Talk:Internet Explorer box model bug

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Microsoft Windows / Computing  (Rated B-class, Low-importance)
WikiProject icon This article is within the scope of WikiProject Microsoft Windows, a collaborative effort to improve the coverage of Microsoft Windows on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
B-Class article B  This article has been rated as B-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.
Taskforce icon
This article is supported by WikiProject Computing (marked as Low-importance).
 

Fails To Mention Height[edit]

Although height is assumed to be in the same issue as width with this box model discussion, it is not discussed in the diagram that goes with the article. This might confuse newbies. —Preceding unsigned comment added by Volomike (talkcontribs) 15:14, 20 June 2009 (UTC)

From bad to worse...then, to 'amazingly stupid'[edit]

This article was bad and I (and a few others) tried to improve the content by providing readers with W3C compliant ways (such as the box-sizing attribute) to ensure their pages don't suffer from common box-model problems. Then an argument about the word "bug" ensued and since I've left the discussion the article has reverted to the Tantek HACK HACK HACK HACK HACK HACK philosophy.

In my opinion, that's extremely bad form and...well...stupid.

I'll summarize the logic of the common authors of this article:

- I don't like Microsoft's box model. - I think it's a bug. - Oh...there's a HACK which enables to me to continue in my ignorance...great!

Any reference to Tantek, in my opinion, should be removed and instead the content (which existed happily in this article for many months) about the box-sizing and -moz-box-sizing attributes should be reincarnated. —The preceding unsigned comment was added by Formation (talkcontribs) 20:55, 13 January 2007 (UTC).

Please read Wikipedia policy. Particularly the bits that focus on Wikipedia:Reliable sources and Wikipedia:Civility. AlistairMcMillan 21:22, 13 January 2007 (UTC)
The real solution is to use a DOCTYPE that forces standards compliance mode to get IE to follow the standards (as opposed to its own flawed box model). The paragraph that describes that work-around should probably come first. The paragraph describing the hack should probably come second, and it should be clarified that the hack is needed only for IE 5 and earlier. -- Schapel 23:33, 13 January 2007 (UTC)
Correction: the hack is totally unecessary in IE 5. Conditional Comments were supported by IE 5.0; therefore the more sensible way of segregating CSS for IE browsers (5.0 and newer) is through the use of conditional comments. --Regards. 17:45, 14 January 2007 (UTC)
Okay. I'm genuinely curious. If the problem with the box model can be solved easily with conditional statements, how come Tantek (a browser developer at Microsoft) felt the need to develop his box model hack? AlistairMcMillan 18:01, 14 January 2007 (UTC)
Yes, the problem can be easily resolved with Conditional Comments. And I have no idea why Tantek "developed", as you say, the hack. I think it's more likely that he stumbled upon it and, like you, he didn't have the sense of mind to investigate Conditional Comments before publishing the ridiculous hack all over the web. (p.s. Isn't it true that he is a 'former' Microsoft employee?) --Regards. 18:58, 14 January 2007 (UTC)
Tantek developed his hack while he was working at Microsoft. AlistairMcMillan 19:55, 14 January 2007 (UTC)
