Wikipedia talk:Article size/Archive 2
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 | Archive 2 | Archive 3 | Archive 4 | Archive 5 |
"Attention Span"
Unless someone has good evidence that the "attention span" of the "average adult" has an "upper limit" of 20 minutes, that section needs to go. I added [citation needed] quite some time ago, but no further evidence has been forthcoming. In fact, the "attention span" of even the same individual (much less the entire population) varies widely, depending on that person's interest in the subject (as well as external factors, such as fatigue). If an article is interesting and well-presented, it could hold your attention for hours. If not, your attention span could be only a few seconds. Trying to come up with a single metric for "attention span", given the wide variance in Wikipedia users, topics, and article quality, is probably not a good way to proceed.
130.126.165.236 22:54, 19 February 2007 (UTC)
- Good work, I thought the exact same thing when I read it sounded like amateur psychology or some kind of new theory someone was trying out here on WIkipedia. Quadzilla99 06:51, 27 February 2007 (UTC)
- Okay, it's gone. If someone actually comes up with a credible cite, put it back -- but I don't think that's going to happen. —The preceding unsigned comment was added by 130.126.165.236 (talk) 19:21, 27 February 2007 (UTC).
- Glad it's gone, never liked it. —Doug Bell talk 19:29, 27 February 2007 (UTC)
- Doesn't matter if a cite comes up, honestly. This isn't an article, and attention span is not the best reason to keep articles down to a reasonable size. Mangojuicetalk 20:05, 27 February 2007 (UTC)
A cite is easy to find if you just search for it. Search google for "lecture attention span" and you will find articles like this that describe studies of attention patterns during lecture. Whether wiki articles are similar enough to lecture for the results to carry over is disputable, and whether attention span is relevant to articles is disputable, but there is a kernel of honest research available if you look for it. CMummert · talk 20:09, 27 February 2007 (UTC)
- If you can find a study done with written material on subjects who were actively seeking out the material (rather than being passively subjected to didactic lecture material chosen by someone else) please put it in. I will note that the Lord of the Rings movies seemed to have no trouble maintaining audience engagement, despite being 3 hours+ in length. Besides, there's no law that says you have to read an entire article in one sitting. That's why we (in theory) have sections, right? Wikipedia is a reference source, not a novel. 130.126.165.236 21:59, 27 February 2007 (UTC)
- I would think that that part in the guideline where it says
- "Readers may tire of reading a page much longer than about 6,000 to 10,000 words, which roughly corresponds to 30 to 50 KB of readable prose."
- should be removed for the same reason. The part that singles out science articles is also unjustified - wouldn't complicated philosophy, or arcane history, be equally difficult to read? CMummert · talk 23:18, 27 February 2007 (UTC)
- I edited that section to remove the specific reference to science articles. I thought that the old wording was choppy and not in a logical order, so I reordered the section and divided the material into paragraphs more logically. I believe that there is no change in the content or intention of the new wording, just a reordering to make the guideline easier to read. CMummert · talk 15:43, 28 February 2007 (UTC)
- I would think that that part in the guideline where it says
Biographies - exception to the rule?
I've noticed several biographies that are 100k
I don't really see how one could justify splitting up a biography into smaller portions, as long as the article is all biography. Some people have interesting and varied lives.
If one accepts the arguments above that "attention span" is not really applicable to the wikipedia (and I think the arguments are well made), and that dial-up, as a reason for short articles, is becoming less of a concern, might the wikipedia not at least make an exception to the rule for well-written biographies?
Andysoh 00:16, 4 March 2007 (UTC)
- This isn't really a hard and fast rule. It's just a guideline which is probably a good idea, but often isn't. If a long article is good for a topic, then it's good. --Gwern (contribs) 07:11 4 March 2007 (GMT)
- Thanks Gwern! When I read the guidelines I found them just a little worrying, hence my slightly defensive comment here... Perhaps someone should add "If a long article is good for a topic, then it's good" into the guidelines for a bit of positive support, or, perhaps as I suggest, expand the 'occasional exceptions' to the guidelines (e.g. add biographies) with any necessary conditions you think important, in some way to make the contributer less worried?
Not to worry, just a thought. Andysoh 21:47, 4 March 2007 (UTC)
How do you find the readable page size?
This page contains the advice: "To quickly estimate readable prose size, click on the printable version of the page, select all, copy, paste into an edit window, delete remaining items not counted in readable prose, and hit preview to see the page size warning."
But where? Where is the page size warning?
qp10qp 05:13, 8 March 2007 (UTC)
- The page size warning will be above the edit window, but only if the raw text exceeds 30k. See User:Dr pda/prosesize.js for a javascript tool. Instructions are on the talk page. Gimmetrow 00:40, 26 March 2007 (UTC)
- I followed the instructions found in reference 2, but no page length warning appears on the preview page. Perhaps some Wiki code needs to be not deleted for it to function? I had to copy and paste it to an email draft to get the Kb size. The article for psychokinesis gives a erroneous warning for 54Kb, but the body of the article (as I write this) is actually only 13 Kb long. So this warning system is not perfect. 5Q5 14:26, 19 May 2007 (UTC)
I took a look at the article psychokinesis, and copied the source code to a text file on my computer. The number of bytes was 56,262. Seems correct to me. --Mr. PIM 02:31, 20 May 2007 (UTC)
- The source code? Try following the instructions given below on this Project page, which is what I flagged as not working. The source code is full of just that: code. Explain why the body of the article quantum physics (41Kb) is at least three times longer than psychokinesis (54Kb)! the Quantum physics article has a small References section, while the one at PK is very lengthy. I think Wiki is miscalculating the extra material in the References on the PK page because it is just outside the {{ }} template tags but still inside the <ref> </ref> tags. The instructions below refer to pasting the material from the Printable version into the Wiki editor and then hitting Preview. When I do that, all I get is the Preview warning, but no length notice above it. Something in these instructions need to be tweaked. Anybody else experiencing this? 5Q5 19:05, 21 May 2007 (UTC)
- The MediaWiki message is based on the size of the code, not on the size of the readable prose. —Centrx→talk • 00:12, 22 May 2007 (UTC)
- Doesn't anyone see the contradiction? The page size warning that appears during a full page edit click refers to the entire source code. But the Project article here says that only the readable prose should be considered for when it's time to split up pages. People can erroneously believe they should begin splitting up an article when the readable prose; i.e., the main body of the article, is not that long because they are considering the source code total, most of which will not print out. This is what is happening at the psychokinesis article, of which I wrote over 95%. It is giving out a consider breaking up warning, even though once all the references and other material are excluded, the readable prose is moderate in size. I just want to point this contradiction out. Being the primary author of the PK page, I will have to deal with people chopping it up when it's far from time considering the length of other articles. Thanks. 5Q5 13:32, 22 May 2007 (UTC)
- That's not a contradiction, that's an ambiguity. Many articles exceed the size at which the too-long message is shown, and it is generally understood that is okay. If you want, the size at which the message is shown can be increased. —Centrx→talk • 02:35, 26 May 2007 (UTC)
Who reads a 20-page article?
I've been looking at the approximate guidelines and the more I think about them, the more they seem excessively tolerant. In the light of many recent FACs with far more than 60k of prose, even about very narrow topics, it seems slightly absurd that 100k of prose in a single article should be tolerated, no matter the subject. When and why is it reasonable to provide the general readership with 20+ pages of text? I mean, most of all that material is pure nerd padding. It's intended for readers who are already interested in the topic. So what's the justification of forcing everyone to read this nitty-gritty? Why aren't we demanding of those we already know to be patient with the topic to click an extra link or two?
Mind you, these are just the reader issues. There are also really very real editorial problems in managing 60k+ of prose: more edits in certain article histories, greater difficulty in coordinating different sections of an article, greater risk for illicit or false additions to pass undetected for long periods of time, etc.
Peter Isotalo 16:10, 2 April 2007 (UTC)
- Pesonally, I don't think there's any "forcing" involved, and I don't really agree with Peter at all - I would suggest it's wrong to say that large articles are mostly "pure nerd padding".
