Wikipedia talk:Article size/Archive 1

Proposed change
Moved from Village pump:

Propose change:


 * "WARNING: This page is X kilobytes long; Ideal length for articles is between 30-50K. Should this page grow longer, please consider breaking the page into smaller sections"  - &#35918;&#30505; p.s. will post to mailist..


 * "Ideal length" would not be in the 30 - 50 KB range. Anything above 20KB of readable text tries the patience of readers and anything above 28KB has technical limitations for users on slow computers and/or use certain browsers. --mav 00:32 Mar 7, 2003 (UTC)
 * I have since changed my mind about this per a very long discussion at Wikipedia talk:What is a featured article?/Archive 2 and common practice. The reasoning is summarized at Summary style. --mav 16:51, 8 April 2006 (UTC)

Slow modems and new text
The policy of putting new text at the bottom tries the patience of anyone with a slow computer. --Susan Mason


 * Brion? on the Wikipedia-l said that 30-50 was the range... I thought it suspect at first, but Im not sure where the range really is... this was an issue on 911.wik where Martin and Elo? fought over how whether the 911 list should be broken up - I suggested why not both? If you cant deal with... you have options still, and that list can be broken up by category... this begs the question of course... lists are better dealt with by software... Either way, it seems like the warning needs some clarifying and consensus... i just signed to wiki-en - ill repost this there... - &#35918;&#30505;


 * 30-50k is way too long. First of all, it's a technical problem in that many users cannot edit pages of such length. Second, I think it's simply uncomfortably long for an article. Can you provide links to some pages that are that long that you think are comfortable to read, easy to edit, and simple to cite? --Brion 03:31 Mar 7, 2003 (UTC)


 * Grammy Hall of Fame Award is about 62 kbs. I'm not done wikifying the list, so I haven't thought about moving it yet (one thing at a time...) but I'm not sure if it should be.  I'd rather have it all on one page.  It'll only need to be edited every year, when they induct new members, and I guess for disambigging as needed, though if I do a good enough job wikifying, that should be rare.  If the consensus is to break up any page too long for anyone to edit, I'll go ahead and move it, but I don't see the point. --Tuf-Kat


 * That's a list, which is a rather different animal than a textual article. It is not subject to the same reading pressures, nor to the same editing pressures. --Brion


 * Having trouble editing a page because it is large is just as bad for a list as for an article. Lists are edited at least as frequently as articles. For a list of awards that is complete and accurate it may be less, but even then, links may have to be added or changed, etc. A difficulty of editing a page means it loses the essence of a wiki page and degenerates to "just a web page". --Patrick 12:29 Mar 7, 2003 (UTC)


 * Longer lists may be easier than having the same on several pages. The 30k may be a good measure for texts, but not necessarily lists.
 * For the days we are to use google to find articles, another limit to keep in mind is the 100k limit of google, thus it may be helpful to trim some of the HTML, e.g.
 * make the absolute links relative ones.
 * replace target=_top in every  with 
 * --Docu

Moved from the pump:
I'm currently engaged in a project describing detailed the history of all the coin denominations of British coinage. At the moment I'm doing probably the largest article, English/British coin Penny, which has some 1200 years of history to recount and at the moment I've got as far as 1422. I'm aware that there's some maximum size to a Wiki article, but I've no idea of how big that is compared to what I've already written -- could someone suggest an article that's near the maximum size so that I can see how much more I can write before I run into problems? By the way, I know some people have complained about the naming style of the different articles for the individual coins, but that was adopted before I ever found Wikipedia - maybe we can rename them in a more elegant fashion once they've been written! Thanks. --Arwel 23:55 Apr 2, 2003 (UTC)


 * There's no theoretical maximum, but our server only has 2 gigs of memory, so don't push it please. ;) In all seriousness though, I very strongly recommend against letting articles grow beyond 32 kilobytes. (This page right now is about 34 kilobytes.) Not only is this uncomfortably long to navigate if you're only looking for a small bit of information, and very annoying to edit if you want to touch on something in the middle somewhere, but it's technically problematic because some browsers (including most browsers on MacOS) cut off the text around this point when editing. If a page can't be reliably edited, it really should be refactored into several smaller, more self-contained sections.
 * When a page reached about 28kb, you'll see a warning message appear on the edit screen letting you know that you're entering the danger zone. See Browser notes et al for more details on browser compatibility. --Brion 00:57 Apr 3, 2003 (UTC)


 * see page size


 * Thanks for the advice. In fact the article hit 30K at 1558, so I'll start writing about Elizabeth I's pennies in part 2! :) --Arwel 23:50 Apr 3, 2003 (UTC)


 * (see posts in previous to block moved from pump)
 * For lists, a good mesure might be 50-60k. We probably should amend accordingly --Docu

finding size
I'm sure this has a very simple answer. I'm deciding whether to split an article. How can I find the size of the article? (it's not a new page). jimfbleak 11:12, 30 Sep 2003 (UTC)


 * Pending a better answer, you can copy and paste the source text to a simple txt file, save it, and then look at its properties to see how big the file is. Hth --Pete 11:19, 30 Sep 2003 (UTC)


 * If you edit the article and it's over 32K, you get a warning about "some browsers", which states the size of the article. --Onebyone 12:03, 30 Sep 2003 (UTC)


 * or count each letter one by one to see the approx byte size. ;-) -fonzy


 * Great idea, but do you count new-lines as one or two bytes? &#922;&#963;&#965;&#960; Cyp 15:04, 30 Sep 2003 (UTC)

Additions
I've added an introductory section, basically saying 32KB isn't that important any more, and a section on restructuring and splitting articles, which basically says don't do it hastily, don't break out subtrivia as new articles, and exercise extreme caution when breaking out controversial sections as new articles.

