Template talk:Navigation bar

moved from user talk:ligulem

Hi - I'm trying to address some issues with template:NavigationBar, one of which is that is looks bad in Opera. As far as I can tell, Opera interprets an explicit "height" style parameter in a div to include the height of a scrollbar that's added (if "overflow: auto" ends up requiring a scrollbar). Looking into this, IE seems to be the only browser that requires an explicit height (Firefox, Mozilla, Safari, and Opera all seem to just "do the right thing" with no height specified). Do you know of a way to have conditional code in a template based on the browser? I'd like to make the height specification only visible if the browser is IE. Any ideas? -- Rick Block (talk) 02:37, 19 October 2006 (UTC)[reply]

"Do you know of a way to have conditional code in a template based on the browser?" No. Sorry for being so short, but I really have no idea how that could be done. To my knowledge, conditions in templates can only be made based on very basic things like "is parameter X of the template empty or not", and "... is it equal/not equal to string Y". You might want to check with user:AzaToth (inventor of template:qif), user:Patrick, or the ultimate authority Tim Starling. You could also post your question to wikitech-l mailing list where the devs are reading. --Ligulem 08:03, 19 October 2006 (UTC)[reply]

I think this would work, but the html comment syntax is apparently independently treated as a wikitext comment (so html comments do not appear in the rendered output!). -- Rick Block (talk) 14:01, 19 October 2006 (UTC)[reply]

Didn't know this trick, interesting :). It's not surprising that we cannot generate comments in the generated html. We only have wiki-text comment (thrown away by the parser). Apperently, that trick is already used on our site: If I look at the generated html of my sandbox I find this inside it:

<!--[if lt IE 5.5000]><style type="text/css">@import "/skins-1.5/monobook/IE50Fixes.css?13";</style><![endif]--> <!--[if IE 5.5000]><style type="text/css">@import "/skins-1.5/monobook/IE55Fixes.css?13";</style><![endif]--> <!--[if IE 6]><style type="text/css">@import "/skins-1.5/monobook/IE60Fixes.css?13";</style><![endif]--> <!--[if IE 7]><style type="text/css">@import "/skins-1.5/monobook/IE70Fixes.css?13";</style><![endif]--> <!--[if lt IE 7]><script type="text/javascript" src="/skins-1.5/common/IEFixes.js?13"></script> <meta http-equiv="imagetoolbar" content="no" /><![endif]--> 

--Ligulem 08:33, 20 October 2006 (UTC)[reply]

IE apparently interprets CSS attributes with a leading "_" (and other browsers don't). I've changed the template to use this hack. -- Rick Block (talk) 02:03, 20 October 2006 (UTC)[reply]

...these are easier to type (see for example template:cite web). However, I seem to be too late already (sigh). --Ligulem 08:45, 20 October 2006 (UTC)[reply]

I changed it boldily. Usage seems to be low enough to do it using the brute force method. Was mainly used in two other templates. --Ligulem 08:55, 20 October 2006 (UTC)[reply]

I iterated trough all calls using m:MWB. All calls are converted to lowercase params now. --Ligulem 09:08, 20 October 2006 (UTC)[reply]

I've changed the name of the template from "NavigationBar" to "navigation bar". I suggest not to use CamelCase for template names nor parameters. --Ligulem 09:18, 20 October 2006 (UTC)[reply]

The name and parameter names were chosen to mirror template:NavigationBox. This template might still be merged into that one (by adding a new parameter, like "Scroll=yes" or something). I'd prefer keeping at least the parameter names consistent between these two templates, and don't really care which way it goes (but suspect it would be tremendously easier to change these back). -- Rick Block (talk) 13:15, 20 October 2006 (UTC)[reply]

I respectfully disagree. Standard on templates for parameter names is lowercase. Don't create a new template using upper case params. I've never heard of the concept of template "merging". --Ligulem 16:22, 20 October 2006 (UTC)[reply]

