Wikipedia talk:Manual of Style (footnotes)/Archive 4

References/Citations/Footnotes - and a 'while you're reading' idea

I really like this system, but the results at the bottom of a well-referenced (and noted) article is a mess. Intermixed references/citations/Footnotes are not really benificial for 'quick' comprehension that should be the web... there must be some way of splitting this same system between different uses. And if you applied the same code between <ref> <foot> and <cite> tags? (and why not <web> while you're at it?) This would be the simplest way of providing users of this system a few more sorting options - and a cleaner pagebottom!

Also, how about a little addition to end this very 'paper-esque' habit of having to flip to the page bottom to see references? Since the tag content is already extracted during the php page assembly, why not add an extra line to insert it into the 'reference number's "title" tag? (As in [<a href="#anchor_five" alt="whatever" title="reference_from_tag_here">4</a>]) - this would make the reference appear while the mouse is over the wee inline reference number that remains small enough to be unobtrusive. Thus, if the reader wants that reference desperately enough he can click down to it, but if not he can continue reading uninterrupted. The wee floating tag may be a bit small (long) to contain all reference text in some cases, but on the other hand I wouldn't suggest using the very obtrusive 'floating div' system. But to each his own tastes. Just my two cents.

--THEPROMENADER 07:25, 17 May 2006 (UTC)

I totally agree with the comments made here. I like the use of parenthetical citations with some sort of <ref> element within the parenthesis. Or perhaps the <ref> element could go in a spearate <cite> element. This way the details could go at the bottom of the page while the summary and page citation could be visible inline or with a mouseover. In any event, adding some structure to this markup would make things more flexible too. Users could control how they wanted citations and references displayed on their preference page.

Perhaps this is better discussion for the development pages, but does anyone know if there's some way to do this with namespaces or some other method using the existing <ref> element? In particular I'd like to create two separate references lists at the bottom of the page: one for footnotes elaborating on the text and one for references proper (I guess in the order that they first appear on the page). I imagine something like <references namespace="footnotes" /> followed by <references namespace="references" />. Any thoughts? --Cplot 21:10, 1 July 2006 (UTC)

While no consensus has been reached on any of the points mentioned in this section, they have been discussed. See the recent issues summary: (3) For differentiating citations from notes, and (5) for having more than one references list at the bottom of the page. Using the title attribute would be benificial, as well, and could be something added to my proposed /List of outstanding ref bugs and issues. --J. J. 14:01, 6 July 2006 (UTC)

If you want a little popup with the text of the reference or footnote, how about using the Navigation popups script? It is brilliant for exactly this purpose! --Slashme 09:43, 10 July 2006 (UTC)

No space between period and ref

Per the most recent edit, I agree that putting a space between a period (or other punctuation) and the following <ref> tag looks bad. At the least, we should not recommend doing so; I'd actually rather specifically advice against it. LotLE×talk 07:37, 17 May 2006 (UTC)

The problem is also that such a space makes the footnote wrap to the next line. To avoid that we would have to use &nbsp; instead of a simple space char. I don't think this is worth the trouble. See also my testpage at User:Ligulem/work/fn1. --Ligulem 09:05, 17 May 2006 (UTC)

It doesn't make it wrap to the next line for me. Perhaps it depends on the browser. If we didn't have the square brackets, it wouldn't look so bad without a space, but with the brackets, it's very crowded. SlimVirgin ^(talk) 09:12, 17 May 2006 (UTC)

It wraps to the next line in Mozilla Firefox 1.5 and Internet Explorer 6.0 (both running on Windows XP Pro, patched with all SP's and security fixes). Which browser does not wrap to next line? --Ligulem 09:33, 17 May 2006 (UTC)

I'm using Firefox on a Mac. SlimVirgin ^(talk) 09:47, 17 May 2006 (UTC)

The problem is that when editing, it looks very messy to have the whole thing running together on one line. This is yet another case of suitability for readers clashing with suitability of editors. :( Johnleemk | Talk 09:22, 17 May 2006 (UTC)

Agreed. It looks messy in edit window. That's really bad. HTML comments could help, but they are a pain to use also. Sigh. --Ligulem 09:33, 17 May 2006 (UTC)

I just noticed that we could insert a space between "<ref" and the closing ">" (example [1]). This would make the text flow to the next line in the edit window only. --Ligulem 09:42, 17 May 2006 (UTC)

It's generally hard to edit with these refs inserted, particularly when there are several in one paragraph. Is there any way of making the words inside the ref tags look different in edit mode from the rest of the text? SlimVirgin ^(talk) 09:45, 17 May 2006 (UTC)

Not to my knowledge. But that would be cool. With an external editor, that would not be too hard to program. Maybe there is even one out there that could be configured to do so. There are a lot of editors for programmers that can do syntax highlighting (html for example). The edit window is generally a real pain to use for larger text anyway (it also laks some important editor capabilites like a decent undo/redo). The inline refs are making that even worser. --Ligulem 09:57, 17 May 2006 (UTC)

I wonder how we could find out whether it's possible. I love these refs but they make copy editing really hard. I'm pasting an example below: if you look at it in edit mode, it's practically impossible to see where one sentence ends and the next begins. SlimVirgin ^(talk) 10:07, 17 May 2006 (UTC)

Some legal scholars have argued that the settlements are legal under international law, including prominent international law expert Julius Stone and Eugene Rostow, Dean of Yale Law School, Under Secretary for Political Affairs under President Lyndon B. Johnson, and one of the dafters of UN Security Council Resolution 242. According to Rostow "the Jewish right of settlement in the area is equivalent in every way to the right of the local population to live there."

In reply to SlimVirgin: I fully agree with you. Editing this kind of stuff is awful. The edit window is technically too limited to help out here (and my programmer knoweldge hints to me that this won't change anytime soon). The only thing that comes to mind is that we could add some spaces/newlines or html comment. For example this would be ok (at least wiki-technically, see wiki source in edit mode):

Some legal scholars have argued that the settlements are legal under international law, including prominent international law expert Julius Stone and Eugene Rostow, Dean of Yale Law School, Under Secretary for Political Affairs under President Lyndon B. Johnson, and one of the dafters of UN Security Council Resolution 242. According to Rostow "the Jewish right of settlement in the area is equivalent in every way to the right of the local population to live there."

Of course, reading this kind of wiki-source is still awful, but the referencing content is somewhat set separate. --Ligulem 11:16, 17 May 2006 (UTC)

That's a bit better, thank you. I'll give it a try. SlimVirgin ^(talk) 22:13, 17 May 2006 (UTC)

Edit window formatting and wrapping

I think a couple things might be unclear in the above discussion. If a space is left between a period and a <ref> tag, the note will not automatically wrap to the next line (in either edit window or as rendered). It's just that it has the opportunity to do so. Exactly when that happens depends on browser, font used, size of screen, size of browser window, etc. You can definitely force the effect by playing with the sizing of your browser window, in pretty much any browser (other than 'lynx', I suppose). But it's not something as simple as IE vs. Firefox.

For example, by selectively resizing my window, I got SlimVirgin's example to render like:

Some legal scholars have argued that the settlements are legal under international law, 
[13] including prominent international law expert [14] [15] Julius Stone [16] and Eugene 
Rostow, Dean of Yale Law School, Under Secretary for Political Affairs under...

Which is definitely not good (it was a bit smaller than I happen to size normally, but not anything extreme; I did, obviously nudge it until it did just the wrong thing I wanted to demonstrate... but this will happen with a certain frequency at any browser size and enough notes).

In any case, I find that a space between the period and the ref looks absolutely awful even when the note doesn't wrap to the next line... and even worse when it does, of course. So I'd really like the guideline to explicitly say not to do that. There are several approaches to improving readability in the edit window, while not leaving a space. Let me mention a couple:

This is main text.<ref >A footnote for the text.</ref>.  And some more text.

By putting a space in <ref >, the footnote may wrap to the next line in the edit window, but won't leave extra space as rendered. If a reference has a name, it achieves a similar effect:

This is main text.<ref name="foo137">A footnote for the text.</ref>.  And some more text.

This lets the edit window wrap before 'name' if it needs to. Adding XML comments also lets you force layout without affecting rendering:

This is main text.<!--
--><ref name="foo137">A footnote for the text.</ref>.  <!--
-->And some more text.

However, I find that the use of citation templates, apart from their basic advantages, also lets you be more presentable with the layout of inlined references. Here's an example I gave in a discussion on my talk page:

In the first few months after the attacks, most representatives from these professions 
who gave statements to media outlets lauded the "performance" of the Twin Towers, 
suggesting that loss of life could have been far worse if design and construction 
of the buildings had been of lesser quality. Radical design decisions made by the 
WTC team were compared to more time-tested [[skyscraper]] designs. A report entitled 
"World Trade Center Building Performance Study"<ref>{{cite book
 | editor=Therese McAllister
 | date=May
 | year=2002
 | title=World Trade Center Building Performance Study
 | publisher=[[Federal Emergency Management Agency|FEMA]]/[[American Society of Civil Engineers|ASCE]]
 | volume=report 403 
 | url=http://www.fema.gov/library/wtcstudy.shtm www.fema.gov
}}</ref> issued by the [[Federal Emergency Management Agency]] (FEMA) in May 2002, 
pronounced the WTC design fundamentally safe and attributed the collapse wholly to 
extraordinary factors beyond the control of the builders.

I think that gives pretty good visual separation of the note, albeit occupying a bit of space in the edit window. LotLE×talk 19:38, 17 May 2006 (UTC)

Nice description. Thanks LotLE. I just would like to add that (for those who like it, no intention to urge anybody to do so) you can add spaces and newlines between "{{" and "cite book" (the name of the template). To the wrapping thing: it can happen also in between multi-footnotes (← like these) if they are separated by normal spaces. If you want to avoid that, &nbsp; must be used instead of a plain space char. --Ligulem 21:37, 17 May 2006 (UTC)

Can someone explain the point of these citation templates? They seem to create extra work for no benefit that I've ever seen. I'm afraid I didn't get the point of Lulu's examples. The first one he said looked bad, for example, looked fine to me. I think we should do what most publishing houses do, regarding whether to leave a space before ref or not, but I'm looking through my bookshelves here, and it isn't obvious whether there's a space or not. If we could get rid of the square brackets, that would help a lot. SlimVirgin ^(talk) 21:43, 17 May 2006 (UTC)

On looking bad: I suppose one part is just an aethetic thing I can't convince you of. Looking at the "Some legal scholars..." quote above, I find "...law expert [14] [15] Julius Stone ..." really jarring to look at with the spaces scattered among the text and notes. But if you don't, you don't. However, look at note "[13]" in the same sample. In that, the note is on an entirely different line than the clause it supports, which is bad beyond mere aesthetics: it's directly disruptive of reading flow. LotLE×talk 22:21, 17 May 2006 (UTC)

There is no need to use those citation templates. Some like them, some not. They do have pros and cons. --Ligulem 21:58, 17 May 2006 (UTC)

It is a somewhat separate issue from footnotes, but I see two main advantages to citation templates:

1. Consistency of reference format across articles. We can change, site-wide, the style guide used for everything that uses, e.g., a "cite book" template. No doubt, any concrete change might raise disagreement; but think also of republishers of WP content. Presumably someone assembling all the WP articles on, say, Geology, into a printed book or CD version (think developing world) might want all the citations to match Chicago Manual of Style. With templates, that's doable, without them it's effectively undoable.
2. Future tool enhancements. This doesn't exist now, so it's somewhat speculative; but not improbable (i.e. I could program it later). If you wanted to know every article that refererenced "A book written by Charles Dickens", or a "A magazine article published in The New Yorker", you'd have a pretty darn hard time extracting that from unstructured textual references. However, some super-duper future tool can much more easily extract such information from the structured format of citation templates. Such a tool might be something like an enhanced "what links here", or it might be some sort of indexing or citation analysis tool (or something I haven't thought of).