- If I'm searching on a subject, a wikipedia article is usually well indexed and I can go directly to the exact part I want. I don't want to be sent round lots of artifically split up sections on different linked pages.
- I bet you find Wikipedia's prominence is at least in part because it is not a few lines of milch, but real substance.
- Personally I think the restrictions should continue to be gradually relaxed, perhaps plus 20k or so on each suggested level.
- Andysoh 19:47, 2 April 2007 (UTC)
- "A few lines of milch" is the encyclopedic core of a topic, and in commercial encyclopedias years of experience and a wide range general knowledge lies behind the selection of that small kernel of encyclopedic relevance. It also happens to be what most people read and in most cases all that is required. Now, the selection of what to include in the summary can be a risky affair, but it's also so much more satisfying for those who don't have the time to read a monster article that includes too much detail. Most of our most devoted editors are nerds of one kind or another, and I'm certainly no exception. I don't see any point in trying to deny this. And those who generally want to have 20-page articles are those with specific knowledge of topics, not the majority of occasional readers who are just look for a quick insight into the subject.
- So, again, why are we making guidelines that are almost guaranteed to encourage articles that cater more to the needs and demands of a minority already blessed either with previous knowledge or patience to read a lot than the majority that doesn't have the time for all those details? And I didn't see a single comments as to some of our biggest and most difficult problems: lack of flow in texts and article cohesion. These problems will only get worst if we tolerate longer and longer articles. Not to mention that the really long articles are usually long only because the editors interested in them are extremely over-represented, now that they're actually more important than other topics. The American Civil War, for example, is not more important than the Middle Ages, even though the main article is almost three times as long.
- Peter Isotalo 13:56, 3 April 2007 (UTC)
- Hi Peter - OK, no disrespect to those who can summarise a complex topic in a few lines. I'm in favour of summaries of one, two or three paragraphs or more. But I think the wikipedia is beginning to stand out above the standard commercial encyclopedias because it is getting more detailed, e.g., more useful to more people.
- I think the opening section of a larger article (that sits under the title alongside the contents listing) should be a very breif summary, covering the content of the article, as far as possible in the order it occurs in the article. This might cause some bigger articles to grow where this is not done, but I think it is good practice.
- I think that wiki lore should say that if an article is big (say 30k plus), the opening section should be a brief summary. And it should say it here, for instance.
- Perhaps if its 100K plus, wiki lore should recommend that editors consider making the second section, below the content list should be a summary or overview of a size fit for purpose.
- Can you assume that a short summary is what most people want? Do you have any surveys which show this?
- Such a survey would have to distinguish between people who are just passing through, and people who really value what they get from the article. These people, who really rate the article highly, are likely to have got just that detail from a lengthy article they wanted, that was not previously easily obtained.
- I don't think you can prejudge what the reader wants, and I'm sure it doesn't follow that only a minority want a more detailed appraisal of the subject, or that the only people who want or who read the longer stuff are people who already know the longer stuff.
- I very much appreciate that the wiki articles I looked up today, (Isaac Newton 48k, Karl Popper 34k) were suficiently detailed to provide the information I wanted, and more accessible, better indexed, with sub heads, etc., as well as linked subjects. There were no text flow problems or lack of cohesion, that you cite. I didn't read the whole article, I was able to easily go to exactly the place I wanted, and found the information I wanted, very well summarised. I haven't seen them, but I think rather than compare the Middle Ages article and the American civil war article as to which one might think is more important, ask whether the article is well written, well presented, the material is relevent, well referenced etc.
- I personally would say to the Newton article editors - its good, write more. Andysoh 19:09, 3 April 2007 (UTC)
When and why is it reasonable to provide the general readership with 20+ pages of text?
When the subject demands more than that? Are you seriously arguing that every single human endeavor can be covered adequately in 20 pages? And why on earth would you want to discourage editors and readers with "specific knowledge of topics" from sharing or expanding that knowledge? If you believe that some topics are "underrepresented" or "more important" than topics with longer articles, feel free to expand them.
This isn't a "commercial encyclopedia", and doesn't suffer from many of the limitations of such (e.g., printing costs and physical size).
P.S. the Brittanica article on the American Civil War is 31 pages long, compared to 21 pages for the Wikipedia article. So, it appears that the article isn't excessively long even by the standards of "commercial encyclopedias".
130.126.165.236 20:59, 9 April 2007 (UTC)
Problems editing a long article
I believe the section Problems editing a long article needs some revision, as the problem is not solely limited to older browsers. I have the very latest, top-of-the-line BlackBerry 8700 and find that it truncates long articles over 32k when doing an edit. The suggestion to avoid the problem by doing section edits is fine, but that solution doesn't apply for an Infobox or lead paragraph edit, nor for vandalism reverts. JGHowes talk - 02:12, 20 April 2007 (UTC)
- Add "non-standard browsers" or "browsers on embedded devices" or something. It's not a problem for PCs. —Centrx→talk • 03:07, 25 April 2007 (UTC)
Probably off-topic...
...but I'll ask it anyway: how can you determine the size of an article without using an external application? --Howard the Duck 03:02, 4 May 2007 (UTC)
- If an article is greater than 32KB a message appears above the edit window if you edit the page. If the page is less than that size, you need to copy and paste to an external program. —Centrx→talk • 20:40, 10 May 2007 (UTC)
Page size messages
Someone has changed the messages we get when we edit so that it no longer gives the page size. It was very, very useful to know the size. Does anyone know who changed it and why, or where it was discussed? SlimVirgin (talk) 02:05, 29 June 2007 (UTC)
- [1]. Rather silly. —Centrx→talk • 23:37, 5 July 2007 (UTC)
Manageability
It seems to me that there's a correlation between the degree of jumble and disorganization in WP articles and the length. So it seems that another reason to keep pages short is to facilitate effective communal editing.Ccrrccrr 02:38, 26 July 2007 (UTC)
- I personally have not seen any evidence of this. Andysoh 18:29, 26 July 2007 (UTC)
Suggest less rigid byte counts
At present, when editing FA status article 'Belgium', one sees "This page is 108 kilobytes long. It may be appropriate to split ...". But let's have a look at the origins of these > 100 KB which according to the guideline "Almost certainly should be divided up":
country infobox: | 4,440 chars. |
section titles, 'main' & 'see also' immediately underneath those titles, image links, and actual article text: |
35,322 chars. |
'see also' section | 602 chars. |
indexed footnote references (mainly for WP:V): | 32,957 chars. |
three roughly equally sized (sub)sections: 'general online references', 'bibliography', 'external links' (together): |
8,892 chars. |
bottom page v.d.e.-boxes, categories, and FA status template: |
892 chars. |
links to the article on other language WPs: | 2,089 chars. |
total: | 85,194 chars. |
This character count total of 85 KB does not match the reported 108 KB well (probably having a technical cause like spaces and/or carriage returns not being counted as characters by my editor program, I did not look into it), but I think the general picture is clear: in particlular the exceptionally large footnotes section (more than 90 differently numbered indexes to well over a hundred linked proper references). Since many index numbers are repeated within the article, and each [??] index takes about 4 bytes, the main part of the article is in fact only about 34.5 KB and the footnotes take 33.8 KB. Whereas other references and a 'See also' section can be assumed to serve as "further reading" and thus a normal part of the article, the indexed footnotes for WP:Verifiability and the links to WPs in other languages, together about 36 KB of the 85 KB total, should not be counted as length of an article (unless if purely wikitechnical problems would be caused). I think the project page should explicitly mention such (possibly also for other elements such as the collapsable v.d.e. boxes at the bottom of an article, which usually occur on articles that belong to a much standardized category like here 'countries'). At an earlier FA review, the size had not been considered a problem precisely because of that extremely lengthy footnotes section. — SomeHuman 31 Jul2007 00:35 (UTC)
