy threshold at some intermediate value. Editors could still argue about the exact weight to apply to a given link, but the stakes would be lower, since readers could elect to display links of any relevancy weight, down to zero---having one's pet link demoted to lower relevance would be less harsh than deleting it altogether. It stands to reason: if editors have diverse opinions about the "best" density of links, readers probably do too. We should try to avoid, if possible, the one-airline-seat-must-fit-all-passengers syndrome, since software need not have the same constraints as physical objects. We can cater to multiple reader preferences. Teratornis 17:08, 27 July 2006 (UTC)

Major overhaul

I just gave this page a major reworking, mostly trying to move content around to be clearer (such as moving a rule from the "considerations" section to the "rules" section). I also tried to make the examples clearer. — Reinyday, 21:11, 16 August 2006 (UTC)

date section

I am restoring the date section. If you would like to discuss its modification, please do so here. However, per the official policy outlined at Wikipedia:Editing policy which says, "avoid deleting information wherever possible..." please do not delete the section again. Please do not be rude and imply that I am pushing some agenda. I am not. I am trying to be comprehensive and clear. I work very hard to improve the Wikipedia and I don't like to be treated poorly for it. As for User:Francis Schonken's incorrect claim that my changes are "Not even correct about what was 'originally' believed," they are taken directly from this 17 February 2005 version of the page. — Reinyday, 23:33, 16 August 2006 (UTC)

Sorry, for style guidelines the template has "Feel free to update this page as needed, but please use the discussion page to propose major changes" - Guidelines follow that logic. No way around it. If you want to see a discussion about that, I can refer you to Wikipedia talk:Manual of Style (dates and numbers)#Wai Wai's changes (accidently Wai Wai also tried impose a rewriting of the "date linking" guidelines, different from yours - all of it was reverted until consensus would have been reached on talk page, which didn't happen yet).
If you want to read more about how guidelines are wrought, see for example these Guidelines for creating policies and guidelines. Also, wikipedians might get a bit wary of everyone developing his/her own standards: this should primarily be a collaborative effort to develop guidelines and policies everyone can work with. See also Wikipedia:Avoid instruction creep which develops some points of the Guidelines for creating policies and guidelines. And you can always join Wikipedia:Village pump (policy) if there are particular guideline-related issues you'd like to see addressed.
Anyhow, on the date linking issue: note that there have been *extensive* (and I mean really, really *extensive*) discussions on this topic. Several of the archives of Wikipedia talk:Manual of Style (dates and numbers) are filled exclusively with discussions on this topic. The total discussion on this topic must be hundreds and hundreds of kB, most of it produced in the last half year. In the end we could agree with what is in Wikipedia:Manual of Style (dates and numbers)#Dates. Nobody likes it. But the words were weighed, word by word, so that in the end we could agree that that was going to be the version in the guideline. If you want to change that, please discuss at Wikipedia talk:Manual of Style (dates and numbers). There are several sections on that page already discussing the date linking issue (again).
I'm not saying you have a particular agenda. Rather, that you're unaware of the ramifications of the issue. No problem, I showed you the way to prior discussion. I could show you the way to some current WP:AN/I discussion that obliquely relates to this issue too, but maybe that would lead a bit too far.
Also, your summary contains errors, e.g. "It was originally believed that unless the user's understanding of the article relies heavily on the surrounding historical context..." Really, incorrect. The link you make to a February 2005 version of this guideline already signals this was a contentious issue at that time. That version has nowhere the intention to write down what was "originally" believed. It proposes an alternative in the ongoing debate. So I'm going to revert again. IF YOU WANT TO SEE THE GUIDELINES ON DATE LINKING CHANGED, PLEASE DO SO VIA Wikipedia talk:Manual of Style (dates and numbers) Not via a guideline fork. Tx. I appreciate your efforts though. And I liked your other changes to this guideline page. --Francis Schonken 09:47, 17 August 2006 (UTC)
Hi Francis. I appreciate your response but I disagree with many of your statements. To clarify, I am not trying to change any guidelines, only to state the existing consensus and practice. So that it is perfectly clear, Wikipedia:Manual of Style (dates and numbers)#Dates currently begins with "This section describes how to link to years, decades and centuries. See sections which follow, regarding when such linking is appropriate." Subsequent sections have a few recommendations on approprate linking and point to Wikipedia:Only make links that are relevant to the context. I have rewritten the dates section. You are free to edit it if you do not feel it is accurate, but it must remain here to avoid the circular linking of both pages pointing to the other for authority, as happened at Wikipedia:Red link. The MoS is for "how to link" (and describes itself as such) and this page is clearly for "when to link". Thanks for your consideration. — Reinyday, 22:46, 7 September 2006 (UTC)

Additional opinions needed

Hi fellow editors. I am writing to request some input regarding the dates section. Francis Schonken has referred to my edits as a "deformation"[1] which I find incredibly rude, so I was hoping some of you could step in and voice an opinion. Here are two versions of the dates section:

This version by Francis Schonken

Because of the date preference formatting MediaWiki software feature, "how" to link and "when" to link dates can not be treated independently one from the other. Unrelated to that software issue, there is no general consensus that the habit of linking separate years (that are date indications that only consist of a "year") should be abandoned, although most Wikipedians disfavour that habit currently.

Details about when and how to link years can be found in Wikipedia:Manual of Style (dates and numbers)#Dates. An overview of the most frequently occuring cases:

  • Dates when they contain a day, month, and year — [[25 March]] [[2004]] — or day and month — [[February 10]] — should be linked for date preference formatting.
  • Stand alone months and days of the week should generally not be linked.
  • Stand alone years do not need to be linked but some users prefer it, and some users prefer to link (with a piped link) to articles formatted as "year in subject" such as 1441 in art, 1982 in film, and 18th century in United States history.
  • Dates in section headers should generally not be linked.
  • Do not link dates inside quotations.
 
This version by myself

Because of the date preference formatting software feature, currently some dates are required to be linked, though some editors may consider the date links irrelevant to the context. For additional information about how to format dates, see Wikipedia:Manual of Style (dates and numbers)#Dates. An overview of the most frequently occuring cases is as follows:

  • Dates when they contain a day, month, and year — [[25 March]] [[2004]] — or day and month — [[February 10]] — should be linked for date preference formatting.
  • Stand alone months and days of the week should generally not be linked.
  • Stand alone years do not need to be linked and there is no consensus on if they should be. Some users prefer to use a piped link to topical articles formatted as "year in subject" such as 1441 in art, 1982 in film, and 18th century in United States history.
  • Dates in section headers should generally not be linked.
  • Dates inside quotations should not be linked.

So could you please voice some opinions on which is clearer and more properly states the situation? There is additional discussion about this section above. Thanks so much. — Reinyday, 07:37, 10 September 2006 (UTC)

I want to report a practical effect of the statement ""Dates when they contain a day, month, and year — 25 March 2004 — or day and month — February 10 — should be linked for date preference formatting". The "Cite web" template followed this advice in developing their template, and now every time we use the Cite-web template parameter for accessdate we get an automated wikilink to the year 2006 (as if that were useful) and to whatever date is put into the parameter for last-accessed-on X/X/XX (as if anyone has reason to know everything that happened on, say, March 25, or June 30, or August 12). If there are fifteen footnotes with cite-web-formatted templates in an article, that means the footnotes are now cluttered with 30 irrelevant wikilinks. The code-writers of this template then refer to Context#date as evidence of consensus for this practice. Maybe it's time to eliminate the wiki-wide assertion that "Dates when they contain a day, month, and year — 25 March 2004 — or day and month — February 10 — should be linked for date preference formatting". Shouldn't date-linking instead be fully dependent on a showing of relevance to the particular topic at hand? I can't imagine any genuine need to automate this kind of auto-linking for dates across the entire wiki. ...Kenosis 20:53, 25 September 2006 (UTC)
I agree with Kenosis.
Have been looking around and I keep wondering this reasoning of linking refered as "requirement" from software (as much I have now understood) There is #time Parser function (see http://meta.wikimedia.org/wiki/ParserFunctions ) with this one can easily format (as much I have experimented in Wikipedia) the date (for example given in ISO) exact format and add in template the linking making code if needed. So I have hard time to see ground for this reasoning when somebody gives support to make irrelevant links to dates --TarmoK 19:50, 27 September 2006 (UTC)

Francis Schonken, please note that you recently reverted my changes stating that I needed to "use discussion page prior to effectuating changes. If you can't find consensus on a change at the talk page, then the change doesn't go through. Period." This is not correct. Not every change needs to be discussed, and in fact the vast majority aren't. Plesse read Wikipedia:Be bold in updating pages: "If you expect or see a disagreement with your version of the article, and you want to change or delete anything substantial in the text, it's a good idea to list your objections one by one in the talk page, reasonably quoting the disputed phrases, explaining your reasoning and providing solid references."

