Wikipedia talk:Easy navigation

Not "EN"

This looks like it might be heading in a good direction. But I must take issue with the shortcut. "EN" already has the meaning "English" on WP, most notably as part of "wiki.riteme.site". With that in mind, perhaps some name other than "Easy Navigation" should be chosen. Maybe "Navigation Conventions"? — Nowhither 18:07, 15 September 2005 (UTC)

Thanks for the comment. Other possibilities for the shortcut: WP:E-N; WP:ENAV - (WP:NAV is already taken by Wikipedia:Navigational templates: a request to re-redirect that to here is possible too, for instance proposing WP:NAVT or WP:NAV-T as an alternative for that other guideline);

Re. the name of the proposal: I'd keep it at "Easy navigation" for the time being - it's no more than a working title. It might become the final title (and I think there are good reasons for that), but more ideas might pop up along the road. --Francis Schonken 18:47, 15 September 2005 (UTC)

How about just WP:EZ? — Your server has been MC MasterChef :: Leave a tip — 06:35, 7 October 2005 (UTC)

Well, in between I already sort of implemented WP:ENAV (that's the one I could remember best) - is that OK for you? --Francis Schonken 10:27, 7 October 2005 (UTC)

Yes!

Okay, this is going to be awesome. It is so what I want from all of the naming convention, category, list, disambiguation, and redirect conversations. Second to NPOV, content, and accuracy, data navigation is what wikipedia is and should be about (otherwise, why use a wiki, and why include all those wikilinks?). So here's an outline proposal (by no means exhaustive) for what this proposal might cover:

1. Ways to navigate:

a. Links

b. Article titles

c. Disambiguation pages

d. Lists

e. Categories

f. Redirects

2. Ways to facilitate information flow:

a. Links

b. Redirects

c. Lists

d. Categories

One thing I'd like to see is an explicit mention of these items as ways to facilitate the flow of information. They should respond to the ways in which Wikipedia is and can be used, and all other considerations should be secondary. Especially with categories, I've seen a lot of talk about them being politically-oriented (which they sometimes are), but the only proposal usually is to delete them (which doesn't fix the problem, since the deletion is itself politically oriented). Changing the way we think about these things may help to fix the problem, though. If we look at categories, for example, as ways to facilitate the user getting the information they want, and try to anticipate possible information paths, while restricting ourselves exclusively to that activity (not exclusively exclusively, but when we're doing information-path oriented categorization), we can maybe push through some of the problem areas in a more NPOV manner. Likewise lists, redirects, and so on. So one thing I'd like to see from this talk page is suggestions regarding how we can increase the flow of information the user wants. Categorization schemes (I've got a few in mind), new ways to use lists and redirects and disambiguation pages, even new ways to write articles are all welcome and encouraged by me. So let's get crackin'! -Seth Mahoney 22:41, 15 September 2005 (UTC)

Despite the enthousiasm of maybe having struck "oil" here, I see that as a sign to go easy on this one. I mean: I think we've got something important here. Let's not rush too much, which might spoil. That being said: I think I'll already import your provisional scheme into the project page. Further, I'll try to find a spot on meta, to notify about this idea growing here. Does that seem OK? --Francis Schonken 08:38, 16 September 2005 (UTC)

Posted on:

• Wikipedia:WikiProject Wikidemia
• meta:Goings-on

--Francis Schonken 09:08, 16 September 2005 (UTC)

Further,

• linked from user:Francis Schonken (FWIW)
• linked from Wikipedia:Categories, lists, and series boxes

Even concocted an intro to the "wikipedia and navigation" section from your comments above. Feel free to hack into the project page! --Francis Schonken 09:57, 16 September 2005 (UTC)

Don't let my enthusiasm fool you - I totally agree that going slowly and involving as many people with different views as possible and gaining consensus are good ideas. Also, well-formed definitions of terms (which I'd be happy to work on), well-researched guidelines regarding visual elements and document links (which I'd also be happy to work on), etc. are good ideas. Four other areas that it would be nice to cover here are: "See also" sections, "Further reading" sections, "External links" sections, and bibliographies/references (including weblinked references). The first should have guidelines regarding when they should be used, and when linking in an article or template should be preferred (and why). The last three are important here because they facilitate and document the flow of information into and out of Wikipedia. -Seth Mahoney 17:27, 16 September 2005 (UTC)

Added some text re. dab pages, including "graphic" example, and re. nav templates (all of this too long, but can be shortened once the thread of the guideline text becomes clearer).