- The recommendations listed on the page here are supposed to be for "readable prose", that is excluding formatting and perhaps footnotes. The software message given when editing a page is only good enough to have a simple bytecount, and oftentimes pages that do have such a large bytecount do warrant splitting up. —Centrx→talk • 03:48, 11 August 2007 (UTC)
Article size not indicative of total bytes downloaded
This user's page gives my computer a lot of trouble even though the page is reportedly only 3,365 bytes. In fact, nearly all of that is code for templates which themselves include more templates and images and scripts totaling untold amounts of data. In this particular case, there are hundreds of images associated with that page. --Mud4t 00:33, 15 September 2007 (UTC)
Calculation of article size
I followed the instructions in your footnotes and it doesn't seem to work. I couldn't see a thing about article size anywhere. Has this process changed? Fainites barley 18:12, 15 November 2007 (UTC)
Proposed Splitting
I'm gonna suggest that List of placental mammals be split into lists based on each order, with the main page retained as a pointer to all the sub articles. Is there a tempalte to formally propose this? Mbisanz (talk) 07:19, 21 November 2007 (UTC)
Shortcuts removed
The shortcuts WP:SIZE, WP:LENGTH, & WP:LIMIT show up as redlinks when they worked yesterday. Any idea what happened? Thanks. -Fnlayson (talk) 22:42, 12 December 2007 (UTC)
- Been added back. Strange.. -Fnlayson (talk) 23:43, 12 December 2007 (UTC)
- There's been some tweaking of the "WP:" pseudonamespace. It's now a namespace alias to "Wikipedia:", see Wikipedia:Village pump (technical)#WP: vs WIkipedia:. There was some slight disruption during the transition, but it seems to have been resolved now. —Remember the dot (talk) 02:05, 13 December 2007 (UTC)
Upper limit
I know the Guidline proposes an upper limit of 100Kb and mentions that pages over 400Kb might have trouble being displayed and should have arbitrary breaks. It might be nice if this was codified more into a rule like "No pages over xxxKb". Some pages like Line of succession to the British Throne and List of Statutory Instruments of the United Kingdom, 1991 are almost certainly in need of an arbitrary break. Mbisanz (talk) 03:14, 13 December 2007 (UTC)
Warning against splitting
Should we warn users against splitting a long article if there's a chance the split article will be deleted? When an article is in one piece, the overall article is governed by WP:N, but the contents are governed by WP:NNC. That means things like Lists of Characters, Episodes, etc. are generally safe if they're part of a whole article that is sourced and notable, but are likely to be AfD'd if they're separated out. I've argued that split articles should be judged for WP:V and WP:N as a whole, but it doesn't look like that's going to happen anytime soon. Torc2 (talk) 08:05, 20 December 2007 (UTC)
- Oh good, I pulled up this discussion page just to see of there was mention of this concern. Perhaps it is discussed elsewhere (if so, someone please be so kind as to link to it here), but there does seem to be a conflict of splitting due to size concerns and concerns of notability (particularly in cases of WP:FICT). Since there is no way to tightly link pages with eachother, content can be removed simply because those watching only the main page aren't automatically informed of an AFD for a forked page so they might provide argument for the page's existence. My point is, Some mention of WP:N should be made in the section explaining how/when to split an article. I'm quite confused on WP policy/guideline/consensus on the subject. -Verdatum (talk) 07:11, 27 December 2007 (UTC)
- Please continue this discussion at [[2]]UnitedStatesian (talk) 20:28, 27 December 2007 (UTC)
Central point
The central point of this article needs some work. It currently states
- Readers may tire of reading a page much longer than about 6,000 to 10,000 words, which roughly corresponds to 30 to 50 KB of readable prose. If an article is significantly longer than that, it may benefit the reader to move some sections to other articles and replace them with summaries (see Wikipedia:Summary style). One rule of thumb is to begin to split an article into smaller articles after the readable prose reaches 10 pages when printed.
The difference between 30 and 50 is large. And there is no guideline as to what significant means. If no one objects I'll try a rewrite of this section in a couple of days. Mccready (talk) 06:23, 25 January 2008 (UTC)
- The 30 to 50 range works fine for all purposes I know of; what changes did you have in mind? I think the page is fine. SandyGeorgia (Talk) 06:34, 25 January 2008 (UTC)
- OK yes 30 to 50 is ok but why not limit it to 50 except for the type of cases mentioned. This would be clearer and avoid the need for the ill-defined "significantly".Mccready (talk) 12:03, 25 January 2008 (UTC)
- Because WP:IAR always comes into play. There are scores of featured articles that have significantly more than 50KB readable prose. User:Dr pda/Featured article statistics I've opposed many of them, but they always pass. SandyGeorgia (Talk) 16:08, 25 January 2008 (UTC)
- I agree, this is a case where setting an explicit guideline detracts from the intent of a good article. It can be decided on a case-by-case basis if a split is warranted or appropriate for a range of article sizes. -Verdatum (talk) 17:50, 25 January 2008 (UTC)
- Agree entirely with SandyGeorgia and Verdatum. WP:NOTPAPER certainly applies. Additionally, I smell some WP:GAME in the air with Mccready (talk · contribs · logs · block log): he has a history of trying to use the article length guideline as an excuse for selective, biased deletions from acupuncture. Most recent example: 25 January 2008: "bold rewrite of head. article is 86kb long and this stuff is covered below.". From March '06:removes large chunks based on V RS's, argues OK because "article is too long". Some things never change. --Jim Butler(talk) 00:08, 27 January 2008 (UTC)
- OK yes 30 to 50 is ok but why not limit it to 50 except for the type of cases mentioned. This would be clearer and avoid the need for the ill-defined "significantly".Mccready (talk) 12:03, 25 January 2008 (UTC)
Slander from acupuncture Jim. And don't flatter yourself, I couldn't care less whether you resign from wikipedia then come back a couple of days latter. Whether it's you or someone else grasping for straws to defend an emotional irrational commitment to whacky altmed makes no difference to me. "Ignore all rules" is obviously tongue in cheek, not an excuse for an article in the order of 100kb. Mccready (talk) 08:18, 28 January 2008 (UTC)
NPOV dispute
What if there's a dispute about how to treat a topic neutrally?
- (current) It may also violate the neutral point of view policy to create a new article specifically to contain information that consensus has rejected from the main article.
In some cases, "editorial consensus" has suppressed one side of a controversy. This, itself, is a violation of NPOV. It can happen when a preponderance of contributors to a controversial article are supporters of one of the sides in the controversy. By violating WP:OWN they can create a "tag team" which operates to eliminate any positive mention of the opposing POV.
I think the above sentence should read:
- (proposed) It can help defuse an NPOV conflict by creating a new article specifically to contain information which is too controversial for the main article.
Note that the arbcom has specifically ruled that "removing well-referenced information" on the grounds that it "advances a POV" is against policy [3]
Some people disagree with that. They disagree with the idea that Wikipedia should include "all significant points of view regarding any subject on which there is division of opinion."
But it would be odd indeed to enshrine this objection to policy in a guideline! --Uncle Ed (talk) 13:23, 16 February 2008 (UTC)
- I think this topic should be wordsmithed elsewhere, not in this relatively obscure guideline. SandyGeorgia (Talk) 16:18, 16 February 2008 (UTC)
Proposed change and factual correction: Large articles DO confused modern browsers
I propose the following changes:
From:
The best permanent solution is to simply upgrade to a more modern web browser, if possible. No major modern web browsers have this problem on their recent versions...
to
The best permanent solution is to simply upgrade to a more modern web browser, if possible. Except for the very longest articles such as older versions of Wikipedia:List of missing journals, which once weighed in at over 400,000 bytes, no major modern web browsers have this problem on their recent versions...
and From:
If you find a section too long to edit correctly and safely, you can post a request for assistance on the Village Pump. Follow the "post" link for the assistance section, which will allow you to post a new comment without editing any existing text. Answering your request may take from an hour to a week, depending on the response of your fellow volunteer editors.