A misunderstanding that has become almost classical: what you quoted is for article (or: "main") namespace. Not for guidelines. For guidelines (in "wikipedia:" or "project" namespace) the following applies:
BTW, Wikipedia:Only make links that are relevant to the context is part of the MoS (as it is a *style* guideline, as indicated by the {{style-guideline}} box on top of the page. All MoS pages (or, if you prefer, all "style guidelines") can treat as well *when* as *how to* apply certain styles. The distinction you made above in that sense is non-existant. MoS (or "style") pages will even generally rather concentrate on the "when" than on the "how-to": there is a category of guidelines that are exclusively of the "how-to" genre. So, Wikipedia:Manual of style (dates and numbers) can and does concentrate on the "when" to link dates, and these issues are discussed there. FORKING the guideline on when to link dates, as you try to do, is not appreciated. "When" to link dates is part of the dates & numbers MoS. The discussion is there, the guidelines are there. The general rule is not to make contradicting guidelines. --Francis Schonken 08:18, 28 September 2006 (UTC)

I didn't not expect any disagreement with my changes, but I will do my best to explain them here. Surely you don't object to me changing "*Do not link dates inside quotations" to "*Dates inside quotations should not be linked" in order to match the sentence structure of all the other bullet points. So I'm still wondering what you do object to. I've removed "'how' to link and 'when' to link dates can not be treated independently one from the other" because it isn't true. They can be treated independently and in fact are on this very page, where there is no discussion of "how" to link dates beyond an example without explanation. I've changed "there is no general consensus that the habit of linking separate years (that are date indications that only consist of a 'year') should be abandoned", which is clearly biased, to "There is no general consensus about linking date indications that only consist of a 'year'." Everything else was just diction. If you have any actual issues with what I have changed, please use exact quotations and explain them here, so that we can discuss them. But do not delete my changes (or the whole section) just because it isn't "your way".

Lastly, please read the official policies of Wikipedia:Civility and Wikipedia:Assume good faith before making any more comments to or about me or my contributions. — Reinyday, 16:51, 26 September 2006 (UTC)

I'd appreciate you stopped trying to take ownership of this page. And tried to be a bit more civil (writing "removing false statement [...]" in an edit summary about a statement that is in no way "false" [2] is in no way "civil"). I'm going to revert, until you're able to demonstrate consensus that the "when" to link dates should be moved from Wikipedia:Manual of style (dates and numbers) to Wikipedia:Only make links that are relevant to the context. --Francis Schonken 08:18, 28 September 2006 (UTC)

In What generally should not be linked?:

  • Individual words when a phrase has its own article. For example, link to "the flag of Tokelau" instead of "the flag of Tokelau". Such a link is more likely to be interesting and helpful to the user, and almost certainly contains links to the more general terms, in this case, Latin.

Sorry, but... huh? Waitak 06:39, 25 August 2006 (UTC)

Sorry. Missed that during the cleanup a few days ago. Fixed. Rossami (talk) 13:42, 25 August 2006 (UTC)

Old talk page at a redirect to this page

For the benefit of those reading this talk page and discussing this issue, another discussion took place at Wikipedia talk:Quotations should not contain wikilinks, the front page of which was recently redirected here. This is mentioned in the archives, but I am bringing it up here to discuss whether the talk page over there should be moved (not copied - that doesn't preserve the edit histories) to a subpage of this talk page - ideally linked to from the archives box at the top of the main talk page. Does that sound reasonable? Carcharoth 12:04, 7 September 2006 (UTC)

  • This makes sense to me. — Reinyday, 22:46, 7 September 2006 (UTC)
  • That talk page is pretty much shut down now. Wouldn't it be easier to leave it where it is but put a link to it in the Archives box at the top of this page? If you pipe the link above, it would achieve the same result but without disrupting anyone who may have watchlisted it at the current title. Rossami (talk) 00:16, 8 September 2006 (UTC)
  • This makes sense to me as well. I also included a link to the talk page from the archive of discussions of links in quotations. — Reinyday, 02:54, 8 September 2006 (UTC)

user:Francis Schonken recently commented out this clause from the page with the comment "Avoid making links within quotations" is, as far as I can see from prior discussion, not an agreed principle.

Avoid making links within quotations, placing links in the surrounding text of the article wherever possible.

I have restored the clause pending discussion here. I participated in one of the prior discussions and have now reread both of them. I think that this represents the consensus opinion of the participants specifically and of community practice generally in a reasonable fashion - that such links should be discouraged but are not absolutely prohibited.

The clause was added after a fairly lengthy debate at Wikipedia talk:Quotations should not contain wikilinks. That was originally a proposal that we should establish an absolute rule against all wikilinks in quotes. There were very strong arguments made against the absolute rule. I would agree that there was no consensus for the proposal as it was originally worded. But just about every editor who commented in depth during the discussion included a disclaimer to the effect that links within quotes should be avoided in favor of a link in the nearby text whenever that was possible and convenient.

The same topic was also discussed at Wikipedia talk:Manual of Style/Archive 37#Wikilinks when quoting where they reached much the same conclusion. They should be discouraged but not absolutely prohibited. Rossami (talk) 00:55, 8 September 2006 (UTC)

"Avoid" as in the phrasing you proposed is too strong, and not warranted at all by the outcome of the discussions at Wikipedia talk:Quotations should not contain wikilinks. I read that discussion too. --Francis Schonken 07:53, 8 September 2006 (UTC)
Read that discussion now too: that discussion doesn't warrant a general "avoid" either. --Francis Schonken 07:56, 8 September 2006 (UTC)
"Avoid" is pretty soft wording, especially when qualified with the "whenever possible" part of the phrase that immediately follows. It doesn't say "can't", "mustn't" or even "shouldn't". It's the kind of wording that's routinely used in our style guides. But if you think it's too strong, what alternate language are you proposing? Because deleting it altogether would be a serious mischaracterization of those prior discussions. Rossami (talk) 13:13, 8 September 2006 (UTC)
  • Perhaps it's best to choose another wording such as "consider links in quotations carefully" or "avoid frivolous links in quotations". Deco 12:08, 8 September 2006 (UTC)
    • How about "Try to avoid making wikilinks within quotations. Often, the destination topic can or should be referred to in text near but outside the quotation, and the wikilink can be put there instead." On a related note: in printed text, sometimes a bracketed gloss is included within a quotation. In Wikipedia, should we discourage those in favour of wikilinks? My feeling is that there should still be a gloss, and the wikilink if needed can go from that. jnestorius(talk) 01:33, 24 September 2006 (UTC)

I recently found a good example where it is helpful to use links in a quotation. See Nobel_Prize_in_Physics. The short statements used by the Nobel Prize Committee to describe the reasons for giving the Prizes are notorious for being filled with precise technical terms that are often impenetrable to the layperson. Wiki-linking from the quotes is a classic casse of making them easier to understand. I still object to normal explanatory links from quotations where they disrupt the flow of a quote, preferring a gloss in the surrounding text if needed. Carcharoth 14:50, 4 October 2006 (UTC)

Parts of discussion below moved from Wikipedia talk:Manual of Style (dates and numbers) Parts of discussion below moved from Template talk:Infobox City
I'm not sure why the wikilink m was removed, but I've reverted the template. Please explain if the wikilink really should be taken out. -- SatyrTN (talk | contribs) 16:35, 3 September 2006 (UTC)

It appears he's using some sort of automated javascript to make these changes. See User:Bobblewik/monobook.js/unitformatter.js and Special:Contributions/Bobblewik. --MattWright (talk) 16:53, 3 September 2006 (UTC)
Thanks for asking the question on his page - I didn't even think to do that :) -- SatyrTN (talk | contribs) 17:06, 3 September 2006 (UTC)
Links add value when they are used for unusual terms. Common terms do not need a link in each article. That is what I think anyway and it is in line with what some other people say. What do you guys think? bobblewik 19:44, 3 September 2006 (UTC)
I can see that the Manual of Style doesn't really address wikilinking units, just which units to use. But it does say that the first instance should be linked, which this is. I guess I'm for keeping it. Others? -- SatyrTN (talk | contribs) 20:23, 3 September 2006 (UTC)
I was sure that I saw somewhere or somebody that said common units did not need to be linked but I can't find a reference, although Wikipedia:Only make links that are relevant to the context is useful. Some units are just plain english words.
With regard to your comment about a requirement to link the first instance of a unit, can you quote it (I can't find it).
bobblewik 21:08, 3 September 2006 (UTC)
The link in this case does not 'remove value' and the infobox is careful to only link the first occurrence of each measurement unit. All four measurements are linked once -- km², sq mi, m and ft. Some are more common than others. If a gradeschooler is reading an article on Denver, why not let them click on 'm' in the infobox if they don't know it means meter. In fact the link adds more value than just the page linked as well, because it is linked on 'm', an abbreviation, and in most browsers, hovering over it will popup 'metre' as the title, explaining the abbreviation. --MattWright (talk) 23:21, 3 September 2006 (UTC)

I agree that common units don't need to be liked. Maurreen 21:33, 3 September 2006 (UTC)

Is "meter" considered a "plain english word"? The quote I was referring to is the third bullet on What generally should not be linked.
I strongly advise against linking common words such as "metre". Nothing stopping a complete ignoramus reader from keying the word into the search box. Otherwise, WP will end up spattered in blue. Tony 02:09, 4 September 2006 (UTC)
Note that the original discussion (although not the title of this section) revolved around linking of the abbreviation 'm' one time in an infobox to the article metre, not linking the word metre. It's slightly different. I agree that in an article there is no reason to link the word metre. --MattWright (talk) 02:13, 4 September 2006 (UTC)
Thanks SatyrTN, for pointing out where the quote came from, I see it now.
I agree that we should have different thresholds for symbolic forms and full forms. For example, the symbol 'ci' in Chrysler Valiant probably needs a link to explain but 'cubic inches' probably does not.
bobblewik 20:00, 4 September 2006 (UTC)

Can anybody suggest what would be regarded as common terms? Consider the following metric terms:

Could somebody propose suitable wording for a bullet on: What generally should not be linked? bobblewik 18:04, 5 September 2006 (UTC)