BTW why not migrate the calls to template:NavigationBox to this template here? ("merging" the other way round?). I could do that with MWB and my bot account. I know, this would be 19,049 pages to edit but upper case params are really ugly. --Ligulem 17:47, 20 October 2006 (UTC)[reply]

Do you mean make all references to template:NavigationBox scrollbar enabled? I think we might want to think about this on a case by case basis, but I don't think it would be appropriate to do it globally. In addition, there isn't universal acceptance of this format (see last thread on this page) and this template isn't exactly ready for prime time yet (still need to resolve the JAWS traversal issue). -- Rick Block (talk) 00:25, 21 October 2006 (UTC)[reply]

moved from WP:VPT

I've created template:NavigationBar, inspired by user:Pengo's template:Panorama. I've verified that it works with IE, Firefox, and Mozilla (on Windows) and Safari (on a Mac). If anyone can think of any technical reasons it should not be used please speak up. Initial discussion about this is at Wikipedia talk:Navigational templates#Compressed templates. template:Places in Bedfordshire uses it, producing:

see Template talk:Navigation bar/examples#Places in Bedfordshire

-- Rick Block (talk) 17:41, 15 October 2006 (UTC)[reply]

I read the initial discussion and understand that answers are here.

Thank you for your idea. I abhor long lists. Places in Bf'shire are many, anyway. What do I expect when I search - or stumble upon - one of them, like Beeston ? Mainly, info about the place.

Other places may not appear at all : there is a "show/hide" template that does the job. For me, it looks easier to see everything at once, when useful, than to scroll an indefinite length of less readable ('cause it moves) text. -- DLL ^{.. T} 20:32, 15 October 2006 (UTC)[reply]

Unlike template:hidden (show/hide), this technique works on all skins (since it's just CSS) although there's an issue with the printable version that I don't know how to fix yet. In general, I think the "title" for any actual use of this will link to an article or list that has all the entires simultaneously visible in some format. As you surmise, the point is to reduce the template clutter at the bottom of articles using large navigation boxes. -- Rick Block (talk) 20:49, 15 October 2006 (UTC)[reply]

This looks intriguing. Thanks for your work, Rick. David Kernow (talk) 01:21, 16 October 2006 (UTC)[reply]

The problem with print is a fairly big one though. We want our articles to be easily printable. These things can also complicate translation info other formats. If a "navigation template" is several hundred items big it's starting to get to the point where replacing it with a category/list of ... article would make more sense than a huge inline list anyway. --Sherool (talk) 05:33, 16 October 2006 (UTC)[reply]

Navboxes aren't necessary in printouts anyway; couldn't we just add something to MediaWiki:Common.css turning off media print? --ais523 08:09, 16 October 2006 (UTC)

The vertical scroll bar looks too tiny, and I can hardly see any text inside. Please verify that the NavigationBar also works in Opera, and I will have no objections to you using it. --J.L.W.S. The Special One 08:11, 16 October 2006 (UTC)[reply]

Per user talk:Graham87#NavigationBar, this causes traversal issues with the JAWS screenreader as well. I'll work on addressing all three of these issues (printing, Opera, and JAWS). I think it's clear this template should not be used until these issues are addressed. -- Rick Block (talk) 14:03, 16 October 2006 (UTC)[reply]

Since a navigation bar is intended to provide hyperlinks to other content it isn't really needed/useful in printed text and thus I'd agree with the suggestion about marking it as 'noprint'. JAWS, Opera, and other possible accessibility issues are a bigger concern. Very nice idea, but needs to work everywhere to be viable. --CBD 15:51, 16 October 2006 (UTC)[reply]

The templates horizontal bar works in Opera, however a vertical bar is also generated. Lcarsdata (Talk) 16:58, 16 October 2006 (UTC)[reply]

I've made a slightly tweaked version of the template in my user space here:

see Template talk:Navigation bar/examples#Ilmari Karonen sandbox

I've tested it with Firefox 1.5 and IE 6.0 on Windows so far, and it seems to work fine. Any comments from people using other browsers (especially Safari, since I don't have access to a Mac) are very much appreciated. —Ilmari Karonen (talk) 13:36, 19 October 2006 (UTC)[reply]