To:
If you find a section too long to edit correctly and safely, you can post a request for assistance on the Village Pump. Follow the "post" link for the assistance section, which will allow you to post a new comment without editing any existing text. Answering your request may take from an hour to a week, depending on the response of your fellow volunteer editors. For extremely long articles that cannot be edited or even viewed in all modern web browsers, you can try splitting the article, as was done with Wikipedia:List of missing journals. You may need to turn off certain features in your web browser, use a different modern web browser, use an older web browser, or using a computer with more memory may allow you to edit and split the article. Before splitting articles discuss it on the article's talk page.
Background: Last year I split Wikipedia:List of missing journals for purely technical reasons. I wound up using a 4.x version of NetScape to make the initial split then I was able to use a modern browser to clean it up. Earlier today I verified that my very-modern web browser still chokes on the old long versions.
The above changes are meant to be consistent with Wikipedia:Article size#Technical issues.
Comments? Objections? davidwr/(talk)/(contribs)/(e-mail) 17:11, 3 March 2008 (UTC) updated davidwr/(talk)/(contribs)/(e-mail) 17:15, 3 March 2008 (UTC)
- What about mobile browsers? —Torc. (Talk.) 19:04, 3 March 2008 (UTC)
- Got numbers? If any major browser hangs on articles above a certain size, we should consider a "hard" cap at that size or below. If it's just inconvenient then it should be a style decision but not a make-or-break issue. By the way, large-number-of-pixels pictures are probably a bigger problem for users looking at tiny cellphone screens than an article that's 350KB long. I expect that the "List of Missing Journals" article would be rendered as well as a typical 30KB article in a mobile browser. davidwr/(talk)/(contribs)/(e-mail) 23:33, 3 March 2008 (UTC)
- No, actually it was a Real Question™ - I have no idea how much data mobile browsers can handle on a single page. I only have an old Sidekick II, and I doubt anybody would use that for a yardstick on how much the average mobile browser can handle. Mainly I just thought I'd bring up the topic because I don't see it mentioned, and I know plenty of people who use their mobiles for looking up answers here. —Torc. (Talk.) 00:18, 4 March 2008 (UTC)
- That's really several questions in one. How much data can they hold in a single web page, how much html can they handle on a single web page, and how big a table or whatever other constructs wikipedia uses can they handle? I'm not sure why 400KB of WikiCode breaks modern browsers, but most modern browsers CAN handle simple web pages in the multi-megabyte range. —Preceding unsigned comment added by Davidwr (talk • contribs) 15:38, 4 March 2008 (UTC)
- No, actually it was a Real Question™ - I have no idea how much data mobile browsers can handle on a single page. I only have an old Sidekick II, and I doubt anybody would use that for a yardstick on how much the average mobile browser can handle. Mainly I just thought I'd bring up the topic because I don't see it mentioned, and I know plenty of people who use their mobiles for looking up answers here. —Torc. (Talk.) 00:18, 4 March 2008 (UTC)
- Got numbers? If any major browser hangs on articles above a certain size, we should consider a "hard" cap at that size or below. If it's just inconvenient then it should be a style decision but not a make-or-break issue. By the way, large-number-of-pixels pictures are probably a bigger problem for users looking at tiny cellphone screens than an article that's 350KB long. I expect that the "List of Missing Journals" article would be rendered as well as a typical 30KB article in a mobile browser. davidwr/(talk)/(contribs)/(e-mail) 23:33, 3 March 2008 (UTC)
True length of article text?
The guideline states we can look to MediaWiki:longpagewarning for the size of a long article. From observation, it measures something other than the raw bytes of data the text constitutes (compare its warning to a version of the article's script saved as a text file). However I have noticed it considers citation templates as raw text. Heavily referenced articles will be rated longer than they are Ninja Gaiden (2004 video game) yields a 73k warning; but if we removed all references (i.e. <ref>[...]</ref>), the unreferenced article yields a 40kb warning. It is not the displayed reference list which contributes to the size (removing the {{reflist|3}} without removing all references still yields 73k). Since the guideline is supposed to be for "readable prose", is there a way to get the script performing the longpagewarning to exclude all references in its calculation? Jappalang (talk) 01:45, 10 March 2008 (UTC)
- The script also considers commented out statements (<!-- ... -->) as part of the text. Jappalang (talk) 13:59, 15 March 2008 (UTC)
- The guideline advising "use a modern browser" and only "readable prose" matters is flat out wrong. There are hundreds of millions of English speaking web users and many are still on dial-up and are using older computers. The most important count actually is just a byte count of the text in the edit window when you hit the "edit this page" tab. I couldn't figure out why the article Barack Obama was hanging up my web browser (and I am using the latest update to Firefox by the way), until I looked at the source (I can't edit it anyway because it is protected), and found that it was 141 kB - which you can see from the byte count in the history[4]. The article has recently been trimmed about 13 kB, and loads a little bit quicker, but still hangs the browser for a looong time. I would like to see a hard limit of 64 kB in place of the hard limit of 32 kB. Over half of the featured articles are less than 32 kB. There is no reason for just eliminating the limit completely, as if "well theoretically someone might be able to edit it". 199.125.109.28 (talk) 16:35, 20 March 2008 (UTC)
- By the way, talk pages can be twice as long - up to 128 kB. For example the much longer Talk:Barack Obama does not hang up the browser. Probably because of the lack of images. The full byte count transferred to the browser for the article is something like 484 k[5] 199.125.109.28 (talk) 16:45, 20 March 2008 (UTC)
- In the "Technical issues" section, is it possible that "over 400 kB" refers to total bytes transferred, not text bytes? 199.125.109.28 (talk) 17:07, 20 March 2008 (UTC)
- I was the one who added (diff) the 400KB info last year. The 400KB refers to the size reported by Wikipedia when you edit an article. This came about because editing List of missing journals (hist) other than one section at a time was nearly impossible before I split it. davidwr/(talk)/(contribs)/(e-mail) 18:36, 20 March 2008 (UTC)
- That's a text only article. Have you done any tests to see if it refers to total byte count or not? Barack Obama is 480kB with the photos, and locks up Firefox, which seems oddly suspicious considering that it could be only because it is over 400 k. Seriously, with the attention span people have on the internet, do you really expect anyone to actually read any article that is over 10k? Wikipedia really needs to go back to a more reasonable guideline. 199.125.109.28 (talk) 22:40, 20 March 2008 (UTC)
- Wikipedia needs at least two guidelines: A guideline for people, a hard and fast rule for typical desktop computers with modern browsers, and possibly another guideline for cell phones, older browsers, and other special cases. I didn't have any problem editing and previewing changes to Barack Obama using a modern Firefox. Editing Barack Obama gets a warning This page is 126 kilobytes long. It may be appropriate to split this article into smaller, more specific articles. See Wikipedia:Article size. When I save the wikicode as plain-ascii, it shows up as 127KB. When I save it as Unicode, it's twice that. By the way, with the new edit-top-section setting you can set in user preferences, big articles are no longer as harmful to logged-in users as they once were. davidwr/(talk)/(contribs)/(e-mail) 02:07, 21 March 2008 (UTC)
- That's a text only article. Have you done any tests to see if it refers to total byte count or not? Barack Obama is 480kB with the photos, and locks up Firefox, which seems oddly suspicious considering that it could be only because it is over 400 k. Seriously, with the attention span people have on the internet, do you really expect anyone to actually read any article that is over 10k? Wikipedia really needs to go back to a more reasonable guideline. 199.125.109.28 (talk) 22:40, 20 March 2008 (UTC)
- I was the one who added (diff) the 400KB info last year. The 400KB refers to the size reported by Wikipedia when you edit an article. This came about because editing List of missing journals (hist) other than one section at a time was nearly impossible before I split it. davidwr/(talk)/(contribs)/(e-mail) 18:36, 20 March 2008 (UTC)
- In the "Technical issues" section, is it possible that "over 400 kB" refers to total bytes transferred, not text bytes? 199.125.109.28 (talk) 17:07, 20 March 2008 (UTC)