I'm curious, too. More to the point, however, do you have a reliable source that demonstrates how to use conditional statements as a work-around? We already have a reference that mentions using conditional statements instead of the hack, but it doesn't really explain precisely how. I don't think it's going to be particularly useful to mention that conditional statements can be used, only to leave it as an exercise for the reader to figure out how. And we can't just write the code ourselves because original research isn't allowed in Wikipedia. -- Schapel 18:25, 14 January 2007 (UTC)
Of course there are reliable sources which demonstrate how to use Conditional Comments. You should start here. Then realize that this article doesn't demonstrate the Tantek hack either; so you're already leaving it as an exercise for the reader to figure it out. Your error is that you're pointing the reader towards a terrible solution (the Tantek HACK HACK HACK HACK HACK HACK HACK) instead of a well-supported and well-documented feature of the web browser (i.e. Conditional Comments). --Regards. 18:58, 14 January 2007 (UTC)
Our article doesn't demonstrate how to use the hack, but the articles that we have linked to do. Can you point to a link that shows how to use the conditional comments to get around the box model problem? AlistairMcMillan 19:55, 14 January 2007 (UTC)
"Can you point to a link that..."  ?? I can point you towards the External Links section of this article where you'll find one called Hack-free CSS for IE.--Regards. 02:06, 15 January 2007 (UTC)
We really don't have a choice in covering the hack because WP:NPOV states All Wikipedia articles and other encyclopedic content must be written from a neutral point of view, representing views fairly, proportionately and without bias. It's a popular view that the hack is used to work around the bug, so we really do need to present it. It would be a personal bias to avoid mentioning it in the article. We could try to emphasize the use of conditional comments better, and mention the sources that recommend them over the hack. Is there a source we can cite that explains in any detail how conditional comments are used to work around the bug? If so, we could add that description to the article. -- Schapel 01:38, 15 January 2007 (UTC)
Actually, I'd say 'yes, you do have a choice' and it's clear that you've made your choice. Tantek's hack is not the popular view nor is it the recommended view; it's clearly the view of yourself and the others who have tried very hard to control this article.
"Is there a source we can cite that explains..." ?? Yes, there's already a source in the External Links section of this article called Hack-free CSS for IE. I might have added the information myself, but I fully expect it would be deleted. Everything I've contributed to this article in the past has since been completely removed and replaced by you, Alistair, and Nigel.--Regards. 02:06, 15 January 2007 (UTC)
The view of myself is to simply use standards and do away with all hacks such as Tantek's and conditional comments. Users using IE 5 and earlier should simply get a degraded browsing experience. However, changing the article to say that would be clearly POV, so I won't do it.
All the source you cite says is that conditional comments can be used to work around the bug. It gives no details as to how to do so. I might guess that I should use a DOCTYPE that forces IE 6 and higher into standards mode, and I should apply a conditional comment that applies to IE versions prior to 6, but that would screw up IE 5 for Mac, wouldn't it? Or would it? Can you find a source that step-by-step explains exactly how to use conditional comments to work around the bug? If so, we can add the description to the article. -- Schapel 02:17, 15 January 2007 (UTC)