In my mind, citation templates are a really good thing. If nothing else, they help remind editors of desired data fields in references. But YMMV. LotLE×talk 22:13, 17 May 2006 (UTC)

Multiple sources cited for one fact - stylistic question

This came up on the (predictably contentious) Hamas page. I may not be exactly right in my account of what happened there, but that's not really the issue. It was something along these lines: some editors wanted to state that Hamas was "best-known" outside Palestine for suicide bombings, and provided five citations supporting that - a reasonable procedure given the controversial and quantitative nature of the contention, which was somewhat diluted in later edits. If I recall correctly, other editors on the talk page remarked that this was "too many citations", and removed some or all of them.

Now clearly, having multiple footnote indicators coming together is unnecessary and ugly, but removing valid citations is not the way of dealing with it.

I would propose that this article recommend, where one fact or contention is being referenced, that all sources being cited for it be combined into one footnote. This is normal procedure in published texts, where I don't recall ever seeing something like this. (That is the result of typing: <ref>Smith, 2005, p. 5</ref><ref>Lammens, 1898, p. 75</ref><ref>Tishrin, 26 August 2004</ref> .) This alternative which results from typing this: <ref>Smith, 2005, p. 5; Lammens, 1898, p. 75; Tishrin, 26 August 2004</ref>, also looks a good deal better. Any comments? Palmiro | Talk 12:17, 17 May 2006 (UTC)

That's what I've always done for multiple citations, for what it's worth. A problem that comes up is the case where one or more of these is a backlink to a previous footnote (<ref name="Foo"/> rather than <ref>Foo, 57.</ref>), which obviously won't work as expected when combined. This is avoided in cases where article editors use a new footnote for each citation—common where reasonably thick books are being cited—but is going to be a problem in articles that have a significant number of backlinks (which tends to happen when heavily citing websites or newspaper articles). Kirill Lokshin 12:23, 17 May 2006 (UTC)

True, though of course this problem could be avoided were we to wean editors off websites and newspaper articles and predominantly on to reasonably thick books, and think of the other benefits ;) Palmiro | Talk 12:29, 17 May 2006 (UTC)

That'll never happen. Most people willing to use book sources already do, and convincing or persuading others who are more comfortable with citing their newspaper or reputable websites to dig up books to cite instead will be hard. (I'm sure there's something at WP:RAUL about this.) Johnleemk | Talk 17:07, 17 May 2006 (UTC)

I propose adding trhe following:

Where possible, combine all citations for one fact or contention into one footnote, rather than adding multiple footnotes to the same place in the text.

Any comments, ideas, or suggestions for better ways of phrasing/explaining this? Palmiro | Talk 12:40, 18 May 2006 (UTC)

Looks fine to me. Kirill Lokshin 12:56, 18 May 2006 (UTC)

In the footnote-dense world of the physical sciences, it is reasonably common to see multiple footnote numbers cited for a single fact. This is often more useful than condensing all those sources into a single footnote, because you may need to cite those sources singly elsewhere in the article. Shimmin 17:45, 2 July 2006 (UTC)

Help:How to use Cite.php references up for deletion at MfD

Just informing, the vote is going on at Wikipedia:Miscellany for deletion/Help:How to use Cite.php references. See also above at #Proposed "How to use Cite.php" section for beginners --Francis Schonken 19:20, 17 May 2006 (UTC)

Comments inside ref tags

I found recently that addition of <!-- and --> delimited comments inside the <ref> and </ref> tags results in an odd error.

User:Ceyockey (talk to me) 16:09, 19 May 2006 (UTC)

Example outside

• gobble
• gobbledy

Example inside

• gobbledygook

References section

I've known about this problem for quite some time now, so I think the developers should know about it as well. You could always check the known issues section of m:Cite... Johnleemk | Talk 14:28, 19 May 2006 (UTC)

Size: some questions by Encephalon

(questions copied from user talk:Francis Schonken:)

• Is there a new template for smaller sizes?

  Yes, see Wikipedia:Footnotes#Helping editors unfamiliar with this system of footnotes. Whether the template is "new" is a relative question, note that it survived a TfD vote a month ago (Wikipedia:Templates for deletion/Log/2006 April 18#Template:FootnotesSmall), but it is "new" in the sense that it has now an updated technology.

• Is it a temporary fix?

  Well, erm, I fixed it, whether that solution is carved in stone or written in sand is too philosophical a question for me. Anyway the resizing technology is now driven to a single template ({{FootnotesSmall}}), though that template is also used by the usual template ({{Footnotes}}, that is the template that adds the help-line when used subst:). PS: currently some 150 articles use the {{FootnotesSmall}} template.

• Have the smaller sizes been standardized?

  {{FootnotesSmall}} "standardizes" in that it allows only 100% and 92%.

  There is a css trick with which resizing to 90% is possible [2] - don't know how much it is used, anyway, currently that ".references-small" css feature is not used by the {{FootnotesSmall}} template. The reason is: anything smaller than 92% looks bad on my monitor. So I suppose there are other monitors, and other people who would have trouble reading such smaller stuff. Note that there is still some continuing "size" discussion at MediaWiki talk:Common.css, even a poll going on right now MediaWiki talk:Common.css#Straw poll for footnote font sizes - which I wouldn't have known about if you hadn't asked me these questions. Anyway any technology can be implemented in {{FootnotesSmall}}, but I object to anything smaller than 92% for the mentioned usability reasons. --Francis Schonken 07:25, 20 May 2006 (UTC)

• Some editors have taken to throwing a <div class='references-small'>...</div> around the reference tag. Is this a 1.) suitable solution, 2.) recommended solution, and 3.) can </references> just default to that class somehow? —Rob (talk) 18:50, 23 May 2006 (UTC)

Period/Full Stop and reference location

See also below #Every sentence should have a reference?

Why does the reference have to come after the period? I think this is a foolish way of doing things. In comparison, the Harvard referencing system and other style guides allow for the reference to come within the sentence. I'm not saying it should be changed the other way, but the policy should allow for both. --Robdurbar 08:28, 20 May 2006 (UTC)

I think it was meant that if the footnote refers to the whole sentence, then it should be placed after the period, not before it. Of course, you can place footnotes in the middle of the sentence. Why not? This is a needed feature. --Ligulem 09:18, 20 May 2006 (UTC)

Hi Rob, the footnote should be placed directly after the point it's being used as a reference for. For quotations, it should be directly after the quote, even if the quote is in the middle of a sentence. If you have three points in a sentence, and only the first point is sourced to that particular reference, the footnote should come after that point, but if the footnote is for the whole sentence, it should be placed at the end, after the punctuation. Footnotes come after punctuation, unlike Harvard references which are usually placed before punctuation (like this: Smith 2006). SlimVirgin ^(talk)

User:Ligulem; thank you, but I wasn't very clear - I meant where it refers to the whole sentence, I understand the mid-sentence feature. So is this footnote after the period a standard and not just a Wikipedia thing? I always use Harvard when referencing outside of Wikipedia and it seems odd to do it out of the sentence. Could the policy be changed to suggest that they come after the period, rather than to prohibit their use before the period? It seems like an unecessary little rule. --Robdurbar 11:34, 20 May 2006 (UTC)

I believe all publishers place footnotes after punctuation, and Harvard referencing before it. SlimVirgin ^(talk) 11:39, 20 May 2006 (UTC)

My personal opinion is that it is good as it is stated (footnotes after the punctuation). No clue how and where that came from. The Chicago Manual of Style (which uses non-bracketed numbers), 14th ed. 1993, writes in clause 15.8 (p. 494) "Note reference numbers. The superior numerals used for note reference numbers in the text should follow any punctuation marks except the dash, which they precede. The numbers should also be placed outside closing parantheses." So as we do it here is in line with that. --Ligulem 12:02, 20 May 2006 (UTC)

It probably just came from the fact that it looks odd to have a sentence, then a number, then a dot, like this . I suppose the syntactical argument is that the number is not part of the sentence. SlimVirgin ^(talk) 12:15, 20 May 2006 (UTC)

I think I've presented the reason even on this talk page: basically, books and magazines are published using high resolution typography that doesn't exist in web browsers. Microspacing, ligatures, 1200 dpi printing, etc. make for different optima than do 72 dpi screen displays with relatively crude font rendering. Given the latter way that a majority of readers will read WP articles, certain things are good compromises for WP that wouldn't be necessary in print. On the web, lacking good quality superscripts and microspacing, it definitely looks a lot better to put footnotes after periods, or anything in superscript (which is why I personally dont' really like the template:ref_harv modification of template:ref_harvard all that much). LotLE×talk 16:52, 20 May 2006 (UTC)

The links don't work, Lulu. My mind is boggling at templates for Harvard references involving periods/full stops. SlimVirgin ^(talk) 17:03, 20 May 2006 (UTC)

Sorry, I abbreviated when I wasn't supposed to. LotLE×talk 18:49, 20 May 2006 (UTC)

A sentence should be written like this.[2] It should not be written like this [2].

If I see a footnote after a full stop, I assume it is a footnote for all the sentences since the start of the paragraph. If I see it before the full stop I assume it is for that sentence. I do not see why this Wikipedia guideline has to be perscriptive on this. --Philip Baird Shearer 16:10, 7 June 2006 (UTC)

{The prescription was added at 05:38, 17 May 2006 by user:SlimVirgin. Before that AFAICT there was no suggestion that the citation ought to go after the stop, just examples where it did. --Philip Baird Shearer 11:11, 16 June 2006 (UTC) }

Gladly the tool allows you to violate that authoritarian guideline! (see below under #Solved). There are also cases where you have 2 or more notes within one sentence, that apply only to certain terms of the sentence. As an example see Macedonia (terminology) and... ask those who support this guideline to figure out how you could do it otherwise!  NikoSilver  ^{(T) @ (C)} 16:20, 7 June 2006 (UTC)

I also belong to the camp that believes the footnote should be within the sentence (i.e., before the closing punctuation). Likewise for quotes. I suspect this is a programmer's professional deformation: periods close sentences, and the footnote is within the sentence. Similarly, parentheses, quotes, and so forth need to be paired correctly.

On the other hand, the usage as expounded within these guidelines has been around for a long time. Who said natural language was logical or consistent?

Urhixidur 12:06, 13 June 2006 (UTC)

I have been moving footnotes to after the punctuation on a lot of articles per the discussion here and what is on the project page. Now Philip Baird Shearer requests me to stop doing this (see his and my talk). The question is: is it forbidden to bring an article at least in a consistent state per this issue? Philip Baird Shearer reverted me twice: [3], [4]. Both articles where he reverted me are again in an inconsistent state: some footnotes are before the periods, and some are after. Can't we at least agree to have a consistent style on a single article? So I could leave then those articles where the majority of the footnotes are before the punctuation as they are. Or is it possible to define areas where articles are consistently in opposition to what's specified on this guideline here? --Ligulem 19:39, 13 June 2006 (UTC)

I am finding User:Philip Baird Shearer's actions disappointing here. Not only did he revert without giving any explanation whatsoever, but when asked to provide a justification, gave this discussion, which at best proves that there is discussion on the topic, but certainly not that there is consensus against the current statement in favor of punctuation. Circeus 20:15, 13 June 2006 (UTC)

This has been debated before; there are many who support this standard who do not frequent this page. On FAC, it is commonly mentioned. See Wikipedia_talk:Footnotes/archive3#citation_location. --Spangineer^[es] (háblame) 21:08, 13 June 2006 (UTC)

Circeus, I did not say there was a consensus against it I said three was no consensus for it. That this discussion is taking place is an indicator of that. As I said above I do not see why this guideline has to be prescriptive on this, and it is not I who is going around articles I do not usually edit changing the footnotes to one style or another.