I wasn't trying to edit the Barack Obama article, I was trying to read it, and it hung up my browser and I had to reboot the computer. I don't have any trouble editing big articles because I just click on any section and change the URL to section=0, which I think is the same thing that edit-top does. That doesn't always work, as sometimes you want to move something from the first section to a later section and want to do it in one edit, or just want to make one edit to the entire article, for example if you are doing a spell-check. So far the Barack Obama article is the biggest article I have ever tried to read, although of the "top 10 biggest feature articles"[6] only Bob Dylan (136k) gave me the same trouble, although it is true that none of the rest are as big as the Barack Obama article. By the way neither the Barack Obama article (127k) nor the similarly sized Byantine Empire (127k) now hang up the browser (but they still take an awful long time to load). But obviously a lot stronger discouragement has to be indicated in WP:SIZE to large (over 60kB articles) as the response I got at Talk:Barack Obama was half a dozen editors chiming in that the article was within limits allowed by WP:SIZE. Less emphasis should be placed on "readable text" as well, as that is not the number that is most readily available. To get "readable text" for Barack Obama, for example one is enjoined to simply cut and paste the text into a text window and then remove the links, see also, reference and footnote sections, and lists/tables. Ok, for me, for the Barack Obama article this means cutting and pasting into wordpad (notepad won't hold it), manually deleting all 226 references (now down to 187) ([1], [2], [3] etc.) plus who knows how many duplicate references ([8], [8] etc.) , then saving the article as a text file and looking to see how big the file is. I didn't do that because it would have probably taken at least an hour and just wasn't of any interest. The byte count (127k) was close enough for me to tell me that it was way way way too big. WP:SIZE should mention "readable text" as one reason for having a size criteria but the entire rest of the article should base it's criteria and counts only on the byte count that one sees when they click the "edit this page" tab, as it is the only size that is readily accessible, and is also visible in the history, except that the count is in bytes there instead of kB (off by 1.024:1). A section should also be added about image size, and number of images, because I really think that a page with a lot of photos is just as problematic as a page with a lot of text. Setting a limit that barely doesn't crash a browser is absurd. Set a limit to half that. Over half of the feature articles are less than 32 kB. Why allow any article to be over twice that big? 199.125.109.75 (talk) 03:18, 21 March 2008 (UTC)
- What web browser do you use and how much spare RAM do you have? If you are already starting to use your memory swap-file or close to it before you open the page, it could crater some browsers. davidwr/(talk)/(contribs)/(e-mail) 07:59, 21 March 2008 (UTC)
- It's not a very modern computer. It is running 98SE, has 480 MB Ram, and performance currently says "System resources: 15% free". I'm using the latest version of Firefox (2.0.0.12). I currently have three programs running, two Firefox windows, with a total of 17 tabs open. Closing everything has little impact on opening large articles. I'm on a slow dial-up that seems to sometimes slow to a few bytes per minute. 199.125.109.28 (talk) 01:05, 22 March 2008 (UTC)
Proposal
Mention "readable prose" only in the Readability section. Change A rule of thumb
From:
Some useful rules of thumb for splitting articles, and combining small pages:
Readable prose size | What to do |
> 100 KB | Almost certainly should be divided up |
> 60 KB | Probably should be divided (although the scope of a topic can sometimes justify the added reading time) |
> 40 KB | May eventually need to be divided (likelihood goes up with size) |
< 30 KB | Length alone does not justify division |
< 1 KB | If an article or list has remained this size for over a couple of months, consider combining it with a related page. Alternatively, why not fix it by adding more info? See Wikipedia:Stub. If it's an important article that's just too short, put it under Article Creation and Improvement Drive, a project to improve stubs or nonexistent articles. |
To:
Some useful rules of thumb for splitting articles, and combining small pages:
Edit byte count | What to do |
> 100 KB | Almost certainly should be divided up |
> 60 KB | Probably should be divided (although the scope of a topic can sometimes justify the added reading time) |
> 40 KB | May eventually need to be divided (likelihood goes up with size) |
< 30 KB | Length alone does not justify division |
< 1 KB | If an article or list has remained this size for over a couple of months, consider combining it with a related page. Alternatively, why not fix it by adding more info? See Wikipedia:Stub. If it's an important article that's just too short, put it under Article Creation and Improvement Drive, a project to improve stubs or nonexistent articles. |
Note that the only change is to refer to the more important "edit byte count" instead of the less important and very difficult to obtain "readable prose size". 199.125.109.100 (talk) 21:56, 27 March 2008 (UTC)
Survey
- I'm going to have to oppose this proposed change for now. Using total byte size as the indicator for cut off is counter-productive to having articles that are well-referenced, particularly those that rely upon the bloat inducing {{cite}} templates, like the Barack Obama, Hillary Rodham Clinton, and John McCain articles do. While the goal is a lofty one (improving access to larger articles for those with older PCs), the result of this proposal would be a decrease in quality on articles that are well referenced and force hundreds of hours of work upon editors of articles where the only issue is that they chose to use the cite templates instead of manually formatting their citations. Now, if someone had a script that could do the conversion, then that's a different story, but as of right now, I haven't heard of one. The templates are there for a reason and the hundreds of thousands of times that they are used in articles is a pretty good indicator of their popularity. --Bobblehead (rants) 22:16, 27 March 2008 (UTC)
- Don't use bloat inducing templates in bloated articles. Why make an already difficult problem impossible? Use simpler <ref>[http://www Subject]</ref> style instead. Half of the United States is still on dial up. What does that tell you about the rest of the world? Why write an encyclopedia in a manner that makes it impossible to use? And as to hundreds of hours? There are only a few articles that are horrendous and they got that way because of a stupid decision to totally eliminate the 32 kb limit instead of to expand it, plus an equally stupid decision to focus on readable prose instead of byte count. Using summary style it is trivial to fix each of the offending articles without losing any content. Make the change suggested and let the minions get to work and you will create a usable encyclopedia instead of the horrendous nightmare that currently exists. With the attention span that people have on the internet, you know that no one is reading more than a few paragraphs anyway. 199.125.109.100 (talk) 02:34, 28 March 2008 (UTC)
- Hey. If you can convince someone to change WP:CITE#FULL, we could go with a simple link to the source rather than adding a full cite, but until then, we use full citations. The bloaty template just makes it easier to get the formatting correct. As far as there only being a few articles over 100k in total size.. There's actually almost 1,000 articles over 100k. Now I can't say that they are all because of sources, but "only a few" is a bit of an under-estimate. And using summary style to reduce the size of articles.. Umm.. What do you think is used in the John McCain, Barack Obama, and Hillary Clinton articles? Yup.. summary style. It's not the text that's the problem, it's the images and the references that's the problem. You aren't going to fix the problem with long load times by trying to hack the article text down. If you want to use full html page size as an example of why article size should be modified, you're talking about trimming the size of less than 10% of the problem in the case of the Obama article. --Bobblehead (rants) 02:47, 28 March 2008 (UTC)
- There are 2,305,594 English articles, and only 932, or 0.04% are over 100,000 bytes. That is what I would call infinitesimal, and most of those are lists. "You" may use {{cite}}, but I certainly don't, and as you can see there is a strong reason to not use it in large articles that have a lot of references. Last time I checked CITE didn't say that you had to use a template. If even says that the ISBN number for a book is optional, but it really should emphasize that it should be included, instead of saying that it is optional. I could easily cut 75% out of the three articles you mentioned and not lose any content, and still use summary style, just by using one paragraph for each summary (instead of the up to 11 that are now included - and 15 and 20 for McCain - you honestly call 20 paragraphs a summary?), and adding subarticles where there currently are long sections that do not have subarticles. 199.125.109.100 (talk) 03:55, 28 March 2008 (UTC)
- Can you point me to some articles that you've "easily" written, that use this style? Wasted Time R (talk) 04:03, 28 March 2008 (UTC)
- There are 2,305,594 English articles, and only 932, or 0.04% are over 100,000 bytes. That is what I would call infinitesimal, and most of those are lists. "You" may use {{cite}}, but I certainly don't, and as you can see there is a strong reason to not use it in large articles that have a lot of references. Last time I checked CITE didn't say that you had to use a template. If even says that the ISBN number for a book is optional, but it really should emphasize that it should be included, instead of saying that it is optional. I could easily cut 75% out of the three articles you mentioned and not lose any content, and still use summary style, just by using one paragraph for each summary (instead of the up to 11 that are now included - and 15 and 20 for McCain - you honestly call 20 paragraphs a summary?), and adding subarticles where there currently are long sections that do not have subarticles. 199.125.109.100 (talk) 03:55, 28 March 2008 (UTC)