It seems that the new NavigationBar now works in Opera - I don't see any more vertical scroll bar in Opera 8.5. Thanks for making this Opera-compatible!

moved from WP:VPT

Since I might be the one guilty of making Rick Block changing his aproach I feel obliged to comment. Rick: This scrollbar aproach is much easier to use than your old concept. Very user friendly. But I am sorry to say I find it somewhat ugly. So I still think my suggestion to split it up into several templates with an alphabet range on top of each is nicer. Although my aproach will cause one extra page load and your scrollbar actually is more easy to use, both as a user and as an editor. By the way, Ilmari's version do look better in both my old IE 5.5 and my Firefox 1.5. (And I don't mean just the colours, I mean how the scrollbar looks.) --David Göthberg 09:03, 20 October 2006 (UTC)[reply]

Well, if it's easier to use as a user and as an editor is there some cosmetic change that might make it "look" better? Or, is the real issue the scrollbar itself? We could tweak the cosmetics (font, color, size, etc.) but the scrollbar is provided by browser so it will look how it looks. -- Rick Block (talk) 13:52, 20 October 2006 (UTC)[reply]

I took out my trusty old paint program so below are two drawn images (as one image file) of how I would like the navigation bar to look.

• The first image is with a scroll bar and is still somewhat ugly. The uglyness stems from the scroll bar and I don't think we can do much about that. But I added some margin space inside the frame which I think makes it look better than before.
• The second image shows how I really would like it to be. As say six different templates, each covering part of the range. Thus making each of those templates small enough to look nice in the articles. In the top of each template there are "alphabet links" which lead to the first article (town) in each range.

So these are my suggestions. What do you people think about it? --David Göthberg 17:24, 27 October 2006 (UTC)[reply]

We could certainly add margin around the current version. Multiple templates is reasonably OK with me, but does not generalize that well. What would you do with Template:Footer Olympic Champions 4x100 m Men (for example)? -- Rick Block (talk) 18:57, 27 October 2006 (UTC)[reply]

Ehm, isn't that obvious? (Perhaps it isn't since you are asking.) So I added it to the example above. When clicking on a range in the "year list" that should link you to the first person for that year range. --David Göthberg 20:04, 27 October 2006 (UTC)[reply]

I've fixed the issues that were previously listed in the template:caution box, specifically:

• Appearance when "printable version" is selected (printing a scrollbar is somewhat less than thrilling)

  ✔ Fixed by adding "noprint" to the class. As far as I know this works for all browsers except for Safari (which doesn't seem to process the "noprint" class correctly and prints anyway, at least in print preview).

• Appearance in Opera

  ✔ Fixed by changing the list height parameter to "_height" since IE is the only browser that seems to need this parameter to be explicitly provided (and other browsers ignore attributes if prefixed with "_"). This is a hack, making the inline CSS not compliant, but since it makes this template work in Opera I think it's worthwhile.

• Usage issue with JAWS ("next element" traversal shortcut not available, so it's difficult to traverse past)

  ✔ Fixed by making the displayed list an HTML list with one element with CSS to make the HTML list be displayed inline. This seems to disable the scrollbar in Mozilla 1.7 (on a Windows XP machine). (checking this again, it seems to be fine in this version of Mozilla)

I'm reasonably happy with the current version. It's not quirk free, but I think the functionality is tremendously useful. User:Ilmari Karonen's version is marginally simpler (no outer DIV). There's a pending question about "ugliness" (see above), but other than that does anyone have any other issues with the current version? Is "works with JAWS" vs. "no scrollbar for Mozilla 1.7" a reasonable tradeoff? (works in both) -- Rick Block (talk) 16:16, 22 October 2006 (UTC)[reply]

I really like it for use with large categories. Great solution.

The only problem I can see with using it as a navbox replacement, is you can't immediately see the link you're already at, in bold within it. However it is infinitely preferable to using the "hide" function, which looks confusingly like an [edit] link.