Comments welcome. Dpbsmith (talk) 14:06, 28 Feb 2005 (UTC)

Recommendation to use bad html formatting?
The article recommends using incorrect html formatting for tables if they get too long? Umm... This may be okay for now, but what happens when xhtml becomes standard? In my opinion, this recommendation should be changed, as most wiki editors will only visit here once, and never again, once they know what it says. We don't want editors five years from now remembering the recommendation to use bad html formatting, and then wondering why it doesn't work anymore.

I'm removing the recommendation, but am willing to hear arguments to the contrary. - Eric Herboso 02:49, 16 Mar 2005 (UTC)


 * Yeah, that's not the best idea. :) I've reverted it. Note that MediaWiki attempts to put in the closing tags, but this won't always come out right. --Brion 07:41, Mar 17, 2005 (UTC)


 * Why add them if they are optional? -- User:Docu


 * Okay, I was going to post this yesterday, but then there was the database hoo-hah and I only got to do the revert. MediaWiki outputs correct, validating XHTML 1.0 Transitional. At least, it tries to. And the wikitext shown in the example (now that I've reverted it), if you view the source of the output HTML, spits out the following:


 * so it's really a wacky hybrid of wikitext and HTML, but it does work, and creates perfectly good XHTML. Is it right to get users in the habit of not writing end tags? Perhaps not, but (a) it's not really Wikipedia's problem, (b) everyone does it; MediaWiki just covers for their problems, and (c) it looks less cluttered. I think it's HTMLTidy or something like that which does the conversion. --grendel|khan 14:53, 2005 Mar 17 (UTC)


 * Additional note: Help:Table shows three ways of doing tables: full HTML, the half-assed HTML with no ending tags, and the pipe syntax. All produce identical output. grendel|khan 14:58, 2005 Mar 17 (UTC)


 * I've reverted it back. Tidy will give up on some pages and we can't guarantee that output will be correct. Our internal HTML fixer-upper may also break things itself sometimes. Aside from desiring general good behavior, missing end tags are known to *crash* Netscape 4.x under many circumstances; though we don't actively support that old browser there are still people using it and we try not to *crash* their browsers. --Brion 01:27, Mar 18, 2005 (UTC)

Cyclic Redirect
"The specific browsers that have problems with long pages are listed at Browser page size limits (...)". This sentence (at Article size) has a link to a page that, now, redirects to Article size, which means, it is a link to the same page... Jotomicron (talk), 29 July 2024 (UTC)


 * That signature displays the current date, the issue was solved about a year ago. --&#160;Omniplex 21:29, 8 April 2006 (UTC)

Change in counting?
Has there been any change in size estimation? Maybe my memory is bad, but several articles I thought were ~80kb and I am sure haven't changed now show as 40-50kb. --Piotr Konieczny aka Prokonsul Piotrus Talk 9 July 2005 19:17 (UTC)

Major biographies
The biographical article William Thomson, 1st Baron Kelvin is getting pretty long and I perceive that there is plenty still to so. This is a major figure.

Are there any examples (good or bad) of biographical articles in summary style?

Ther first step would seem to me to move notes/ bibliography/ external links off to a separate page. There might be some technical issues in using the ref and note templates. What should this page be called?
 * William Thomson, 1st Baron Kelvin - notes and bibliography

Any other thoughts on this? --Cutler 10:17, July 26, 2005 (UTC)

Lists?
I'm currently working on a list of Formula One drivers (here), which will have a lot more information than the current one (here). As you can see, I've only got up to C (179 of 783 drivers - see talk page), and I've reaced the 32KB mark. Is there a policy for splitting up lists? Making several pages like List of Formula One drivers (A to D) or something? I'd prefer not to split it up at all, but obviously there is the issue of people not being able to edit them. Thanks AlbinoMonkey (Talk) 04:43, 30 September 2005 (UTC)

List of SCOTUS clerks
I have this question too regarding a page I've started, which due to the great and overwhelming contribution of others is becoming large already and will become larger still, the List of law clerks of the Supreme Court of the United States. The list is much more informative than a simple list of names would be, it contains the information people are interested in when trying to find out about a particular clerk, who they clerked for before the SC, their law school, and years of each. However even with the template reducing the size of formatting elements a bit, the article is already large and will grow. According to What is a featured list?, desireable attributes of a list include comprehensiveness, but making this list comprehensive will make it quite large. Currently there may be up to 37 names added a year. Although many current and former clerks do not have seperate pages now, I believe that will change as the list matures. These represent the top-tier of legal minds and many go on to become SC Justices, Federal Judges, and influential legal scholars or government attorneys. Feel free to leave thoughts and ideas here or on my Talk page. Phil 05:03, 17 June 2006 (UTC)


 * Why not simply split'em apart alphabetically? --maru  (talk)  contribs 01:46, 21 June 2006 (UTC)

Sizes incorrect?
* >3232KB - May eventually need to be divided (likelihood goes up with size; this is less critical for lists) * >2230KB - Might need to be divided (make sure sections are <20K - preferably much smaller) * <2230KB - Probably should not be divided * <86541K - If an article or list has remained this size for over a couple of months, consider combining it with a related page; this does not apply to a redirect or disambiguation page. Alternatively, why not fix it by adding more info? See Wikipedia:Stub. If it's a really important article that's just too short, put it under COTW, a project to improve stubs or nonexistent articles.

So anything less than 86MB is a stub? And anything less than 2MB should not be divided? It seems like these are VERY large, considering that there used to be a 3MB limit.