Many people reading an article here have different exposure to imperial units vs. metric units. A single link to most of the above units does not cause any problem on articles and I don't think the MOS needs a rule on it. I would agree that duplicate links to these terms can be removed, but that is true of most any links on an article. Also, as WP:MOSNUM says now, some non-metric units (such as gallon) have more than one possible value, in which case linking can actually be important. Places where links generally wouldn't be needed is if both imperial and metric versions are shown (in a conversion) or it is a very common unit across the globe (hour, minute, second, day, etc.), but this is already handled via common sense. --MattWright (talk) 18:54, 5 September 2006 (UTC)
Thanks for your response. I am paying attention to your main point, but let me make some minor points.
  • Units with more than one value
The solution is to make them explicit (I have done so with many) rather than leaving them ambiguous and adding a link.
  • Very common unit across the globe (hour, minute, second, day, etc.), but this is already handled via common sense.
Whenever I read the expression common sense my alarm bell starts ringing. I don't have confidence that it is common and sometimes doubt reliance on a universal view of what is sensible. Sure enough, now that I check 'What links here' for the units you listed, there is no apparent sense to the thousands of pointless links to them. But it does seem that we agree on the principle that they are generally not needed, we merely need to agree on which are very common across the globe.
Regards bobblewik 19:25, 5 September 2006 (UTC)

MattWright makes a very good point in saying that a grade-schooler that is reading an article may not know what is meant by m or sq mi in the infobox or in the article. I think that linking the unit symbol the first time it is used certainly wouldn't hurt. On the other hand, it should be said that over linking of common symbols for meausurements or anything for that matter certainly could hurt. I would suggest linking the symbol only once in the infobox and only once in the article. —MJCdetroit 13:43, 6 September 2006 (UTC)

If the article is long, it may be appropriate to link the symbol again, as is true of any links in a long article. -- Donald Albury 14:38, 6 September 2006 (UTC)
I agree that some symbolic forms may be less commonly understood than the full forms (indeed, I gave the example of 'ci' as obscure). We can first decide which *full* forms might be regarded as 'common units', then secondly decide the symbolic forms that are 'common units'. What do people think of the list of full forms above? bobblewik 19:25, 6 September 2006 (UTC)

Discussion above moved from Wikipedia talk:Manual of Style (dates and numbers)

What do people here think about the above discussion? bobblewik 08:20, 9 September 2006 (UTC)

Yes, someone who doesn't know that a foot is can always type it in the search box - but that's true of ANY link, isn't it? Why link anything? Isn't it equally valid to demand that users type in "Vegetable fats and oils" instead of clicking on a link? And if you're not going to link to the Foot article because it's a common unit, shouldn't we just delete the article as being unnecessary? ClairSamoht - Help make Wikipedia the most authoritative source of information in the world 15:27, 9 September 2006 (UTC)

What generally should be linked

I expanded on the Technical Terms theme with two more items that relate especially to non-natives. I think this is consistent with the context, and hope it is not too controversial. The geo-names linking is pretty well universal, so this is probably stating the obvious. The non-native speaker help may not be obvious to those who have not studied a foreign language. Many words have multiple meanings, and in another language, they translate to different words, each of which may have other multiple meanings. So one of the advantages of a wiki, as opposed to a flat text, is that we can make it easier for everyone to understand by using links intelligently.

  • Word usage that may be confusing to a non-native speaker (or in another dialect). If the word would not be translated in context with an ordinary foreign language dictionary, consider linking to an article or Wiktionary entry to help foreign language readers, especially translators. Check the link for disambiguation, and link to the specific item.
  • Geographic place names, since many places have similar names, and many readers may be from a distant place. Link to the most specific available article, or create a stub or redirect if the place deserves a new article (check similar nearby places for naming conventions and category tags).

Dhaluza 01:57, 9 December 2006 (UTC)

Question RE: Linking Source Access Dates in a Citation

For an FA-candidate, one reviewer continues to object on the grounds that the dates on which a cited source was accessed as mentioned in the footnote/citation should be linked up. I refuse do to this, as the nominator and chief contributor to the FA-candidate because it's just ludicrous given that this is just a "guideline" and that it would violate Wikipedia:Only make links that are relevant to the context in that linking up such dates will not be relevant to the article's context or aid or facilitate in any way a reader's understanding of the article. So I ask the following:

  • What is the consensus regarding this guideline being in force on access dates in citations?
  • If there hasn't been a consensus, what is the opinion of the contributers and other people who debate this MOS provision?

I thank you in advance for your answers and assistance. —ExplorerCDT 18:04, 1 January 2007 (UTC)

My observation Quoting Wikipedia:Manual of Style (dates and numbers), "If a date includes both a month and a day, then the date should almost always be linked to allow readers' date preferences to work, displaying the reader's chosen format."
Also, "This Manual of Style, like all style guides, attempts to encourage consistency and ease of reading. The guidelines here are just that: guidelines are not inflexible rules; one way is often as good as another, but if everyone does it the same way, Wikipedia will be easier to read, write and edit." Please comment. Regards.--Dwaipayan (talk) 18:24, 1 January 2007 (UTC)
But what has to be asked, what the core question here is: What does linking source access dates have to do with the context of the article? And what would compelling such linking do to aid a reader's understanding of the article? Are we to sit back and expect a reader to ponder..."what happened on the day User:so-and-so accessed this source he cited?"—ExplorerCDT 18:32, 1 January 2007 (UTC)
Haha! That's the question that bugged me when I first saw this guideline. Anyway, let's see what other experienced wikipedians have to say.--Dwaipayan (talk) 18:38, 1 January 2007 (UTC)
Linking of dates serves two purposes - one relevant to this guideline and one not. I would agree with you that linking the dates you describe has little to do with "context" and would not significantly impact "the reader's understanding of the article" however, wikilinking of dates also serves to transform the date from your preferred format into the reader's preferred format (as set in the user preferences). For example, I am most comfortable with dd-mmm-yyyy as my standard format. Many readers are uncomfortable with that and strongly prefer mmm dd, yyyy. Wikilinking the day and month is a mechanism to ensure that we both can have our way. The Manual of Style on dates can give you lots more background and detail.
Now, in a perfect world the WikiMedia software would give us some alternate tag by which we could denote dates so the auto-formatting could work but without simultaneously creating an extraneous link. But that's not available to us (yet) so we do the best we can with what features we have now. Rossami (talk) 01:17, 2 January 2007 (UTC)
  • I know the reasoning behind the wikilinking of dates, it doesn't need to be repeated...what the core issue is: Do you think it really matters that someone sees a date as 17 August or August 17. I don't. Do you think this MoS provision negates that given the circumstances? Do you think an FA-candidate should be held up on what is seemingly ambiguous and contradictory? —ExplorerCDT 01:39, 2 January 2007 (UTC)
    Sorry, but your comments above made that unclear. The short answer is that a lot of people do think that it matters if someone sees a date as 17 Aug or Aug 17. This page follows the MoS guideline and explicitly calls out dates as an exception to the "context" rule. I don't see anything ambiguous or contradictory about it. Now, whether the MoS recommendation is a good idea or not is another question - but that should be resolved on the MoS page, not here. If you can get consensus to change the rule on the MoS page, then we'll change the rule here to follow suit. Rossami (talk) 05:27, 2 January 2007 (UTC)

Nutshell summary

Do we have to have a "nutshell" template on this page? I think they are ugly, usually pointless and get in the way of the reader. While the current nutshell summary is an accurate statement of the page, it is redundant with the pagename. It is merely another box for the reader to have to struggle past in order to get a real understanding of the page. Rossami (talk) 21:54, 15 January 2007 (UTC)

The precept to "Only make links that are relevant to the context" can be construed narrowly as "If some part of text represents an entire phrase which has (or is explicated in) a separate article, then link to that article; don't link its individual elements (words or subphrases)".

However, in doing so, the entire wording (and also graphic content) of an article may be considered the relevant context upon which appropriate linking is to be applied. Therefore also single words, not least verbs, could be recognized as relevant and linked. Consider the​ ​​example from a present article Abraham Lincoln:

" ... his Gettysburg Address rededicated the nation to freedom and democracy."

The (mere) verb "rededicate" is even relevant enough to appear in the present Gettysburg Address itself. Indeed, a reader may expect the entire wording of every article to be appropriately linked (not even to dwell on annotations), and thus to be fully structured in its relevant components; either to visit the corresponding article, or, no less useful, to review the structuring "OnMouseOver".

Plain words may be usually subsumed. (However, note for instance Bernard's famous (albeit presently inexcusably underrepresented) "Yes and No, Minister!") Subsidiaries ought to be linked as well, to affirm that they are subsidiary.

One counter-argument which seems more debatable than, say (possibly temporary) limitation of resources or (surely user-customizable) varying preferences how fully linked text ought to appear, is the "spandrel issue" that, if only comparatively few terms are linked, they appear highlighted against the "background text" (at least with suitable user skin customization), thus allowing very efficient speed-reading of such articles. Perhaps such desired emphasis of few terms can be achieved by other means as well (say by emboldening or by Font color). However, this would necessarily interfere with the actual textual makeup of the article (which linking itself obviously does not).