213.249.159.244's solution[edit]

tantek's box model hack is big and clunky, IMO it's much more elegant to just do

padding: 50px;
width: 200px;
width: "300px";

the final rule is only read by IE in quirks mode as it ignores the quotes, providing they correctly wrap the value. you can also narrow the rule to be truly IE-only by prepending an underscore:

_width: "300px";

as only IE (in quirks mode) will treat "width" and "_width" as the same thing.

obviously, though, this does not validate against the w3c standards but is much easier to use in a css file as it is inline, all that is needed is a comment explaining WTF it does, final result:

div.mydiv
{
    padding: 50px;
    width: 200px;
    _width: "300px";    /* hack: IE 5 box model width */
}

i've been plugging away at explorer and have a test page showing various "inline" hacks at my website http://www.cyclomedia.co.uk/?9

213.249.159.244 14:21, 21 June 2006 (UTC)

Not a Bug?[edit]

Given that IE does not properly follow the CSS specification's bounding box model, why is this article claiming that this behavior is not a bug? Jonabbey 00:13, 28 October 2006 (UTC)

How can different implementation be a bug ? Both models (IE and W3C, aka border-box and content-box) have their own pros and cons.82.207.5.214 06:28, 4 October 2007 (UTC)
It's by design - Microsoft's own design that doesn't conform to the standard, but by design nonetheless. Unless Microsoft created the bug and now they're too proud to admit it and fix it ;) David 18:16, 3 November 2006 (UTC)
Ok, so it isn't a bug, but it is also not a "unique implementation of the World Wide Web Consortium's box-model" - it's unique, for sure, but it's not an implementation of the W3C box-model, because then the browser behaves differently than the W3C's box model requires and does so by design instead of mere mistake. So we have failure to conform and failure to even try. Which means this IE quirks box-model is an implementation of something almost, but not quite, entirely unlike the W3C box-model. Aragorn2 22:31, 15 November 2006 (UTC)
Actually, it's not Microsoft specific but IE+Windows specific since IE on the mac uses the w3c model.
The CSS Specification could not be properly followed. It seems that there's a lot of confusion about what is the "correct" box model -- the fact is that there was no "correct" implementation (until 1999!!!). MSIE added support for Cascading Style Sheets in version 3.0B2 (July 1996) - and therefore they had to implement a box model (this is a necessary aspect of any CSS implementation). The CSS 1 specification wasn't even created until December of that year and therefore, at the time of the browser's release there was no specification for Microsoft to follow. The W3C continued to make amendments to the specification and specifically to the box model in CSS1 until as late as 1999 (see Appendix F of the CSS1 Spec and you'll note that corrections/revisions to the box model continued into 1999). Through those three years, while the W3C were revising the CSS1 specification, Microsoft's web browser proceeded through two major versions and was already in version 5.0 (released in March 1999) and, obviously, by that time Microsoft had to support it's own decisions regardless of which recommendations/decisions the W3C eventually followed. And now (finally) in version 7.0, Internet Explorer supports both of the box models described by the W3C's proposed CSS3 working draft (content-box and border-box).--Regards. 19:41, 14 January 2007 (UTC)
CSS 1 wasn't "created" in Dec 1996, but published then as a final recommendation after a full W3C consultation and drafting process. Please read http://www.w3.org/Style/LieBos2e/history/ . Håkon Wium Lie and Bert Bos say that Microsoft's Thomas Reardon was participating in the W3C workshop program during 1995. They also say that IE3 did "not implement much of the box model", "IE4 [did] not fully support CSS1" and they quote the WaSP project listing "The box model" as number one under "The top 5 bugs" of "IE's Top 10 CSS Problems". --Nigelj 20:57, 14 January 2007 (UTC)
I believe you've missed the point. "created" vs. "published" -- there's no need to mince words. The point is that W3C's specification was still a working draft when Microsoft released IE 3.0B2 (which was their first browser with a CSS box model). If you were a developer working for Microsoft at that time...what box model would YOU have followed? ('NONE' is the answer because there wasn't a complete specification to describe a "correct" box model - you'd have had to forge your own path or predict the future.) And, to further illustrate the point, W3C were still making revisions to the formatting (box) model of CSS1 well into 1999. So again, if you were a developer working with Microsoft in 1999 (working on MSIE 5.0 for example) -- would have have believed that the W3C specification was a 'finished' document. ('NO' is the answer because even the W3C members were still arguing about their own defition of a box.) --Regards. 02:28, 15 January 2007 (UTC)
There's no point in discussing this until you've read the article, rather than speculating about what "you" would have done. --Nigelj 21:23, 15 January 2007 (UTC)

Re: Counterpoint[edit]

"Since the inner box must exist within the border and padding of the outer box, using the IE box model, the inner box's 50% width will take up more than 50% of the content space available to it."

After some analysis, it turns out this is not true, though by rights it certainly should be. QuirksMode (via no doctype) in IE results in:

http://www.pinkgothic.com/padding-percentage.gif

It's a screenshot of a page render of the following source code (with the '25 pixels padding' comment added manually and the pink line to denote what the counterpoint paragraph claims):

<html>
 <head>
  <title>blah
  </title>
 </head>
 <body>
  <div style="width:100%; border:1px solid #f00;">
   <div style="width:50%; height:100px; border:1px solid #00f;">
    &nbsp;
   </div>
  </div>
  <br>
  <div style="width:100%; border:1px solid #f00; padding:25px;">
   <div style="width:50%; height:100px; border:1px solid #00f;">
    &nbsp;
   </div>
  </div>
 </body>
</html>

The source code is as simple as it gets, I realise.