- Hey. If you can convince someone to change WP:CITE#FULL, we could go with a simple link to the source rather than adding a full cite, but until then, we use full citations. The bloaty template just makes it easier to get the formatting correct. As far as there only being a few articles over 100k in total size.. There's actually almost 1,000 articles over 100k. Now I can't say that they are all because of sources, but "only a few" is a bit of an under-estimate. And using summary style to reduce the size of articles.. Umm.. What do you think is used in the John McCain, Barack Obama, and Hillary Clinton articles? Yup.. summary style. It's not the text that's the problem, it's the images and the references that's the problem. You aren't going to fix the problem with long load times by trying to hack the article text down. If you want to use full html page size as an example of why article size should be modified, you're talking about trimming the size of less than 10% of the problem in the case of the Obama article. --Bobblehead (rants) 02:47, 28 March 2008 (UTC)
- Don't use bloat inducing templates in bloated articles. Why make an already difficult problem impossible? Use simpler <ref>[http://www Subject]</ref> style instead. Half of the United States is still on dial up. What does that tell you about the rest of the world? Why write an encyclopedia in a manner that makes it impossible to use? And as to hundreds of hours? There are only a few articles that are horrendous and they got that way because of a stupid decision to totally eliminate the 32 kb limit instead of to expand it, plus an equally stupid decision to focus on readable prose instead of byte count. Using summary style it is trivial to fix each of the offending articles without losing any content. Make the change suggested and let the minions get to work and you will create a usable encyclopedia instead of the horrendous nightmare that currently exists. With the attention span that people have on the internet, you know that no one is reading more than a few paragraphs anyway. 199.125.109.100 (talk) 02:34, 28 March 2008 (UTC)
Actually it was trivial. But bear in mind that you have to limit yourself to only one paragraph, not ramble on for 20 paragraphs, in a summary. Just simple cut and paste, making sure that you keep all the references. 199.125.109.100 (talk) 04:23, 28 March 2008 (UTC)
- Oppose too. The less-than-32Kb featured articles are probably old ones, back from when citing requirements were much less stringent. For modern WP:BLPs of high-profile figures, especially political figures as are being mentioned here, sentence-by-sentence citing (sometimes even denser than that) is an absolute must. Any time you try to skimp on the citing, editors will challenge or try to remove material. That's just the way reality is right now. Wasted Time R (talk) 23:17, 27 March 2008 (UTC)
- I don't mind if you cite every word. Just split it up into readable sections of less that 40-60 thousand bytes. Don't make me wait for three minutes while a page is loading so that I can find out for example how old someone is, or where they went to college. 199.125.109.100 (talk) 04:03, 28 March 2008 (UTC)
199.125.109.100's suggestion at Talk:John McCain that the entire early life and military career set of material be reduced to one paragraph, and replaced by a discussion of whether he is constitutionally eligible to be president, makes me wonder about this IP's sincerity. And this IP's frequent readings and editings of Talk:Hillary Rodham Clinton and Talk:John McCain, which are 130K and 100K in size, makes me doubt the dialup woes as well. Wasted Time R (talk) 12:57, 28 March 2008 (UTC)
- I never suggested it be replaced. That is a separate discussion. Talk pages are text only and they load a whole lot easier than their associated article pages. This thread started when I went to Barack Obama to try to read it and it hung up my computer and I had to reboot. I did some research to try to find out why. At first I thought there must have been some hidden javascript or something, but I found out that the only reason it wasn't loading and was hanging up the computer was because at the time the page was about 140 kb and with images was about 480 kb. When I complained I got some lame excuses that wp:size says it's ok to make pages that are unreadable. Which is why I am here. To fix an easily fixed problem with Wikipedia that makes it unusable in its present form. Change three words and the problem will be fixed. How hard is that? Change "readable prose size" to "edit byte count" and the problem will be solved. 199.125.109.100 (talk) 14:59, 28 March 2008 (UTC)
- Because the three word change would result in certain articles being of very poor quality simply because they cover a contentious topic and have to have pretty close to one reference per sentence. The Barack Obama articles is currently getting over 500 visits per minute and you're literally the first person to complain about not being able to bring up the page. Now, there are complaints that it loads slowly, but I haven't noticed an appreciable difference in loading the Obama page than any other article of similar size. I would think that there were would have been more complaints if it was a systemic issue. --Bobblehead (rants) 15:48, 28 March 2008 (UTC)
- First there would be no change in content, so how anyone can say that quality would be affected is beyond me. Most people who view Wikipedia don't even know there are talk pages, and there have been many, many who have complained about long articles. How fast a page loads is also affected by how many images it has, so a text only page like a list of page tends to load a lot faster than an ordinary article. Most of the long pages are list of pages. 500 per minute is not a average over a 24 hour period, but I'm sure that the number is that high for some minutes out of the day. The average for February works out to 62.865 per minute. Most of those people have no interest in reading the entire article, but are only looking for one small piece of information, which they would find a lot easier if the article was segmented properly. Split it and find out what happens. Be bold. 199.125.109.122 (talk) 02:37, 29 March 2008 (UTC)
- Because the three word change would result in certain articles being of very poor quality simply because they cover a contentious topic and have to have pretty close to one reference per sentence. The Barack Obama articles is currently getting over 500 visits per minute and you're literally the first person to complain about not being able to bring up the page. Now, there are complaints that it loads slowly, but I haven't noticed an appreciable difference in loading the Obama page than any other article of similar size. I would think that there were would have been more complaints if it was a systemic issue. --Bobblehead (rants) 15:48, 28 March 2008 (UTC)
- I also oppose. Obtaining the amount of readable prose is a simple process that can be done in less than a minute. My concern is making it too easy to justify slicing up appropriately long articles (because of the nature of the material) by editors with little grasp or interest in the underlying subject. Tom (North Shoreman) (talk) 01:29, 2 April 2008 (UTC)
- Long articles are also popular topics and are quickly fixed by someone who does have a grasp and interest in the subject. Just for the fun of it randomly chop any long article inappropriately and watch to see how fast it gets fixed. You won't have to wait very long. Don't throw anything away. Keep everything but put it into subarticles. Your concern is unfounded. Only long lists that need to be sortable and therefore have to be in one article are "appropriately long articles". All others can appropriately be split up. 199.125.109.102 (talk) 01:29, 6 April 2008 (UTC)
- Take United States for example. It is the most viewed non-topical article on Wikipedia other than Sex. Currently it is 167,707 bytes and has a tag at the top saying that since it is over 100 kB some browsers may have difficulty rendering the article. Yet if you look at the article, all but the first section, Etymology, not only has subarticles but all of the subsections have subarticles, for a total of 62 subarticles. That's a lot of content. Assuming that each has 30 kB that adds up to 1.8 Mb of content about one subject. Why painfully shove 167 kB of that into the main article when you can trivially move all but a few hundred words from the main article into the subarticles and leave the main article readable? You haven't changed anything but make the information accessible. Now isn't that the point of writing an encyclopedia, to make it available? What other example of bloatware can I think of? Oh yes, I remember. Don't make wikipedia become bloatware when you can make it become usable. 199.125.109.102 (talk) 03:43, 6 April 2008 (UTC)
- You asked me on my talk page, “Do you realize what you have written?” My answer is yes. You then go on to write, “... I can just look in the edit window and see that it says "This page is 38 kilobytes long." for example which tells me everything I need to know about the page length - that it is too long.” It is this type of simplistic thinking that I am concerned about -- ignoring the content of the article and start slicing.