In a nutshell: there is no consensus that Wikipedia:Only_make_links_that_are_relevant_to_the_context#What generally should not be linked should contain any entries other than "Individual words when a phrase has its own article". Frank W ~@) R 18:43, 17 January 2007 (UTC)

p.s. There seems to be presently no instance of the phrase "Relevant background can be found in ..." in all of Wikipedia except in the archive of this discussion and in Wikipedia:Only_make_links_that_are_relevant_to_the_context#What generally should not be linked itself. This indicates that relevant background (on any particular subject term) can be found already through a link of that term, or of text in direct context of that term. If there is no text "Relevant background can be found in ...", then the potential akwardness of having to link it doesn't arise. —Preceding unsigned comment added by Fwappler (talkcontribs)

Fwappler, I've read your comment three times now and can't figure out exactly what you're trying to say. Of course, it doesn't help that most of the section-links you included as examples are broken. What is your point? Rossami (talk) 23:11, 17 January 2007 (UTC)
If you've read the above paragraph In a nutshell: there is ... three times and haven't gotten the point yet, then ... thanks for asking me to restate it:
I (strongly) agree with the statement, plain and unadorned, to "Only make links that are relevant to the context". Also, I have (strong) reservations to any limitation on making (what I'd consider contextual) wikilinks of any (accordingly appropriate) piece of artilce text (or graphic). (Instead I'd be very happy for instance with the linking expressed in Wikipedia_talk:Build_the_web#What Allwiki is not.)
Therefore I consent only to a subset of what's stated in Wikipedia:Only_make_links_that_are_relevant_to_the_context#What generally should not be linked. Do you (or anyone) disagree, or can I start soon editing away to what I won't consent? Frank W ~@) R 05:08, 18 January 2007 (UTC)
If I understand your proposal correctly, then I and many other people do disagree. The Allwiki concept was evaluated early in the project and rejected by the community. It was proposed again several times (usually at the Village Pump and the earlier decision was endorsed each time. Linking every word or phrase has several serious negative consequences:
  1. Attempts to implement it created endless bickering over the right level of linking. For a concept like "joint and several liability", should you link joint and several liability, joint and several liability or joint and several liability? Any of them could conceivably be useful to some reader but the software can not support all of them. We were unable to come up with any useful rule that gave workable guidance.
  2. Overlinking dilutes the value of all the links for the reader by making it harder to find the truly informative links. This argument is based on the presumption that our readers expect our editors to exercise editorial judgment.
  3. Consistency of link density is important to maintaining readability. So if you're linking every word or phrase in the first three paragraphs, you have to continue the pattern. That inevitably creates redundant links. Readers don't expect to see redundant links. They expect that each link will take them to a new place where they can learn more. They get confused when they follow a new link and it takes them to the page they just read. They get frustrated and start posting nasty message on the Village Pump that our linking engine must be broken.
You can check the archives for more reasons why Allwiki was rejected. Those are just the arguments that stuck closest for me.
Having said all that, see my second comment below. I'm not yet sure that I do understand your proposal correctly. But until we do understand it, please do not start removing anything from the page. Rossami (talk) 20:42, 18 January 2007 (UTC)
p.s. The only one broken (i.e. marked red, leading "nowhere", and thus not displaying the annotation either) link above is that to Fractiousness. (p.p.s. Realizing now that this point, too, might have been missed, let me spell it out as well: Of course I had left this jibe broken on purpose. (After all, AFAIU, broken links are admissible outside of article space.)) That can be filled in promptly. All (many) other links were not broken. However, I'll now apply two (fairly small) annotational edits. —Preceding unsigned comment added by Fwappler (talkcontribs)
They are "broken" links in the sense that they do not take the reader to a section name that matches what pops up when you mouseover the link. For example, a link you created above [[USA#Is a ::Nation::|the nation]] takes the reader to the top of the USA page, not to a section titled "Is a ::Nation::". The popup only shows USA but the hyperlink in the bottom left of the browser shows the full linkpath including the section link. There is no such section on the page. You clearly created those links with great deliberation but their function in your example is unclear. The example you posted at Wikipedia talk:Build the web#What Allwiki is not is equally unclear. I still don't think we understand what you're trying to describe. Rossami (talk) 20:42, 18 January 2007 (UTC)


$ If I understand your proposal correctly [...]
Your immediate question ($$$ What is your point?) and my attempt at addressing it was concerned whether a point was to be made and understood. (Has it?? -- Well, your subsequent remarks are at least more or less "on topic".) A proposal goes considerably beyond a (mere) point; for definiteness I'll try to make my proposal concrete below.
$ The Allwiki concept [...]
What of it? Surely it's wise to avoid "that" ("discarded hat", i.e. having each and every piece of writing, jointly or severally, linked up, come hell or high water); being instead concerned with "this" (e.g. identifying which article topic is matched to a particular contextual element in a given text, and linking it down).
Now, yes: either way, the full text of an article may thereby end up being linked. Your corresponding objections are therefore relevant (and "obstacles to my proposal draft"):
1.
$ [...] should you link joint and several liability, joint and several liability or joint and several liability?
That's to be determined by context, i.e. by where and how the phrase appears. The fact that Allwiki lacks this ingredient is the distinction I propose to be drawn, as the only distinction between what's rejected already and what's debatable in order to "make links that are relevant to context". Also, context may surely require yet other link structure, e.g. (for as little as I know of such matters to respond smoothly to your example):
"Then Stan took over this joint and several liability cases were dropped."
Also (and here my limited familiarity with your example starts to become a hindrance in illustrating further cases), I'm surprised that the phrase "joint and several liability case law" seems extremely rare. As a hypothetical, the corresponding context might be "#Tort cases ::joint and several liability::|joint and several liability case law".
Consequently: Since the "joint and several liability" expression has that article, this type of link will often be applicable.
"joint and several" redirects to joint and several liability. That's very surprising and seems absolutely inappropriate to me because, AFAIU (and already kinda used above), "joint and several" is a notion quite independent of Law#Legal ::Liability::, having to do with Individuality and Community, Distinguishability and/or Interconnectivity and/or What_either_is_not instead. (I can't quite name that notion itself at the moment -- for reference below I'll designate it X for now; but I intend to supply that later, after perusing, in particular, Steven Weinberg's "Field Theory".)
(That provides Cluster Decomposition and Cumulativity - bingo! <tt>F</tt>rank <tt>W ~@) R</tt> 20:09, 21 January 2007 (UTC))
I can refer to The Joint and Several Effects of Liquidity Constraints, Financing Constraints, and Financial Intermediation on the Welfare Cost of Inflation as example matching my expectation.
So: "X#joint and several liability" will surely come in handy within the article joint and several liability which (surely) contains and aims to define the phrase "joint and several liability".
"joint" (not what I thought! &) is surely "way too Allwiki" in context of your example. joint might be applicable perhaps in (the text of article) Joint venture.
$ [...] the software can not support all of them.
First off: would you please put a link that curious phrase: "the software"? (Might just be the very software that started breaking my signature, "<tt>F</tt>rank <tt>W ~@) R</tt> 16:12, 19 January 2007 (UTC)", couple of years ago ...)
Now, hopefully you don't mean "all of them" in terms of quantity, i.e. resources being required and not to be had, in principle, for decades.
Otherwise, meaning "all of them at once" in terms satisfying all link examples on the very same piece of article text -- that would of course be impossible upfront. However ...
$ We were unable to come up with any useful rule that gave workable guidance.
My responses to your examples were mainly designed to show that context granularity other than "Atomic Allwiki" or "Status quo - Keywords only" can be discerned and expressed. Would th' thriggin' software still not allow to have optional user preferences concerning the application of wikilinks to given text?? (See again below.)
(Also: is there any record of what, if anything "we" (present company not necessarily included) might have come up with yet?? There seems nothing retrievable, for instance, via Wikipedia:Build the web#Allwiki or Wikipedia talk:Build the web; and sure as hell nothing else through WikiPEN.)
2.
$ Overlinking dilutes the value of all the links for the reader by making it harder to find the truly informative links. This argument is based on the presumption that our readers expect our editors to exercise editorial judgment.
As far as I understand the second sentence: it seems important to separate editorial judgement in writing article text from editorial judgement (or, as far as "the software" will permit, readers' personal judgement) in applying links to that context.
Further (while wondering whether your second sentence was (also) meant as disclaimer against what I'm about to ask): Do (have?) (ought?) articles contain any text that is not "truely informative", to any reader, at that reader's level of familiarity with the subject? Thus, considering (perfectly naively) each reader a novice, reading Wikipedia on a topic to become an expert, is there any piece of CONtext not "truely informative" to all readers?
As w:user and editor as well as expert on some article topics, I'd much rather have prerequisit links easy to find, as well as Where is this employed as ::Definition:: vs. Where is this ::applied or referenced::.
3.
$ [...] inevitably creates redundant links.
Then, on first blush, context (i.e. article text) had to be inevitably redundant. Well, let's suppose for instance that Abraham Lincoln#as ::Orator:: and Abraham Lincoln#as ::Nonconformist:: - one of ::man::y whose legs were just long enough to reach the ground appeared in the same article, redundantly.
$ [...] Readers [...] expect that each link will take them to a new place where they can learn more.
Such readers might especially like Which ::distinct:: articles are linked here. I might grow to like even that as well. So far, the Which articles linked here have I already visited (i.e. "dark-purple-ish") works marvellously (for me; hopefully for us as a whole, too).
$ [...] You can check the archives for more reasons why Allwiki was rejected.
Again, how/what do you mean, please? Is what you mean retrieved when you (as I just did) type "Allwiki" in the search field, then check all namespace marks on, then search again? (This returned 9 (nine) hits ...) Frank W ~@) R 16:12, 19 January 2007 (UTC) (planning to treat "p.s." and to copy-edit on 21-Jan-07)