Anyway, please add what inspiration tells you. The hardest work will be the compacting anyway.

Re. the "navigation back and forth to the external world": I'd put that in a separate section. The section that is now started, I'd keep that to internal wikipedia navigation; the "external navigation" section below it. --Francis Schonken 19:14, 16 September 2005 (UTC)

Recent edit

My recent edit mainly changed the structure of the proposal and added some short definitions and a shorter pre-intro intro. Just so its been said, feel free to change it - it is by no means my intention to hijack your proposal by reframing the goals in my terms. -Seth Mahoney 19:02, 16 September 2005 (UTC)

For the time being I moved my original (longer) intro back up. Yours is probably OK too (and anyways has the advantage of being shorter), but is rather technical.

What I think possible (but the future will have to make clear if that is indeed so), is to describe small steps of what wikipedians can do to improve navigability. Each small step (when writing an article; when making a category; a dab page;...) contributing to an enhanced information flow. Most wikipedians (and surely not starting wikipedians) don't have complete wikipedia information paths in their head (nor should we "instruct" them to encompass that). But they can get advise on how to improve their contributions. When that advise (which I see as the topic of this guideline) is written cleverly, it will contribute to navigability of those larger-stretch information paths. The theoretical analysis of these concepts, I'd see rather as the subject of a Wikidemia/m:Wikiversity project (you know at meta there's a vote going on to get Wikiversity started as a separate project!). --Francis Schonken 10:20, 17 September 2005 (UTC)

One of the things I'm hoping can be spun off this proposal is a group to discuss and implement overall strategies for improving information flow, which is why I got into the technical style/language. But yeah, you're probably right - as a general proposal it is a bit much. -Seth Mahoney 01:11, 18 September 2005 (UTC)

Is the overall intent is to be a style guideline?

Is the long-term intent that this guideline becomes part of the MoS? There are a number of current overlapping, sometimes conflicting, guidelines and policies that it touches upon. I'd like to know what particular niche is envisioned. I'm not at all saying that something like isn't needed, but I think it would be helpful to know where in the maze of helps and guidelines and policies it would end up. -- Rick Block (talk) 23:19, 17 September 2005 (UTC)

Ha, interesting question. The {{proposal}} template mentions these three "types" of thing it could eventually become:

• policy
• guideline
• process

->"Policy" as in the meaning of "those constitutional fundamental things that drive wikipedia" (like the NPOV policy), would not be the right category - I don't think this guideline would ever get that fundamental.

->"Guideline" as in the sense of MoS, was not my intention when I started the idea, rather: one of the philosophies behind good MoS writing. So it's a bit more meta than most of the existing MoS guidelines. Note that I announced the development of this proposal on meta. Why I have it here nonetheless, is because it develops as something specifically for English wikipedia: spin-offs for other languages and types of projects might occur in a later stage, but that's too soon to say yet (first it'd need to prove it can be made into something operational for en-wikipedia, if that doesn't work, I wouldn't be interested in philosophising about it on meta). But yes, here on en-Wikipedia it might become part of the MoS somewhere, eventually.

->"Process": that's very close to what I intended it to be too. I also think Seth (see above) sees it much in this direction. But also here, rather the over-all view on web-navigation processes @ wikipedia, than the guideline to any particular of these processes (which would be more the usual MoS type of guideline).

I really, really would like to ask you to notify here about *particular* conflicts with existing guidelines & policies. That's where it would become interesting: some things that are tried to be solved here, were maybe already solved in other places much better than what I'm proposing, so please let me know about these things you know about, that can only enhance it (or show it is redundant while someone else already had the idea). On the other hand, it might help to solve some "eternal conflicts" in wikipedia guidelines. Some recurring naming conventions discussions are about how much "content" can be transmitted with the name of a page. Some content of article discussions are about how wikipedia relates to web search engines. The principle proposed is to make a clearer distinction: when article content is in question, don't worry about how it will turn out in wikipedia's internal search engine, nor about how it turns up in Google or Yahoo (makes it easier to come to decisions on some of these issues) - and when page names are in question, the page names that most enhance web navigation are preffered (which also allows to cut the cake, for some lingering discussions in that area).