Ligulem they are not in an inconsistent state, please read what I wrote above. I am sorry if I did not make myself clear, but I think the positioning of the "note", before or after the full stop indicates if the reference is specific to the information in the sentence or to all the sentences before the reference. So the articles are not in an inconsistent state.

I have let a number of other similar edits of pages I have contributed to by Ligulem pass before I raised the issue here, having raised it here and had two other editors contribute to the conversation I think it is reasonable that Ligulem, does not make these changes to pages (s)he does not usually edit while we are discussing the issue. --Philip Baird Shearer 08:11, 14 June 2006 (UTC)

Spangineer I read the link you gave and I do no think the discussion was about this issue. At first I could not think of a way to show you that many editors do not agree with this rule. But the number of pages which Ligulem has changed in the past few days is a good indicator [5] [6] .That is 654 pages changed between 18:17, 3 June 2006 and 15:50, 13 June 2006, ranging over a large area of interest e.g. Software patent debate, Regional Assistance Mission to the Solomon Islands, Bow Wow Wow and Argentina and England football rivalry. --Philip Baird Shearer 08:44, 14 June 2006 (UTC)

I agree to stop doing these kind of edits. I agree that there is obviously no consensus to have it as it is stated on the project page. Although I find the articles look lousy with footnotes before punctuation and there is also no semantical benefit to have it in one sentence after the period and in the other before the period, all in the same article. But well, if there is consensus to have Wikipedia look lousy, well then. If the footnote is at the end of the sentence it always refers to that sentence, not the paragraph (or at least it cannot be distinguished). BTW I'm a male. And to Philip: you could have talked to me a bit earlier about this. And once again: nearly all English books use the format as described by the Chicago Manual of Style (foonotes after punctuation except for the dash), which is compliant with what's currently in the guideline here. --Ligulem 09:03, 14 June 2006 (UTC)

I've changed the project page to reflect the current state of non-consensus. --Ligulem 09:28, 14 June 2006 (UTC)

Thanks for seeing the other POV. I could have mentioned it to you earlier, but I did not know that you were altering so many pages, (because I only see those edits to pages I watch). Once I realised that you were making systematic changes to lots of pages I raised the issue on this talk page. I assumed as you had already added comments to this section and were quoting this guideline as a source for the changes that you were watching this article and talk page. One lives and learns. Next time this sort of thing occurs I shall leave a notice on the editors talk page as soon as I make such an observatiuon on the guidline talk page. I apologise for my presumption that lead to me not informing you earlier. --Philip Baird Shearer 12:19, 14 June 2006 (UTC)

One thing left: what is the range of articles I edited an indicator for, if not at best, that a lot of wikipedians obviously agree with what I did or at least accept it as I did? All feedback I had when doing this was positive/agreeing. One wikipedian explicitly stated that — despite his personal preferences dictate otherwise — he agrees to have it consistent on wikipedia and that refs after puncts are the majority. I also didn't have any reverts, or complaints on my talk, besides yours. And please note that there is no policy on wikipedia that requires editors to significantly contribute to an article in order to be allowed to edit that article. It rather smells like some WP:OWN problem to me (No offense intended. I fully understand that. I would probably have that myself if I had written large good articles). Also due to the fact that the diffs look horribly sparse if I only move punctuation. Nevertheless these are complicated edits I have done. And it is normal that I have to try doing this in some amount or articles to see what happens. I have 12'000 articles on my list (created from an xml-dump), which I not yet did. And which I am not going to do. And a last one (I said that other occasions already): Wikipedians do copy the style of good articles. That's why I try to change good articles from good editors, also on — for some editors — minor things like fontsizes or footnote styles. Those styles get copied by less seasoned editors. --Ligulem 12:37, 14 June 2006 (UTC)

It's clear that there are many people who put citations before punctuation, but I'd argue that that is often an ease of use and/or inexperience issue, not an indication that it is the "right way" to do it. I am not here referring to those here who have reasoned arguments for putting citations first; I'm addressing the argument of "everyone is doing it" as support for putting citations first. First off, the 654 thing doesn't really help--I could probably find scores of words that are misspelled on Wikipedia more frequently than that. Furthermore, in my area, hardly anyone says "I should have gone to the store". They all say "I should have went to the store." Does that make it OK? Should I write like that in a paper for school? There's a difference between what people typically do without thinking and what is "correct". Obviously here, the dichotomy isn't as strong between right and wrong, but it's extremely well accepted in virtually all publications using footnotes that footnotes follow punctuation. That's just the way it is—the Chicago Manual of Style, one of the only style books still advocating superscript footnotes, says to do it that way, and professional writers agree. Even here, it's clear that few FACs pass through the process without the style getting changed to have citations follow punctuation. The editors consistently writing the highest quality articles (as judged by participation in WP:FAC) tend to agree that superscript citations follow punctuation. Ligulem isn't the only one converting the things. It's well established. As for Ligulem's point of consistency within articles, that is essential. Even as bad as it looks to put citations first, it's worse to have them haphazardly placed. And, as far as I'm concerned, he who is willing to put the effort into making the citation location consistent gets to decide which way they go, if indeed there is no consensus on a guideline (which I still highly doubt). --Spangineer^[es] (háblame) 13:50, 14 June 2006 (UTC)

There are a number of people who have contributed to this talk page who hold the same views on this as I do. I am not implying that every citation on every page which Ligulem altered was intentionally structured that way and that the count should be one editor for every page. Just that some (perhapse many) of them are structured that way intentionally, and clearly there is no consensus for prescribing citations after the stop. If someone is concerned that the locations are inconsistent within an article then I would encourage that person to use the talk page to discuss how consistancy can be reached without the bludgen of a prescriptive guideline. I don't thing the ease of use argument is correct, because I find it more difficult to place the ref before stop than after it. Also please see the section #Every sentence should have a reference? below, having a mixture of some citations before the stop and others after it is not necessarily inconsistent. --Philip Baird Shearer 15:40, 14 June 2006 (UTC)

Footnotes in double columns

To my mind, footnotes in double columns looks better to the eye, and make better use of space. I've taken the liberty of demonstrating this in the double layers article, though you require Firefox. Comments? --Iantresman 19:06, 23 May 2006 (UTC)

Looks pretty good with the browser window expanded, but the columns are just a bit narrow in my normal browsing width of about 850px. Font-size:95% doesn't seem to perceptibly reduce the size in my browser; I think it would look better if it was still smaller. It may also benefit from tighter leading overall, or a bit of line space between list items.

Is it possible for it to switch to one column when the parent element is narrower than a certain width, without resorting to Javascript? —Michael Z. 2006-05-23 19:26 Z

Too bad the citations start with the floating carat character and the floating a and b's. The left margin of the list would be emphasized much better if each simply started with the capital letter at the beginning of the entry. —Michael Z. 2006-05-23 23:47 Z

Have a look at Double layers/temp, for an example with the back-links moved to the end of the line, and an added 1/2-line of space below each entry. It's much easier to read. —Michael Z. 2006-05-24 00:16 Z

Pretty neat hack. It may be too complicated for most editors, even though it looks like all you've done is add <div style="font-size:87.5%; -moz-column-count:2; column-count:2;">. I'm sorry that your previous comment about back-links (more commonly link-back) was ignored. Maybe people will respond with reasons why the way it is.

FYI, "Notes" instead of "Footnotes" is a more standard heading. I try not to change existing ones, though, as it breaks links.

For your temp example, I'm assuming you've manually edited the layout to move the back-links and add extra space. Could you document the process on the /temp talk page? It may be good to centralize the discussion there. You may want to check these links for help, as well. --J. J. 13:51, 26 May 2006 (UTC)

• Wikipedia:WikiProject Usability/Areas#Design
• Wikipedia:WikiProject User scripts#Requests (see Lincher's column request)

References at bottom

Is there any success in solving this problem yet? I believe LotLE had a patch, did it work? - FrancisTyers 10:12, 26 May 2006 (UTC)

Which problem are you referring to? Section ordering? If so, see Wikipedia:Guide to layout#Standard_appendices. --J. J. 13:02, 26 May 2006 (UTC)

I mean, e.g.

<ref name="foo">

==References==
<references>
<ref name="foo">this is a reference</ref>
</references>

Basically having the spellout of the reference at the bottom of the page. - FrancisTyers 13:08, 26 May 2006 (UTC)

See various discussion in Wikipedia talk:Footnotes/archive3 and m:Talk:Cite/Cite.php. LotLE patch here. - FrancisTyers 13:12, 26 May 2006 (UTC)

Another option would be to separate footnotes from references/bibliographic entries. Having footnotes inlined is not such a bad thing, but having references inlined definately is. - FrancisTyers 23:53, 29 May 2006 (UTC)

Have you tried:

foo foo foo<ref name="foo" />
goo goo goo<ref name="goo">this is a goo reference</ref>

== References ==
<ref name="foo">this is a foo reference</ref>
<references />

Would that work for you? The only disadvantage is that it repeats the indexing number right above the ref-texts.  NikoSilver  ^{(T) @ (C)} 22:47, 29 May 2006 (UTC)

You could try HTML comments, though that would be completely manual and for organizational convenience only. Armedblowfish 02:35, 30 May 2006 (UTC)

Another option would be using templates, but I can't see many people going for that :/ - FrancisTyers 08:11, 30 May 2006 (UTC)

Every sentence should have a reference?

After reading in the guideline that Citations should always follow punctuation and the thread above (Period/Full Stop and reference location) it seems that a reference can't be attached to a paragraph (or series of paragraphs).

Does this mean that since (apart from the introduction) every sentence I write is sourced from a reference, every sentence needs a reference?

An example of my problem (abridged from Seikan Tunnel):

The area is folded into a nearly vertical anticline. It consists of volcanic rock^[8].

The first sentence is unsourced (I will find a reference soon I hope). The reference shown, [8], only applies to the second sentence. I now fear that someone will move the [8] after the full stop and it will look like it applied to the first sentence too.

