I read Wikipedia a lot and I always enjoy clicking on date/year links, or using popups on them, because I am interested to place the information I read within a historical context, which is rarely provided by the article text.
Often I am interested in historical context not directly related to the article itself, for example if I read about a person born in 19th century Europe I may be interested to learn how 19th Africa was like. The reason is that as a reader I read in a nonlinear way and I like to integrate information in new ways.
Furthermore, I use dates and years (as well as geographical links) to escape from a set of interrelated articles when I lose interest in them. This is also why I like links that are not relevant to the context: if I am reading about pop music and all links are also about pop music and I get bored about pop music then I cannot escape easily, unless there is a link not relevant to the context so that I can jump to another theme of articles. Linked years, geographical terms, and "irrelevant" links provides me with the opportunity to read Wikipedia in a nonlinear way, hopping from theme to theme and returning back as I see fit, and integrating the information I read in my way.
However, delinking dates and years, or only allowing links directly relevant to the context, denies me the opportunity to read Wikipedia's historical information in a nonlinear way easily, as I have to use the search box, which I very rarely do (and with the skin I use, Nostalgia, is not very easy to use because I can't see easily the Ajax suggestions, it's bug 15902 for which I have provided a patch on bugzilla). Without year links and "irrelevant" links, I only have geographical links to escape from a theme of articles (and I don't use the random page feature too often becaue it usually only returns me pop music stuff or American villages), and if an article doesn't have them then I am stuck with reading in the oldfashioned linear way, therefore my reading experience as a reader diminishes and I get less value from Wikipedia compatred to what I could get if the years and "irrelevant" links were left in place.
Hoping this will provide an insight into why some people want years linked and why they so like to build the web.
On 10/10/08, nsk nsk@karastathis.org wrote:
Hoping this will provide an insight into why some people want years linked and why they so like to build the web.
You may be preaching to the choir here, though I can only speak for myself.
—C.W.
On Sun, Oct 12, 2008 at 8:59 AM, Charlotte Webb charlottethewebb@gmail.comwrote:
On 10/10/08, nsk nsk@karastathis.org wrote:
Hoping this will provide an insight into why some people want years linked and why they so like to build the web.
You may be preaching to the choir here, though I can only speak for myself.
—C.W.
Cool, I've always wanted to be in a choir! I agree as well. A year should be linked in its first appearance in the article- even modern dates will be "history" in just a few years. Historical context is *important* and this total de-linking of all dates is damaging.
However, the wiki is resilient and I've been watching people re-linking stuff after a script delinking has been done. It should be interesting to watch how the *real* consensus pans out on the general wiki rather than on a single talk page.
On Sun, Oct 12, 2008 at 9:19 AM, Elias Friedman elipongo@gmail.com wrote:
I agree as well. A year should be linked in its first appearance in the article-
[snip]
Why the *first* appearance? We've been able to link into sections for a long time, there are many articles I've read whos intros I've probably never seen. Do you never become curious about some date on the Nth occurrence rather than the first?
On Sun, Oct 12, 2008 at 12:04 PM, Gregory Maxwell gmaxwell@gmail.comwrote:
On Sun, Oct 12, 2008 at 9:19 AM, Elias Friedman elipongo@gmail.com wrote:
I agree as well. A year should be linked in its first appearance in the article-
[snip]
Why the *first* appearance? We've been able to link into sections for a long time, there are many articles I've read whos intros I've probably never seen. Do you never become curious about some date on the Nth occurrence rather than the first?
Sorry, I was trying to be succinct; my first draft of the email actually did contain a sentence or so about "distant" sections and lists/tables. These things are already addressed in [[WP:OVERLINK]] and [[WP:BUILD THE WEB]], I just didn't want to clutter up my note with stuff everyone already knows. On a side note, it seems there's a discussion to merge those two guidelines into [[WP:MOSLINK]].
2008/10/12 Elias Friedman elipongo@gmail.com:
I agree as well. A year should be linked in its first appearance in the article- even modern dates will be "history" in just a few years. Historical context is *important* and this total de-linking of all dates is damaging.