My angle on this is coming from naming conventions discussions. Recently there was a WP:RM vote whether a page name (referring to the same person) would be Philip the Good or Philip III, Duke of Burgundy. While no consensus on either of these, also Philip the Good, Duke of Burgundy was added as an option. The vote resulted in "no move" while all of these options (and still some more that were eventually added) all received about the same amount of votes. The present guideline is intended to maybe cut down on options in such a vote, that is, barring those options that never would lead to good navigation. And then vote results would maybe become more outspoken, having the number of options reduced. If you want to see how something like that is currently going on, take a look at the vote going on at talk:Victoria, Princess Royal and Empress Frederick - the original 10 or so alternatives were reduced to four. I would further reduce the options to maybe two if I could let "navigability" play as an argument, but I can't let that play as an argument yet, while this guideline does not exist yet. Another thing, not touched upon in these votes, neither yet in guidelines as far as I know, is trying to define what the ideal "minimal" set of redirects to such pages would be. Thus far redirects are defined by their upper limit: what is "allowed" as redirect. Not: what is the minimal set of redirects (...and of disambig pages) one would need to have to allow an acceptable level of navigation quality towards "difficult" to name pages. We're already doing better than Britannica on-line in that respect, but that doesn't mean no further improvement would be possible!

So, what would you think about qualifying it a "MoS meta-guideline" for the time being (but as much working title as the name of the proposal itself). As you can see on the project page of this proposal, I had the idea of at least one other of such MoS meta-guidelines, the coverage guideline, giving helicopter view on inclusion criteria, article structure (across biography articles, comics articles, and whatever other type of WikiProject-formatted articles there are), level of quality for articles, even lay-out issues etc --Francis Schonken 15:17, 18 September 2005 (UTC)

Issues related to the fundamental concept of navigational convenience come up nearly constantly on WP:CFD and WP:TFD. I believe this is one of the issues behind user:Radiant!'s WP:CENT. The comment about conflicting guidelines and policies was not directed toward this page, but other pages that already exist. Policies, guidelines, and processes that are related (feel free to add more to this list) are listed below. I believe many folks agree this whole area is ripe for a significant refactoring effort. -- Rick Block (talk) 18:07, 18 September 2005 (UTC)

Hi, thanks! With that naughty character of mine I can't avoid to add some short comments (these are well-meant, believe me) --Francis Schonken 21:43, 18 September 2005 (UTC)

(in no particular order)

• Wikipedia:Categories for deletion - Yeah, know this one. It triggered me to start writing a guideline the first time, which is still up and running.
• Wikipedia:Templates for deletion - Less known by me, except that single time the template connected to the guideline mentioned in the previous point was up for TfD. It survived. I reverted the change to the template text that had resulted from the vote, and put the new text in a separate alternate template. Further, I changed the TfD deletion criteria as a consequence. The change is still there.
• Wikipedia:Categorization - Yeah, of course, the guideline mentioned in the previous points was about categorisation.
• Wikipedia:Categories, lists, and series boxes - Again, one of the old friends. I'm happy this one improved a lot since the last time I worked on it (in its very early days).
• Wikipedia:Redirect - Indeed, as mentioned above: very complete guideline, apart from not yet defining the "minimal set" of redirects very well. That point is however hard to get a grip on, without general "navigation" concepts - which the present Easy navigation guideline is about.
• Wikipedia:Article series - This is indeed grouping several things. But note that some of these are hanging around as "dormant" guidelines (also, again recognising some old friends, long ago I tried to group the several proposals for navigational templates that had been popping up at short intervals in the second half of 2004). This ensemble of guidelines might benefit from some freshing up, to which the present meta-guideline might help, in sorting out what should be kept, and what becomes obsolete.
• Wikipedia:Embedded list - this is definitely one that should be affected by the Easy navigation guideline. The tension between fluent text and easy lists is one of these things that recurs in discussions in several forms. The present guideline would, as seems most of the time the option, a clean separation between the text (of which paragraphs should not be too long, but that's for the coverage guideline), and lists that are navigational aids, and so should not take too much room on text pages.
• Wikipedia:Disambiguation - yeah, of course. presently the main issues seem to be whether or not to use piped links on disambig pages; whether or not to allow additional links apart from the things that are disambiguated; and whether or not each disambig bullet should start with the link to the item that is disambiguated. "Easy navigation" will have it's angle at each of these issues. Other disambig issues include, top-of-the-text-page disambig notices (see for instance the horror of three of such sentences at Elizabeth of Bohemia - I put them there, and, alas, still see no possibility of reducing them: anyway: that's what this easy navigation guideline is about, trying to find ways to ease the navigation in such cases).
• Wikipedia:Manual of Style (disambiguation pages) - see previous
• Wikipedia:Manual of Style (links) - that's definitely something hooked to this new guideline proposal too.
• Wikipedia:Naming conventions - I got stuck with the wikipedia:naming conventions (people) proposal, which is nearly finished: just the final touch, which depends from this "easy navigation" guideline is missing.
• Wikipedia:Incumbent series - layout has improved, compared to what is still used in most articles of royals. There's a tiny remark to that one too, but which is quite important w.r.t. page names of people. I aleady applied that improvement to some of the pages of Belgian kings, but will come back to that. The proposed improvement is already mentioned in my talk with John K. on talk:Frederick II of Prussia if I remember well.