The solution seems to be that every sentence gets a reference, I don't mind doing that but haven't seen it done before. What to do?--Commander Keane 06:36, 28 May 2006 (UTC)

A footnote comes always after the punctuation. This is just a matter of style. The position of the footnote after the period does not imply that it refers to the first sentence too. The idea that every sentence should have a footnote is very bad. Don't do this. --Ligulem 09:28, 28 May 2006 (UTC)

So really we have no way to tell where the footnote applies, we just guess? --Commander Keane 10:56, 28 May 2006 (UTC)

Well, if you doubted the statement, you'd be checking the cited source itself, wouldn't you? It should be fairly easy to determine what it applies to then. Kirill Lokshin 13:21, 28 May 2006 (UTC)

The citation needed template should ONLY be used for when you need to add a reference there; not to indicate that another reference doesn't belong to that sentence. That's nonsense. —Michiel Sikma, 05:52, 31 May 2006 (UTC)

If the first sentence is actually unsourced—rather than merely not being footnoted—you can just add {{citation needed}} to it, and that will make clear that the later footnote doesn't apply to it. Kirill Lokshin 13:21, 28 May 2006 (UTC)

Yeah, I can see my problem. I was really aiming for 1:1 brilliant sourcing, but wasn't actually providing a decent level of accuracy (eg, Kitamura, p7, paragraph 18) in the references themselves so the whole thing is pointless. I will settle for the current way (seen in featured articles etc). Thanks for your thoughts.--Commander Keane 13:42, 28 May 2006 (UTC)

It is rather ambiguous, but when I write articles, I include citations such that everything between citations one and two comes from the reference marked by citation two. That rule, of course, doesn't work when citations are added to an article after it was written. --Spangineer^[es] (háblame) 14:48, 28 May 2006 (UTC)

When I write, if its within the sentence then it applied only to that sentence, if it is outside, it applies to the whole block. - FrancisTyers · 12:36, 14 June 2006 (UTC)

This is the way I do it as well. Taking the example "The area is folded into..." above, while there were sentences before [8], not referenced by [8], I would rather keep to the format used by Commander Keane, then having found a second reference I would move both citation to the end as it saves cluttering up the paragraph with citations on lots of sentences:

The area is folded into a nearly vertical anticline. It consists of volcanic rock.^[9][8]

Using this method means at a glance one can see which sentences are referenced and which ones are not, and using the presumption of goodwill, it means one can see at a glance which parts of a paragraph are covered by citations and which ones are not, without having to read the document cited which may or may not be easily accessible. In a collaborative effort like this, it frequently happens that an editor adds a paragraph which is sourced and then someone else comes along and adds more sentences which are not covered by the citation. I think it is important that it is easy for a reader of the article to be able to judge what is covered by citations and what is not. I do not think that putting {{citation needed}} is desirable unless one is disputing the verifiability of the sentence which in the example Commander Keane gave is probably not true. --Philip Baird Shearer 15:57, 14 June 2006 (UTC)

Removing the message "This template is deprecated." msg from {{Ref}} and {{Note}}

I suggest the message "This template is deprecated." be removed from {{Ref}} and {{Note}} since Cite.php is still only a beta version with bugs and limited functionality and can not replace the templates yet. (see m:Talk:Cite/Cite.php#Grouping the references + m:Talk:Cite/Cite.php#Allowing sub-references and Requests for references at the bottom). Thank you. --V111P 20:51, 2 June 2006 (UTC)

Please, please... if any admin sees this, please remove the vandalism by User:Cyde who has repeatedly abused administrative powers to push a personal agenda on references... including most recently this damage to the templates. If ever there was an admin in desperate need of desysop'ing, this is the one. LotLE×talk 21:25, 2 June 2006 (UTC)

I've replaced it with a more accurate message. LotLE, please try and stop being so dramatic :P - FrancisTyers · 14:16, 7 June 2006 (UTC)

I'm not sure of the exact wording that you added in replacement, but the passage that relates to this isn't quite accurate I think: "Before Cite.php footnotes were available, the same functionality was achieved with a system called Wikipedia:Footnote3..." I don't think there are many people here who would say that the two systems deliver the same functionality. Perhaps the passage could be revised to read something like "Before Cite.php footnotes were available, similar functionality was achieved with a system called Wikipedia:Footnote3..."? User:Ceyockey (talk to me) 12:44, 11 June 2006 (UTC)

Superscripting

I would like to know if I could turn off superscripting the reference number and tag. Superscripting messes with intra-line spacings and makes paragraphs look ugly. If this option is not available, I would like to make a feature request. Thanks. Chaipau 15:37, 6 June 2006 (UTC)

Issues

Solved

• Refs not restricted at end of sentences. You can force users to place the <ref>text</ref> after the period if you wish, but don't make the tool do it on its own.(discussed above)
  • Does this have anything to do with actual technical features? As far as I know, this is entirely an issue of style; cite.php will work just fine anywhere in the sentence. Kirill Lokshin 13:25, 7 June 2006 (UTC)
  • You are right. It works already, I was confused. Apparently it was fixed, or maybe I am mistaken. Consider it stricken.  NikoSilver  ^{(T) @ (C)} 13:56, 7 June 2006 (UTC)

Rejected

• Separate footnotes from citations. In LaTeX/BibTeX this is done by differentiating between \footnote{This is a footnote.} and \cite{Smith2006} which is a citation that can be found in the citation list.
  • See my comments to the first point. In many cases, the reason why footnotes are chosen over Harvard-style citation is precisely to allow the mixing of raw citation information with commentary. Kirill Lokshin 14:39, 7 June 2006 (UTC)
    • Ok, fine, thats not the biggy anyway. :) - FrancisTyers · 14:59, 7 June 2006 (UTC)
• Be able to choose whether to have harvard style references, or footnotes style.
  • Automated conversion between footnotes and Harvard-style references is, in many cases, impossible, for a number of reasons (most obviously because a single footnote may contain multiple author/page sets, which would then need to get parsed into multiple Harvard refs). Kirill Lokshin 14:39, 7 June 2006 (UTC)
  • Why would that be impossible? I think it can be done automatically.  NikoSilver  ^{(T) @ (C)} 15:15, 7 June 2006 (UTC)
    • See my example for the separate note/ref sections discussion. How would you convert that to a Harvard-style reference? Kirill Lokshin 15:18, 7 June 2006 (UTC)
    • Right. Rejected.  NikoSilver  ^{(T) @ (C)} 15:41, 7 June 2006 (UTC)

Outstanding

We need: (Empty list) :-))

Additional comments

If anybody wants to add anything here, feel free to add to the list above. Developments or further comments on these can be posted below each respective comment.  NikoSilver  ^{(T) @ (C)} 13:06, 7 June 2006 (UTC)

Added a few comments inline. Kirill Lokshin 13:25, 7 June 2006 (UTC)

...more below each... NikoSilver  ^{(T) @ (C)} 13:56, 7 June 2006 (UTC)

...and more! Kirill Lokshin 14:39, 7 June 2006 (UTC)

This sounds like an ideal system to me! InvictaHOG 14:28, 7 June 2006 (UTC)

Proposals to developers after consensus

The "consensus" here is apparently between four users only. Namely: User:FrancisTyers, User:Kirill Lokshin, User:InvictaHOG and me. If anyone disagrees, please comment. I suggest we leave this note here for a week, until 14/6/2006 15:00UTC.

Summary of proposals

(For details/explanation on each of the proposals, check the respective proposal numbers in the "Extensive discussion" section right below)

1. Option to separate the bibliographic references from the editor's notes or other notes.
   1. By two different functions: <ref> and e.g. <note> and by creating two sections in the bottom of the article (e.g. under ==Bibliography== to have <references/> and under ==Notes== to have <notes/>).
   2. And/or by different numbering systems (eg Latin or Greek, or letters)
   3. And/or different formats (eg ⁽¹⁾ rather than ^[1] or ^b[1] rather than ^n[1]).
   4. And/or, even better, the same ref function could be modified to include the option for multiple sections (and/or numbering systems). (e.g. <refX="noteref">note text</ref> could produce ^X[1] and the result under <referencesX/>, allowing multiple options for X).

2. Turn off tag superscripting at will (only on a global site-wide basis)

3. XML comments within refs

4. Notes/refs that have separate paragraphs without using <br>, but by just pressing [enter]. Also, some formatting option to help keeping the text flow while in the edit window: The [enter]s right after the <ref> tag and right before the </ref> tag should not print. See #Example format below.

5. Use of refs within templates.

6. Be able to separate the bibliographic information from the article body by:
   1. Having a hidden section where all bibliographic notes can be placed together and
   2. Canceling the truncation of the text contained in the second (or following) instance of the same ref name provided the first one(s) is(are) empty, which will also facilitate article re-organisation, paragraph moving etc.

Extensive discussion

(Numbers here match the points in summary above)

1. Two different functions: <ref> and e.g. <note> to separate the bibliographic references from the editor's notes by creating two sections in the bottom of the article (e.g. under ==Bibliography== to have <references/> and under ==Notes== to have <notes/>).

• • Depending on how this is implemented, it could cause problems for printing articles, which would remove the links and make it impossible to determine which section a footnote number was pointing to. Kirill Lokshin 13:25, 7 June 2006 (UTC)
  • Most articles already have a separate bibliographic from note section (manually) anyway. For the numbering, I would suggest different numbering systems (eg Latin or Greek, or letters), or different formats (eg ⁽¹⁾ rather than ^[1] or ^b[1] rather than ^n[1]). Even better, the same ref function could be modified to include the option for multiple sections (and/or numbering systems). (e.g. <refX="noteref">note text</ref> could produce ^X[1] and the result under <referencesX/>, allowing multiple options for X). In any case, it is not logical to mix bibliographic refs with notes, creating a reference salad [sic], so let the editors decide how they will solve this per case. NikoSilver  ^{(T) @ (C)} 13:56, 7 June 2006 (UTC)
    • Actually, it's sometimes more confusing to try and separate them, since a footnote may contain both citation details and further commentary, often regarding the source itself, or disagreements between various sources (e.g. "1000 to 2000^[1] → ^[1] Smith, 500; Jones, 302. Jones cites the lower number, while Smith argues that the higher is more likely; see also Abrams, 409."). I suppose it could be useful in some cases, but I think that any system we adopt should allow both separate and combined citation/discursive footnote sections. Kirill Lokshin 14:39, 7 June 2006 (UTC)
    • I agree. Sometimes putting them together is better. We shouldn't exclude the option of splitting them though. What do you think about refX, refY, refZ proposal above? NikoSilver  ^{(T) @ (C)} 14:47, 7 June 2006 (UTC)
    • That could work quite nicely; I'm not sure how difficult it would be to implement, though. (I suspect changes to this code are not exactly at the top of the developers' to-do lists.) Kirill Lokshin 14:55, 7 June 2006 (UTC)
    • Result: Request for the same ref function if it could be modified to include the option for multiple sections (and/or numbering systems). (e.g. <refX="noteref">note text</ref> could produce ^X[1] and the result under <referencesX/>, allowing multiple options for X)

• Disagree I have never (or at least so rarely that I don't recall) seen a scholarly book or journal that creates two classes of footnotes. Even Science and Nature intermingle bibliographic citations with other textual matters. Wikipedia should follow convention, not try and create new conventions.Thatcher131 04:15, 9 June 2006 (UTC)
• Disagree Having too separate numbering schemes is just too confusing, both for editors and readers.--Srleffler 15:59, 10 June 2006 (UTC)
• Agree. See Mediazilla:6271, and take a look at (for instance) Comparison of operating systems if you think multiple sets of footnotes are worthless. —Simetrical (talk • contribs) 04:42, 11 June 2006 (UTC)

2. Ability to turn off tag superscripting at will (discussed above)

• It would be nice to have the option of using parentheses or square brackets in normal text instead of being restricted to superscripts.Thatcher131 04:20, 9 June 2006 (UTC)
• This can be done by a simple edit to the PHP file. —Simetrical (talk • contribs) 04:42, 11 June 2006 (UTC)

3. Inclusion of xml comments within refs (discussed above)

• This was a bug, that's been fixed. If the fix isn't live yet, it will be live soon. —Simetrical (talk • contribs) 04:42, 11 June 2006 (UTC)

4. Ability to have notes/refs that have separate paragraphs without using <br>, but by just pressing [enter]. (eg. see ref#5 in Macedonia (terminology) where even <br>: is used for indenting the second par!) Also, some formatting option to help the text flow while in the edit window: The [enter]s right after the <ref> tag and right before the </ref> tag should not print. See #Example format below.