Full support. :) --Quiddity 09:45, 28 October 2006 (UTC)[reply]

Sorry if I'm raining on your parade, but this horizontal-scroller idea is a terrible solution for the problem of over-crowded navboxes. Scrollbars should be left to the main browser window and nowhere else on the page, especially in areas where more content or navigation would exist, as it is more often than not a usability nightmare. I do appreciate your effort, but this is not the way to go — it would be like using a steering wheel to function as a mouse. ...Don't just take my word for it; read what no-nonsense expert Jakob Nielsen has to say about this development: "Scrolling and Scrollbars".

Perhaps this idea can be spared to serve as an overflow limitation for overtly-long lines of text, but it would be smarter to explore other methods of displaying text content/navigation without masking it or using overflow:hidden. I would instead suggest using columns, or even some variation of the tab dividers (as seen on the Special:Preferences page.) Please avoid applying any form of horizontal scrolling for Wikipedia the future, for sanity's sake. —Down10 ^TACO 05:23, 26 October 2006 (UTC)[reply]

I wonder if we could get Jakob or some other genuine usability expert to comment on this. I took a look at the article you cited, and while the points it makes are good, I don't think most of them apply to this template. Let's look at the checklist he presents:

1. Offer a scrollbar if an area has scrolling content. Don't rely on auto-scrolling or on dragging, which people might not notice.
   • ✔ We do this.
2. Hide scrollbars if all content is visible. If people see a scrollbar, they assume there's additional content and will be frustrated if they can't scroll.
   • ✔ We do this too.
3. Comply with GUI standards and use scrollbars that look like scrollbars.
   • ✔ We do this as well.
4. Avoid horizontal scrolling on Web pages and minimize it elsewhere.
   • ✘ This we do fail. However, his later comments in the article are particularly targeted against having both horizontal and vertical scrollbars on the same element, which we don't do. It'd be nice to have him (or some other expert in the field) comment on the usability of horizontal-only scrolling, and of scrolling in-line elements in general, since I can't really tell from the article how he feels about that particular case.
5. Display all important information above the fold. Users often decide whether to stay or leave based on what they can see without scrolling.
   • ✔ The stuff in these navbars isn't part of the actual article content, and is not particularly important.

So, we fail one out of four by having a horizontally scrolling in-page element, which has to be navigated separately for users to see its entire contents. On the other hand, most users probably will not want to see it all; the part that is visible without scrolling, including the title, is enough to establish it as a themed list of "see also" links, in which the most readers are unlikely to be interested at all, and the ones that do are presumably motivated enough to move their mouse to scroll the list (they'll have to do so anyway to click the link).

All in all, I'm more concerned with the interactions with screen readers and other non-graphical browsers. If those are acceptably solved, I'm personally quite willing to slightly annoy the minority of users, who want access to these links without extra scrolling, for the benefit of the majority who just want to scroll (vertically!) past them. —Ilmari Karonen (talk) 13:25, 26 October 2006 (UTC)[reply]

I couldn't have said it better myself. I've checked out interaction with the JAWS screen reader and the initial version had an issue (the content was not "skippable", forcing the user to hear the whole dang thing before traversing to the remainder of the page) which has been fixed (by making the content a single element HTML list). I distinctly agree that scrolling should nearly always be avoided. However, for this specific purpose I think it's a whole lot less evil than the monstrous navigation boxes (effectively inline categories) folks seem to want at the bottom of articles. My actual preference is that we should use categories or lists for this sort of thing, but TfD attempts for these monstrosities regularly fail because of some vocal subset of users who like the "convenience" of having all the links right there on the same page as the article. -- Rick Block (talk) 18:42, 26 October 2006 (UTC)[reply]

Hmm, was that fixed in the version of Jaws that costs USD $900 to buy? Handicapped Wikipedia readers aren't made of money. Both Nielsen and the WCAG say to accommodate current technologies.