----

• WP:CENT - new to me. Seems like a market place for these kind of proposals. Will go look in that a bit deeper. On first sight I don't see the difference with Wikipedia:current surveys very well; sort of looks like the less official version of the same.

(comments added by --Francis Schonken 21:43, 18 September 2005 (UTC))

"Type" of guideline

Finally, this morning I started a new subcategory of category:wikipedia guidelines: the category:wikipedia coherence - If this ENAV guideline makes it to guideline, then I suppose it will be a wikipedia:coherence type of guideline. --Francis Schonken 11:46, 7 October 2005 (UTC)

New template

Just made this new template:maindab template, and used it on Victoria (disambiguation):

For example, {{maindab|WP:CENT}} produces:

{{maindab}}

(for use immediately under section headers on dab pages that become too long to be effective)

--Francis Schonken 11:14, 19 September 2005 (UTC)

Redirects, minimal set

Two quick proposals regarding the minimal set of redirects:

1. The set of redirects should include all other names the person goes by. This includes, for example, last names only (Hegel), initialized names (G.W.F. Hegel), first and last names (Eve Sedgwick), and first, middle/maiden, and last names (Eve Kosofsky Sedgwick). It also includes titles, and title-identifiers (where the title completely replaces the person's name).
2. The set of redirects should include all common misspellings of a person's name (where common can be defined as, say, 30% - I'm not particularly attached to this number - of google hits).

Regarding the first proposal, it only makes sense to have, as part of a minimal set of redirects, all possible correct ways of looking up a person article at the user's disposal (without resorting to the tedious article content search). Regarding the second, we shouldn't be in the business of penalizing users for not knowing how to spell a person's name, especially if it is commonly misspelled (plus, the redirect provides a gentle corrective reminder). -Seth Mahoney 17:19, 19 September 2005 (UTC)

Re. proposition 1: wouldn't do that: could clog up the search engine with too many irrelevant hits if looking for someone near to, but not exactly the same as, the one who has hundreds of redirects. Perhaps you didn't think about "hundreds of redirects"? Maybe try to crack this nut: British, German, Prussian, Vicky, Victoria, Viktoria, V., Adelaide Mary, A., M., Louise, L., Princess, princess, Royal, royal, of the United Kingdom, of the UK, of Great Britain, Ireland, Queen, queen, of Prussia, Empress, empress, of Germany, of Saxe-Coburg,-Gotha, Wettin, Kaiserin, Frederick, Friedrich, and: if throwing together these names, even if limited to correct sequences, immediately there are over a hundred possibilities. Adding popular errors, I wouldn't be surprised ending up with near a thousand.

Then you go searching for this poor soul's mum, and you get all the combinations of the following (for your convenience I bolded the differences): "British, German, Victoria, V. Alexandrina, A., Princess, princess, of the United Kingdom, of the UK, of Great Britain, Ireland, Queen, queen, Empress, empress, of Germany, of Kent and Strathearn, of Saxe-Coburg,-Gotha,-Saalfeld, Wettin, of India, and".

Maybe I missed a few, but the result would be obvious: looking for either of them by the wikipedia search engine would list several pages of search results, and you would be very happy to discern the one you were actually looking for, seen the limited info you get when a redirect page is listed.

So, here's my invitation: could you list the minimal set of redirects for Vicky, supposing that her page would stay at Victoria, Princess Royal and Empress Frederick? And yes, I would limit that to some ten redirects. Alternatively, have a look at her talk page: talk:Victoria, Princess Royal and Empress Frederick: there are some ten likely (and some "less likely") choices for renaming her page there.