• Disagree/unsure I think for purposes of an encyclopedia, if a footnote is long enough that it needs paragraph breaks, the material should probably find its way into the main article. Thatcher131 04:20, 9 June 2006 (UTC)
• Disagree. <br> is not such a big deal. I strongly disagree with the suggestion that newlines before and after the reference tags should not print. That would break a lot of pages, where people have put a reference at the end of a paragraph.--Srleffler 15:59, 10 June 2006 (UTC)
• Disagree. There's a reason MediaWiki typically ignores single line breaks; if you disagree with that, change it for everything, not just refs. —Simetrical (talk • contribs) 04:42, 11 June 2006 (UTC)

5. Ability to use notes/refs within templates, and the actual note/ref to appear in the article that includes the said template. (eg. see Template:Geographical Macedonia within article Macedonia (terminology) -where "Note a" is forced within the template, rather than in the ref section in the article. Relevant test failed (see history of both articles).) Useful diffs: template article

• This would be excellent. Sadly, I suspect it would require a total rewrite of the code. —Simetrical (talk • contribs) 04:42, 11 June 2006 (UTC)

6. Be able to separate the bibliographic information from the article body (discussed here).

• • Summary:

foo foo foo<ref name="foo" />
goo goo goo<ref name="goo">this is a goo reference</ref>
== Wish this was a hidden section ==
<ref name="foo">this is a foo reference</ref>
== References ==
<references />

• • This workaround can separate some refs to the specific section. The only disadvantage is that it repeats the indexing number right above the ref-texts. Can we format that number to dissappear, or otherwise blank it?  NikoSilver  ^{(T) @ (C)} 15:15, 7 June 2006 (UTC)
  • Only this one is left, care to comment? Accept as it is or are there any better thoughts?  NikoSilver  ^{(T) @ (C)} 15:43, 7 June 2006 (UTC)

I don't think that will work because the full text ref has to come before the subsequent id ref. - FrancisTyers · 15:55, 7 June 2006 (UTC)

I just tested it and it works fine :-). See here, before I reverted. Even if it didn't work we could ask that it should work (that's the point of this discussion). NikoSilver  ^{(T) @ (C)} 16:11, 7 June 2006 (UTC)

I just tested it and it didn't work :( Please see here and tell me what I am doing wrong. If this worked, I would be happy. - FrancisTyers · 16:43, 7 June 2006 (UTC)

You are right. :-( It says so too in the actual project page that all subsequent texts in the next instances are removed. We can ask for it not to be so, though. :-) Ok? NikoSilver  ^{(T) @ (C)} 16:53, 7 June 2006 (UTC)

That would be great :) - FrancisTyers · 17:03, 7 June 2006 (UTC)

• • Result: See text within <pre></pre> tags above. We wish that (a) we can have a hidden section where all bibliographic notes can be placed together and (b) that we can place this section in the beginning of the article, so as to workaround the later instances of the same ref text truncation -or- even better to cancel this truncation and allow for the text to exist in the second (or following) same ref, provided the first one(s) is(are) empty.
• Disagree The whole point of the new referencing system is to put the text of the reference with the text that refers to it. This ensures that the references still work as text gets moved around or deleted. If you delete the text, you delete the reference as well, so you don't leave broken references that are never referred to. Putting a separate reference in a hidden section just to clean up the edit window is a bad idea. --Srleffler 15:59, 10 June 2006 (UTC)
  • Disagree - some notes/references are needed at multiple places, so it makes no sense to include the text only at one of them. Then what happens with the inline references when "text gets moved around or deleted"? In many cases it's better for the refs to be in the ==References== section. --V111P 20:50, 10 June 2006 (UTC)
• NikoSilver, you should read m:Cite/Cite.php and it's talk page. There are many good suggestions there already. --V111P 20:50, 10 June 2006 (UTC)

Sampling

In "Do not want", I present an example of a paragraph converted to use Cite.php. It is largely unreadable due to the fact that there is a four line citation stuck right in the middle of it. I'm willing to accept that I am just really bad at reading, but does anyone seriously think that this is pleasant to read. If the text within the ref tag was a much lighter shade of grey, that would make all the difference, but I'm not sure what would be harder, making ref texts either very light grey (or even white) or putting references at the bottom. In "Do want", I present some examples of the same paragraph with non-crazy referencing. - FrancisTyers · 14:37, 7 June 2006 (UTC)

Do not want

In 1845 the [[Russia|Russian]] scholar [[Viktor I. Grigorovič]] travelled in the Balkans in order to study 
the south Slavic dialects of Macedonia. His work announced to the world for the first time the existence of
two separate [[Bulgarian language|Bulgarian]] dialects: Eastern and Western. According to his findings, the
Western Bulgarian variety, spoken in Macedonia, was characterized by traces of Old Slavic nasal vowels 
<ref>[http://www2.unil.ch/slav/ling/recherche/biblio/97macedTK.html Seriot, P. (1997) "Faut-il que les 
langues aient un nom ? Le cas du macédonien", in Andrée Tabouret-Keller (éd.) ''Le nom des langues. L'enjeu 
de la nomination des langues'', Vol. 1, pp. 167-190 (Louvain : Peeters)] (in [[French 
language|French]])</ref>. It wasn't until the works of [[Krste Misirkov]] that parts of what had been 
regarded as West Bulgarian dialects were defined as a separate 'Macedonian' language. Misirkov was born in 
a village near [[Pella]] in [[Macedonia (Greece)|Greek Macedonia]].

Do want

In 1845 the [[Russia|Russian]] scholar [[Viktor I. Grigorovič]] travelled in the Balkans in order to study 
the south Slavic dialects of Macedonia. His work announced to the world for the first time the existence of
two separate [[Bulgarian language|Bulgarian]] dialects: Eastern and Western. According to his findings, the
Western Bulgarian variety, spoken in Macedonia, was characterized by traces of Old Slavic nasal vowels 
{{ref|seriot1997}}. It wasn't until the works of [[Krste Misirkov]] that parts of what had been regarded as
West Bulgarian dialects were defined as a separate 'Macedonian' language. Misirkov was born in a village 
near [[Pella]] in [[Macedonia (Greece)|Greek Macedonia]].

or

In 1845 the [[Russia|Russian]] scholar [[Viktor I. Grigorovič]] travelled in the Balkans in order to study 
the south Slavic dialects of Macedonia. His work announced to the world for the first time the existence of
two separate [[Bulgarian language|Bulgarian]] dialects: Eastern and Western. According to his findings, the
Western Bulgarian variety, spoken in Macedonia, was characterized by traces of Old Slavic nasal vowels 
<ref name="seriot1997"/>. It wasn't until the works of [[Krste Misirkov]] that parts of what had been 
regarded as West Bulgarian dialects were defined as a separate 'Macedonian' language. Misirkov was born in 
a village near [[Pella]] in [[Macedonia (Greece)|Greek Macedonia]].

Using indentation

As per past discussion, html tags can neaten (somewhat) the appearance in edit mode and also help identify the refs whatever method of citation is used (direct html links, Harvard, < ref> < note> or cite.php):

In 1845 the [[Russia|Russian]] scholar [[Viktor I. Grigorovič]] travelled in the Balkans in order to study
the south Slavic dialects of Macedonia. His work announced to the world for the first time the existence of
two separate [[Bulgarian language|Bulgarian]] dialects: Eastern and Western. According to his findings, the
Western Bulgarian variety, spoken in Macedonia, was characterized by traces of Old Slavic nasal vowels.<!--
  --><ref>[http://www2.unil.ch/slav/ling/recherche/biblio/97macedTK.html Seriot, P. (1997) "Faut-il que les
langues aient un nom ? Le cas du macédonien", in Andrée Tabouret-Keller (éd.) ''Le nom des langues. L'enjeu de
la nomination des langues'', Vol. 1, pp. 167-190 (Louvain : Peeters)] (in [[French language|French]])</ref>
It wasn't until the works of [[Krste Misirkov]] that parts of what had been regarded as West Bulgarian
dialects were defined as a separate 'Macedonian' language. Misirkov was born in a village near [[Pella]]
in [[Macedonia (Greece)|Greek Macedonia]].

or

In 1845 the [[Russia|Russian]] scholar [[Viktor I. Grigorovič]] travelled in the Balkans in order to study
the south Slavic dialects of Macedonia. His work announced to the world for the first time the existence of
two separate [[Bulgarian language|Bulgarian]] dialects: Eastern and Western. According to his findings, the
Western Bulgarian variety, spoken in Macedonia, was characterized by traces of Old Slavic nasal vowels.<!--
  --><ref>[http://www2.unil.ch/slav/ling/recherche/biblio/97macedTK.html Seriot, P. (1997) "Faut-il que les
langues aient un nom ? Le cas du macédonien", in Andrée Tabouret-Keller (éd.) ''Le nom des langues. L'enjeu de
la nomination des langues'', Vol. 1, pp. 167-190 (Louvain : Peeters)] (in [[French language|French]])</ref><!--
-->It wasn't until the works of [[Krste Misirkov]] that parts of what had been regarded as West Bulgarian
dialects were defined as a separate 'Macedonian' language. Misirkov was born in a village near [[Pella]]
in [[Macedonia (Greece)|Greek Macedonia]].

David Ruben ^Talk 15:38, 7 June 2006 (UTC)

No offense, but it is only a microscopic improvement, the reference still completely blocks the flow of the text. :( - FrancisTyers · 15:49, 7 June 2006 (UTC)

Example format

Fran, would you settle for the following option (provided we can have [enter] rather than <br> as requested above and that [enter] right after the <ref> and right before the </ref> tags wouldn't print?):

In 1845 the [[Russia|Russian]] scholar [[Viktor I. Grigorovič]] travelled in the Balkans in order to study the
south Slavic dialects of Macedonia. His work announced to the world for the first time the existence of two separate
[[Bulgarian language|Bulgarian]] dialects: Eastern and Western. According to his findings, the Western Bulgarian
variety, spoken in Macedonia, was characterized by traces of Old Slavic nasal vowels.<ref>

[http://www2.unil.ch/slav/ling/recherche/biblio/97macedTK.html Seriot, P. (1997) "Faut-il que les langues aient
un nom ? Le cas du macédonien", in Andrée Tabouret-Keller (éd.) ''Le nom des langues. L'enjeu de la nomination des
langues'', Vol. 1, pp. 167-190 (Louvain : Peeters)] (in [[French language|French]])

</ref>It wasn't until the works of [[Krste Misirkov]] that parts of what had been regarded as West Bulgarian dialects
were defined as a separate 'Macedonian' language. Misirkov was born in a village near [[Pella]] in [[Macedonia
(Greece)|Greek Macedonia]].

Huh?  NikoSilver  ^{(T) @ (C)} 16:04, 7 June 2006 (UTC)

Still trying to digest this?  NikoSilver  ^{(T) @ (C)} 16:13, 7 June 2006 (UTC)

This would be very confusing for a reference appearing at the end of a paragraph - there will need to be two blank lines (one in your proposal is ignored as part of </ref>, then other needed to indicate a new paragraph).David Ruben ^Talk 16:49, 7 June 2006 (UTC)

No, you could just place [enter] after the </ref> tag. Like this:

In 1845 the [[Russia|Russian]] scholar [[Viktor I. Grigorovič]] travelled in the Balkans in order to study the
south Slavic dialects of Macedonia. His work announced to the world for the first time the existence of two separate
[[Bulgarian language|Bulgarian]] dialects: Eastern and Western. According to his findings, the Western Bulgarian
variety, spoken in Macedonia, was characterized by traces of Old Slavic nasal vowels.<ref>

[http://www2.unil.ch/slav/ling/recherche/biblio/97macedTK.html Seriot, P. (1997) "Faut-il que les langues aient
un nom ? Le cas du macédonien", in Andrée Tabouret-Keller (éd.) ''Le nom des langues. L'enjeu de la nomination des
langues'', Vol. 1, pp. 167-190 (Louvain : Peeters)] (in [[French language|French]])

</ref>It wasn't until the works of [[Krste Misirkov]] that parts of what had been regarded as West Bulgarian dialects
were defined as a separate 'Macedonian' language. Misirkov was born in a village near [[Pella]] in [[Macedonia
(Greece)|Greek Macedonia]].<ref>

Suppose we needed a ref in the end of this paragraph and this was the text.</ref>

Continue with rest of the text...

