User talk:Jnc/Disambiguation

Um, slap me if I'm missing a point, but can't this be done by moving pages around? Move {foo} dismb pages to {foo} (dismb); move {foo} pages to {foo} whateverTheMainMeaning; fix all the links. It's a lot of work, but not a technical problem, right? Are you just trying to think of an easier way to do it? &mdash; Xiong (talk) 03:06, 2005 Mar 18


 * If I understand your comment (I'm not 100% sure I'm fully understanding your point), yes, there's nothing technical which stops us from doing it this way. My proposal is mostly couched in terms of new disambiguation pages because I felt that people would reject a proposal that called for changing disambiguation is done with existing pages (e.g. tree). Noel (talk) 18:47, 12 Jun 2005 (UTC)

I may overstepped accepted bounds, but having thought about it, I liked your view/approach enough to change disambiguation.

It's just right. but now I think I'm gonna get slapped down.

Josh Parris &#9993; 07:07, 18 May 2005 (UTC)

I'm all for any proposal that makes all disambiguation pages consistent. Now after some period of inactivity when I start doing some work on the 'pedia, I immediately realise that the naming conventions (yes, not just naming of pages, but the naming conventions themselves) have become so inconsistent that I can no longer guess whether there should be a disambiguation page or if so, where the disambiguation page is supposed to be. I've been slapped here and here already. -- Paddu 18:07, 17 July 2005 (UTC)

redirect
I think that the redirect doesn't neccessary need to go on the Foo page; if Foo (disambiguation) redirects to Foo you can get the same effect of being able to distinguish between the intentional and unintentional links to Foo. Josh Parris &#9993; 01:23, 21 July 2005 (UTC)


 * "Casual" links are usually made to Foo and not Foo (disambiguation). With Noel's approach we can blindly assume all such links have to be changed to point to either Foo (some topic) or Foo (disambiguation).


 * With your approach, some people would intend to link to the disambiguation page, and upon seeing that the Foo page is a disambiguation, without checking for the existence of a Foo (disambiguation) page would link Foo. Hence their links would get mixed with casual links in Special:Whatlinkshere/Foo. -- Paddu 21:17, 21 July 2005 (UTC)


 * Hmmmm... I suspect you'd get a similar set of behaviours with a redirect - because redirects are almost transparent. And remember: no one checks what they've linked to. Josh Parris &#9993; 00:10, 22 July 2005 (UTC)


 * All "wrong" links point to Foo and not to Foo (disambiguation) whatever be the approach. If we declare Foo (disambiguation) to be the "right" page to link to, there would be no "right" links to Foo, so whenever Special:Recentchangeslinked/Foo gets even a single link we can be assured that needs to be changed and go ahead and change it. But if we declare Foo to be the "right" page to link to, Special:Recentchangeslinked/Foo will always be non-empty due to a few "right" links and it is difficult to detect when someone has "wrongly" linked to Foo. I think what we declare to be the "right" link is more important than where the redirect lies and where the disambiguation page lies. -- Paddu 05:15, 22 July 2005 (UTC)


 * I understand and agree. My point is different.  I say the dab text doesn't need to go at Foo (disambiguation), it can go at Foo and have Foo (disambiguation) redirect to Foo.  Everything else, including nothing ever linking to Foo, holds. Josh Parris &#9993; 06:24, 22 July 2005 (UTC)


 * Hmmm. IIRC that was the policy on disambiguation months/years ago. Disambig pages called Foo have no links except for a manually generated list (since replaced by categories) and any explicit link must be at Foo (disambig). Somehow all the inconsistencies that I've seen or something else made me forget that. But the main problem is to tackle the partisans of the "if one meaning dominates there shouldn't be a disambig at Foo" rule (whoever started such a rule... &mdash; I personally think what is said to dominate is pretty much POV in some of the cases and might vary widely in the cosmopolitan group of Wikipedians). -- Paddu 20:07, 22 July 2005 (UTC)