p.s.
$ They are "broken" links in the sense that [... for] instance [...] takes the reader to the top of the USA page, not to a section titled "Is a ::Nation::".
Allright - that's manifestly not "(plainly) broken links".
$ The popup only shows USA but the hyperlink in the bottom left of the browser shows the full [...]
That's what I find as well, and that works as intended given the present limitations of (presumably) "the software".
$ [...] linkpath including the section link.
Rather: an annotation, which takes on the responsibility to explain (partially) the sense of having USA being pointed to by the phrase "the nation" in its above context. The curious reader is directed to learn (by following the link) that (whether?) the United States of America are indeed a nation (or were, when Lincoln gave the context); or, by following the secondary annotational link, what a nation might be in the first place. The curious (link-)editor is directed to learn which (if any) contents, links, or sections are provided within USA to help out the curious reader. (Linking "the nation" in the above example instead as "the nation" would seem less contextual and less helpful.)
I might (should ?) have already objected to your earlier characterization: ...
$$$ [...] most of the section-links you included as examples are [...]
... No, those are not meant to be section-links as such, but annotated links (to a contextually appropriate article). Only if the article happens to have a section which even sharpens the context, and if the wording of the annotation allows to link to such a section directly (rather than the article as a whole), then the annotated link also constitutes a section-link.
$ There is no such section on the page. [...]
Surely the total number of distinct relevant context elements in all articles is larger than the number of all sections (of all articles). Therefore, yes, annotations will frequently not link to a particular section of an article, but merely point out that certain terms are used in that article; such as Is a ::Nation:: points out what fits (AFAIU) the phrase "the nation" within the context "His Gettysburg Address rededicated the nation to freedom and democracy."
In the specific example, the present article of United States (to be precise) doesn't have any (sub-)section such as "The USA as a Nation", or "The USA being referred to as a Nation", or somesuch. As a consequence, including the word "nation" in the annotation for a link to this article (or for readability and further reference rather, ::Nation::) will create a link that functions the way you above called "broken". Some might instead call the article broken, for lacking the required (sub-)section. Or perhaps it's "just a matter of th' software".<tt>F</tt>rank <tt>W ~@) R</tt> 19:40, 21 January 2007 (UTC)


Disambiguation Reference

I just took the liberty of adding a reference to disambiguation pages as a place to specifically NOT link extensively, based on the style guidelines already in place for those pages. I know I didn't come in here and ask for consensus on this page first, but having just cleaned up a ton of dabs with linkorrhea, I figured that wider dissemination of the guideline had to be a Good Thing. I decided to be bold. The Monster 19:15, 30 January 2007 (UTC)

I agree with the (literal) statement which you added; and (opportunistically? &) I applaud the boldness expressed and tolerated.
I just took the liberty to point out where the statement of the (related?) guideline which was quoted (further down in the article) from Wikipedia:Manual of Style (disambiguation pages)#Individual entries (and is presently still there) may contradict the (literal) precept of this project. Frank W ~@) R 16:49, 7 February 2007 (UTC)
I have just reverted that addition because even after reading it three times I couldn't figure out what the paragraph was trying to say. Please re-draft the comment here (preferably in simpler language). Thanks. Rossami (talk) 21:00, 7 February 2007 (UTC)

See also

I frequently see people remove links I place in a See Also section at the end of an article because they are linked somewhere in the text. This policy does not preclude repeating links in another section, especially if the first appearance was near the top. But what is the consensus? I think that it is useful to repeat links with a major connection because people don't always read the whole page, and even if they do, a prioritized list of where to go next is helpful. Is there any pre-existing guidance on this, or does it need to be addressed? Dhaluza 17:02, 19 February 2007 (UTC)

This page is the only pre-existing guidance on the topic that I know about. The page generally says that duplicate linkings are a bad idea because they make future maintenance harder. For example, if you move a page, you have to go back and fix all the double-redirects. You find the inbound redirects via "what links here". Go to the page, find it, fix it and move to the next page. If the link is duplicated on the page, there is a very good chance that the second (or third, ...) occurrence will be overlooked. This leaves the reader with a dead link which is more than a little frustrating.
There is also a general presumption about indexing that each link will take me, the reader, to a new topic where I can learn more. Users have expressed frustration when they find that a second link takes them to a page they've already read. (This is especially problematic when one of the links is piped and it's not obvious to the reader that he/she was about to go to a page already viewed.)
The guideline is quiet on the "See also" section specifically. Personally, I prefer the "See also" section to be limited to just pages that I haven't already had a pointer to but I can understand the value of a synopsis of major links at the bottom. I am more tolerant of that than I am of duplicate links within the text.
It's more a style question than a strict policy question and style issues are always interpreted based on the specifics of the page. For example, I would expect to see more tolerance for a recap link on a very long page than I would for a one-sentence stub. I'm not seeing enough disagreement to be worth expanding the guideline but if you see a lot of ill will over this topic, we could consider it. Rossami (talk) 19:57, 19 February 2007 (UTC)
I agree with you on double links in the text for readability. But when you go to change them after a rename, if you don't get them all, the page will just stay on the what links here list, and you have to go back. You can avoid this by searching for repeat occurrences when you edit, so it's a minor issue.
If you think about it, there shouldn't be anything in the see only that wasn't covered in the text — if there is, the article should be expanded to cover them. Anyway, I see this as analogous to "fixing" redirects to {{R with possibilities}} pages. Not earth-shattering, but conunter-productive, and should be discouraged. Dhaluza 20:15, 19 February 2007 (UTC)

Disambig

[Do not link]: "Words in a disambiguation entry other than to the disambiguation target itself. The general rule is "one link per entry"; additional links tend to confuse the reader."

I'm not keen on this section for a number of reasons:

  • "additional links tend to confuse the reader" - I think this is generally untrue.
  • Additional links can be "useful in context", for example (made up), suppose I want to know where the "Scott Memorial" is, I get a disambig page telling me there's one in Edinburgh and one in Atlantis (now lost), I may want to go straight to Edinburgh. Rich Farmbrough, 10:39 10 March 2007 (GMT).
That's not a change that should be first proposed or resolved here. That line is derived from the [[Manual of Style rules on disambiguation pages. If you can convince editors to change it there, we will pull the change through to this page. Rossami (talk) 14:47, 10 March 2007 (UTC)

Discussion about template:NYCS and overlinking

I have begun a discussion at Template talk:NYCS. --NE2 00:09, 24 March 2007 (UTC)

Several levels of geographical names

Is the preferred style "Copenhagen, Denmark", "Copenhagen, Denmark", or "Copenhagen, Denmark"? –Henning Makholm 07:19, 5 May 2007 (UTC)

Depends whether you think the reader should have the option of clicking through separately to the articles on Copenhagen and Denmark. If you think Denmark is not relevant to the article, don't link it, but if you think the reader needs to know where Copenhagen is, still mention Denmark. It is all about the context of the article, as the title of this guideline says. Carcharoth 21:13, 6 May 2007 (UTC)

WTF does this one mean?

"[In general, do create links for ]Geographic place names, since many places have similar names, and many readers may be from a distant place. Link to the most specific available article, or create a stub or redirect if the place deserves a new article (check similar nearby places for naming conventions and category tags)."

The logic escapes me ("since") and it's very poorly worded elsewhere. I intend to remove this point unless someone can explain what it means. Tony 10:06, 3 June 2007 (UTC)

The idea here is that what is obvious to one reader may not be obvious to another. A casual mention of "Paris" in an article may mean "Paris, France" to someone from New York or Los Angeles, but to someone from nearby rural Texas, may mean Paris, Texas. Hope this helps. Robert K S 10:23, 3 June 2007 (UTC)
The disambiguation motive is much better served with simply adding ", France" in the text itself. Then readers who don't think they need to follow the link will also get correct information. I agree that place names should be linked if possible: Either it is relevant to the context to link the place name, or it is not relevant in the context to mention it at all, and it should be excised from the sentence completely (which should be easy, as place names are rarely essential to the sentence structure). But the disambiguation rationale alone does not sound convincing. –Henning Makholm 11:23, 3 June 2007 (UTC)
I won't work to "convince" but I will state my disagreement with the above supposition as a philosophy for linking articles (that such links have no place in, or cannot or should not be used to, disambiguate). Take for instance the example of biographical articles about people who were computers. It would be superfluous and bizarre to refer to a person as a "human" (and in fact it would be entirely incorrect in context), but many (most?) readers may not realize that "computer" was until extremely recently a human occupation. Providing the wikilink as a disambiguator is, to me, the most sensible, most efficient, and most helpful way to inform the reader. The good and great advantage of a hyperlinked encyclopedia is that any term defined in its own article should not need to be laboriously defined in its instances in other articles! Robert K S 11:47, 3 June 2007 (UTC)
It's not simply disambiguation. If a place is mentioned in an article, it is relevant to the article (otherwise it should not be mentioned as you say). If the reader is not familiar with the place, they should be able to click on the link to find out more, like where it is located, and other contextual info. If it is not linked, they may have trouble finding the right place, since place names are commonly duplicated. The idea is to make the article content accessible to as many readers as possible. One of the benefits of the wiki is that we can cast a wider net by using wikilinks to allow readers unfamiliar with a topic to get more information, without overburdening a more familiar reader with excess detail. Place names are a special case of this. Also, if the place's article is renamed, a redirect will take the reader to the new name, while a plain text link will not. Many place names are disputed, and they are often changed; without the link, editors won't be able to follow the 'what links here' to find only the affected articles.
As an example, when working on the BOAC Flight 911 article, a reference listed Oshima as a waypoint in the flight plan. From the context of the reference, I was able to determine the correct link was to Izu Oshima, but if I didn't link to the WP name, it might be confusing to someone wanting more detail, especially if they are not familiar with Japanese geography. Dhaluza 12:07, 3 June 2007 (UTC)
What you're describing is disambiguation. :-) Robert K S 12:20, 3 June 2007 (UTC)
Yes, but it's not simply disambiguation, there is more to it, especially for place names unfamiliar to the reader. But why is simple disambiguation a problem anyway? Dhaluza 12:38, 3 June 2007 (UTC)