As an aside, I like the arguments for year-linking. This doesn't require full date-linking, though - why not delink the "May 17" part and leave "1973"?
On Fri, Oct 10, 2008 at 6:42 AM, nsk nsk@karastathis.org wrote:
I read Wikipedia a lot and I always enjoy clicking on date/year links, or using popups on them, because I am interested to place the information I read within a historical context, which is rarely provided by the article text.
[snip]
However, delinking dates and years, or only allowing links directly relevant to the context,
If you want almost everything linked then you want something that could be accomplished by machine. Better to run some userscript that autolinks dates/etc than to fill the article with low value links: The machine linking can do it with perfect consistency and does not require per-article effort to maintain, while the manual linking allows the identification of the significant, a task the machine is ill suited for, and preserves the ability for people who do not care for everything being linked to still browse with the most significant links provided.
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
If you want almost everything linked then you want something that could be accomplished by machine. Better to run some userscript that autolinks dates/etc than to fill the article with low value links:
To me a "low value link" is one that pipes a noun into a verb, adjective, adverb, etc. e.g.
*[[malice|maliciously]] [[injury|injured]] by a [[baldness|bald]] [[table tennis|ping pong player]].
I remove this sort of crap if I see it but I can almost guarantee they weren't added by any machine.
The machine linking can do it with perfect consistency and does not require per-article effort to maintain
Yeah, uh huh. *Sherman ordered them to [[March 2000]] furlongs eastward. *On channel [[12 June]] Cleaver is cooking supper. *Every [[December 30]] relatives show up uninvited. *On [[January 6]], [[1680]] homes lost electricity.
while the manual linking allows the identification of the significant, a task the machine is ill suited for, and preserves the ability for people who do not care for everything being linked to still browse with the most significant links provided.
Frankly I don't trust the community to decide which links are significant.
http://en.wikipedia.org/w/index.php?title=Switzerland&diff=prev&oldi...
If dates, sovereign nations, and their official languages are really "low-value links" Wikipedia might as well be written on paper.
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
Why the *first* appearance? We've been able to link into sections for a long time, there are many articles I've read whos intros I've probably never seen. Do you never become curious about some date on the Nth occurrence rather than the first?
I will agree with this. Surely one becomes more curious when information is repeated, not less. Once per "screen height" is probably idea. I'd want more if using sortable tables.
—C.W.
On Sun, Oct 12, 2008 at 1:45 PM, Charlotte Webb charlottethewebb@gmail.com wrote:
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
If you want almost everything linked then you want something that could be accomplished by machine. Better to run some userscript that autolinks dates/etc than to fill the article with low value links:
To me a "low value link" is one that pipes a noun into a verb, adjective, adverb, etc. e.g.
*[[malice|maliciously]] [[injury|injured]] by a [[baldness|bald]] [[table tennis|ping pong player]].
There are all kinds of low value links.
Your verb example is a sub-case of my own linking pet peeve: Linking some specific instance of something to a 'generic' article. In your example I'd expect "injured" to link to some article about this specific instance of injury, or at least the specific class of injury. It's not just noun->adfoo, it happens with perfectly noun-noun things as well. In your example I would have expected "ping poing player" to link to an article about the specific involved ping pong player.
Generally I see that as a pattern of linking to something boring and generic that I could easily look up if I were interested in it and in which reading about would not enhance my knowledge of the subject at hand very much.
I, and no doubt many other people, see many instances of date linking as examples of exactly the same class of bad linking: If I'm reading about the discovery of [[Uranus]] the material on "1781" and especially "March 13" are of no more value to me than "injury", "malice", or "baldness" are in your example. I do not intend to say completely without value, but not more than other generic things in the article text. Colors, common verbs, and other things I could easily search for if I were interested.
It seems that your valuation of links differs from mine, which shouldn't be a surprise: People can hold considered opinions which none-the-less differ.
Yeah, uh huh. *Sherman ordered them to [[March 2000]] furlongs eastward. *On channel [[12 June]] Cleaver is cooking supper. *Every [[December 30]] relatives show up uninvited. *On [[January 6]], [[1680]] homes lost electricity.