Regardless of that, the horizontal scroll bar is a terrible idea, and by deciding to use it at all, we would be ignoring and misinterpreting Nielsen's advice. Firstly, in that article he says not use horizontal scrolling, which he also rated number three in the Top Ten Web-Design Mistakes of 2002. The scrolling article is a list of ways to ameliorate the disaster, if you choose to ignore his initial advice. Ignoring even one of his points would be a second failure. In case you haven't been paying attention, he writes, with bold text, "users hate horizontal scrolling and always comment negatively when they encounter it. Customer satisfaction is surely reason enough to avoid horizontal scrolling."

Finally, the question of design—a horizontal scrolling list? Mystery meat navigation, with an arbitrary number of items always hidden? There's a reason the Mac OS and Windows don't implement horizontal scrolling lists as a standard interface widget—they're inefficient. We don't have to ask what Nielsen would think of this particular interface: we can just make note of the fact that no one uses it in their software or their web sites. If someone wants to argue the merits of this design further, please show us a few examples of well-designed functional web sites (not novelty concept designs) which use such a navigation element.

One more thought: how would different web browsers print this? —Michael Z. 2006-10-26 19:53 Z

The JAWS fix is with the current version (not the $900 one). The template is of class "noprint", so printing is inhibited (at least by browsers that honor the css media stuff). If anyone has a different practical solution to the issue of the visual clutter brought about by monstrous "inline category" navigation boxes, please speak up. The only other approach I've seen (given that we aren't going to refrain from using them) is the show/hide approach, which doesn't work across all skins. -- Rick Block (talk) 20:20, 26 October 2006 (UTC)[reply]

I strongly oppose horizontal scrollbars, or scrollbars within elements. The only scrollbar should be the main vertical one, easily scrollable with the keyboard (up and down, page up and page down, or spacebar). — Knowledge Seeker দ 00:50, 27 October 2006 (UTC)[reply]

Do you have an alternative solution to the issue this template is intended to address? Have you looked at Category:Living people? I absolutely agree horizontal scrolling is generally a very bad thing. However, as Ilmari explains above, this is content that most people really don't care about and even if we assume it's an inconvenience for those that do, is reducing the "clutter" in the articles where this template might be useful worth this inconvenience? -- Rick Block (talk) 03:58, 27 October 2006 (UTC)[reply]

How about this: If most people don't care about the info in the template, remove it. Link to a relevant category or list of articles instead, no point in listing every related article inline in every article! So for example instead of adding things like {{Footer Olympic Champions 400 m hurdles Men}} to every biography on a person who are an Olympic Champion in 400 m hurdles, just add a category, that's what they are there for. Alternatively link to a relevant list like List of Olympic medalists in athletics (men) and those interested can go there to explore. Much better than adding humongous "navigation" templates and then hiding the info using scrollbars, show/hide links or whatever. If a "navigation" template need a scrollbar it's too big to be usefull anyway, replace it with a category and keep navigation templates small and sweet like the sucession templates for presidents and royalty (previous next) that sort of thing. --Sherool (talk) 07:46, 27 October 2006 (UTC)[reply]

Some people do care, and will defend "their" template at TfD and will add it back if it's deleted from the articles. I agree we should link to a list or use a category, but there are a signficant number of folks who disagree with us. I previously created a "small and sweet" succession-ish version of template:Places in Bedfordshire that produced (for example)

{{Template:Places in Bedfordshire/small|Little Barford}}

allowing "direct" traversal to the "nearest" 4 places and "virtual scrolling" traversal to all (5 at a time). The template had a different appearance on each page (sort of "self scrolled"), but bascially required a script to create and was effectively not directly editable. Including the scrolling in the template itself is (IMO) roughly the same operation while allowing access to all of the items in the list (and preserving an editable template). -- Rick Block (talk) 14:31, 27 October 2006 (UTC)[reply]

Mostly for demonstration purposes, I've created template:LargeCategoryTOC, used at Category:Living people. Is there another approach that might work for this? -- Rick Block (talk) 20:32, 26 October 2006 (UTC)[reply]