For your proposal #2, yeah, that would work most of the time, but again not for all. Justine Henin, Justine Hardenne and Justine Hennin maybe make sense for redirecting to Justine Henin-Hardenne. However: Justine Hardenne Henin doesn't IMHO, allthough that would be the usual order of last names after marriage. --Francis Schonken 20:32, 19 September 2005 (UTC)

You are of course talking about an extreme example that is likely already covered under more specific policies. There is also room for more specific guidelines for name redirects, such as limit redirects to name combinations the person is known as, excluding certain groups (royal families, maybe, or people with non-Europeanish names who are likely already covered under other policies). The purpose, you're right, should be to make finding articles easier, and when the consequence of applying any policy actually results in making the article more difficult to find, the policy should be modified in that case. I think, though, for the vast majority of articles this would lead to a sensible number of redirects. -Seth Mahoney 21:12, 19 September 2005 (UTC)

I admit, the example is extreme: I gave the most extreme one I knew. But check Elizabeth of Bohemia too, nearly as extreme. And I have a few other examples, e.g. the emperors/empresses of the Latin Empire of Constantinople (and many of their relatives) are each of them nearly as complex

And no, it's not covered by the guidelines (and/or when applying the entangling rules of the guidelines this leads to questionable, unrecognisable and/or at least awkward and ambiguous results), that's why endless move operations, never-ending votes etc... keep going on. Well I have of course big hopes for this ENAV guideline (and for the wikipedia:naming conventions (people)) to bring that back to reasonable proportions. Doing the naming of one of such complex people pages presently takes more time and energy of more wikipedians than writing a new guideline, including all the consultation that takes.

And what is certainly not yet covered anywhere is the minimal (or in this case more likely: maximal) set of redirects. So please, it's a very good example, that can learn a lot in that respect.

You have to know I have uncharacteristically few interest in royalty, so for me they're just examples like any other to test improvements to navigation issues. --Francis Schonken 23:16, 19 September 2005 (UTC)

I didn't mean that it was a bad example, and if its not covered by other guidelines, maybe one of the things we should consider here and at the people naming page is when groups of pages should have their own specific set of guidelines and when those guidelines supercede what is decided here (I know you've already started on that work at the people naming page), and maybe one of the next projects would be to start setting up those more specific guidelines.

On a side note, one other thing we should decide here is how to handle articles that are about people who share a name with another person with the same or a similar name (check out Jonathan Katz for a good example of similar names; I can't think of one off the top of my head where people share the same name, but I know they're out there), including who gets the primary article (without the (actor) or whatever after the name), if anyone, and how to establish consistent secondary articles (with the (actor) or whatever after the name). I'm sure we already have guidelines regarding this, and maybe just copying them here would be sufficient, but they should at least be noted here and at the people names page as well.

Back to the subject at hand, though, with ease of navigation in mind, what minimal set of redirects would you support? -Seth Mahoney 16:10, 20 September 2005 (UTC)

Yeah, the double-named people are out there. I start to doubt whether you read the guideline proposal I was working on, wikipedia:naming conventions (people). It is specifically about that sort of things.

Royals & other nobility are presently handled by wikipedia:naming conventions (names and titles). So Vicky-with-the-endless-set-of-attributes is covered by that guideline, or should be: you see, I had expected you would've read that guideline by now too. I referred to it several times as the "set of instructions", remarking every time the instructions are so complex they rather entice endless debate than solutions. That's when I decided to start writing the "...naming conventions (people)". That guideline does not solely handle royals and nobility, but in general: "people", in a fashion derived from wikipedia:naming conventions (common names), as opposed to "...naming conventions (names & titles)" that refutes the "common names" principle, except for nearly impossible names. And even then, most adherents of the "names & titles" NC usually prefer artificial complexity, above the most common name.

And no, I don't think the ENAV guideline should copy-paste more detailed guidelines to here. If there is something more to be said about "braketed" disambiguation of persons than what is already in wikipedia:naming conventions (people)#Qualifier between brackets or parentheses, it should be there and not here.

"Minimal set of redirects": without having a clear answer yet, the only thing I'm quasi-convinced about is that it's not going to come out like an instruction that can be applied with mathematical precision. Rather something like: try to imagine what the names are people would type in the search field, to try and reach the subject a page is about. Further, maybe, something like that Google test can help in finding more forms of the word defining that page. The only mathematical instruction would rather be the "upper" limit: normal pages: maximum 5 redirects arriving; really, really complex pages: maximum 10 (but I don't want to pin myself on these numbers yet). --Francis Schonken 19:50, 20 September 2005 (UTC)