OK?  NikoSilver  ^{(T) @ (C)} 17:02, 7 June 2006 (UTC)

I find it (personal preference I agree) less clear to quickly parse through the text and identify one paragraph from another, without slowing down to determine where the </ref> tag is (end of line means a new paragraph, at start of line means paragraph continues). But I suppose I could get use to it... :-( Rather than having spare blank lines to distinguish references, it would be easier to have such markup shown in a different colour. I'll mention the obvious counter arguements: wiki-server load, wikimedia reprograming, issues of which colour for those with colour blindness or poorer visual acuity. David Ruben ^Talk 17:27, 7 June 2006 (UTC)

Ha ha! I really don't know if we can have different colors within the edit window anyway. So let's not push for things that may not be feasible. I think we are now complete, as to what we want. The Wikipedia talk:Footnotes#Proposals to developers after consensus are ready. Care to give them a look? I'll post a poll below and hope you agree!  NikoSilver  ^{(T) @ (C)} 17:36, 7 June 2006 (UTC)

when will this be solved?

this issue is the killer for me (i first raised it months ago - its now been incorporated onto the page in the "disadvantages and future improvements" section):

Only the first <ref> note with the same name attribute (e.g. <ref name="foo">) is used to determine the content of a footnote. Edits to an article can inadvertently put an empty named reference before one with content and/or put differing content in multiple named references (hence hiding all non-first note contents).

any idea when/if it could be solved? Zzzzz 17:22, 7 June 2006 (UTC)

Maybe if we all SHOUT loud enough it will be solved faster. Check below:  NikoSilver  ^{(T) @ (C)}

should i put the above issue into the "Proposals to developers after consensus" section then? Zzzzz 17:50, 7 June 2006 (UTC)

It's already there in the last proposal.  NikoSilver  ^{(T) @ (C)} 17:54, 7 June 2006 (UTC)

Solution just needs to be acted on

Note that I proposed a specific code fix to address most of this, about a month ago. If anyone wants to stir the pot and knows active MediaWiki developers, the fix is <http://bugzilla.wikimedia.org/show_bug.cgi?id=5885>. Moreover, if this fix is made active, it will give the technique discussed at User:CitationTool/Hybrid referencing more flexibility. Basically, the idea there allows hiding the verbose references, and refering just to the name; but with the mentioned bugfix in place, we could move those hidden blocks to the "References" footer instead of sticking it at top. LotLE×talk 20:05, 7 June 2006 (UTC)

Great! Now all we need is a LOUD consensus that will help push implementation. I understand those guys who miraculously create all those funky tools for us, though. Sometimes it is harder to understand what users want than to actually do it. That's why all of these proposals were hanging around here and there for so long without anyone taking the responsibility to fix the darn bugs! I hope that after this poll everything will be crystal clear for the developers!  NikoSilver  ^{(T) @ (C)} 10:45, 8 June 2006 (UTC)

Poll

Do you agree with all of the improvements suggested above in Wikipedia talk:Footnotes#Proposals to developers after consensus for the <ref> function?

Sign below, indicating your preference. Poll closes in one week, at 14/06/06 15:00 UTC. All users are eligible to vote, apart from anons and users with less than 50 edits before June 6th 2006.

Please avoid comments along with your vote. Instead, place them below in the proper section.

This poll is closed.

Final result: Support: 32 [33], Oppose: 5

- FrancisTyers · 20:29, 14 June 2006 (UTC)

Agree

1.  NikoSilver  ^{(T) @ (C)} 17:45, 7 June 2006 (UTC)
2. FrancisTyers · 18:19, 7 June 2006 (UTC)

   LotLE×talk 20:14, 7 June 2006 (UTC) On reflection, I'm not quite sure what this is endorsing: I like parts of the above discussion, and am neutral about others; I few parts I am moderately inclined against, though don't per se oppose. (6) is the only point I emphatically endorse (fix non-first ref content bug).

   I think the main points worth discussion here are #6 (fix non-first ref content bug) and #1 (split bibliography from other notes). The others are to my opinion details. If you disagree with having the ability to split the refs in some articles, then you did correct in retracting your vote.  NikoSilver  ^{(T) @ (C)} 22:26, 10 June 2006 (UTC)

   As I said, I emphatically agree with #6. I think #1 is pretty good, but in most cases I'd rather use Harvard referencing for the bibliographic section than a different numbered style (but I'm not opposed to the option for editors who want it. It's hard to boil that down to just agree/disagree. LotLE×talk 00:38, 11 June 2006 (UTC)

3. InvictaHOG 21:10, 7 June 2006 (UTC)
4. Chaipau 21:45, 7 June 2006 (UTC)
5. Telex 22:48, 7 June 2006 (UTC)
6. MatriX 06:34, 8 June 2006 (UTC)
7. Aldux 11:59, 8 June 2006 (UTC)
8.  Earth KING  ^(Talk) 13:20, 8 June 2006 (UTC)
9. Jkelly 16:10, 8 June 2006 (UTC)
10. Slambo (Speak) 16:12, 8 June 2006 (UTC)
11. AlabamaboyAlabamaboy 16:15, 8 June 2006 (UTC)
12. --HolyRomanEmperor 17:00, 8 June 2006 (UTC)
13. --Hectorian 17:05, 8 June 2006 (UTC)
14. Ian13/talk 17:09, 8 June 2006 (UTC)
15. Todor→Bozhinov 17:40, 8 June 2006 (UTC)
16. --Mallaccaos, 8 June 2006 (UTC)
17. -- «₪Mÿš†íc₪»  ^(T) 18:59, 8 June 2006 (UTC)
18. --user:Panosfidis 11:25 (Greek time) 8 June 2006
19. --Nakos2208 22:50, 8 June 2006 (UTC)
20. --Miskin 22:54, 8 June 2006 (UTC)
21. --   Avg    00:52, 9 June 2006 (UTC)
22. Samir धर्म 01:02, 9 June 2006 (UTC)
23. Fastfission 01:50, 9 June 2006 (UTC)
24. --Greasysteve13 05:54, 9 June 2006 (UTC)
25. --FocalPoint 15:19, 9 June 2006 (UTC)
26. Dr. Manos 19:53, 9 June 2006 (UTC)
27. Siva1979^{Talk to me} 20:14, 9 June 2006 (UTC)
28. Titoxd^(?!?) 06:03, 10 June 2006 (UTC)
29. --Yannismarou 09:56, 10 June 2006 (UTC)
30. --Kalogeropoulos 22:19, 10 June 2006 (UTC)
31. --Spangineer^[es] (háblame) 04:00, 11 June 2006 (UTC)
32. --TheArchon 10:13, 11 June 2006 (UTC)
33. --FlavrSavr 23:56, 13 June 2006 (UTC)

Disagree

(comments made by opposing users were moved right below in comments section)

1. David Ruben ^Talk 18:38, 8 June 2006 (UTC) (in part, see discussion)
2. Thatcher131 04:25, 9 June 2006 (UTC) (in part, see discussion)

   Titoxd^(?!?) 06:51, 9 June 2006 (UTC)

3. --Srleffler 16:02, 10 June 2006 (UTC) Proposal needs more work.
4. Agree with 1.1 and 6, obviously agree about 3 and 5 (both bugs, at least 3 already fixed), disagree with 2 and 4. --cesarb 17:31, 10 June 2006 (UTC)
5. Two sets of intermingled references in the text is just a bad idea. No one does it that way. It would be complicated to implement, confusing to use, and frustrating for the reader. If a bibliography is separate from notes, then only the notes have in-text references. —Michael Z. 2006-06-10 17:48 Z

   Ok, but nobody forces you to split them anyway. Actually it is the other way around: It is you who is trying to force everyone to join them! Oh? an why should only the notes be priviledged to have in-line references? Can't we have the ability to choose if the one or the other or both will? This is ref-racism! :-)  NikoSilver  ^{(T) @ (C)} 22:15, 10 June 2006 (UTC)

Comments on vote

• I think implementing Proposal 1.1 above ("Option to separate ... references from ... notes ... [using <ref> and <note> tags]") would be a significant improvement. I need to make some time to digest the other intriguing ideas. Regards, David Kernow 01:15, 8 June 2006 (UTC)
  • Look at 1.4 too. It covers it better. Actually 1.1 is a subset of 1.4, since in 1.4 you can separate the footnotes in more than two categories, and even specify a descriptive character that disambiguates them from each other. A very important feature is also 6.2, which allows you to move paragraphs while re-organising an article, without having to check if the ref with the text actually comes first. To help your digestion even more, all proposals are not mandatory. If they are fixed, you use them only if you need them (like Alka-Seltzer!)  NikoSilver  ^{(T) @ (C)} 10:16, 8 June 2006 (UTC)

(following comments made within poll were moved here below:)

• Disagree - with parts 1.2, 1.3 (suggest subsume into a modified 1.4) and 2, unsure re 6 - whilst flexibility to have separate lists for 'Reference' and 'Footnotes' sections would be an improvement and now seems obviously required (rather than completely separate systems), wikipedia is a body of work and as such there should be some standardisation of formats. Hence the use of multiple single quotes for italics or bold, rather than the user forcing the issue with html markup, or even worse using html 'font' to redefine the font size. Options for altered appearance are then a matter for the individual wikireader via their selection of skins or for a publisher (given that wikipedia0.5 and eventually wikipedia1.0 might appear in hardcover paper versions).--David Ruben
  • Increased attributes do not prohibit the standardisation of formats. On the contrary, decreased attributes prohibit the differentiation of formats, should they be needed for some reason. The whole thing can be well defined through a proper guideline. We both have all 26 letters on our keyboard, "f", "u", "c" and "k" included, but we don't use them consecutively because of the WP:NPA policy. The purpose of having different options separated by "and/or" is only to facilitate developers, since we are not in a position to know what is actually feasible technically. The end result is what counts: To be able to separate the biblio-refs from the note-refs. --NikoSilver
• (1.2 & 1.3) - Yes different cite.php sections will need different numbered lists, but I think their number style should be formally defined, rather than being totally different on each wiki article and dependant on the whim of the article's creator. Hence lets have <refR>, <refF> & <refB> or better <ref-ref>, <ref-note> & <ref-book> to separate out links for References, Footnotes & Bibliography citation lists (is there any need for any other class of linked lists ?) - but each of these should have a fixed standard link style (eg ^{(1) <1> {1}}).--David Ruben
  • I agree. We can use refB, refN and refR as a standard. Let's define that in a proper guideline. Please don't use software to restrict other uses that can be useful in cases we don't imagine now. After all, this is an intention vote (hence "and/or" between the said proposals). We have the same intentions, so we agree! --NikoSilver
• (2) As for turning off superscripts, again this both removes standardisation and will be very confusing with any in-line links such as [7], which also appear as normal non-superscripted text.--David Ruben
  • Again, yes to standardization, but through guideline, not through software. --NikoSilver
  • Update: I modified the proposal according to User:Titoxd's suggestion. --NikoSilver
• (6) I like the current cite.php system in which I can edit and expand a section of an article adding citations as I go (rather than having to perform a separate edit of a 'Footnotes' section too). I would not though vote against the whole proposal if only this issue. David Ruben ^Talk 18:38, 8 June 2006 (UTC)
  • A matter of standardisation vs differentiation again, I presume. There appear to be many editors who want the bibliography section separate for various reasons. It is all over the talk above, with many arguments and counter-arguments. Personally, I don't care either way, but it is irrational for the software to support only half of those users. --NikoSilver
• Moved David Ruben's comments here, sorry poll rules. My rsp's are inserted between your different points and I've used short sigs for both of us (especially me). I suppose that we agree in the end result as this poll only serves to point out the general intentions of the users. Therefore, please reconsider your vote as we can always sort out the details later.  NikoSilver  ^{(T) @ (C)} 22:44, 8 June 2006 (UTC)