Category:Living people doesn't need a navigation bar with over 700 items in it. The 26 letters of the alphabet, plus the 26 two-letter combinations beginning with the current letter are enough. Honestly, if your navigation has more than ten items, it's time to rethink what is really necessary. —Michael Z. 2006-10-27 09:11 Z

I don't understand how you're getting the 26 two-letter combinations beginning with the current letter. The previous TOC has a link to the beginning of the entries with a given first letter. The bold single letter section headers repeat this same letter. There was no way to get to, say, the page including Phil Mickelson other than get to the "M" page, click "next 200" way more than 20 times (20 gets me to the "May"s), or get to the "N" page, and click "previous 200" scores and scores of times, or figure out how the URL works and directly go to the page with the "Mi"s. There are someting like 70,000 entries in this category. CategoryTOC provides access to 26 spots in this (first entry with each different first letter), meaning on the average there are about 2600 entries per letter. I agree the 26 initial letters, plus the 26 two-letter combinations beginning with the current letter are enough. LargeCategoryTOC does exactly this. And it's 700 items. -- Rick Block (talk) 14:29, 27 October 2006 (UTC)[reply]

I don't understand what you mean by "I'm getting" them. I'm saying that the following bad navigation bar with 700 items is way too complicated:

see template:LargeCategoryTOC

Scrolling around in that to find the right two-letter combination of over 700 links is frustrating. Just showing the current letter's combinations in the second level is enough. If I want to get to Phil Mickelson, I first click on the M, and then click Mi in the 26 second-level links.

In this example I can see all of the links at once. The hierarchical relationship between the first and second row is clear (although perhaps the A can be formatted to be a bit more prominent, to reinforce it). Clicking twice to find the right spot in 700+ category pages is not unreasonable.

In the example, I've incorporated the prev/next navigation into the navbox too: still not sure if this is the best way. Perhaps those should be in a third line. Or perhaps there should be a third line above, containing a heading and back-and-forth nav: "← A →".  —Michael Z. 2006-10-27 17:21 Z

Categories display with the same "fixed" content (what you see when you edit the category page) regardless of where you are in traversing the category members. So, when you click "M" from the TOC on the first page, you go to a "new" page but you see the very same TOC you saw before. The "from=" URL parameter is (at least currently) not available to a template. I completely agree having a smaller, non-scrollable, TOC would be better - I just don't think it's (currently) possible. If this category were split into 26 categories, one for each first letter, we could do this, but with it being a single category there's only a single TOC that displays identically wherever you are in the category. -- Rick Block (talk) 18:43, 27 October 2006 (UTC)[reply]

I see; that's too bad. I guess a solution would have to tie in to whatever hooks at the Wikimedia software level create the prev/next 200 links.

I wonder if there's any way for the template to determine which letter in the template itself was clicked to reach the current page: perhaps by using a page anchor link? —Michael Z. 2006-10-27 19:56 Z

If Cat:LP is the problem child that leads to this "solution", why not try something similar to Special:Allpages, where the user can input their own "start string" to jump to Mick" or "Mi" or whatever they like. I don't think this kludge is the answer, as well-intentioned as it may be. Please address the problems of: a) getting a hardcopy of the information you've now hidden from printing, b) navigating the new navbox without using the mouse (in Firefox, you can use the right/left arrow keys when the main window pane overflows, but your navbox can't get the focus to do this). -- nae'blis 14:43, 27 October 2006 (UTC)[reply]

Cat:LP is perhaps the poster child for large unwieldy categories, but there are others (like category:Albums by artist). This solution is targeted at any place where there's a navigation issue of the form "lots and lots of links that 1) have a natural and obvious ordering and 2) most readers don't care to see or use".

a) Since it's navigation, not content, not printing it seems like the correct behavior. I view this as a feature, not a problem.