At any rate, it seems to be undeniable that IE (whether it should by it's own logic or not) uses 50% of the content-width that is available in the buggy box model.

I started investigating this briefly because the paragraph confused me - I first misunderstood, even; then thought, "So why haven't I ever seen that alleged behaviour? It makes sense that that is what would happen," and had to check. The result is that screenshot.

What's even more confusing: the moment you add a padding of 25px to the inner box, it is the W3C model that causes this kind of behaviour, though on purely visual levels (in its own logics, it is perfectly correct) - the width from border to border is then 50% + 25px + 25px, meaning you cannot fit another 50% box into the outer box without extending it's content width. ( http://www.pinkgothic.com/padding-percentage2.gif )

Perhaps the paragraph can be reworded along the lines of 'The expected behaviour of the IE box model is [...]' - as it is worded now, I find it not to be a very good counterpoint under scrutiny.

I hope this observation will help someone. :) -pinkgothic 16:57, 24 November 2006 (UTC)

Whether we should have "bug" in the title[edit]

Google Search for Internet Explorer box model bug excluding results from Wikipedia or Wikipedia mirrors = 621

Google Search for Internet Explorer box model excluding results from Wikipedia or Wikipedia mirrors = 201

Aside from that, an article titled "Internet Explorer box model" should be about the whole thing, not this single issue.

And please don't misunderstand our NPOV policy. The idea is not to insert our own POV. Stating that other people have a POV on a certain subject is not against our policy. AlistairMcMillan 17:21, 4 December 2006 (UTC)

  • I haven't misunderstood anything. The term 'bug' has been perpetuated by the points of view of ignoramuses who figured that it was Internet Explorer's fault that they couldn't layout their page properly in multiple browsers. Microsoft implemented the box model differently than the W3C suggested; I'll agree that's been terribly annoying but it's not a bug.
  • Your Google searches are irrelevant. YOUR LINK (which you claimed excludes Wikipedia mirrors) includes an amazing amount of this kind of garbage wherein the word "Wikipedia" has been omitted but are obvious scrapes of previous versions of this page. (So your Google links above are as annoying as your persistence that the word 'bug' actually belongs in this page title.) How many of those 639 search results at Google do you think are attributed to the misinformation promulgated by this page?
  • This article I'd say has been one of the reasons that the issue continues to be advocated as a 'bug'. If others share your opinion, it's likely because they've read this page! This article is neither accurate nor helpful to webmasters who are earnestly trying to learn about Internet Explorer's box model and how it differs from other browsers. (Hmm...what's your agenda?)
  • I agree that an article entitled "Internet Explorer box model" should be about the whole thing, not this single issue. I suggest that the content on this page would be the best start towards an article dedicated to an examination of the box model in Internet Explorer (and its various versions). So we can start now by renaming this disgraceful page to "Internet Explorer Box Model" and then concentrate on the relevant content. (Regards. 02:16, 5 December 2006 (UTC))
Two points:
  1. You do realise that everyone can read your Talk page right? That when you say "I also admit that I gave the article a decidely pro-IE slant"[1], that everyone can read that.
  2. You should probably hold off on calling other editors "ignoramusus"[2] [3], until you've learned to spell the word.
When I get the time I'm going to re-write this article in an NPOV manner under the title "Internet Explorer box model bug", removing your bias as well as everyone else's. Enjoy your "pro-IE" version while it lasts. AlistairMcMillan 03:09, 5 December 2006 (UTC)
You said: "in an NPOV manner under the title 'Internet Explorer box model bug'" -- that's an oxymoron. The article cannot be void of point of view so long as you use the word bug in the page title. The title alone (regardless of page content) FAILS to be without point of view. (Regards. 06:21, 5 December 2006 (UTC))
Again, you are misunderstanding the meaning of our NPOV rule. The idea is that we are not supposed to insert our own POV. Reporting the POV of someone else is perfectly okay. If a bunch of people say something is a "bug" then writing an article stating that a bunch of people say something is a "bug" is perfectly fine. Adding a comment that "...this page does not describe a software bug..."[4] to the article is inserting your own POV into the article and thus breaking the NPOV rule. Please read the WP:NPOV page and try to understand the rule before editing any further. AlistairMcMillan 17:45, 5 December 2006 (UTC)
As far as I can tell, the bunch of people you're referring to includes yourself and Schapel.
Oh and before I forget, we have this little policy here called Wikipedia:Civility. Perhaps you should try reading that. AlistairMcMillan 03:14, 5 December 2006 (UTC)
Just in case there's still any doubt as to whether IE's box model is a bug or not, I'll leave a quote from Web Design in a Nutshell, Third Edition, by Jennifer Niederst Robbins (page 450):