I think #1 is a solution for a problem that does not exist. I have never seen a scholarly publication that had multiple classes of endnotes. Footnotes vs endnotes, yes, but since wikipedia pages are limited to a single height we will never really have footnotes anyway. It will create reader problems, with people looking in the wrong section. There will be standardization (I guess, I would call it internal consistency) issues with some articles using a single class of endnote and other articles using multiple classes.Thatcher131 04:41, 9 June 2006 (UTC)

The problem exists in many articles. Most editors split the biblio-refs from the note-refs manually. NikoSilver  ^{(T) @ (C)} 09:04, 9 June 2006 (UTC)

I was orignally fine with #2 until I read Ruben's comments. Unless someone goes through and converts every inline link to an endnote, the added confusion will not be worth the tidier page display. When I read an article and see a note[1], I don't want to have to think about whether that is to endnote 1 or to an inline link that will take my browser to who knows where. (In fact, I never use inline links and recently saw on WP:ANI where someone was altering links to take people to a spam/p0rn site that infected them with a virus.) People like inline links, though, so in that case the endnotes need to be supered.Thatcher131 04:41, 9 June 2006 (UTC)

You are right about this. See my comment for User:Titoxd below. I hope it fixes it for you. I think this is a trivial issue anyway and wouldn't mind if it was dropped alltogether. NikoSilver  ^{(T) @ (C)} 09:04, 9 June 2006 (UTC)

6.1 is a good idea I think. I don't understand 6.2. The cite.php module should be able to pick up the appropriate ref info no matter where on the page the parent reference is. If one group of editors wants to put all the parent refs in one place and hide them, and use the nametags in the article, that should work. It would also have the benefit that an outside editor visiting the article could add a ref in the body of the article without breaking the ref system or getting them out of order (without the outside editor having to conform--the regular editors can move it later if they really want).Thatcher131 04:41, 9 June 2006 (UTC)

6.2: The function as it is now, looks ONLY in the first instance of the "ref name"X in the article body and prints ONLY this. If this is empty then it prints NOTHING, regardless if the ref text exists in a lollowing paragraph. We ask that if the ref-text is empty, it should look for the first one in the body text which is FULL, and print that instead. This will also help people move paragraphs around in articles, without moving the ref-text from the paragraph that used-to-be-first-but-now-is-second to the paragraph that used-to-be-second-but-now-is-first. NikoSilver  ^{(T) @ (C)} 09:04, 9 June 2006 (UTC)

4 is another solution without a problem. In an encyclopedia (which generally is a much shallower coverage of a topic than a scholarly book, and properly so) if information is important enough too require a multi-paragraph endnote, it should probably make its way into the main article. Thatcher131 04:41, 9 June 2006 (UTC)

See Macedonia (terminology) ref#5 and advise how we could do it otherwise. My main point in these issues is: Let us not limit the system by software, because there are things we can't imagine being useful to particular instances. We can always have it done by a guideline. Also, see the #Example format which could help editors separate the footnote from the body text and make it easier to read while in the edit window.  NikoSilver  ^{(T) @ (C)} 09:04, 9 June 2006 (UTC)

I copied your sig to every paragraph above and inserted comments. Hope it's ok! :-)

(more comments moved here below from within poll)

Too much lumped together, some are better than others. Thatcher131 04:25, 9 June 2006 (UTC)

Responses posted above. Keep in mind that all suggested improvements are not to be mandatorily used. If you disagree with anyone of them that doesn't help you personally, then just don't use it in the articles you edit.  NikoSilver  ^{(T) @ (C)} 10:23, 9 June 2006 (UTC)

Everything looks right except turning off superscripting. After reading the arguments, I don't really see a point to doing that on a per-article basis. It might be a good idea to include a global variable can be turned off and turned on in a site-wide basis only. I agree with everything else. Titoxd^(?!?) 06:51, 9 June 2006 (UTC)

I agree to that too. ("Global basis"). I'll add it to the proposal. It will slightly modify the result of the previous 25 supporting votes, but since it can be done anyway globally, and since this is a trivial issue, I suppose it doesn't matter. Please don't disagree on this procedural detail, so as to save time and effort.  NikoSilver  ^{(T) @ (C)} 08:45, 9 June 2006 (UTC)

• What's the point of 6.1? How is it any different from what I've done at Operation Auca? Why would we want to hide bibliographic information? --Spangineer^[es] (háblame) 15:46, 9 June 2006 (UTC)
  • It's the summarised product of a lengthy discussion. I saw what you did in Operation Auca and the end result has a similar effect, only the ref functions link to what you call in that article "References" section, and the "notes" section doesn't exist. So the article can have the following elements:
    • Body text with reference names only and no text after the names.
      i.e. <ref name="Book1" />
    • A hidden section where all those names are defined having their text right beside
      i.e. a list of lines like this: <ref name="Book1">Book title, Author, Publisher, Year etc</ref>
    • A normal section for all references with <references />, so that the above text appears.
  • So the combination of (a)being able to have a hidden section and (b)being able to write the ref texts after the first refs of the body text, separates the bibliography from the body text. This can be useful to help the flow of text while in the edit window. See the #Extensive discussion section under #6 and FrancisTyer's SECOND sample in #Do want section for more details.  NikoSilver  ^{(T) @ (C)} 16:20, 10 June 2006 (UTC)
    • OK, that makes sense. But then there's no way to cite page numbers, correct? Of course, sometimes they aren't needed, but this method doesn't allow them, correct? --Spangineer^[es] (háblame) 17:21, 10 June 2006 (UTC)
    • Correct in everything but one: This particular method does not allow them, BUT you can alternatively choose to have your preferred method even in the same article if needed. The one method does not annoy the other one under no circumstances. For example:
      • Body text with both methods of reference:
        names only i.e. <ref name="Book1" />
        and normal i.e. <ref name="Book2">Details of Book 2</ref>
      • A hidden section where all the names that have not been defined within the body text are defined having their text right beside
        i.e. a list of lines like this: <ref name="Book1">Book title, Author, Publisher, Year etc</ref>
      • A normal section for all references with <references />, so that both the above texts appear from both methods.
    • So we can have both showing page numbers, and not showing them, for different books, even in the sane article! Ok?  NikoSilver  ^{(T) @ (C)} 22:07, 10 June 2006 (UTC)
      • Gotcha. I figured that both would be options, but didn't realize that both would work in the same article. Good stuff. I don't think #2 is necessary or even beneficial (since it causes potential confusion with existing inline external links), but as long as it's in the preferences and not set as the default, doesn't matter much to me. Adding my name under support. --Spangineer^[es] (háblame) 04:00, 11 June 2006 (UTC)

• What's the point of voting on 3 and 5? Both are nothing more than bugs, and nobody would oppose fixing them (particularly, 3 is bugzilla:5384, which has already been fixed for more than a week). --cesarb 05:51, 10 June 2006 (UTC)
  • You're right about both #3 and #5. But I think that all points -more or less- can be considered bugs. The purpose of this poll is to illustrate that apparently there are many users who want those bugs fixed.  NikoSilver  ^{(T) @ (C)} 16:20, 10 June 2006 (UTC)

• Feature requested should be proposed on Mediazilla. People who support them can vote for them. —Simetrical (talk • contribs) 04:43, 11 June 2006 (UTC)

Get in touch...

...with the developer who wrote the damn Cite extension in the first place. Special:Version is your friend, and it tells you that that's Ævar Arnfjörð Bjarmason. Who happens to also be User:Ævar Arnfjörð Bjarmason. robchurch | talk 14:57, 9 June 2006 (UTC)

I spoke to him about it around three weeks to a month ago. - FrancisTyers · 15:00, 9 June 2006 (UTC)

Comments here and there

I think these proposals need more discussion, before being presented to the developers as a "consensus".--Srleffler 16:00, 10 June 2006 (UTC)

Your comment was moved here from within the poll proposal! I suggest you discuss more then, and obstruct an ongoing vote less! To your point: Nobody alleged consensus (without quotes) before obtaining it. On the contrary, I put scary quotes in "consensus", wrote that it is only by four users who happened to be discussing (where were you?) and introduced a poll before I transferred the results to the developers without asking anybody (as I definitely could). Aslo, (duh) why would there be a poll if there was a consesus in the first place anyway? Now is there anything productive you want to discuss? Do you have any better thoughts? Any proposals?  NikoSilver  ^{(T) @ (C)} 22:49, 10 June 2006 (UTC)

NikoSilver, I agree with Srleffler. In the Proposals to developers after consensus section above you are suggesting that 'we leave this note here for a week, until 14/6/2006'. Then you start the pool a few hours later - that's way too early. --V111P 09:56, 11 June 2006 (UTC)

Guys, the way I see it, we wouldn't even have this discussion in the first place had I not started a poll. People respond to polls, but don't respond to talks. We had 4 users here, and now we have 40! Now this is a solid basis for a consensus (without scary quotes anymore). See the history of the talk. A week wouldn't have been enough anyway! We'd probably get a couple of more opinions. Let's please drop the technicalities and focus on the issue, now that it has become hot (because of the poll). The result so far proves that it wasn't so bad an idea. It also proves that the proposal didn't need as much more work as some imply here. In any case, you don't drop a note within a proposal which has been put to a vote, when there is a specific section called "#Comments on vote" for that purpose.  NikoSilver  ^{(T) @ (C)} 09:44, 13 June 2006 (UTC)

I agree the poll was a good idea (a good way to attract more discussion), as long as nobody starts counting votes (counting votes is evil). --cesarb 02:14, 14 June 2006 (UTC)

The "consensus" here is apparently between four users only. Namely: User:FrancisTyers, User:Kirill Lokshin, User:InvictaHOG and me. If anyone disagrees, please comment. I suggest we leave this note here for a week, until 14/6/2006

As an aside from the previous discussions on where it is easier to place full citation details (in-text making it hard to read, or at the end but needing edit to a separate 'footnote' section),:

• There is the issue that too many articles state facts without providing citation details, or almost worse, articles with long technical discussion that then have a huge list of bullet-listed references to research papers that are not linked to the relevant discussion in the main text. A system that allows simple conversion of a bulleted lit to a hidden-end section < Ref> system is hardly likely to encourage sorting out and attributing the references into the text.David Ruben
• Having a referencing system with <ref> in the text helps ensure editors thinks about verifying statements made at the relevant location.David Ruben
• As for the argument raised previously that a hidden end-section for reference definition, would facilitate later editors to add additional references - please no ! The dumping of large number of journal references would be very hard to process if not assigned to specific comments made within the text. We would end up with a hidden section containing loads of additional references for which no link is provided up in the main article – at least having the definition up in the text forces at least one link to appear.David Ruben
• Yes I know these problems are of how editors might wrongly use a referencing system, but any of the previous depreciated footnote systems can (with enough manual effort) do most of what we now seek - the point is to develop a system that is both easier, more automated and needs less editor effort in correctly structuring. The important word in this is 'structuring' – I suggest that this is not just the specifics of linkages, sequential numbering and multiple-links, but ALSO a little (paternalistic) nudging of editors to write well cited articles as a whole. Whilst I agree footnotes & references should have different styles of links (eg ^{(1) <1>}) this should be defined by cite.php not the editor (one can specify preferred options all one wants, but we will end up patrolling and editing those who wish to have ^{A1B A2B A3B} as a system !)