b) I think navigation within the scrollbar without a mouse is fundamentally a browser issue. I'm happy to try to address it, but ultimately I'm not sure I care. Again, it's navigation not content. If you can't use it with your browser without a mouse, I'm sympathetic but perhaps not very. Can you use an existing 100 element navigation box without a mouse, like the one at the bottom of Carl Lewis Veronica Campbell? -- Rick Block (talk) 18:43, 27 October 2006 (UTC)[reply]

Changed from Carl Lewis to Veronica Campbell since I've condensed Template:Footer Olympic Champions 4x100 m Men. -- Rick Block (talk) 16:20, 29 October 2006 (UTC)[reply]

This template is a usability horror, mostly due to the horizontal scrollbar, and the fact that you can only see a tiny fraction of the content of the navbox at once.

Therefore I will replace it with:

Maybe that has its issues too, but it's way more practical than any horizontal scrollbar solution can ever hope to be. Shinobu 13:47, 27 October 2006 (UTC)[reply]

What you're suggesting is already available as template:NavigationBox. This is meant to be an alternative. If the consensus is that we should not use horizontal scrollbars (ever, under any circumstances) this template should be deleted. -- Rick Block (talk) 14:36, 27 October 2006 (UTC)[reply]

Actually, you mean {{navigation}}. By the way, I did it using divs because I saw it first as divs, and only later on using templates. The template {{navigation}} is only a fairly thin wrapper for the div method. Its main function is to add v-d-e-links. Demo:

see Template talk:Navigation bar/examples#Navigation

I guess support for an image like in {{NavigationBox}} can be added, if it's decided we want images in navboxes. Shinobu 18:15, 30 October 2006 (UTC)[reply]

Previous to the creation of this template, Rick Block and I brainstormed on ways to create a workable TOC for very large categories. True, this would be much better handled by a software upgrade. But until a better solution is found, this is better than nothing. I also think that this template is often very well suited for displaying panoramic images. I don't think a blanket dismissal of its use will improve the project. Perhaps there should be guidelines for when this should be used, and when it shouldn't. --Samuel Wantman 04:23, 1 November 2006 (UTC)[reply]

I'll copy my comment from way above, as I too think this is a great solution in a number of circumstances.

I really like it for use with large categories. Great solution.

The only problem I can see with using it as a navbox replacement, is you can't immediately see the link you're already at, in bold within it. However it is infinitely preferable to using the "hide" function, which looks confusingly like an [edit] link.

Full support. :) --Quiddity 09:45, 28 October 2006 (UTC)[reply]

I tried an alternative for large categories, with subpages by first letter transcluding the main (category) page. Turns out this does not work. Without some sort of software change I don't think there is a reasonable alternative. I've worked up alternatives for the current non-category uses of this template. Instead of:

template:Places in Bedfordshire

the following could be used (no scrolling, but two clicks to get to a given place)

And, instead of

template:Footer Olympic Champions 4x400 m Men

something like the following could be used (no scrolling, but two clicks to get a given year or a given athlete)

These "substitutes" are less general and harder to code, but are nearly as functional. For very large categories I'm not aware of ANY alternative. The question here boils down to whether horizontal scrolling, within a navigation template, is so inherently evil that we should never ever use it. My opinion on this is that it is not THAT evil. I mean, it's not like something that only works with one specific browser. This works with ALL browsers, including screen readers like JAWS. Navigation without a mouse is apparently difficult (with at least some browsers and some OSs), but I suspect the alternatives are not very accessible given this constraint either. -- Rick Block (talk) 05:15, 1 November 2006 (UTC)[reply]

I agree.

As for the navboxes, I'm more in favour of just linking to the list page itself (eg List of places in Bedfordshire) in the "See also" section, and removing the navbox (or like {{Places in Bedfordshire/small}}). But I'd be completely happy to see the horizontal scroll used for these too. --Quiddity 05:58, 1 November 2006 (UTC)[reply]

The demo on pump works for me but I hate it; it's poor ergonomics.