Those are all terrible sentences which would confuse or mislead readers even in the absence of links.
I have no doubt that you could cook up some examples which were were not ambiguous, but rejecting a 99.999% solution because it's not a 100% solution is a perfect way to abandon being good entirely.
There are two reasonable ways I see to close the gap to 100% there:
1) Have a 'grouping markup' in the text, or multiple link types. I.e. [[{12 June}]], it's not a link but it would indicate that the contained text is an atomic unit. This would assist all forms of machine parsing.
and/or
2) Instead of auto-linking text use JS to allow readers to highlight ANY contiguous span of text and click to get an "I'm feeling lucky" search result. This would allow article authors to link on the "you really should know about this" text, but still provide readers with the ability to explore generic concepts with which they may lack familiarity.
(as a side benefit the same linking JS could be provided to third party websites which might find it useful and which would drive more traffic to Wikipedia)
while the manual linking allows the identification of the significant, a task the machine is ill suited for, and preserves the ability for people who do not care for everything being linked to still browse with the most significant links provided.
Frankly I don't trust the community to decide which links are significant.
Then you're wasting your time arguing about this, aren't you? There are millions of pages on Wikipedia, so you couldn't maintain the linking personally even if the community would let you be supreme-god-of-linking. If the community can't achieve your definition of success then your definition of success simply can't be met. Our attention would be better spent on achievable goals, no matter how good we decide your criteria to be, and the mass linking of dates is by no means a universally agreed good thing.
In your example I would have expected "ping poing player" to link to an article about the specific involved ping pong player.
Ouch, that's *my* pet peeve! Don't link to an article using text that could well be the title of another article. A generic phrase like "ping pong player" should only link to a generic article about ping pong players (or the player section of the ping pong article if we don't have an article dedicated to ping pong players). If you want to link to a specific player, use their name as the link text. Piping is great when using an alternative name for something or for using something in a different part of speech than the title of the article (redirects work just as well), it shouldn't be used for hiding what article is actually being linked to. You should be able to tell from the link text what to expect when you click on it.
On Sun, Oct 12, 2008 at 3:30 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
In your example I would have expected "ping poing player" to link to an article about the specific involved ping pong player.
Ouch, that's *my* pet peeve! Don't link to an article using text that could well be the title of another article. A generic phrase like "ping pong player" should only link to a generic article about ping pong players (or the player section of the ping pong article if we don't have an article dedicated to ping pong players). If you want to link to a specific player, use their name as the link text. Piping is great when using an alternative name for something or for using something in a different part of speech than the title of the article (redirects work just as well), it shouldn't be used for hiding what article is actually being linked to. You should be able to tell from the link text what to expect when you click on it.
I'd argue that a link used in the direct object of a sentence should be a link about the direct object or not a link at all. If the sentence is about a general topic, it should be a link to a generic topic. Something specific, to something specific. Or in other words the linking should be at my conceptual level of understanding, which is usually a much higher level than 'just the text'. I can't say with confidence that I am even aware of the exact text that I'm reading in most cases unless I pay special attention.
My complaint is not so much that I am pro-linking-the-generic-to-the-specific, so much as I am *against* linking to a generic thing when the sentence has me thinking of a specific thing. I find it upsetting, frustrating, and confusing. I alone am not a statical study, but I've heard from others with the same complaint.
I think this means our positions are not mutually exclusive: I suspect both of us happy by not linking the "generic words" at all and always including specific words when you intend to link. Although this might cause some twisty wording in cases where the specific name of something had to be needlessly repeated just for the purpose of anchoring a link.
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
Although this might cause some twisty wording in cases where the specific name of something had to be needlessly repeated just for the purpose of anchoring a link.
Well, nobody ever said writing was easy.
—C.W.
2008/10/12 Gregory Maxwell gmaxwell@gmail.com:
My complaint is not so much that I am pro-linking-the-generic-to-the-specific, so much as I am *against* linking to a generic thing when the sentence has me thinking of a specific thing. I find it upsetting, frustrating, and confusing. I alone am not a statical study, but I've heard from others with the same complaint.