And when was there a 32MB recommendation? Thus the 32223KB recommendation is considered to have stylistic value in many cases

there may be a decimal missing somewhere... Techgeekxp 00:40, 25 October 2005 (UTC)


 * Looks like someone fixed in the meantime. BTW I also restored the suggested differenciation between articles and lists. The recent versions didn't quite state that and the change wasn't discussed here. -- User:Docu

Kilobyte = kB and kilobit = kb confusion
In the computer world generally a consensus seem to have been just about reached that kb means kilobit and kB means kilobyte. The small k is universally used for kilo, and forms part of official recommendations.

But throughout Wikipedia there is total confusion, and I believe some of the statistics use kb when they mean kB. Here people keep using KB when it should be kB. It seems a good place for me to bring the point up. Mention of 30MB pages has got me totally confused! --Lindosland 15:28, 27 December 2005 (UTC)


 * Actually, these should all be in kibibytes (KiB). ;-)  See Manual_of_Style_(dates_and_numbers).  This is certainly a place "where the precise byte capacities are important to description". — Omegatron 15:14, 24 April 2006 (UTC)

A 'Root page' suggestion
While working on noise I came across a fundamental problem. Some wanted to turn it into a 'disambiguation page' because they perceived noise to mean many things. But doing this took away all the well thought out common structure of the article, leaving a collection of fragments that you would't go to unless you knew exactly what you were looking for. Furthermore, the germans wanted two pages because of problems over the fact that they use two words for noise. And then there was the problem that 'environmental noise' is a form of 'acoustic noise', which 'radio noise' is too, but it isn't if you are looking at the electronics! There's a complex hierarchy, based on two meanings (random and /or unwanted) and its futile to try to explain this on every page.

I put the original page back, calling it a 'root page' and linked to specialist pages for all the detail. They each have a link taking you back to the root page. It taught me that 'disambiguation' is best kept for truly different multiple meanings, not for branching out. Is there a case for promoting the 'root page' concept here? --Lindosland 15:46, 27 December 2005 (UTC)

Article length: judge by book standards?
The assertions stated here about article length ("Anything above 20KB of readable text tries the patience of readers", "I think it's [30-50K] simply uncomfortably long for an article") reek of a violation of supposed Wikipedia policy, as they lack any kind of supporting evidence.

But rather than ask people to document these assertions, I have a question which may be useful to help establish how long really is too long. That question is: how long are articles in "real" encyclopedias (Americana, Funk & Wagnalls, Brittanica, etc.)? How long are the longest articles in print? Seems like this should be pretty easy to determine. (After all, weren't a lot of articles here just lifted from Brittanica?)

I think this would be a better metric to judge length by, rather than the very computer-ish 32K (as if some kind of psychological research has established that MEGO(1) sets in at 32,768 bytes!). It seems the 32K limit has become, for the most part, obsolete because of technical reasons (mobile browsers and a few other cases excepted): the question now should be, how long can an article become before people lose interest?

(1)MEGO: My Eyes Glaze Over --ILike2BeAnonymous 06:52, 5 January 2006 (UTC)


 * I think there are some cases where 32K just doesn't make sense. Say you look up an article on World War II. Would you really expect the text to be limited to 32K? I would say that 100-200 K is easily justified for topics about which there is so much to say. --MvH Jan 13, 2006.


 * OK, but again: can somebody give some meaningful comparisons to real-world printed examples (like real encyclop[a]edias)? Otherwise, it seems we're just pulling arbitrary numbers out of our butts. --ILike2BeAnonymous 17:58, 13 January 2006 (UTC)


 * Britannica has 63 pages on the World Wars. And the articles "Islam" "Islamic arts" "Islamic world" together are 132 pages. Of course, in wiki you can still find quite a lot of the same material because it's divided over numerous articles, and you can still find quite a bit in wiki that's not in Britannica. But in any case, I'd think 100K for WWII would certainly not be so long that it would have to be split. --MvH Jan 13, 2006.
 * PS. That doesn't mean that I'd like a lot of articles to become longer. My point is that if an article is very well written (and there are many jewels here at wiki) then the length of the article should not be an issue. MvH Jan 13, 2006.


 * Thanks for the research. I agree with your conclusions; as I said, it seems to me that the recommended 32K article size limit is both arbitrary and very "computerish" and therefore not a good measure of what "too long" is. (Of course, given how many people are idiots with short attention spans in the Computer Age, maybe it is a good limit, but that's a whole 'nother rant.) --ILike2BeAnonymous 02:52, 14 January 2006 (UTC)


 * Just read Summary style. No reason to repeat all that here since it has all already been worked out. Also, most articles in EB are small and the large ones are huge for reasons relating to the fact that EB is not a hyperlinked work where detailed treatments of subtopics can be treated. Thus all aspects are treated in one place whereas we can summarize various points and provide links to more detailed treatments of those subtopics. This is far more useful to the reader, who then has a choice of the amount of detail he/she reads. In short, we are not limited to the way things are done in print and can thus try to keep articles at a size that can comfortably be read in s single sitting (7000 to 10000 words long, depending on the topic). Get a copy of EB and look for yourself. It's coverage is highly uneven and often does not make sense. --mav 00:07, 17 January 2006 (UTC)