Josh, when you say the dab text doesn't need to go at "Foo (disambiguation)", it can go at "Foo" and have "Foo (disambiguation)" redirect to "Foo", you seem to be assuming that nobody would ever want to link to either "Foo" or "Foo (disambiguation)", so whichever name one puts the actual disambig text at (with a redir from the other), any link to either is inherently suspect. To put it another way, you are implying that the two names are effectively identical, so it doesn't matter which one holds the text, and which one is the redirect. However, that assumption is incorrect. Sometimes, for example, dab pages contain definitions which aren't worth making pages for, and then one does get valid links to dab pages; there are other cases, too. Which is why I say to put the dab text at "Foo (disambiguation)": you can pretty much bet that anyone who links to "Foo (disambiguation)" really did want the dab page, whereas with links to "Foo", it's most likely just someone being lazy and not checking their link targets. Noel (talk) 17:14, 25 August 2005 (UTC)

Efficiency, etc
Absolutely NOT. Other than alternate spelling/capitalization redirs, if a reader types "foo" they should get either an article on foo, or a disamb page. Passing them thru a redir to either is a pointless waste of Wikipedia DB servers and the reader's time. Any links to foo should be checked regardless of whether it's a redir or a disamb page. Niteowlneils 20:09, 21 August 2005 (UTC)


 * Your point about waste of Wikipedia DB servers is incorrect. First, if efficiency were the sole measure of a way of doing things, the Wikimedia software would be written in C, not PHP. In this case, we need to ask "what is the benefit of going through a redirect to get to the disambiguation page, and what does it cost", and I think the benefit (being able to find links where people have not bothered to check to make sure they go to the right target) outweighs thev cost.
 * Second,, because of the way redirects are implemented, the cost is not as high as you think. To contrast, every page that contains a template (and many contain more than one) is already doing several database lookups, and nobody suggests getting rid of templates. Because of the way redirects work, they are actually more efficient than templates; with templates, you have to look up the 'article' body in the database, whereas with redirects you don't: the name entry contains a direct pointer to the article in the database, it doesn't have to retrieve the 'article' body - which is why redirects to targets (i.e. subheaders) don't work (and won't, anytime soon).
 * Your second point, and the reader's time, might be more on target. I'm assuming you mean cases where one meaning is far more common than the others (although now that I read your edit comment, Passing them thru a redir to either is a pointl[ess], perhaps that's not what you meant). So, for example, putting "tree" at "tree (plant)" and making "tree" a disambiguation makes people look at an extra page first. You do have a point there, I'll have to think about that for a while. Noel (talk) 17:09, 25 August 2005 (UTC)
 * Now that I think about it, the way to handle cases like the one above is to put the article on the plant at "tree (plant)", have the disambig at "tree (disambiguation)", and make "tree" a redirect to the first one. That still allows us to find people who link to "Foo" without checking to see if they got the right kind of "Foo", and it means that people who go to "Tree" get the article on the plant first. Noel (talk) 17:32, 25 August 2005 (UTC)
 * I think that your proposal for to be a redirect in some cases is a good one and if everyone was willing to play by the rules a workable solution. I am going to mention some concret examples because I find it easier to make a point that way. As we are debating at "Talk:William of Orange" if "William of Orange"->"William III of England" instead of a redirecting to "William of Orange (disambiguation)" then because most English usage of the phrase "William of Orange" without qualification are to William III, less links will be broken between clean ups. However this case happens by chance because of a peculiarity of the way European monarch pages are called and it has caused an edit war as to where the redirect should point. There has just been a WP:RM debate on the page Cricket. If the page Cricket was to be moved to "Cricket (sport)" then it opens up the potential for a three way edit war for the redirect of "Cricket" to "Cricket (sport)", "Cricket (insect)" and "Cricket (disambiguation)" which can be even more time consuming (and destabalising) than fixing incorrect links which point to the current article Cricket (which is about sport) when meaning the insect. Philip Baird Shearer 21:20, 25 August 2005 (UTC)