The form of this that always gets me is linking to a single word in a "proper noun phrase" - a set of words that might reasonably be the title of something.
"...at [[Cambridge]] University..."
"...the 17th [[cavalry]] regiment..."
"...[[John Smith]], a Member of [[Parliament]]..."
(exaggerated examples to make it clear)
I'm never entirely clear why these partial-links get made; perhaps to avoid a redlink at the time of creation, or perhaps because it isn't realised that the phrase is a single unit.
Although this might cause some twisty wording in cases where the specific name of something had to be needlessly repeated just for the purpose of anchoring a link.
Isn't this the same problem we sometimes get in the first sentence of an article, where we try to avoid linking title words?
"The '''1972 Ruritanian federal election''' was a [[Federal elections in Ruritania|federal election]] held in 1972 in [[Ruritania]].
I'm not sure if we ever found an elegant way of getting around that sort of problem.
On 10/13/08, Andrew Gray shimgray@gmail.com wrote:
"The '''1972 Ruritanian federal election''' was a [[Federal elections in Ruritania|federal election]] held in 1972 in [[Ruritania]].
Shrug, this may mildly confuse the average reader who either: A. clicks "federal election" expecting to find information summarizing federal elections in general, or: B. assumes the link isn't specific enough, and ignores it.
—C.W.
On Tue, Oct 14, 2008 at 10:15 AM, Charlotte Webb charlottethewebb@gmail.com wrote
On 10/13/08, Andrew Gray shimgray@gmail.com wrote:
"The '''1972 Ruritanian federal election''' was a [[Federal elections in Ruritania|federal election]] held in 1972 in [[Ruritania]].
Shrug, this may mildly confuse the average reader who either: A. clicks "federal election" expecting to find information summarizing federal elections in general, or: B. assumes the link isn't specific enough, and ignores it.
At least you would be able to easily get from [[Federal elections in Ruritania]] to whatever generic pages exist (as there should be links in the intro). The converse, finding the specific from linking to the generic, is not equivalently easy.
On 10/14/08, Gregory Maxwell gmaxwell@gmail.com wrote:
At least you would be able to easily get from [[Federal elections in Ruritania]] to whatever generic pages exist (as there should be links in the intro). The converse, finding the specific from linking to the generic, is not equivalently easy.
I agree, it's a helpful link to have, but hiding it behind a generic phrase violates the principle of least astonishment, though not as badly as [[Glenn Cowan|ping pong player]].
—C.W.
2008/10/14 Charlotte Webb charlottethewebb@gmail.com:
On 10/14/08, Gregory Maxwell gmaxwell@gmail.com wrote:
At least you would be able to easily get from [[Federal elections in Ruritania]] to whatever generic pages exist (as there should be links in the intro). The converse, finding the specific from linking to the generic, is not equivalently easy.
I agree, it's a helpful link to have, but hiding it behind a generic phrase violates the principle of least astonishment, though not as badly as [[Glenn Cowan|ping pong player]].
Mmm... though isn't that offset by the desire to define "Ruritania" in the opening line?
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
Your verb example is a sub-case of my own linking pet peeve: Linking some specific instance of something to a 'generic' article. In your example I'd expect "injured" to link to some article about this specific instance of injury, or at least the specific class of injury.
No it would be better to explicitly mention the type of injury as an undisguised link, never like [[bone fracture|injury]].
It's not just noun->adfoo, it happens with perfectly noun-noun things as well. In your example I would have expected "ping poing player" to link to an article about the specific involved ping pong player.
No, here it would be better to mention the ping pong player's name as an undisguised link, never like [[Glenn Cowan|ping pong player]].
—C.W.
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
Generally I see that as a pattern of linking to something boring and generic that I could easily look up if I were interested in it and in which reading about would not enhance my knowledge of the subject at hand very much.
Sounds like the MOS guys' post-de-linker advice, which seems to be "copy what you think might be an article title, and paste it in the search box". I only wish I was grossly exaggerating.
If I'm reading about the discovery of [[Uranus]] the material on "1781" and especially "March 13" are of no more value to me than "injury", "malice", or "baldness" are in your example.