David Ruben ^Talk 01:06, 11 June 2006 (UTC)

• • Point taken. Why not then force the hidden refs not to appear at all (or even post a red error cite.php notification in preview/save) when there are no non-hidden refs in the body text to match them with!  NikoSilver  ^{(T) @ (C)} 10:02, 13 June 2006 (UTC)

• Davidruben, don't you understand that in some articles the some reference are used in multiple places? --V111P 09:46, 11 June 2006 (UTC)

• Of course, hence the <ref name="x"/> tag. What I was trying to point out is that the current system ensures that the first full <ref name="x">details</ref> ensures at least one link is provided in the article. Whereas allowing for hidden definitions in a separate section opens up the possibility of entries for which there are no links at all in the article. This might be because the full reference was added but no link in the first instance was inserted, or that a link is removed by a later edit. Note for the latter that currently the removal of a duplicating <ref name="x"/> is ok because somewhere in the article the full reference remains, but if such a tag now becomes used for the first link too (rather than just subsequent duplicates), then its removal leaves the article with an orphaned full reference. David Ruben ^Talk 11:01, 13 June 2006 (UTC)

• As I said just 2 paragraphs above, we will force the hidden refs that don't correspond to a non-hidden ref within the body text to produce an error, so you will not be able to include un-linked refs.  NikoSilver  ^{(T) @ (C)} 12:05, 13 June 2006 (UTC)
• And what happens if somebody removes that first full ref tag with the details? With the current system if the "details" are included in every <ref name="ref1">details</ref> then removing the first instance won't be a problem, but some of us think that the "details" should be included in only one place, and that one place should be in the references section. (There's no need to reply to this - it's just my opinion). --V111P 00:03, 14 June 2006 (UTC)

On a lighter note, can you imagine correcting/modifying/adding to a ref that appears e.g. 10 times in the body-text with the current system? (btw I have no problem the hidden section being between <references> and <references/> and think that nobody here would be WP:DICK enough to argue about it)  NikoSilver  ^{(T) @ (C)} 00:30, 14 June 2006 (UTC)

• A "hidden end-section" is not needed, since the text in <ref>s is hidden anyway. What we need is a <ref> parameter (like the name parameter) that tells Cite.php not to add [1] at that place and b at the references section. I am sure this would be very simple for the developers to do. On the other hand we can have a hidden section - between <references> and </references> --V111P 09:46, 11 June 2006 (UTC)
  • Indeed, it is not needed for the machine. It is needed for the human who reads the rext in his edit window and looses himself in the text/code/note intermingled flow. No other use than that. Can't say it isn't important.  NikoSilver  ^{(T) @ (C)} 10:09, 13 June 2006 (UTC)

For me, this is the most important part. It is (I think) the only current advantage of Footnotes3 wrt. Cite.php - FrancisTyers · 10:24, 13 June 2006 (UTC)

Forget what I said above. I think the "hidden section" should be between <references> and </references> --V111P 00:03, 14 June 2006 (UTC)

• There is one other problem that was not mentioned here. Please, read m:Talk:Cite/Cite.php#Allowing_sub-references and m:Talk:Cite.php#Suggestion:_sub-reference_attribute_for_ref_tags! --V111P 09:46, 11 June 2006 (UTC)
  • Wow! Very good proposal! I suggest we include that too later on (or even at the same time if that is possible)!  NikoSilver  ^{(T) @ (C)} 10:18, 13 June 2006 (UTC)
    • The name of the book should be printed on every line - otherwise when you click on the ref-link you will only see something like "page 23" at the top of your screen, and you will have to scroll up to see the name of the book. So it should probably be "* Name of the book" on the first line, and ":* Name of the book - page 23" for the sub-refs. But you should not have to enter the name of the book for every subref, you should be able to set it in the "hidden" references section. --V111P 00:03, 14 June 2006 (UTC)
      • That, or we should force the link to direct to the first line of the sub-reference group (surely the reader can find the corresponding sub-note from there).  NikoSilver  ^{(T) @ (C)} 00:01, 15 June 2006 (UTC)
• About the references/notes section. In order Cite.php to replace the ref/note templates we need one more thing - to be able to have as many groups of footnotes as we want (like footnotes, references, subsection in the references section). Don't tell me 'they don't do that anywhere' - in a book they can't do it, but we can do it in a webpage, and it looks way better that way! We need another parameter - group. It would be used like that: <ref group="" name=""> and <references group=""/> This will be more difficult to implement though. The numbering is another question - it may have to be 1,3,4,7 in the first section and 2,5,6,8 in the second. --V111P 09:46, 11 June 2006 (UTC)
  • That's the idea of the refX, refY, refZ proposal (1.4), plus there numbering issue is solved, like X1, X2, X3... and Y1, Y2, Y3... There is also a proposal to standardise through a guideline the use of particular letters, such as B=Bibliographic ref, F or N=(Foot)note etc. Or perhaps do you mean something else which isn't covered by that?  NikoSilver  ^{(T) @ (C)} 10:18, 13 June 2006 (UTC)

I just meant that there should be a parameter, not different tags, which limits the number of the "groups" we can have. But if we want different numbering for every "group" then we probably need different tags. Do we really need different numbering? --V111P 00:03, 14 June 2006 (UTC)

Gotcha. You were speaking about group numbering in the above sense (example at m:Talk:Cite/Cite.php#Allowing_sub-references) and I was talking about what you would call super-groups (or sections). i.e. you could have one ref section for all your books (with groups for each book showing pages etc) and one more for e.g editor's comments, or footnotes, or whatever. I liked also your note about the top of the screen (repeating book title in sub-refs). I like it even more now!  NikoSilver  ^{(T) @ (C)} 00:24, 14 June 2006 (UTC)

Why called "references"?

Why is this called "ref" and "references" instead of "foot" or "note" and "footnotes"? It is used for and serves the purpose of general notes. —Centrx→talk 09:06, 13 June 2006 (UTC)

It's callled "legacy code". It was like that in the original, and now it's beyond any reasonable limit to change it. Circeus 20:00, 13 June 2006 (UTC)

They could however allow both "ref" and "references" and "note" and "footnotes" to work the same way. --V111P 00:06, 14 June 2006 (UTC)

...whatever... NikoSilver  ^{(T) @ (C)} 00:17, 14 June 2006 (UTC)

Hiding references from the body text

This will allow the separation of references from the main text, and hide the second set of references.

foo foo foo<ref name="foo" />
goo goo goo<ref name="goo">this is a goo reference</ref>
== Wish this was a hidden section ==
<span style="display:none;">
<ref name="foo">this is a foo reference</ref>
</span>
== References ==
<references />

--Iantresman 09:57, 14 June 2006 (UTC)

Thanks. Great tip Ian! I've tested it and it works perfectly! Only two minor details:

• The hidden section must be before the body text (for now, since 6.2 has not been enforced yet)
• The hidden section title must be substituted with an XML comment, or it must be included within the span section (otherwise it prints)

An example of the final code is below:

<!-- Ref definitions -->
<span style="display:none;">
<ref name="foo">this is a foo reference</ref>
</span>

=== Body text ===

foo foo foo<ref name="foo" />

goo goo goo<ref name="goo">this is a goo reference</ref>

=== References ===
<references />

Anyone who wants to see the above in action, please visit my talk at User talk:NikoSilver#Test for footnotes section. NikoSilver  ^{(T) @ (C)} 12:34, 14 June 2006 (UTC)

Ooops! There's still a glitch though! The refs that are defined in the hidden section (foo in this case) are printed below in the ref section like this: 1.^{a b}, where "a" corresponds to the hidden ref (and you can't see it even if you click it) and "b" corresponds to the first instance of the ref in the body text!  NikoSilver  ^{(T) @ (C)} 12:39, 14 June 2006 (UTC)

Darn! --Iantresman 13:28, 14 June 2006 (UTC)

Please don't. Using display: none to hide text is bad for accessibility. See Wikipedia:Don't use hiddenStructure for more information on a similar hack. --cesarb 18:13, 14 June 2006 (UTC)

So perhaps the generating code can be modified to have a method of hidding a footnote, eg.:

<ref name="foo" type="hidden" />

--Iantresman 13:38, 14 June 2006 (UTC)

Would work, but too complicated, I think. Maybe we should stick to the initial proposal... NikoSilver  ^{(T) @ (C)} 13:48, 14 June 2006 (UTC)

References & Footnotes

Is it possible to have both inline refs and footnotes in an article, without the footnotes being listed in the references section. The only idea I have is using a depreciated template for the footnotes, but the usage info seems to have been removed. smurrayinchester^{(User), (Talk)} 20:59, 15 June 2006 (UTC)

Notes or Footnotes?

I've noticed this example page uses the heading "Notes" for the footnotes section. I think "Notes" is misleading. It confused me when I first got to Wiki, because I kept thinking it was stuff I was supposed to read.

The obvious heading for a bunch of footnotes is, like, Footnotes, isn't it?

Therefore, I think this example page should have its example reference heading entitled "Footnotes", not "Notes", to encourage people to use that expression. Gatoclass 16:39, 18 June 2006 (UTC)

I am inclined to agree. —Centrx→talk • 22:53, 21 June 2006 (UTC)

Strictly speaking, if the article is printed, we get endnotes, not footnotes; whereas if the article is viewed on-screen, they can be regarded as footnotes. The usage of "Notes", while not necessarily optimal, at least avoids having something utterly nonsensical in either case. Kirill Lokshin 23:53, 21 June 2006 (UTC)

Agreed, using the less specific 'notes' is adaptive to the multiple output/viewing modes the articles are subject to. User:Ceyockey (talk to me) 00:36, 22 June 2006 (UTC)

Even if the article is viewed on-screen, it is still "Endnotes", they are not at the bottom of a screenful and they are at the end of the article. However, this may be less commonly understandly terminology. —Centrx→talk • 00:52, 22 June 2006 (UTC)

A way so footnotes do not clutter text?

Is there a way to define footnotes in their own section, then using the reference with just a short name, rather than defining them in the body of the article? —Centrx→talk • 05:43, 24 June 2006 (UTC)

See User:CitationTool/Hybrid referencing for a semi-solution. For the complete solution, we need to convince the MediaWiki developers to implement a submitted patch in m:Cite.php. Or you can use the older style {{ref}} notes. LotLE×talk 08:25, 24 June 2006 (UTC)

Summary of poll needed

Would it help if I summarized the #Proposals to developers after consensus (particularly the Results)? I noticed that there were a couple comments to User talk:Ævar Arnfjörð Bjarmason afterwards, but I haven't seen any kind of final proposal. I can try and round up the most recent comments over the past week or two, as well. --J. J. 13:52, 26 June 2006 (UTC)

Sure, any help is good :) - FrancisTyers · 13:56, 26 June 2006 (UTC)

Hopefully I'll get that done in the next day or two, then. Sorry that I missed out on the voting. I come back to this footnotes discussion every few weeks, but I definitely missed an important poll this time! --J. J. 20:42, 26 June 2006 (UTC)

Hey JJ, would you like to also include the bug links as they appear here please? :NikoSilver: 21:11, 26 June 2006 (UTC)

I'll do that. Working on it right now! Hopefully will have it done tonight, assuming I don't get lost trying to catch up with all the little issues and bugs mentioned. --J. J. 21:38, 27 June 2006 (UTC)

This is a lot of work! I've worked through a lot of the discussion and am editing a summary on my local computer right now; I'll post the completed version when I'm done (posting the working draft somewhere would just confuse you, trust me). Sorry to those who are waiting, but "you can't rush quality" ;-) --J. J. 02:45, 28 June 2006 (UTC)

Your efforts are deeply appreciated. Take your time, we've already wasted a lot of it anyway, since nobody dared take the initiative.:NikoSilver: 12:29, 28 June 2006 (UTC)

Seconded. David Kernow 15:13, 28 June 2006 (UTC)