Not only that, ha ha, I can't even load this talk page. It breaks my browser. John Reid ° 10:05, 13 November 2006 (UTC)[reply]

That above issue is now fixed; all the demos and weird mutant snakes have been moved to a subpage -- one I'll never stick my hand into. I always could see the demo on Pump; I'm not claiming a technical issue against the template. I just want it dead.

Now, I hate to bite the hand that didn't feed me to the snakes but yetchh. That's my professional opinion. As somebody points out above, the template only fails Nielsen on one out of five essential usability guidelines: it's a side-scroller. But boy, what a failure!

• "We know from user testing that users hate horizontal scrolling and always comment negatively when they encounter it."
• "...this abomination."
• "...one of the top-ten Web design mistakes of 2002."

Isn't that enough to sink a battleship? Wikipedia is here for one reason only: for our readers. This isn't SourceForge any more than it's YooToob. We aren't here to amuse ourselves. Everyone's worked hard and been very clever. Congratulations on making it work. Now go stick it on your user page. Okay?

This is not a solution to a problem but an encapsulation of a problem: Huge nav templates.

There is a burning desire that beats within some breasts to collect. It is, at root, a neurotic disorder; I sympathize because I share it. I want to have all the state quarters, all the stuffed animals, all the rare music, all the arcade tokens, all the obsolete computer gear, all the hand tools, all the eletronic components, all the stuff. And ideally, I'd like to live at San Simeon, where I'd have enough room to store it all and display it. But even Hearst could not build fast enough to house his collection.

I see the most jackassed stuff. There are templates for newspaper chains. These are constantly being obsoleted as individual papers and even whole chains are bought, merged, renamed, and sold. Most such templates are bigger than the stubby articles they emblazon. What's the point? A simple link from paper to chain is correct; a list of papers in the chain belongs on the chain's page (or list page) and nowhere else. It is not conceivable to me that a reader of the Akron Beacon Journal needs or wants a direct link on that page to the Hilton Head (S.C.) Island Packet. Yes, he should be able to get from one to the other -- and he can do so in two clicks, without the obtrusive Template:Knight Ridder Newspapers.

Burn the huge nav templates. If a nav template has more than about a dozen main entries, it's probably wrong for articlespace; more than about 24 and it's not useful to most editors, either. The only virtue of existing huge nav templates is that they look huge; we can take aim on the slow-moving beasts and drop them. I'm not fooled by a side-scrolling Yellow Pages, though.

• If a reader can get from A to B in two clicks and there's no clear expectation that many readers will want to make the same trip (the same specific trip, not the general case), then leave B out of any template put on A.

Hmm, good rule. I'll have to work on that. John Reid ° 06:46, 20 November 2006 (UTC)[reply]

Burn the huge nav templates? Fine with me, but are you going to nominate them for TfD? There are lots and lots and lots of them (I'd guess thousands if not tens of thousands). Even if they're nominated at the rate of 3 a day I'm not sure we could keep up with the creation rate. My thought is that it would be a whole lot easier to show folks how to create a template that isn't a monstrous eyesore than to go out and pick a thousand fights. And, do you have any solution for a very large category TOC? I agree horizontal scrolling a window is evil, especially when there is also vertical scrolling but that's not what we're doing here. My professional opinion is that a horizontal scrolling navigation bar is the most practical path to the most benefit. Let them have their huge nav templates - just don't let the nav templates overwhlem articles. If you don't want to use it, no one is forcing you to. If you don't even want to see them we could create a CSS style so you could make them invisible in your monobook.css. -- Rick Block (talk) 02:24, 21 November 2006 (UTC)[reply]

I'm not a fan of the huge nav templates but I find {{LargeCategoryTOC}} to be very useful and I love the scrolling panorama pictures. If nothing else, I'd like to keep it around for the panoramas. Hopefully, the software will be upgraded at some point to have intelligent TOCs, automatically generated. Until that happens, LargeCategoryTOC is the best tool we have. If we want to discourage the use of this template for use in nav templates, I wouldn't object. I'd strongly object to deleting it entirely. -- Samuel Wantman 20:36, 25 November 2006 (UTC)[reply]