I have read both, actually. When I say something like, "I'm sure there's a guideline covering X somewhere" that's really just shorthand for, "I don't feel like looking up an exact reference at the moment." When I suggest covering here something we've covered elsewhere, I'm suggesting that we cover it here with the focus on navigability, maybe with the intention of preserving existing policies and maybe with the intention of revising them. When I say that, wherever they exist, we should consider maybe just copy-and-pasting them here, its as a point of reference, and only insofar as the guideline impacts navigability and/or should be modified with respect to navigability. One of the implicit claims I'm making here (and I don't really think its arguable) is that these pages all tie together and should work in concert and that we should facilitate that working in concert as much as we can. So yes, referencing other regulations is worthwhile, because people are going to be reading these pages (hopefully) and having that information handy is a good thing. And, hey, why not try to use Wikipedia's strengths (for example, linking) on this page to make it as navigable as possible?

Back to redirects, I hope you're not reading me as having said that we should be aiming for mathematical precision here. We're talking about guidelines, not strict rules. So if I mention, say, 30+% of a name's google hits as being significant enough to warrant a redirect, that's a guideline, not a strict rule. It should be taken as an indicator of what qualifies as significant, not as a rule that must be followed.

I'm wondering, too, why you are opposed to large numbers of redirects. When I use the search box at, say, the side of this page and type in, say, Hegel it takes me to the page about Hegel. It doesn't pull up a list of articles that mention Hegel, unless it can't find a page or redirect called Hegel (which it can, because someone was kind enough to create a redirect from Hegel to the actual article about Hegel). If it can't find such a page or redirect, then I get the list of articles that contain the word Hegel. Its exactly that list I'd like to see users not run into, and if that means creating extremely large lists of redirects, that's fine with me. In fact, as I see it, one of the major goals of this page should be helping the user to not have to use the current Wikipedia content search ever, because, frankly, its broken as a tool for finding articles with useful content. Categories, redirects, lists, and article links can and should all be used to further that goal, and their potential should be maximized. So when I'm talking about increasing information flow, on one level, this goal is exactly what I have in mind, because information flow, as it pertains to this project, is all about increasing the amount of useful information, which is exactly what the current Wikipedia content search does not provide. -Seth Mahoney 18:35, 21 September 2005 (UTC)

OK, I understand better. Sorry for the misunderstanding. Anyway, I just did:

• BG James, a wrestler with 7-odd stage names, and overlap with Brian James, the musician, and B. G. James, the politician. The move request was at WP:RM for some time, seen the backlog there I implemented the result of the vote (which in this case didn't need admin intervention). (As if I would be interested in wrestlers). Anyway, BG ended up with 9 redirects, not yet covering all the stage names mentioned on the BG James page. You can check these redirects here: Special:Whatlinkshere/BG_James. Brian James is a disambig page sorting out the wrestler and the musician. If you're weary of royals, i'd really like your idea on this one: are there redirects that still should be added? Could some of them be speedy deleted? etc.
• I added BG James and the two Engelbert Humperdincks as examples to Wikipedia:naming conventions (people), the last ones as examples to "bracketed disambiguation", while both are "(musician)".

I understand your referencing of other guidelines better. Another ideal for a guideline is, of course, too that it is short & clear. That was a bit of my fear which I maybe overexaggerated. But trying to cover too much in detail is definitely a danger for a high end guideline that touches many quarters of wikipedia guideline & policy processes (I estimate: about half of them; and that's a lot). --Francis Schonken 19:48, 21 September 2005 (UTC)

Still didn't reply to the search engine issue: one of the usability rules (can look up the reference, and will do some day, it's at http://www.useit.com): see users that users always can fall back on a search engine, if the rest doesn't work (what it does more often than we admit), the search engine should be helpful, not just a feature. Having said that,

software of the MediaWiki search engine has always been problematic, before it was too often not available, some time ago it went from bad to nearly useless at the moment. Yeah that's a problem. but that problem's rather something for m:Wikimedia Research Network to which I finally subscribed today.

the other side is: whatever the search engine, it has to be "fed" with the content of wikipedia. Search engine is part of navigability, so it's part of this guideline for the "content" part. If I type "Victoria lake" in the search field and get not anything near Lake Victoria in the search result, then that's probably partly a SW problem; but maybe also a page name problem. --Francis Schonken 21:55, 21 September 2005 (UTC)