- As far as your example of the United States article, an appropriate use of section titles and sub-titles allows a reader to browse through the article and fasten on the information they are looking for. Do we really need to so dumb down our processes that we refuse to acknowledge that the ability to browse through written materials is both a desired and readily attainable intellectual skill? BTW, let’s keep the discussion on this page. Tom (North Shoreman) (talk) 12:20, 6 April 2008 (UTC)
- What you do need to do is limit the amount in one article to 30 kB so that the material can be found quickly. It is impossible to browse through the article to find the information one is looking for when the main article takes 2 minutes to load because it is 167,000 bytes long. Ditch everything about history for example and move it all to the subarticles other than one or two sentences. You have not lost anything other than gained the ability to find the information you are looking for. You may be on broadband and it may load quickly for you, but half of the US is on dial up and what about the folks in Australia or New Zealand that want to find out something about the US? Why make it almost impossible for them to find? You obviously just don't get it and I'm failing to find the right size 2x4 to whack you aside the head with so that you do. The only time an article needs to be over 30 kB is if it has no subarticles to slough material off into. Then you can let it grow to 40 kB edit byte length before you start thinking about splitting it, just like the guideline says (remember to think edit byte count though), and can wait until it grows to a total of 60 kB edit byte size before it probably should be split. However once the sub-articles have been created there is never ever any excuse to let the main article be over 30 kB, or 40 kB, or 60 kB if you really insist, although that would be really really stupid. As I said before, your concern is totally unfounded. 199.125.109.102 (talk) 23:21, 6 April 2008 (UTC)
- Did you know that there is a template that lists 169 "major articles" about the United States? I didn't, because I didn't have time to wait two minutes for the article to load, plus it isn't included on that page, but if it was I probably would have found the factoid I was looking for quicker by looking through that list than I would have from looking through the article. Grrrr. 199.125.109.102 (talk) 00:34, 7 April 2008 (UTC)
- Interesting - it is on the US page, but it is hidden half way up the page in the See also section so I didn't see it. 199.125.109.102 (talk) 00:40, 7 April 2008 (UTC)
- As far as your example of the United States article, an appropriate use of section titles and sub-titles allows a reader to browse through the article and fasten on the information they are looking for. Do we really need to so dumb down our processes that we refuse to acknowledge that the ability to browse through written materials is both a desired and readily attainable intellectual skill? BTW, let’s keep the discussion on this page. Tom (North Shoreman) (talk) 12:20, 6 April 2008 (UTC)
Discussion
Please note that Barack Obama while meeting size guidelines if you only count readable prose size balloons to 484kb[7] and is very difficult to read for anyone on dial up. This would be remedied by simply referring to the more important and more easily obtained edit byte count in classifying whether the article should be split. 199.125.109.100 (talk) 21:56, 27 March 2008 (UTC)
- It looks to me like this entire discussion has more to do with someone wanting to cut down Obama than with actual practice on Wikipedia articles. I Oppose the proposed change to this guideline, which does exactly what it's intended to do; hold article size to something within the average reader's attention span, which is 6 to 10,000 words, or 30 to 50KB of readable prose (while referring still to the older, technical limitations). Please take the Obama discussion over there; it's not even remotely close to being one of Wiki's longest featured articles, and has plenty of room to grow. SandyGeorgia (Talk) 16:04, 28 March 2008 (UTC)
- Nope. I'm not suggesting removing anything. I'm only suggesting making it readable by moving it into separate manageable sections. And it has nothing to do with Barack Obama. Barack Obama just happens to be the first horrendously long article I tried to read. I could talk only about Manhattan, or Vereinigte Staaten, or Bob Dylan, or Billie Jean King, or Paul Wolfowitz, or General relativity, European Union, or Byzantine Empire, none of which I have ever tried to read, but all of which are in the top 250 biggest pages on Wikipedia.[8] And you must read faster than me, unless you meant that the average attention span was from 6 words to 10,000 words. Since the average attention span on the internet is from about 20 seconds to 3 minutes, you would have to read 3,000 words per minute to read 10,000 words. If you really want to know the average attention span you can look at the server logs and see how long each person stays at one page before they move on to another page. It isn't very long. There are four reasons for splitting an article, as outlined on this project page, reader issues, editor issues, contributor issues, and technical issues, and within them a total of 11 items, and attention span is only one of them. Someone has lost sight of the forest for the trees, and it isn't me. 199.125.109.122 (talk) 02:06, 29 March 2008 (UTC)
- Wikipedia is written for people who still read. It doesn't have jazzy visual layout and its fair use doctrine means that it often doesn't have a lot of pictures. It doesn't look like a modern magazine with changing fonts and lots of colors. It's not meant to capture flighty-minded people hoppin' round the net. Consider Wikipedia a throwback, a digital version of the shelf-long encyclopedia sets people used to buy for their homes. It's for real readers, not the people you describe. Wasted Time R (talk) 02:24, 29 March 2008 (UTC)
- Actually one source emphatically states that "the average attention span of most Internet users is 8 seconds". And yes, most people do not read at all. The average person does not own even one book, unless it is a bible. I think it's great that Wikipedia is providing a repository of information, and while there actually are people who read dictionaries cover to cover, and encyclopedias the same way, most people use them as a reference source, to look up some factoid they need for something. Instead of creating long articles that take three minutes to load you can allow them to find that factoid four times quicker by splitting up that article into pieces and letting them click on the one that has the information they are looking for. Clearly some editors seem to think that the goal is to create humongous articles, but it isn't practical to do that. You make the material much more accessible to take an article that has ABCDE sections and split it into A B C D E subarticles. 199.125.109.122 (talk) 03:22, 29 March 2008 (UTC)
- Wikipedia is written for people who still read. It doesn't have jazzy visual layout and its fair use doctrine means that it often doesn't have a lot of pictures. It doesn't look like a modern magazine with changing fonts and lots of colors. It's not meant to capture flighty-minded people hoppin' round the net. Consider Wikipedia a throwback, a digital version of the shelf-long encyclopedia sets people used to buy for their homes. It's for real readers, not the people you describe. Wasted Time R (talk) 02:24, 29 March 2008 (UTC)
- Nope. I'm not suggesting removing anything. I'm only suggesting making it readable by moving it into separate manageable sections. And it has nothing to do with Barack Obama. Barack Obama just happens to be the first horrendously long article I tried to read. I could talk only about Manhattan, or Vereinigte Staaten, or Bob Dylan, or Billie Jean King, or Paul Wolfowitz, or General relativity, European Union, or Byzantine Empire, none of which I have ever tried to read, but all of which are in the top 250 biggest pages on Wikipedia.[8] And you must read faster than me, unless you meant that the average attention span was from 6 words to 10,000 words. Since the average attention span on the internet is from about 20 seconds to 3 minutes, you would have to read 3,000 words per minute to read 10,000 words. If you really want to know the average attention span you can look at the server logs and see how long each person stays at one page before they move on to another page. It isn't very long. There are four reasons for splitting an article, as outlined on this project page, reader issues, editor issues, contributor issues, and technical issues, and within them a total of 11 items, and attention span is only one of them. Someone has lost sight of the forest for the trees, and it isn't me. 199.125.109.122 (talk) 02:06, 29 March 2008 (UTC)
Here is a history of changes:
- 7 March 2003 Copy from FAQ 10/10/20/30 [9]
- 7 March 2003 Add comment that Size means byte count, not readable text [10]
- 1 June 2003 Change 30 must be divided to 32 should be divided [11]
- 1 June 2003 Add size of lists [12]
- 6 April 2005 Change 10/20 to 20/20 [13]
- 17 January 2006 Change 32 to 50 [14]
- 21 February 2006 Change 20/20/30/50 to 20/20/30/60 [15]
- 24 February 2006 Change 20/20/30/60 to 20/20/30/50 [16]
- 6 April 2006 Change 20/20 to 20 [17]