Is it just me? Or do others feel the word "American" is overlinked all over Wikipedia? (This isn't an anti-American rant; I'm an American.) It almost always seems irrelevant and unhelpful. Here's an example. Seeing that Bob Dylan is an American songwriter, I might jump to the linked article wanting to learn how American songwriters are different from the other sorts. But no; this is a long article on the United States, its history, its government, its population. (Of course it is. That's what it should be.) Perhaps if I were reading about a singer from Ghana, I might wonder where Ghana is and take a moment to find out. But--I'm sorry--about the last thing that would help me understand Bob Dylan, or Johnny Cash, or Leonard Bernstein is a treatise on the United States of America. (BTW, Lenny's article links to United States twice in the first two lines!) I'm inclined to start knocking these links out, but wonder if there's a guideline or consensus I'm not aware of. Anyone? Hult041956 23:41, 29 August 2007 (UTC)

Go for it. But expect some push-back on the issue of consistency. It could be seen as a US-centric view to assume that everyone knows how to find the article on the US but not on Ghana. I'd start with a few pages and see what the reaction is. (The duplicate link definitely needs to go, though.) Rossami (talk) 23:50, 29 August 2007 (UTC)
If someone creates a page that specifically talks about American songwriters, then you could replace those links with links to that page, but until then, the link is not as specific as one might hope. The essential quality of links is that they create a web - so even these general links are not unhelpful (or at least - removing them would be at least as unhelpful as keeping them). --Joy [shallot] 00:20, 31 August 2007 (UTC)
Here's a counter-example. I just came across (at random) William Carlos Williams, who was an American poet. The link goes to a list of American poets(!), not surprisingly to the history of the United States. I am grateful for small blessings. ;-) Hult041956 00:03, 19 September 2007 (UTC)
In general, I'd link to America (and any other country) only in articles or sections about global geography or politics; the links someone expects to see depend on what they're looking up. Feezo (Talk) 20:14, 19 September 2007 (UTC)

'overlinking' not considered harmful

It is correct that too many links have disadvantages, but I'd like to point out that many WP users surely enjoy to surf, and I'm one of them. If I am on an article about the Geology, for example, I may click on a link pointing towards Poetry, not because it has any relevance to that, but specifically because it doesn't. Unfortunately, due to the WP:OVERLINK policy, I often end up in articles that only have links to similar articles in their context, eg a linguistics article may only contain links to a language, a linguist, and perhaps a date or two. The only way to escape from such articles (that act as deadlocks to surfing) is then to click on a date, click on a category to find something with more links, or to click on the main page or on the random article link (which isn't good for some surfers for various reasons). What we need is not to discourage 'overlinking', but to encourage or at least tolerate it while managing it to minimise its disadvantages. We need a way to distinguish between links put in for their surfing and web-building value and links that were put for their direct relationship to the article they are in. We could, for example, put all web-building links under a CSS class that round render them in a colour very similar to the surrounding text, and important links in the usual colours we use now. We could also put some preference bits to the MediaWiki software to enable users to choose whether they wish to see the web-building links or not. Another solution would be to add some functionality to MediaWiki to optionally enable (in the prefs) all words to become links automatically no matter whether there is a wikitext link or not. NerdyNSK 15:14, 16 September 2007 (UTC)

Links on random words that are not related to their context, do not serve any meaningful "surfing" or "web-building" function. Most browsers have back buttons which may be clicked in case of dead ends. –Henning Makholm 00:57, 18 September 2007 (UTC)
If you just want the links there as a means to switch to a new, unrelated article, try Random article on the sidebar to your left.--Father Goose 03:54, 18 September 2007 (UTC)
And NerdyNSK, there's absolutely nothing stopping you tapping an item into the search box: easy as pie, and takes four seconds longer than the mouse click on a link. Saves the messy bright blue, and allows the high-value links to breathe. You can, in essence, pursue your own web tree. TONY (talk) 16:16, 6 July 2008 (UTC)

New subsection for sortable tables

Proposal for a new subsection under "Other considerations":

Sortable tables (proposal "A")
In the case of a sortable table (class="sortable"), each row should be considered its own separate entity for linking, even when this gives the appearance of overlinking the table. This is because changing the sort order reshuffles the table and makes too hard and inconvenient to find back the single instance of a bluelink in a given column. (Example of a not-overlinked sortable table: List of mountains in Peru.)

It could probably be reworded a bit. Comments before I add it?

— Komusou talk @ 14:48, 7 October 2007 (UTC)

That sounds reasonable, but I don't think it goes far enough: why limit it to sortable tables? In most cases where it is appropriate to present some material in tabular form at all, it must be because we expect each row of the table to be individually useful to a reader. Tables are not made to be read sequentially from beginning to end. You locate the row you're interested in, and then you look left and right to find the information you need. Nothing's gained from requiring the reader to look upwards in the table for a bluelink. I'd prefer wording such as
Tables (proposal "B")
In a table, readers will often jump quickly to a particular row and expect it to make sense even without the context provided by the other rows. Therefore it is appropriate to repeat the same link in many rows of a table even though this may at first look like overlinking. For example the table of host cities in the Olympic Games article links all city and country names, even in cases where the same city or country is linked earlier in the table.
Henning Makholm 15:38, 7 October 2007 (UTC)
I completely agree. In thesis-antithesis-synthesis fashion:
  • On the one hand, I'm 100% with you for linking in all tables, and even further in table-like data. (For instance, in the ==Personnel== section of an album, I will not-overlink each instance of an instrument on a per-musician basis, because if I quickly lookup the 9th musician credited and it just mentions an unlinked "shakuhachi", I don't want to be forced to go back and parse the whole personnel list in order to discover the shakuhachi bluelink among the instruments of the 4rd musician – that sort of thing.)
  • On the other hand, I was under the impression that deletionists and delinkers currently have a majority at Wikipedia, and that I may have a hard time proposing a guideline allowing such massive (not-)overlinking. So I had decided to go with the much more easily pleaded case of sortable tables, thinking consensus would be easier.
  • That being said, I fully endorse your extended proposal "B" (I took the liberty to tag them so that others editors can more easily refer to what they would accept and not). And since we go this way, I'll even propose trying to complete the scope of the topic by appending three sentences to your text, please see proposal "C" below.
— Komusou talk @ 18:34, 7 October 2007 (UTC)
Tables and table-like data (proposal "C")
In a table, readers will often jump quickly to a particular row and expect it to make sense even without the context provided by the other rows. Therefore it is appropriate to repeat the same link in many rows of a table even though this may at first look like overlinking. For example the table of host cities in the Olympic Games article links all city and country names, even in cases where the same city or country is linked earlier in the table. (This is even more necessary for sortable tables, for example List of mountains in Peru needs to link each instance of regions and mountain ranges.)
The same rationale applies to table-like data such as a bulletted list where each entry is relatively independant, where a reader can directly jump to a given entry and expect it to make sense on its own, such as a list of lists of instrument credits in an album page's Personnel section. (For instance, the musical credits section of Peter Gabriel's album Passion is unpractical in its delinked form, because the reader who came interested in song #20 faces an unlinked "arghul drone" and has to parse the whole personnel list to find the "arghul drone" link back in song #3.) —Preceding unsigned comment added by Komusou (talkcontribs)
  • I understand the points being made above but I worry about the increase in maintenance that this will drive. This proposal will make it far harder to find and fix all the instances of the link when one of the destination pages is moved or when the link needs to be updated. I'll grant that if you link every line in one particular column, you might make the page slightly easier to read because whatever column you've decided to link will be consistently shown as a link but I'm not sure that's enough to justify the maintenance costs. I don't see it as that much trouble to scan back up the list to find the link I need. Rossami (talk) 04:11, 8 October 2007 (UTC)
    • The only time I can think of that a link needs to be updated without the actual text changing (which would require editing it whether or not it's linked) is when it becomes a disambiguation page. With a tool like Lupin's popups, you can select the right target for one link and they'll all be changed. I agree with the proposal to exempt tables and the like, though I figured it was common sense. --NE2 10:35, 8 October 2007 (UTC)
Rossami:
  • For the maintenance, I think there's no problem about that. For one thing, WP:REDIR#Do not change links to redirects that are not broken is clear that links to pages that become redirects need not maintenance, only double redirects need it. For another, as NE2 pointed out, when updates do become necessary, it's more a matter of software, there are tools and bots for automatically updating these things which can be done by a dumb bot and don't even need human supervision. (Similarly, renamed categories need to be updated in each and every article, but it's done by a bot after a CFD decided "Rename".)
  • For the "scan back up the list", I see two problems with that. First, it really doesn't work in sortable tables, and it's often a PITA with large static lists, the Peter Gabriel album I mentioned above would be a good example of that. Second, most of our readers actually have no idea about our secret little rule "only the first instance is linked, go back up to find it", it's an editor's thing, but it's not that obvious for an average reader who doesn't already know the trick – that why the rule actually say to make links at the place(s) where the reader is most susceptible to want to know more, rather than mechanically linking only the first occurence.
NE2: About "I figured it was common sense", the problem as I see it is that with such a large project and so many different people of different ages, common sense often fails ;-) When you catch someone linking every instance of years in an article, it's very effective to be able to just undo/revert and mention WP:MOSDATE in the summary. Similarly, if you catch someone massively delinking a table or a list, it would be most useful to be able to simply refer him to a section of WP:OVERLINK that spells out loud that it's not overlinking in tables and lists. Another big reason to have it here is that it "proves" that it's indeed Wikipedia's accepted common sense, rather than one editor making it up according to his own private definition of common sense. — Komusou talk @ 10:01, 19 October 2007 (UTC)