 * The 1911 Encyclopedia Britannica article on the Bible is approximately one megabyte's worth of ASCII text. I don't know the actual figures but a quick inspection of the current Encyclopedia Britannica will confirm that the typical Wikipedia article more closely resembles those in the 12-volume portion known as the "Micropedia" than the long articles in the 18-volume "Macropedia." Our article on Encyclopedia Britannica notes that Micropedia articles are mostly "from one to five paragraphs" while Macropedia articles "range from a few pages to over three hundred pages." Given the relatively large page size and relatively small print, I don't actually have one at hand but I'd guess that a Britannica page probably weighs in at about 10K, which would mean that the shortest Macropedia articles are in the 32K ballpark. Dpbsmith (talk) 13:17, 9 May 2006 (UTC)

length of a section
My comprehension of the article is that it talks about the length of a section or sub-section, not the total lenght of all those sections together. When it says 'article', I think it means then, an article without any section. But I'm not sure if I understand it right. Help someone! --Aïki 16:56, 13 January 2006 (UTC)


 * Um, I don't think so; my understanding is that it's the total length of the article. Could be wrong, but that's my take. (Bumped your comment up a level for formatting.) --ILike2BeAnonymous 17:54, 13 January 2006 (UTC)


 * In the introduction of the project page, it is said: 'With the advent of section editing, ..., this hard and fast rule has been softened.', and further in the section 'Technical issues': 'The section editing feature lessens some page length problems as long as each section of a page is within the limit.'

I think that it is now clear that the lenght's limit is applied to each section individually, not for the totality of the article, except, evidently, if the article has no section at all. But then, we just have to make sections to regulate the problem. Do you understand it the same way? --Aïki 16:51, 14 January 2006 (UTC)


 * There are also issues of readability. See above. --mav


 * Thank you for your judicious remark, but, initially, the question was: 'Is the lenght's limit applied to each section individually, or for the totality of the article?' --Aïki 01:02, 17 January 2006 (UTC)


 * The whole article.-- mav


 * Thank you. --Aïki

Practical example

 * ''Editing Talk:Centrifugal force.
 * This page is 128 kilobytes long. This may be longer than is preferable; see article size.

How can we resolve this problem? Somebody knows? --Aïki 17:26, 21 January 2006 (UTC)


 * Archiving every discussion older than 7 days helps greatly. I have reduced the size to 44 kilobytes. --Ec5618 17:47, 21 January 2006 (UTC)


 * Misunderstanding: Your demonstration is very spectacular! But it is not what I asked. I did not ask to do it, but just tell how it is done. (Now I know.) And this question was in the context discuss here i.e. to know if the over-lenght is for the article entirely or for sections.
 * Secondly, you archive edits less than 7 days. Thirdly, this archiving procedure put the internal links out of order, so I don't like it. Can it be another way, like creating a subpage instead of an archive-page?
 * Fourth, can you put back the archive part in the talk-page, as it was before? ( OK, I did putted it back. --Aïki 00:41, 22 January 2006 (UTC))


 * Thank you. --Aïki 18:29, 21 January 2006 (UTC)

Justification?
The justification for this division was originally technical, but now it seems to be a policy without a guiding principle, and a bit arbitrary. Why are some 70kb+ articles (Hugo Chavez, History of the Jews in Poland, etc.) able to make FA status, but others are not? Why should World War II be reduced to 30-50kb, when it already is in summary style, but still obviously covers complex history? I see these questions have been asked before, but I can't find any defenders of the present policy on the page. I will make a couple of edits here, but think we should re-evaluate this. --Goodoldpolonius2 04:44, 30 January 2006 (UTC)


 * This is all about reading time and making Wikipedia most useful to the largest number of people. Providing summaries and linking to more detailed treatments is the way to do that. This is all covered in far more detail on the Summary style page. --mav 03:40, 2 February 2006 (UTC)


 * The persons providing the justification for limits on article size are predominantly "techies" for whom the writing part is a chore compared to the joy of formatting pages, blocking miscreants and otherwise engaging in the plumbing aspects of html page production. These are the folks who theorize that readers will get bored with articles that are longer than x kb (notice how the limits are in kb and not words - very instructive) often because of their own inadequacies in that department.  What is lost in this discussion is that some articles are well written and can hold the readers' interest far longer than much of the mediocre prose found in other entries.  One look at the Harry Potter books belies the belief that people will not read long stories.  What I often see happen is many editors in obescience to the size limits create daughter articles that end up on AfD.  Once the daughter article is deleted, the text in the main article rises to compensate.  Then you find yourself back on the same merry-go-round.  Of all the rules on Wikipedia this one is the most counter-productive.  Its unsoken assumption is that people writing cannot do so in a literate manner.  That assumption is simply wrong.  Jtmichcock 14:44, 29 April 2006 (UTC)


 * Reading time is still a very important part of readability of an article but, as you state, good writing can help mitigate for that (but only to a point ; wiki is not paper, but Wikipedia is not Wikibooks either). Also, some topics simply require more space than others. Thus the sliding scale used at Summary style along with the mention of readable prose and approximate word count. Encyclopedia articles cannot grow to be book length. So some soft limits and stylistic guidelines are needed. The beauty of hyperlinks is that we can have summaries that lead to more in-depth treatments. Thus we can serve those who need varying levels of detail. --mav 15:21, 14 May 2006 (UTC)

I updated the discussion of what to include when determining page size for purposes of style. The previous policy was closely tied to the technical limitations that are not particularly germain to stylistic concerns. —Doug Bell talk•contrib 07:34, 20 March 2006 (UTC)

Articles split over many pages
Is splitting a long article such as the History of PR China in subarticles (History of the People's Republic of China (1976-1989), scroll down for the "previous" and "continued" links) considered to be good style? --Abdull 18:24, 30 April 2006 (UTC)


 * Only if History of the People's Republic of China sucks as a result. Yep, looks like that happened. That article should be in the 30-50KB of prose range but it is little more than a stub. Nothing wrong with breaking out detail into daughter articles, but good-sized and functional summaries must be left. --mav 15:11, 14 May 2006 (UTC)