- 5 October 2006 Add "of prose" [18]
- 5 October 2006 Change to "Prose size" [19]
- 22 February 2007 Add 100k [20]
- 22 February 2007 Change 100k from "Should be divided immediately" to "Almost certainly should be divided up" [21]
- 4 March 2007 Increase 20/30/50 to 30/40/70 [22]
- 4 March 2007 Reduce 70 to 60 [23]
- 7 January 2008 Change Prose size to Article size [24]
- 3 March 2008 Change Article size to Readable prose size [25]
I would have to say that the sizes have always meant byte count and not readable prose, and that it was an error to pretend otherwise. Otherwise at the time the switch was made from byte count to prose the numbers should have been reduced by three to four times, which was not done. It isn't prose count that is the most important, it is byte count that is the most important, anyway, and since prose count is not readily accessible, it should not be used. In the beginning the only count mentioned was edit byte count, and that is still the count that most editors think of when they refer to the size of an article, because that is the count that shows up in the browser for anything over 32 kb. 199.125.109.122 (talk) 06:56, 29 March 2008 (UTC)
- It seems obvious from the above that the change to "readable prose" was made without prior consensus and should be removed. Ottava Rima (talk) 03:36, 23 April 2008 (UTC)
- I couldn't agree more. 199.125.109.100 (talk) 21:46, 28 April 2008 (UTC)
Erledigt Oakwillow (talk) 17:14, 14 May 2008 (UTC)
Support readable prose as the measure; it is no longer so much a matter of technical limitations as reader attention span. Readable prose has stood here as the measure for several years; it should continue. SandyGeorgia (Talk) 17:41, 14 May 2008 (UTC)
- The above facts suggest otherwise. Oakwillow (talk) 17:47, 14 May 2008 (UTC)
- However, it is much better to discuss things instead of just changing them, as was done in March. I agree that the first edit adding "overall" was ok, but changing "article size" was not. Oakwillow (talk) 17:54, 14 May 2008 (UTC)
- Oakwillow, you seem to be focusing on the March 2008 change.. What about the October 2006 change to "prose size" that remained the active guideline for well over a year until the January 2008 edit back to "Article size". Even if you acknowledge the 2 month period it existed as "Article size" on this article, the operating understanding of this guideline has been that it is readable prose that determines if/when an article is split up, not the article size. It's also pretty obvious based on the result of the above proposal that readable prose is the determining factor...--Bobblehead (rants) 18:05, 14 May 2008 (UTC)
- You can read all about it in the above discussion, but to summarize, there are four main issues affecting size and twelve factors, some of which relate to readable prose and some of which relate to edit byte count plus other factors. Oakwillow (talk) 18:17, 14 May 2008 (UTC)
- Oakwillow, you seem to be focusing on the March 2008 change.. What about the October 2006 change to "prose size" that remained the active guideline for well over a year until the January 2008 edit back to "Article size". Even if you acknowledge the 2 month period it existed as "Article size" on this article, the operating understanding of this guideline has been that it is readable prose that determines if/when an article is split up, not the article size. It's also pretty obvious based on the result of the above proposal that readable prose is the determining factor...--Bobblehead (rants) 18:05, 14 May 2008 (UTC)
- Also oppose change to edit byte count. Readable prose and size of the final HTML page are measures directly affecting readers. Edit byte count only affects editors, and only if they do not use the feature to edit sections. Also, edit byte count grows as the article gets more well referenced, which would mean that improving the verifiability of an article would required unrelated changes to the content to keep it within those limits. HermanHiddema (talk) 09:41, 15 May 2008 (UTC)
- Further, I would like to note that I find the actions of Oakwillow somewhat inappropriate. We had an ongoing discussion at Talk:Go (board game) about the size of that article, and when pointed to the guidelines at WP:Article size, Oakwillow made the change from "prose size" to "article size" above, despite the fact that the majority of responses had opposed that change, and the fact that prose size had been the guideline for a long time. HermanHiddema (talk) 09:41, 15 May 2008 (UTC)
- You are misreading the proposal. The proposal was to change from Readable prose to Edit byte count, which is still under discussion. It was later determined that Article size was inappropriately changed to Readable prose, and that revert was requested, unapposed, by Ottava Rima. I simply implemented it. Oakwillow (talk) 13:19, 15 May 2008 (UTC)
- Sorry, but you're wrong. It was never determined that Article size was inappropriately changed to Readable prose. Only suggested. The text was changed from "Article size" to "Readable prose size" on March 3. But before that, it was changed from "Prose size" to "Article size" on January 7, without discussion. It is the first change that is inappropriate, not the second one, which only reinstated the correct formulation of a policy that had been in effect for years. These limits have been on prose size since basically forever. Eg, see this version of exactly 3 years ago, where the lead section identifies two issues. a) technical issues, and b) considerations of readability and organization. These technical issues were with older browsers that could not edit text over 32kb in length. No current browsers have this limitation, and the ability to edit sections invalidates most "dialup speed" issues with editing long text. So what matters is b) readability and organisation. There was discussion on that over 3 years ago over here. Clearly this policy is and always has been about readable prose. HermanHiddema (talk) 14:03, 15 May 2008 (UTC)
- Actually, I believe I read every revision of the guideline. That particular version, in my opinion was clearly referring to edit byte count except for in the one occasion that it called out a number for readable prose, saying that "readers may tire of reading a page in excess of 20-30 KB of readable prose". They may also tire of reading a page that is a lot shorter than that too. Since most people use encyclopedias for reference (duh), making information more accessible is more important than writing 500 page articles that only a few have the attention span to read. Oakwillow (talk) 15:06, 16 May 2008 (UTC)
- Sorry, but you're wrong. It was never determined that Article size was inappropriately changed to Readable prose. Only suggested. The text was changed from "Article size" to "Readable prose size" on March 3. But before that, it was changed from "Prose size" to "Article size" on January 7, without discussion. It is the first change that is inappropriate, not the second one, which only reinstated the correct formulation of a policy that had been in effect for years. These limits have been on prose size since basically forever. Eg, see this version of exactly 3 years ago, where the lead section identifies two issues. a) technical issues, and b) considerations of readability and organization. These technical issues were with older browsers that could not edit text over 32kb in length. No current browsers have this limitation, and the ability to edit sections invalidates most "dialup speed" issues with editing long text. So what matters is b) readability and organisation. There was discussion on that over 3 years ago over here. Clearly this policy is and always has been about readable prose. HermanHiddema (talk) 14:03, 15 May 2008 (UTC)
- You are misreading the proposal. The proposal was to change from Readable prose to Edit byte count, which is still under discussion. It was later determined that Article size was inappropriately changed to Readable prose, and that revert was requested, unapposed, by Ottava Rima. I simply implemented it. Oakwillow (talk) 13:19, 15 May 2008 (UTC)
Please address the issue of why you claim that the March 3 edit was inappropriate, but the January 7 edit was not. HermanHiddema (talk) 17:59, 16 May 2008 (UTC)
- Article size has always been edit byte count. It is a perversion to think otherwise. However, it allows anyone who wishes to think so, and is the preferred wording for that reason. Oakwillow (talk) 18:23, 16 May 2008 (UTC)
- So the March 3 edit is inapproprioate because it conflicts with your opinion on the issue, while the one on January 7 is not inappropriate because it agrees with your opinion on the issue? HermanHiddema (talk) 18:36, 16 May 2008 (UTC)
- It conflicts with the content. My opinion is neither here nor there. This is an encyclopedia, not a blog. Opinions are not very important. Oakwillow (talk) 18:54, 16 May 2008 (UTC)
- On the contrary, it is supported by the content. HermanHiddema (talk) 22:09, 16 May 2008 (UTC)
- It conflicts with the content. My opinion is neither here nor there. This is an encyclopedia, not a blog. Opinions are not very important. Oakwillow (talk) 18:54, 16 May 2008 (UTC)
- So the March 3 edit is inapproprioate because it conflicts with your opinion on the issue, while the one on January 7 is not inappropriate because it agrees with your opinion on the issue? HermanHiddema (talk) 18:36, 16 May 2008 (UTC)