When it comes to CSS hacks, Internet Explorer 5.x on Windows (IE 5.x/Win) really started it all. This browser version has caused many problems for CSS designers, and perhaps the worst was its flawed implementation of the box model, which led to the first CSS hack. (emphasis mine)

I would say that a "flawed implementation" is also described by the term "bug". It's not just a different implementation; it's flawed, it's incorrect, it's a bug. -- Schapel 04:43, 5 December 2006 (UTC)

Response:

  • You're making a giant leap in logic to conclude that a "flawed implementation" is a "bug". We've all reached consensus long ago that Internet Explorer's implementation did not follow the W3C spec -- I suppose THAT is why Robbins called it "flawed". But, did she say "bug"? Nope.
  • And if you'd like to discuss "Civility" then know this:
    • you should look up the spelling of [ignoramuses] (plural) before implying that I've spelled it wrong.
    • and that's precisely what you're being if you insist that this page title should include the word "bug".
  1. This page title should NOT include the word "bug" for same reasons that [this page title] does not include the word "Holocaust" and [this page title] does not include the word "gay" -- because it would be ridiculous!

I'm going to move this page (again) to the new page title and -- hey I've got a unique idea -- when you can cite [verifiable proof] that the box model implementation in Internet Explorer is in fact a "bug", then you can rename the page to whatever you choose! (Regards. 06:11, 5 December 2006 (UTC))

From the Software bug article: "A software bug is an error, flaw, mistake, failure, or fault in a computer program" (emphasis mine) A flawed implementation is the very definition of a bug. It's proven. -- Schapel 13:29, 5 December 2006 (UTC)
I can't believe you're being so obtuse about this but thank you for referencing [this page] where it actually says: "a software bug is an error, flaw, mistake, failure, or fault in a computer program that prevents it from behaving as intended (e.g., producing an incorrect result)." (emphasis mine, as you say) Microsoft's implementation was intentional and its implementation is flawless. Their decision to NOT follow the W3C spec was their error, but Internet Explorer's box model works exactly as they intended it and exactly as they documented it. The browser isn't without bugs, some which relate to the box model such as the [doubled float-margin bug] and the [phantom box bug], but those are flaws in the float model and the behaviour of overflow -- not bugs in the box model. (Regards. 02:21, 6 December 2006 (UTC))
As if we needed yet another source, Sams Teach Yourself CSS in 10 Minutes (Sams Teach Yourself) shows the correct box model, as compared to Internet Explorer's box model. Because Internet Explorer's box model is different from the correct one, it is not correct, or in other words, incorrect. Behavior of a program that is incorrect is a bug. -- Schapel 13:50, 5 December 2006 (UTC)
Just in case, let's make sure we have a reference that literally says the problem is a bug. That would be CSS Hacks and Filters: Making Cascading Stylesheets Work. -- Schapel 14:17, 5 December 2006 (UTC)
Thanks again! Sam uses the word "bug" loosely to claim that somehow the entire box model is the bug. But then spends a few pages explaining exactly how the IE box model works (which is exactly as Microsoft intended it). I therefore don't accept your opinion that Sam's article is verifiable proof of your point of view. (Regards. 02:21, 6 December 2006 (UTC))