Makes it easier to confirm that the discovery of Uranus is mentioned those pages (as it bloody ought to be). I believe systematically de-linking the day/month/year pages will sharply impede their development.
If somebody wants to write about events of [[1699]] they probably depend on whatlinkshere to learn that [[William Gustav of Anhalt-Dessau]] was born that year and [[Hortense Mancini]] died that year. I'm not sure how else they'd realistically find that out, google maybe?
I do not intend to say completely without value, but not more than other generic things in the article text.
I agree that "injury", "malice", and "baldness" are near-useless in my example, and I'm am frustrated that the de-linking scourge is targeted toward links which actually are useful, including but not limited to days, years, months, countries, languages, oceans, U.S. states except for Gerogia, provinces of Canada, famous cities which require no disambiguation (such as Toronto, Moscow, etc.), all based on an incredibly narrow estimate of what the mythical average reader will intentionally click on.
—C.W.
On Sun, Oct 12, 2008 at 4:51 PM, Charlotte Webb charlottethewebb@gmail.com wrote:
If somebody wants to write about events of [[1699]] they probably depend on whatlinkshere to learn that [[William Gustav of Anhalt-Dessau]] was born that year and [[Hortense Mancini]] died that year. I'm not sure how else they'd realistically find that out, google maybe?
My linking pet peeve is related: linking all the access dates for online references seriously dilutes the value of whatlinkshere for date articles. The only reason for linking them is to enable auto-formatting of the date, but they are among the lowest value links per se. We need a way to apply date formatting without creating links, and use that in all the citation templates (and encourage its use in manually formatted citations).
-ragesoss
2008/10/13 Sage Ross ragesoss+wikipedia@gmail.com:
My linking pet peeve is related: linking all the access dates for online references seriously dilutes the value of whatlinkshere for date articles.
As an aside, what's the consensus on internal links in references? If we have an article on the author, should we link it? How about one on the source itself?
(I can see arguments both ways)
On 10/13/08, Andrew Gray shimgray@gmail.com wrote:
As an aside, what's the consensus on internal links in references? If we have an article on the author, should we link it?
Yes.
How about one on the source itself?
Yes. If it's a news article you want to link to the title of the newspaper/magazine/etc., as we probably will not have an article about the article.
{{cite news | title = Netcraft confirms Wikipedia is dying | url = http://whatever/lol/document_1.htm | last = Gray | first = Andrew | authorlink = Andrew Gray (wikien-l user) | work = [[The Picayune Intelligence]] | etc = etc. }}
Fortunately I noticed [[Andrew Gray]] is a disambiguation page. For a book you would want to link to the title. The actual text is less likely to be online, and even if it is we would rather link to it from our article about the book.
{{cite book | title = [[Horse-ripping for Dummies]] | last = Gray | first = Andrew | authorlink = Andrew Gray (wikien-l user) | publisher = [[Paladin Press]] | etc = etc. }}
—C.W.
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
Those are all terrible sentences which would confuse or mislead readers even in the absence of links.
I have no doubt that you could cook up some examples which were were not ambiguous,
You are correct, but that doesn't mean such sentences don't appear "in the wild" so to speak.
but rejecting a 99.999% solution because it's not a 100% solution is a perfect way to abandon being good entirely.
It wouldn't be a 7% solution or even a 3/5 compromise as "client side" links would be all in our heads, not existing outside of the user's screen.
However, we have no clue what screen size and text size the client is viewing with, and it varies rather dramatically. So, once per "screen height" is simply not achievable in the text, though it could be accomplished via scripting on the client side.
You mentioned scripting earlier. The point I was trying to make is that it's disproportionately easier for a machine to reliably remove links than restore them. This applies to content in general but that's another matter.
I dunno that I'd find once per screen would be much of an improvement though: I'll see a word that I want to click on, but now I have to scan up to find a prior linked instance. Maybe one exists, maybe it doesn't. I'd be better off just searching.
That was shorthand for:
once per ==section==, unless
The sections are "too short", in which case once every other section, or The sections are "too long", in which case once every paragraph, unless:
The paragraphs are "too short", in which case once every other paragraph, or The paragraphs are "too long", in which case more than once per paragraph, but not "too much"
The article doesn't have sections to begin with, in which case FAIL
Which of course would be just as subjective as screen height. What I really mean is I don't like having to jump through ridiculous hoops to navigate from one article to another.