Mobile devices
Large pages are a nightmare when you're trying to access Wikipedia from a mobile phone or PDA. I think we need to be firmer about the 32K limit for this reason. Jammycakes 20:12, 8 May 2006 (UTC)


 * Enfored limitations are a nightmare when you're trying to access a full, comprehensive Wikipedia from any regular device. I think we need to be firmer about expanding stubs for this reason. --maru  (talk)  contribs 02:43, 9 May 2006 (UTC)


 * Note of sarcasm there, judging from your edit summary? I don't think mobile users are too much of a minority to be honest. And 32K is quite a generous limit too (4000+ words) so by the time you get to that size it does begin to get a bit large from a readability perspective too. OTOH what I did find to be more of a problem was trying to access articles of any size with the default stylesheet, though you can of course change that in your preferences. Jammycakes 13:14, 29 June 2006 (UTC)

Section size
What are the minimum and maximum limits on section size? I don't find anything at Help:Section. Perhaps there was more information at Guide_to_layout? Hyacinth 19:09, 15 May 2006 (UTC)

Puh-leeze
I'm just going to toss in my $.02 worth here without slogging through all of the previous discussion: It's my experience that essentially every non-trivial article in Wikipedia triggers the "may be longer than is preferable" warning. Ergo, it's time to change the criteria for triggering the warning. Obviously, most editors (including me) disagree with it and aren't heeding it. It feels like a glaring anachronism to me, and as such it harms WP's credibility. KarlBunker 02:12, 7 June 2006 (UTC)


 * As this message is invisible to readers and only seen by editors, I disagree with the credibility problem. Would you be happier if the warning trigger was moved to 50k, which is what the guideline suggests as maximum article size? Any article longer than 50k should definitely trigger a warning, because it is in danger of becoming too tiring to read, a Bad Thing for an encyclopedia article, which is supposed to give the reader a quick overview over a topic. Personally, I really think 32k text plus whatever the references, onfoboxes and similar nonsense take should be the limit above which articles should be shortened. The only problem is that this is a bit harder to measure than the length of the wikicode. Kusma (討論) 02:38, 7 June 2006 (UTC)


 * "Convenient though it would be if it were true, Mozilla is not big because it's full of useless crap. Mozilla is big because your needs are big. Your needs are big because the Internet is big. There are lots of small, lean web browsers out there that, incidentally, do almost nothing useful. If that's what you need, you've got options..." -JWZ


 * Calling for a 32k limit is myopic and foolish. Perhaps Britannica doesn't mind limiting itself to its self-imposed small articles, but we should not. Some of my articles are big, like Fujiwara no Teika, or Palpatine, or The Encyclopedia of the Brethren of Purity. Why are they big? Not because of worthless verbiage, but because their subjects are big, and require a big article to do them justice.
 * Incidentally, moving the warning limit to 50k is a good idea. I'm gonna hunt around over on Bugzilla to see whether there's such a modification I can vote for. --maru  (talk)  contribs 05:14, 7 June 2006 (UTC)


 * People make a lot of sweeping statements on this subject, often about "what readers want." Something that would be interesting and useful would be some actual data on that topic, but I suppose it would be hugely difficult to--say--(temporarily) add a little feedback form to each article that asks "was this article: [] too long   [] too short   [] just right."
 * One form of feedback WP does have is the talk page to each article. Any reader can leave a comment complaining that an article is too long. Offhand I can't recall ever seeing a complaint that an article was too long, except from editors who are reacting to the "longer than is preferable" warning, and wondering just what it means. KarlBunker 11:31, 7 June 2006 (UTC)


 * Well, I do sometimes hear (offline) the complaint "I don't like your Featured articles, they are too long. I would prefer a short article like in a printed encyclopedia". Perhaps we'd need a micropedia (short overview, articles below 5k) and a macropedia for that, but forks are evil. Kusma (討論) 15:45, 7 June 2006 (UTC)


 * Yeah, how about a new "Wiki", the "Short Attention Span 'Encyclopedia' for the Modern Person On The Go Who Doesn't Have Time for Anything More than a 15-second Sound Bite", aka "The New Global Dumbed-Down Information Resource". That way people can get Wiki-information while driving their SUV and applying makeup at the same time. Just what the world needs. Yeah, that's the ticket! ==ILike2BeAnonymous 18:00, 7 June 2006 (UTC)

Two things: --mav 20:54, 14 June 2006 (UTC)
 * 1) As the content page states, it is the size of the *readable prose* that matters most. The total size of the wikitext is no longer as much of a concern due to fix browsers and very widespread use of broadband to access the Internet (even dial-up users can start to read the article while it is downloading). Any properly footnoted and referenced article with lots of intwikis, internal links, external links, see alsos and tables will often be two to three times the size of the readable prose. However, articles that do not have all that will have a total page size that is comparable to the size of the readable prose. Thus to catch the later group of articles when they may be getting too long, we still must keep the current trigger size for the long page warning (what we really need a feature that counts readable words outside of the list-like sections).
 * 2) The lead section is supposed to be able to stand alone as a concise version of the article. Thus no reason to have a separate micropedia.

Articles being split over last couple of days
I've run into this problem (the edit box is not filled with the entire article, but I get the buttons and everything looks normal) several times in recent days, on 2 different computers and not necessarily with particularly large articles either. Both are running Firefox 1.5.0.4 (the latest version). What gives? Has a bug crept into Mediawiki or into Firefox? Some examples:   --kingboyk 13:44, 13 June 2006 (UTC)
 * I'm surprised there was no followups to this, as it appears to be a problem with Firefox and the Google toolbar, hardly an unusual combination one would have thought. Anyway, here's the Bugzilla case: --kingboyk 12:09, 23 June 2006 (UTC)
 * The problem seems to be fixed, at least according to bugzilla. someone might change the text to reflect that. i don't do it myself, because i don't know if i should delete the sentence altogether or change it to say "some versions..."--  ExpImp talk con 23:44, 23 September 2006 (UTC)

Article size and footnotes
Some articles have extensive footnotes, which is good in some subjects for verification purposes and solid grounding.