Recap: The only objection was by Rossami 10 days ago, answered by NE2, without counter-objections afterwards (I've added another layer of answer just for the record), so it looks like it's okay to update the guideline. Are there last-minute objections before I add "proposal C" as a new section of the guideline? — Komusou talk @ 10:01, 19 October 2007 (UTC)

Actually, I'm a bit concerned about instruction creep here - it's a rather large block of text for such a minor point. Perhaps we could achieve the desired effect by just amending the "what generally should not be linked" bullet to:
  • The same link multiple times, because redundant links clutter up the page and make future maintenance harder. It is not uncommon to repeat a link that had last appeared much earlier in the article, but there is hardly ever a reason to link the same term twice in the same section. (Table entires are an exception to this; in general each row of a table should be able to stand on its own).
Then afterwards if somebody questions that part of the guideline we can point to this discussion and consider whether we need a more elaborate rationale in the guideline itself. –Henning Makholm 22:10, 19 October 2007 (UTC)
  • Hey, hold on a minute. While I concede that links in tables do not carry the same disadvantages as in the main text, could be please get the opinion of the two Featured List Directors and the two delegates who close the FLRC list? They're the experts, and there's recently been a lot of change there. They may agree, but may also have suggestions for better wording, etc. I'll post a note there. TONY (talk) 16:19, 6 July 2008 (UTC)

Please contribute to the ongoing discussion at wp:mosnum Links to common units of measurement. Lightmouse (talk) 21:52, 9 April 2008 (UTC)

Wikifying a rare word?

What level of wikifying is expected for a great article. If an article contains a word that is very rare would you expect the article to wikify the word so the reader could understand what it means? If further more the rare word had no article on wikipedia to explain it, what should happen? Should it be redlinked, left without being wikified or something else? SunCreator (talk) 16:58, 16 April 2008 (UTC)

It sounds like your asking about a link when you want to provide a definition, not a link to another encyclopedia article.
First, I would recommend against underestimating our readers. Most of them have surprisingly broad vocabularies and, even when they don't, are very capable of using a dictionary. I might link a word if it were highly specialized and had a specific and non-obvious meaning in the context but not just because it was "rare".
If you do think it's so rare that it needs a definition, though, link to the Wiktionary page. You can do that via the format [[wikt:foo|foo]]. (The link is displayed as foo. Note the slightly different color.)
If the definition does not exist at Wiktionary, be bold and create it. The experienced Wiktionarians will help clean up the format. I would definitely not create a mere dictionary entry in a Wikipedia page. See WP:WINAD for more. Rossami (talk) 23:05, 16 April 2008 (UTC)
Thank you Rossami. The word does not exist in the concise dictionaries I've so far checked, it also does not yet exist in Wiktionary. In the article it's quoting someone who used it, so not easy to remove it as the quote is important to the article. To comment in the article about it's definition would be go into original research. However it does exist in much literature, so the word was intended in the original quote. Find it in G books and G scholars here - Find sources: Google (books · news · scholar · free images · WP refs· FENS · JSTOR · TWL. SunCreator (talk) 23:33, 16 April 2008 (UTC)
In one such case (the term "hyphenate" in 2008 Hollywood strike), I created a short entry in Wiktionary, then linked to it. Actually, in that case, the word was already in Wiktionary, but I added the noun form, which is radically different from the verb.--Father Goose (talk) 02:58, 17 April 2008 (UTC)
I'm somewhat confused by that post, what has hyphenate got to do with hypertheoretical? SunCreator (talk) 12:45, 17 April 2008 (UTC)
I was saying that in a case like what you describe, I added an entry to Wiktionary then linked to it. You could do the same for hypertheoretical. Rossami offered the same advice ("be bold and create it").--Father Goose (talk) 02:16, 18 April 2008 (UTC)

Proposal at Talk:Build the web

Since WP:CONTEXT and WP:Build the web are tightly related, and in dynamic tension, I figure that, at the risk of falling afoul of WP:CANVASS, I should solicit opinions of people here who might have some thoughts that they'd like to contribute to this discussion, about what to do with the "Wikipedia does not use Allwiki" section.--Aervanath's signature is boring 21:17, 24 May 2008 (UTC)

What about those who create idiosyncratic piped linques.?JeanLatore (talk) 03:30, 1 June 2008 (UTC)

Not sure I understand. What are you referring to? Could you give an example?--Aervanath's signature is boring 10:52, 1 June 2008 (UTC)

Autoformatting purposes

I have taken the liberty of taking the discussion to wp:mosnum. I hope nobody objects but the issue is relevant to both places and that is a more active forum. Lightmouse (talk) 17:00, 29 June 2008 (UTC)

Indeed. For the record, MOSNUM clarifies that date autoformatting is optional. The reason this has evolved over the past year is partly based on the realisation that no matter how much huff and puff in our entreaties to the WikiMedia developers, our arguments for renovating the system are met with a brick wall. Don't hold your breath. In any case, who cares about seeing the two major date formats, as long as consistent within an article? We accept far more variety in spellings. TONY (talk) 01:02, 30 June 2008 (UTC)
So, based on one possible interpretation of the current section, date autoformatting should not be used unless it is relevant to the context? If that is indeed the case, it should be stated here explicitly. siℓℓy rabbit (talk) 03:07, 30 June 2008 (UTC)
I don't think that today's edit was the way to go about it ("do not apply to..."), because we don't generally say that in style guidelines...if the recommendation is different in different circumstances, we say what it is. But I do support some kind of change of wording to make it clear that either is okay...any ideas? - Dan Dank55 (talk)(mistakes) 13:16, 30 June 2008 (UTC)
  • Autoformatting, unfortunately, was programmed to use the pre-existing code for linking, with all the hideous disadvantages we now see. It has no place in a page called "Only make links that are relevant to the context". That is about links, not the autoformatting function. It would be appropriate to remove all mention of autoformatting, but I haven't gone that far—I've merely removed the redundant statement that silly rabbit has just inserted, and updated the guideline on autoformatting. TONY (talk) 13:21, 30 June 2008 (UTC)
Yes, I fully agree. The statement is redundant, but its purpose was to shore up the evident conflict between the two points: that autoformatting is alright (even if the resulting links are not part of the context), and not to link dates unless they are relevant to the context. I hope the present version achieves the same result in a more suitable way. We'll see. siℓℓy rabbit (talk)
Thanks, Rabbit. - Dan Dank55 (talk)(mistakes) 13:59, 30 June 2008 (UTC)

Wikilinking of metric units

Metric units are not nearly as well known in the U.S. as elsewhere, so I believe that there is a legitimate need for them to be linked. I believe that it is damaging to the goal of wikipedia to remove wikilinks of metric units from scientific articles. It appears to be a type of reverse bias.—RJH (talk) 03:38, 5 July 2008 (UTC)