—C.W.
On Sun, 12 Oct 2008, Charlotte Webb wrote:
You mentioned scripting earlier. The point I was trying to make is that it's disproportionately easier for a machine to reliably remove links than restore them. This applies to content in general but that's another matter.
This is about the point where I mention spoiler warnings again. (And I'm not interested in technicalities about machine-assisted humans not counting as machines.)
Reading all this over, I can't see the harm (I think) in delinking [[1776]], lets say, but for common words? It depends on how common. If I write,
"Samuel Adams was hit in the [[face]] causing an [[Consussion|injury]] in [[1776]] by a frying pain wielded by a [[British]] loyalist," I could see delinking Face (obvious), Concussion (depends, was it an actual concussion or some really interesting valid link there?), and 1776, but why British? That's the sort of link that I'm prone to follow when I read an article. Compare these:
-> '''Bob Jones''', born 1956, was raised in [[England]]. -> '''Bob Jones''', born 1956, was raised in [[Namibia]]. -> '''Bob Jones''', born 1956, was raised in [[Mypos]].
On my recent FAC, the MOS stuff came up to delink Australia and England (it was about a band composed of two Australians and an Englishman). Would Namibia count as a common phrase? How about Azerbaijan? I don't know if it's for the MOS to decide commonality on things like place names or nationalities, as an example. Sure, I know lots about England. But maybe the person reading my article (or my Bob Jones example) doesn't. I know nothing about Namibia, so I'd be likely to clink that to see where Bob grew up. Maybe someone else reading about Bob is FROM Namibia, but never heard of Mypos.
- Joe
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
It might be better than "first use" which places a lot of links in the intro, since that is the place I am least likely to want to follow links except when I need definitions in order to understand the text. But I do not believe that once per rough screen would resolve the "link not available where I want it" problem. … but the ability to highlight-search would.
Highlighting a few words to trigger an ajax script and colors them based on whether they match an article title?
Haha, that would be WYSIWYG Editing 101 if you could also set it to save afterwards!!
I like this idea better:
There are two reasonable ways I see to close the gap to 100% there:
- Have a 'grouping markup' in the text, or multiple link types. I.e.
[[{12 June}]], it's not a link but it would indicate that the contained text is an atomic unit. This would assist all forms of machine parsing.
This would be great if it includes: 1. Database table to track these non-link links. 2. Various whatlinkshere modes to list them apart from or together with regular links. 3. Extend date format prefs with option to make dates coloured and clickable. 4. Checkbox for #3 enabled by default. 5. Some flag/parameter to prevent month/day inversion within quoted passages.
Make it happen, Greg!
—C.W.
On Sun, Oct 12, 2008 at 1:45 PM, Charlotte Webb charlottethewebb@gmail.com wrote:
On 10/12/08, Gregory Maxwell gmaxwell@gmail.com wrote:
Why the *first* appearance? We've been able to link into sections for a long time, there are many articles I've read whos intros I've probably never seen. Do you never become curious about some date on the Nth occurrence rather than the first?
I will agree with this. Surely one becomes more curious when information is repeated, not less. Once per "screen height" is probably idea. I'd want more if using sortable tables.
Sorry to split-reply, I hit send a bit too soon.
I agree with links becoming *more* interesting with repetition.
However, we have no clue what screen size and text size the client is viewing with, and it varies rather dramatically. So, once per "screen height" is simply not achievable in the text, though it could be accomplished via scripting on the client side.
I dunno that I'd find once per screen would be much of an improvement though: I'll see a word that I want to click on, but now I have to scan up to find a prior linked instance. Maybe one exists, maybe it doesn't. I'd be better off just searching.
It might be better than "first use" which places a lot of links in the intro, since that is the place I am least likely to want to follow links except when I need definitions in order to understand the text. But I do not believe that once per rough screen would resolve the "link not available where I want it" problem. … but the ability to highlight-search would.