I'm not convinced that footnote length should be counted in article length. Footnotes are optional, and read as supplements to specific points. A well cited article should be of readable length, with footnotes as needed. Comments? FT2 (Talk 12:43, 22 June 2006 (UTC)


 * See above "As the content page states, it is the size of the *readable prose* that matters most. The total size of the wikitext is no longer as much of a concern due to fix browsers and very widespread use of broadband to access the Internet (even dial-up users can start to read the article while it is downloading). Any properly footnoted and referenced article with lots of intwikis, internal links, external links, see alsos and tables will often be two to three times the size of the readable prose. However, articles that do not have all that will have a total page size that is comparable to the size of the readable prose. Thus to catch the later group of articles when they may be getting too long, we still must keep the current trigger size for the long page warning (what we really need a feature that counts readable words outside of the list-like sections)." --Anon

Long article handling
I've updated the intro slightly, for slight readability and clarity. I notice though that although we discuss how to split long articles, and how not to rush it, there isn't a real discussion about them.

Such a discussion should, I feel, cover these issues:
 * Some articles -- not many -- are by nature, sometimes long.
 * If so, do what can be done to keep them short, and pay above average attention to readability. But if it genuinely would not make sense as two articles, leave it as one.
 * Readers of such articles will probably expect them to be long. If they are easy to read, and well structured, and justify the length, then that helps.
 * Excessively long articles are not encouraged and should not be the rule.
 * Readability is the key criterion.

Realistically there are many articles over 30-50 kb, and will continue to be. So a section that at least covers when and why and when not, would be valuable. It's not long so I added it.

FT2 (Talk 17:43, 27 June 2006 (UTC)

A remedial question
I agree the length of an article is important. But how is the length measured? If the article should not be more than 32kB, how do I know when it has reached that level? This is frustrating for me because everyone else seems to know and I have looked seemingly everywhere for this information and cannot find it. I would like to be able to look at an article or section and say "That text is only 17kB, so we can add more!" or something. I have no clue. Can someone help me? RonCram 13:05, 8 July 2006 (UTC)

30KB
I've just seen a warning on a page that was only 30KB. 32KB is already too soon to be warning people, in my view. Are we reducing the recommended length still further? SlimVirgin (talk) 23:33, 15 July 2006 (UTC)

Unsigned quote
Articles are like grass, they need to be trimmed periodically, so that new grass can grow – yet we save the old grass by planting it into different areas. —Preceding unsigned comment added by Sadi Carnot (talk • contribs) 23:38, 28 November 2006

50 to 110 kb long-article breakup proposals
Hi, in the past I've broken up four over the limit pages. People seem to be fairly happy about the change after it is done, e.g.: Moreover, for comparison, I have both the computer versions of Encarta and Britannica and they break their articles up into sections of 2-4 pages long; then you click along if you want to read step-by-step through the article or click on related topics (usually as side links or bottom links). Now, a few articles in Wikipedia are growing without limit. Some of these are very tension-thick articles.
 * Thermodynamics
 * Entropy
 * Love
 * Gravitation.

Long article examples
Here are a few current examples:


 * Entropy - 41 kb (13 printed pages); feels about right, but getting tight around the belt.
 * Caffeine - 52 kb (14 printed pages); feels ok, most of the page is references.
 * Photon - 67 kb (15 printed pages); feels like a heavy read for one sitting; could use some trimming?
 * Article is too long


 * Evolution - 92 kb (27 printed pages); feels way over the limit; I don't even want to attempt to contribute to this page because it's so big. (one of the Wikipedia's top 500 longest articles as well as one of the top 100 most edited articles).
 * 90 kb article
 * Call to trim
 * Call to trim (part 2)
 * Evolution of the evolution article


 * George W. Bush - 96 kb (most visited article)

Most articles over 100 kb are lists (Wikipedia’s longest article is 270 kb). A related website with ridiculously long pages is MySpace.com. We do not want to turn into a MySpace.

Page tension issues
To compound this problem, many editors in Wikipedia are very "attached" to certain pages and are not favorable towards the idea of breaking up their lovely work. This compounds the issue. An example is the evolution page; I brought up this point but there was resistance, i.e. one editor said “that’s being working on”. Big articles tend to stagnate in growth. Everyone, for example, loves contributing to stubbies or even better yet "starting a new article". Typically, if you find a long page put a long page "tag" at the top of the article:

Many editors, however, will remove these unpleasant looking headers, because they are so attached to their beautiful articles. This is another tension issue. Once the page is split up, however, things are again peaceful, and the talk page tensions subside.

Readibility and tension spans
I do a lot of website design and I have read many articles on recommendations of page-length and the psychological reasons behind them. Here's an example from WebStyleGuide.com:


 * Page length
 * Determining the proper length for any Web page requires balancing four factors:


 * 1) The relation between page and screen size
 * 2) The content of your documents
 * 3) Whether the reader is expected to browse the content online or to print or download the documents for later reading
 * 4) The bandwidth available to your audience


 * Researchers have noted the disorientation that results from scrolling on computer screens. The reader's loss of context is particularly troublesome when such basic navigational elements as document titles, site identifiers, and links to other site pages disappear off-screen while scrolling. This disorientation effect argues for the creation of navigational Web pages (especially home pages and menus) that contain no more than one or two screens' worth of information and that feature local navigational links at the beginning and end of the page layout. Long Web pages require the user to remember too much information that scrolls off the screen; users easily lose their sense of context when the navigational buttons or major links are not visible.