I totally agree about not clogging this page up with lots of text that isn't necessary, but maybe we can use the page as an experiment in increasing navigability, as a way to show off how the proposal that ends up being developed here can work. That would have to include useful linking and perhaps a host of other ideas not yet developed, but if we can show, with this page, that really, really good navigability is possible by making the proposal perfectly clear and referencing other proposals in ways that make sense and are totally transparent to the reader, then we've done our job, and we can show everyone who sees the page what good navigation looks like, bolstering the strength of the proposal on two levels. -Seth Mahoney 16:41, 22 September 2005 (UTC)

Oops! Royals vs. wrestlers! I'm not tired of royals - I just don't know enough about the naming conventions used regarding them to make a coherent reply. As far as the wrestler goes (and anyone similar to whom this would reply), I'd say that every known stage name should be a redirect, as should every variation on his actual name that was fairly often used (and isn't used by another article), and as should every common misspelling of his name or his stage names, as well as variations (if applicable) in the way his name is spelled in English. As far as the Humperdincks go, Engelbert Humperdinck should be a disambiguation page, and the articles on specific Humperdincks should be bracketed. This aids in navigation by providing consistency, and also adds possibly interesting info to people who are interested in one Humperdinck and might not know about the other, information that might be missed otherwise. -Seth Mahoney 16:48, 22 September 2005 (UTC)

Search engine

I totally agree that the search engine is a part of navigability, and that problems with it should be directed to software people. However, it is still a problem, and one that we can patch rather easily by using those things that Wikipedia is good (and getting better) at: Categories, redirects, links, etc. (And you made my point for me: The search engine gives us back what we feed into it, so we should be feeding it good - useful - stuff!) Even if the search engine gets fixed, it should be a sort of last resort. Why type, when you can click? Then again, I've always preferred (and been better at) browsing to searching. Navigating a well-designed category tree is easier than typing what you hope to find into a search engine and scrolling through pages of stuff until you find what you want (for me, anyway), which is why I generally oppose the deletion of well-intentioned categories. Category trees based on areas of study (literature, history, physics, gender studies...), different conceptions of knowledge (philosophy, science, religions...), and different ways of looking at the same article(s) are going to help us out there. But when the search engine becomes necessary, broken as it is, I think well-designed (and I'm not saying my proposal is the best-designed) redirect systems and disambigs that help to ensure consistency and usefulness are going to be our best asset. I should be able to type in Hegel and get Hegel (which is a redirect), just as I should be able to type in Katz and get a disambig page that shows all the Katzes we have in Wikipedia. I don't generally see large numbers of redirects and disambiguation pages as a mess, but as a really handy thing when I absolutely have to do some searching (which is why I created the redirect Eve Kofosky Sedgwick - there are tons of people who switch that 'f' and 's' around). So I'm trying to make the case here for using redirects, lists, and disambigs to bolster the capacities of the search engine by creating pages that will be immediately pulled up when their title is entered into the search box. Likewise well-constructed article titles. One of the ways to deal with this on a title level, for people accustomed to Wikipedia, is to go for consistent titles. One of the ways to deal with it for those not accustomed to Wikipedia, is to use as many redirects as are useful. Categories and links, on the other hand, I'd like to see as alternatives to searching, ways to get away from ever having to use that search box if you don't want to. -Seth Mahoney 16:41, 22 September 2005 (UTC)

I won't go in the details, because we basically agree. Just a few points:

• Last names are often taken. For classical composer's that's a fact (I think before any guideline in that sense was written); for the early Roman emperors it's in the guideline (that is to say: the "single name", which for Ancient Romans is not always what we call a "last name"). A few weeks ago I had to redo Augustus, while someone had taken it for the honorific, which I put at Augustus (honorific) as per guideline. Beethoven was and is exclusively redirect to the composer, forgetting the film with the same name (for which I made a disambig notice on top of the Ludwig van Beethoven page). I suppose Humperdinck redirecting to the classical composer is still a remainder of that (I suppose unwritten) rule of having classical composers always having a claim on their last name (which for the Strausses is a problem of course, there are a host of them, not all relatives: so in this case it leads to a disambig, where eventually even Lévi-Strauss got a fair chance - but as you see by the red link he could do with a few extra redirects) - In short, the last name redirecting to the most obvious one has, I think, always been an implied rule, only some made more work from it, and the result is not always equal. But it's hard to catch in a rule. I don't think "Beethoven" should necessarily, as a rule, be more claimable by a person than a film. Nonetheless I used this composer and his derived names as an example for the rule I concocted in wikipedia:naming conventions (people).
• For search engine, for me there's no doubt some of it will be in v. 1.0 of the ENAV guideline, I mean, I already have some things in my head, just chewing a bit still for putting them into words. Usability is explicit: whatever all the other navigation options, a search engine should always be available, and work. Better have a rudimentary one that works than a high tech one with natural language analysis etc that doesn't produce anything near to an acceptable result half of the time.
• About large numbers of disambig pages and even about large numbers of redirects I think we really agree. But still it's not all that easy to catch in a good guideline proposal: as I tried to explain above: navigation can get a bit out of balance when some group of wikipedians experienced in a topic builds a whole web of redirects only supporting their topic, with many maybe even redundant redirects, and sort of claiming them even when the fringes of these redirects have overlap with other topics (claiming "Humperdinck" for the classical music, and the like). That sort of unbalance is also topic of ENAV guideline, even more than just the fact of supporting "excess redirects".
• Re "consistency of page name titles" I still think I have another type of consistency in mind than you do. For me the only real consistency would be capturing what people would be most likely to type in a search engine window when looking for a topic. With minor corrections re. orthography; avoidance of abbreviations; not adding qualifiers unless between brackets at the end, that sort of thing. As few "fixed schemes" per topic as possible, because such naming by naming scheme sooner or later unavoidably clashes with topics of another group (and other problems: the more rigid the rules, the more chance of endless dispute and people attracted to gaming the system). Did you know that presently Queen Elizabeth II redirects to a totally different page as Queen Elizabeth 2? Each of them a correct "redirect" for the type of topic they belong to, nonetheless you can see how easy things start to overlap.
• I'd like to hear your comments on my last additions to the ENAV proposal. It's too long, I know, but just want to know whether as an approach of how I elaborated some examples it might work?

Anyways, sometimes I forget to say: many thanks for your efforts here! --Francis Schonken 23:26, 22 September 2005 (UTC)

Detailed notes on current content

Content vs. navigation

I don't think there is quite the split you suggest here. Good navigation should include some rules of thumb for content (and the other way around, which you actually give in the templates section). I like what you hint at earlier, rather than later: that coverage and navigation go hand in hand, not that they are separate entities. Then again, this could just be an issue of language use, an unintended side-effect rather than the meaning you want to get across. So maybe there's just some tweaking in order.

Disambiguation pages

I really like what you've said here about what the appropriate content for disambiguation pages is (that is, little-to-no content).

Template locations

I'd like to see more about which locations will facilitate what kind of information flow. My initial perception is:

• Templates on the bottom are good for "this is the previous X" (on the left) and "this is the next X" (on the right).
• Templates in the upper right (or upper left, I guess) are better for displaying a series, especially when that series appears in some specific (say, chronological) order.

Which seems to be pretty much in agreement with what you say, except that mention of templates in the upper right area isn't as well developed.

Navigation for editors

I too like this idea of bringing together guidelines aimed at improving navigability of Wikipedia for readers looking to find or browse the content. What I am also looking for is a similar project, or set of pages, that helps editors navigate around wikipedia and the huge numbers of areas, portals, projects and so on. I came here hoping that this page would have something on it. I know about the Community portal (linked from the sidebar), but that is not quite the overview of the areas within Wikipedia that I was looking for. Does anyone know if such an overview or 'map' of wikipedia exists for editors? The closest I've come is looking in Category:Wikipedia. 194.200.237.219 14:41, 11 October 2005 (UTC)

Hi 194.200.237.219,

Of course if over-all navigability improves I think & hope that will also have an effect in "wikipedia:" namespace.

Some other tips and tricks:

• well, er, take a login, then there is a "welcoming committee" that posts a short list of links on your new user talk page - quite handy (one of these can be found here, but I'm not sure that's still very up to date);
• There's of course wikipedia:WikiProjects (see also Category:WikiProjects);
• Most of what you're looking for is in "wikipedia:" and "help:" namespace I suppose;
• There's also m:meta, taking a broader look on all Wikimedia-related projects (so also wiktionary, etc...)

Don't know whether these are things I help you with, but here you are --Francis Schonken 18:22, 11 October 2005 (UTC)

Policy maze

Since we're not very formal on Wikipedia, if something doesn't exactly fit in any of the given categories, it's better to take the cat descriptions with a grain of salt than to create a new subcategory. Honestly, this will only confuse others. If in doubt, please keep relevant pages simply in Category:Wikipedia guidelines, which contains all guidelines that aren't strictly about deletion, behavior, etc. Radiant_>|< 10:39, 29 October 2005 (UTC)