What, every single one in an article? Perhaps the first one, but let me see, if I'm unfamiliar with what a kilometre is, and I encounter it in a science article ("the Moon is 400,000 km from the Earth"), does the article really help me? Have a look, and you may agree that it's not a good use of that person's time, because it gives lots of irrelevant info. What the person really wants is a quick and easy way of visualising what a km is, in relation to a mile.
"every single one in an article": no. "you may agree that it's not a good use of that person's time": no. I have encountered readers of scientific articles who are not familiar with metric, so clearly some type of explanation needs to be available. Having no wikilink at all is therefore clearly unhelpful, and so the reader is forced to search for it. You may agree that it's not a good use of that person's time.—RJH (talk)
This gives me an idea. Why not provide just such a method of visualisation in a centralised article, where you might have the major units (km, kg, hectares, etc) each occupying a linear length, with the US equivalent immediately below and aligned. It could be linked to on the first occasion of all of these metric units, and would give US readers who are relatively unfamiliar with the size of metric units a much better way of automatising their conception of them.
Having pop-up unit conversions may be helpful, and it's been suggested before.—RJH (talk)
Linking US units is never necessary. We don't need to know, since they're always converted in US-related articles.
I disagree here as well. An English-speaking reader may be unfamiliar with old English units. So they need a means to convert it to a scale they understand.—RJH (talk)
They're always converted, whereas metrics are not always converted in scientific articles. And my point is that these link-articles are not useful for that purpose. Yet people go through this knee-jerk process of linking, thinking that the problem is solved. Not so. TONY (talk) 03:14, 6 July 2008 (UTC)
Well deleting the link certainly doesn't solve it; that only makes the problem worse. Now the non-metric reader is left wondering what is an "m". I also think you're way off base calling it a "knee-jerk process".—RJH (talk) 17:09, 6 July 2008 (UTC)
What do you think? TONY (talk) 04:19, 5 July 2008 (UTC)
I don't think we need a special rule for this. If the first instance in the article is linked, then if they need (or even care) to know what a kilometer is, then they can look at the article, and the conversion to Imperial units is right up there in the corner. I think that's enough. Maybe we can add a hyperlink to that box that links to a metric/imperial converter site. However, I think that in the vast majority of articles, people honestly won't care. Who cares whether the distance to the moon is in km or miles? "Far" is about as specific as most people need to get.--Aervanath lives in the Orphanage 02:23, 6 July 2008 (UTC)
I think that the vast majority of links of units are unnecessary and unhelpful. Standard units of length, mass, volume, time, etc are widely recognized by anyone with even a basic science (read elementary school level) education. I may not remember exactly how long it is but I do know what the unit is and can easily look it up if I need to. I don't need my hand held with a clutter of hyperlinks just for that.
Links for specialized units such as erg can be helpful and I support the routine conversion within the article of specific measurements. Rossami (talk) 16:24, 7 July 2008 (UTC)
I agree entirely with Rossami's point. TONY (talk) 17:14, 7 July 2008 (UTC)
Is this the same elementary education system that now teaches math one hour each week? I reiterate that there are wikipedia viewers that do not know the metric system. You can't just extrapolate your local experience to the entire world.—RJH (talk) 17:34, 7 July 2008 (UTC)
Neither do we write for the lowest common denominator. (That would be simple.wikipedia.) At en.wikipedia, we have to try our best to write for the center-mass of our audience. I believe that the center-mass does recognize basic units. Yes, even with our deplorable state of science education. Rossami (talk) 22:22, 7 July 2008 (UTC)
Linking the first instance of the unit is enough. Readers who need to know what it is can click that link and read the article about the unit. Readers who already know are not confronted with a sea of blue units. Everybody's happy.--Aervanath lives in the Orphanage 02:48, 8 July 2008 (UTC)
Well then we have a fundamental difference in opinion. The fact is that the majority of U.S. citizens do not use metric units in every day life. We can not just assume that they can immediately relate to a m, cm, or km.—RJH (talk) 15:53, 11 July 2008 (UTC)
  • I don't want to see "miles" linked on first occasion. Nor "kilometres", since it will be converted to miles in parentheses. Look, any reader who hasn't learnt what these words mean should go back to kindergarten. It's completely unnecessary, as CONTEXT already says. TONY (talk) 08:19, 8 July 2008 (UTC)
  • It's true that the letter of CONTEXT does say that, although I'm not sure that the letter follows the spirit in this regard. We're only talking about linking the first instance, not every one. I think that still follows the spirit of CONTEXT. Isn't a link to the unit of measurement is a relevant link?--Aervanath lives in the Orphanage 13:05, 8 July 2008 (UTC)
  • No one has informed me who is going to follow these links, and whether someone who doesn't know what a mile or a kilogram is will be helped by those articles. Looks like the old diversionary browsing, to me, like linking the names of such obscure countries as the US, Canada, the UK and Australia. TONY (talk) 13:44, 8 July 2008 (UTC)
  • Who is going to follow the links? People who want to know about those units, whether they've seen them before or not. I'm not sure what your last statement means, though. We routinely link to those countries the first time they appear in an article.--Aervanath lives in the Orphanage 14:26, 8 July 2008 (UTC)
    • Exactly. Even though Tony dislikes seeing those links to nation states, I have on occasion wanted to follow them in order, for example, to view some information from a particular national perspective. I don't consider that diversionary browsing; we can't assume we know in advance how those links will be used. As long as they are somewhat relevant, then they fulfill a useful purpose.—RJH (talk) 15:56, 11 July 2008 (UTC)

Response: Thanks for your comments, Aervanath and RJHall. I believe that the principle of disciplined linking that is oh-so-gradually gaining more traction on WP concerns the balancing of:

  1. the disadvantages of adding to link-clutter in an article (reduced readability and appearance, and the dilution of high-value links) with
  2. the advantages of hitting a bright-blue splotch over simply typing an item into the search box.

While judging which should prevail in every context often involves subjective judgement as to context, and even individual taste, there are aspects of disciplined linking that WP should probably be encouraging within an environment that still allows for the occasional functional exercise of such editorial judgement. In relation to the linking of units and the names of commonly known countries among English-speakers, the likely utility of a link to the reader should figure strongly. In the case of units, I've pointed out above my suspicion that the current unit articles are not a good way to assist readers who are unfamiliar with units to conceive of them in relation to other units (for example, pounds vs kilograms), which I think is the original reason for encouraging such links. The article on mile, for example, starts this way:

A mile is a unit of length, usually used to measure distance, in a number of different systems, including Imperial units, United States customary units and Norwegian/Swedish mil. Its size can vary from system to system, but in each is between one and ten kilometers. In contemporary English contexts, mile most commonly refers to the international mile of 5,280 feet, 1,760 yards, or exactly 1,609.344 meters. However it can also refer to either of the following for specific uses:

  • the U.S. survey mile (also known as U.S. statute mile) of 5,280 survey feet which is slightly longer at approximately 1,609.347 219 meters (1 international mile is exactly 0.999 998 survey mile).[1][2]

The swamp of technical and historical details is just fine for the article per se, as is the historical account further down, but the information that is likely to be sought by a reader concerns the comparison of miles with metric units and with smaller US customary units. I think you'll agree that there has to be a better way of delivering such information, and better still to make it visualisable in the reader's mind. At the moment, the article is highly diversionary and readers must work jolly hard for it to "significantly increase [their] understanding of the [current] topic", a key requirement of MOS with respect to links.

In the balance between the disadvantages and advantages of linking country-names such as the United Kingdom, India, Canada and the United States on first appearance, I think we have to ask ourselves what specific knowledge a reader is likely to be seeking through such links. As pointed out above, it's hard to second-guess this precisely, but I put it to you that WP should no longer be that our readers don't know where or what such well-known countries are. This is rather similar to the argument that kitchen and horse should generally not be linked, since this is not Wiktionary and a certain basic knowledge of the language and, indeed, the world should be assumed of our readers—even grade students. Again, hitting such links swamps the reader with information and either encourages diversionary browsing or forces a process of massive information filtering: country-articles tend to be large and comprehensive. The person who is snagging on the names of commonly known countries might instead be encouraged to read those articles first. TONY (talk) 16:46, 11 July 2008 (UTC)

  • I think you are introducing a personal bias when you assume what the readers will want, whereas I'm constantly surprised by what viewers actually seek when they are looking at a topic. In this case I find the links to be helpful and not diversionary. The reader is always free to ignore the links and just view the content. You shouldn't be trying to control what the reader wants to see.—RJH (talk) 17:04, 11 July 2008 (UTC)
  • Now you are putting words into my mouth. Do not promote falsehoods about what I'm doing: "trying to control what the reader wants to see"—evidence, please; I could just as easily twist your line by saying you're trying to control what readers actually see. Of course, you don't have a personal bias, do you ("I find the links to be helpful and not diversionary"). When you assert that "the reader is always free to ignore the links ... well, that's not true and you know it: readers cannot ignore the bright-blue and underlined splotch; if they could, I wouldn't care what was linked. I'm learning to discount what you say, since it seems to come from a very POV basis. I've put logical arguments as to why the links are unlikely to be helpful, but you haven't engaged with or responded to them, one by one. TONY (talk) 11:00, 12 July 2008 (UTC)
    • "I'm learning to discount what you say". Ah well, I've already come to feel the exatly same. No point in continuing this discussion. Bye.—RJH (talk) 21:25, 17 July 2008 (UTC)

dynamic tension?

Someone reinserted "This guideline is in dynamic tension with the the goal of building the web.". I though discussion here was going in the other direction—that there's little or no tension now, and that the pages might even end up being merged into MOSLINK. TONY (talk) 13:05, 11 July 2008 (UTC)

That someone was me. Until we have merged the three guidelines, I don't want WP:Build the web to languish in obscurity.--Aervanath lives in the Orphanage 06:09, 12 July 2008 (UTC)
P.S. It still has a long way to go before it needs to be even mentioned as a possible proposal, but anyone who wants to help me write it, or comment on it, can go to Wikipedia:Build the web/MOSLINK merge.--Aervanath lives in the Orphanage 06:13, 12 July 2008 (UTC)
Isn't the issue to merge both CONTEXT and BTW into MOSLINK—Your title suggests otherwise. TONY (talk) 10:54, 12 July 2008 (UTC)
Then I guess the title is also under construction.  :) Feel free to rename it.--Aervanath lives in the Orphanage 12:27, 13 July 2008 (UTC)

User:Lordvolton and I are in a dispute regarding possible overlinking. A few days ago the article memristor was linked in the "See Also" sections of the following articles: Programmable logic device, Electronic component, Neural network software, Biological neural network, Hodgkin-Huxley model, Action potential, resistor, Hewlett-Packard Company, and many more related articles, 23 in total. I believe this is linkspam, and reverted all edits, which was then undone. I have not reverted again and promised not to do so until a consensus is reached among people familiar with computational neuroscience, though I think anyone with an opinion should chime in. SamuelRiv (talk) 18:31, 6 August 2008 (UTC)

We do tend to grant a little more leeway in the "see also" sections at the bottom of an article as long as there is a reasonable connect. A few of those links above seem reasonable. I do not, however, see any direct connection between the man-made memristor and the concept of a biological neural network. I can see a weak connection to the concept of electronic component and action potential but no stronger than for every other kind of electronic component. If we bloated the "see also" of the electronic component with a link to every article on any electronic component, the section would quickly become useless. I think in those cases, a category may be closer to the user's intent than a massive listing within "see also" sections. I would encourage you to pull out the ones with only a tenuous connection between the topics. Where there is dispute, post the question on the individual article Talk pages and get advice from the specialists watchlisting that specific page. Rossami (talk) 19:36, 6 August 2008 (UTC)
  1. ^ A. V. Astin & H. Arnold Karo, (1959) Using values in this document, international yard divided by former U.S. yard is (9144×3937)/(3600×10000) = 499999/500000 = 0.999998 exactly.
  2. ^ Tina Butcher et al. ed. (2007), Appendix C, p. C-13, footnote 11.