There is also the problem for those with dial-up (I use high speed and dial-up), and for those who use hand-helds, and for those who use laptops with a cellular connection, etc. If a page gets so long that it deters editors, we can only image how it will deter readers.

Mandatory breakup proposals
In conclusion, I propose the following:


 * Any article 12-15 pages printed (or 50-70 kb) is call for strongly recommended article breakup.
 * Any article over 15 printed pages (or 70 kb) is call for mandatory article breakup.

I may possibly plan to assemble the WikiProject Extra-Long Article Committee project page. If you would like to join, please leave your name and comment there. I will need at least three administrators to back up the project by stopping reverts:

Breakup committee members
If you are sick of seeing long bulky articles, would like to help break up some long articles, and would like to join the break up committee please leave your name below. Thanks and comments are welcome: --Sadi Carnot 09:39, 9 November 2006 (UTC)

Members
 * 1) – experience breaking up four over the limit articles (likes trimming science articles)
 * 2) Samsara (talk • contribs) – admin who started the WikiProject Modular Articles.

Discussion

 * Note:Up to second note copied from discussion at WP:ANI
 * Good idea, in theory. But I don't really like the fact that you seem to think that instead of establishing a consensus that a page should be broken up, it is better to recruit some admins to prevent people from reverting such a thing. If this idea is implemented well, there should be absolutely no need of admins. -Amarkov blahedits 00:43, 29 November 2006 (UTC)
 * No, my plan is that:
 * First, an editor tries to establish consensus: the issue is brought up on the talk page, and it is suggested that the regulars break up the article into subtopics, with short summary paragraphs (w/ main article attachments), see thermodynamics as an example, so that the main page gets below a certain limit.
 * Second, if plan #1 stifles out in argument and indecision to act, for a number of consecutive weeks, than an breakup arbitration committee notice is placed on the talk page, putting an ultimatum deadline, such that either the regulars break up the page to below a certain limit by that date or an external breakup committee, enforced by a team of administrators, will do so.


 * Without a group project such as this, then Wikipedia talkpages and articles will become like Congress: lots of arguing but little action. This will need to be a team action if it is to be successful.  Here is a recent example in which I placed a "long article" tag on a page but it was quickly reverted; for this situation I would have needed administrative assistance. --Sadi Carnot 01:16, 29 November 2006 (UTC)


 * See WikiProject Modular Articles here. I've been meaning to do this for weeks. Will you help? Many thanks. - Samsara (talk • contribs) 00:46, 29 November 2006 (UTC)


 * Yes, we seem to be on the same page; I would be glad to help with this, especially with the science-related articles, time permitting. We just need a bigger team.  If we can get at least three core administrators, to help with the potential revert wars erupting between seasoned page editors connected to those pages, then I can scavenger up more regular editors to join the team who also like to see smaller articles.  For now, I added your name here. I'll wait till the group gets up to about 10 people.  --Sadi Carnot 01:16, 29 November 2006 (UTC)
 * I'd hope we wouldn't need to push decisions down anyone's throat, and that administrators will support what is reasonable without being associated with the project. - Samsara (talk • contribs) 02:03, 29 November 2006 (UTC)


 * That would be a worst case scenario. Ideally, if an article “breakup team” existed, then the mere placement of a "talk page notice" would be enough to compel the regulars to break up the page on their own.  When one works on a page for more than a month, then article beer goggles tend to develop, wherein the page seems perfectly fine no matter how long it gets.  Presently, the “32 kb warning” tag that pops up on long pages is completely useless, because editors will unconsciously justify their “unique” long pages for so and so reason, and argumentitively attack anyone who questions them about page length. ---Sadi Carnot 02:22, 29 November 2006 (UTC)
 * My problem still exists. A breakup committee should not have the power to decree that a page will be broken up by the involved editors, or someone else will come and do it for them, with admins preventing any reversion of changes. With the exception of the fact that they are openly acknowledging that they have that power, that is called a cabal, and it is bad. -Amarkov blahedits 03:15, 29 November 2006 (UTC)

Well what alternatives do you recommend? Should we let bloated 100 kb page articles linger around for months or years on end because of a few hard-minded editors; while, in the mean time, hundreds of thousands of readers get turned off and give up reading or better yet can’t load the page in the first place because they have dialup or are using a Blackberry, etc.? I don’t see what harm can come from this. We open up some new pages, cut and paste, everyone does some cleanup work, and than instead of having one 100 kb page, we now have, for example, three 33 kb pages. The process takes a day or two. It’s not that complicated. But a project team is needed for a “pressure-effect” and administrators may be needed to give user warnings to seasoned editors. In the end, everyone is happy. I am certainly open for other ideas? --Sadi Carnot 03:43, 29 November 2006 (UTC)
 * Um... The current method of establishing consensus before changing anything seems to work just fine. Obviously, the "few die-hard editors" think there is a good reason why their page should be 100 kb long. Why is this committee assumed to know better than the editors who have actually worked on the article? -Amarkov blahedits 04:25, 29 November 2006 (UTC)
 * On another note, this discussion should be somewhere else, but I don't know where would make sense. -Amarkov blahedits 04:36, 29 November 2006 (UTC)
 * Copied from ANI.-Amarkov blahedits 04:46, 29 November 2006 (UTC)
 * Note:Above discussion copied from WP:ANI


 * I don't see a serious problem with long articles per se. The technical reasons have all but gone - I know of no browser that cannot handle long pages. Loading times via dialup are problematic, but that cuts both ways. In much of the world, phone access is metered, and I'd rather download a few longish articles and go offline for reading than downloading a short one and then notice that I have to go online again and again to get sub-articles. Lack of structure and fluff are problems, but not necessarily connected to article length alone. Calling all long articles "bloated" is wrong - some are long for a reason. Breaking up an article for no other reason but length is a bad idea. If there are no identifiable subtopics that can be summarized and off-loaded, then the topic may justy be inherently rich and complex. --Stephan Schulz 12:29, 29 November 2006 (UTC)


 * Breaking up articles because they are long is sometimes a good idea, since people don't have infinite attention spans. The problem is more that the proposal means that once an article gets to a certain size, it MUST be split up, or a big scary committee will come and break it up, with no possibility for dispute, as there will be admins to prevent reversion. -Amarkov blahedits 02:59, 1 December 2006 (UTC)

Essentially, the long and the short of the problem is this:


 * "Wikipedia is now a top-12 website in the world with almost two million articles. That people read these articles is obvious.  How much of each article the average person reads, however, is not so obvious.  The average person stops reading, of course, when his or her reading tension span is broken.  In this direction, Wikipedia articles are now only broken up when more than 50% of “editors” feel they are too long.  The real question is, at what point do more than 50% of “readers” feel an article is too long?  Hence, the famous motto “ignore all rules”, with respect to article size, is presently favored towards the editors rather than the readers."

With more than one-thousand administrators at Wikipedia, I don’t see why a non-optimal situation like this should exist? --Sadi Carnot 09:52, 4 December 2006 (UTC)

64kb extra-long article proposals
I would also like to recommend implementing the use of “long-page warning” banners unique to each category of long page, e.g. 32-63kb (extra-long), 64-96kb (super-long), etc. Presently, the only warning pages have is the following, which pops up for pages longer than 32kb when users click on the edit button:

Now, for articles in the range of 32-64kb, I will admit, there is room for some gray area as to article. For articles more than 64kb, i.e. twice the recommended maximal size, however, I would recommend that we code a page such that “readers” (not editors), in a quick and easy manner, can cast their vote as to whether or not they feel an article is to long, such as:

In this manner we can collect data as to which articles readers feel are getting too long. Editors, conversely, usually have a thorough knowledge of the topics they edit and may never potentially feel that an article is too long. Please comment. Thanks: --Sadi Carnot 00:42, 5 December 2006 (UTC)

Archiving
I've re-archived this talk page because 1) the old page included a dead/inaccurate link to the first talk page archive, and 2) part of the talk page archive was moved to another article page, which is at MfD. Also, an unsigned quote had been added to the top of the page, which is now archived. Here is the talk page version before I archived. Sandy (Talk) 10:46, 21 December 2006 (UTC)

Current status?
It's unclear whether there is still any discussion on this topic at all, or where to contribute if there is. The archived pages suggest severe disagreement with suggested 'enforcement' of guidelines, with apparent consensus to "delete" the entire matter.

My interest has to do with the 'warning' that we receive when working on the Contract_bridge_glossary, that piece being about 153 KB in its socks, as of today. The glossary is a key feature of the Contract bridge Project, and a natural starting place for someone looking into the subject, or simply looking for a reference source. It also operates, to a degree, as a table of contents to the project's work. And it relies heavily on internal linking, by (if I may) definition.

Over the past week I've re-written the in-page links of the glossary, so as to link directly to the terms in question, rather than to alphabetical section heads (e.g., 'ruff' now links to 'Ruff' rather than to 'R'). I've also indicated which links (bold-faced) lead to new pages, whole articles devoted to specific terms and topics, as a notice to dialup viewers.

In practice, this makes the page 'smaller', easier for viewers to work with by far.

As for the subject here—size of pages, especially with regard to browser and internet connection limitations—breaking up the glossary would in practice make the page 'larger', since proportionally more of the links would require loading separate pages, and these when loaded would interact more slowly than by this method.

The question of size here, I think, should focus on the writing, its clarity and conciseness, with clear links to more expanded discussions. If well done, it can result in a form of 'browsing' that loses track of time, or makes the most use of it, as preferred. FutharkRed 05:08, 6 January 2007 (UTC)
 * It would be helpful if you would provide a link to the article you're discussing. Sandy Georgia  (Talk) 05:21, 6 January 2007 (UTC)


 * Sorry, I do get carried away. Obviously, I should "focus on the writing, its clarity and conciseness", as per my own advice! And obviously, after a while, people lose track of where I started, which was with the Contract_bridge_glossary article way back in the second paragraph. Should I just mention it, as above, or actually link to it, as Contract bridge glossary?


 * As a comparable matter, The Bridge World online glossary, much more extensive (so far), is split into 24 separate alphabetical pages, with no links at all between definitions, which seriously limits its usability/usefulness/usage--all three. They are considering a change. FutharkRed 09:39, 12 January 2007 (UTC)


 * A glossary is not intended to be read from start to end. Thus there are no stylistic/readability concerns for the above example. My only concern would be download time for those with slow modems and cell phone browsers. --mav 23:24, 2 February 2007 (UTC)

32KB page size limitations
The article reads that by June 2006, Firefox is the only commonly used browser which cannot handle the bug with pages with more than 32KB of content. At the moment, I don't seem to have such a problem with Firefox (version 2.0) so I think that statement can be removed. Please reconfirm me about this prior to editing the text. huji— TALK 14:26, 29 January 2007 (UTC)
 * According to the fact that the article itself ruled out the mid-2006 bug with Firefox with Google Toolbar, I removed the statement in question. huji— TALK 18:28, 29 January 2007 (UTC)