Responding to Formation's last comment. We don't need to prove that IE's implementation is a bug. We only need to prove that people call IE's implementation a bug. Which Schapel (thank you Schapel) has done. Again, if you don't understand why this is happening, then please study Wikipedia policy. WP:NPOV, WP:VERIFY, etc. AlistairMcMillan 00:20, 6 December 2006 (UTC)

Again, Sam is so far the only citation provided which actually uses the word "bug" -- and this citation alone doesn't provide verifiable proof of anything at all (except that the IE box model works exactly as Microsoft intended it!) The other quotes and references that have been provided herein have not used the word "bug" but instead have used loose terms like "flawed" and all argue that the reason it is flawed is simply that Microsoft didn't follow the W3C spec. This in itself is not proof of a flaw in the box model but arguably a flaw in Microsoft's judgement. So, you need to ask yourself these two questions:
  1. Are you and Schapel the only people who you're referring to in "We only need to prove that people call IE's implementation a bug."?? That's like saying, "Me and Schapel are right because we're right and because we said so".
  2. If others are naming this box model's behaviour a "bug", is this article the cause? Is this article perpetuating this misinformation? (uh...yes, as [you so eloquently proved with these search results at Google].
I'm not sure what in your mind would constitute "proof" that the Internet Explorer box model is indeed a bug. You seem to simply rationalize all our evidence away. Why ask for proof when you'll simply dismiss anything we could possibly come up with? Tell us, what in your mind would constitute enough evidence that the problem is actually a bug? Remember that Wikipedia's official policies are that the content of articles must be verifiable by citing reliable sources and no original research is allowed. What would constitute a reliable source, if not published books on CSS? Alternatively, can you cite a reliable source that explicitly states that IE's box model is not a bug? -- Schapel 02:36, 6 December 2006 (UTC)
Look at your own words. You're asking me to decide whether "the Internet Explorer box model is indeed a bug". The whole thing is bug, is it? The "box model" -- as a single entity -- is a bug? That's like saying "Wikipedia is a bug" or "The Eiffel Tower is a dent". I'm sure Wikipedia has bugs, but it isn't one. Likewise, the Eiffel Tower probably has a few dents in its metal construction, but the entire thing cannot be described as a "dent". The concern I have about this page title is the language as much as anything.
Your "evidence"?
I'm not asking for proof.
What problem (specifically) are you calling the "bug"? What aspect of Internet Explorer's box model is your "bug"? Nobody has yet to explain what "the bug" is exactly?
At most, some authors here have claimed that it doesn't behave like other web browsers. I understand that has bothered a lot of people, but that's not a bug. Go start a topic called Criticisms of Microsoft's Unwillingess to Implement W3C Standards if you want a venue to complain about it but I'm waiting for a good reason that "bug" should exist at all in this page title.
You said "if not published books on CSS?" But I assume you meant published BOOK (singular) -- all other references thus far have been online sources and most of those have quoted this article as their source!
"can you cite a reliable source that explicitly states that IE's box model is not a bug?" Do you want me to also find a citation that explicitly states that the sky is not red? or that the table is not sitting on the ceiling? or that coffee is not pink? —The preceding unsigned comment was added by Formation (talkcontribs) 03:08, 6 December 2006 (UTC).
Authors are not simply stating that Internet Explorer's box model is not the same as other browsers'. They are stating that it is flawed, incorrect, a bug, a problem, and it does not work as intended. We have provided ample reliable sources to verify this claim. On the other hand, you claim that Internet Explorer works exactly as intended, and is simply different from that of other browsers. If you cannot produce even one source that makes this statement, you cannot claim it to be true on Wikipedia. Stop asking us to provide evidence to support our claims, which we have, and supply some to support your claims. -- Schapel 03:50, 6 December 2006 (UTC)
Agreed. No software manufacturer can say, "We know there's a published, internationally agreed standard for this, and that the interoperability of the whole structure depends on everyone keeping to agreed standards, but we just decided to implement it differently, that's all." To claim that they did, that that's OK, and that it's their choice, is not really a supportable argument - even if they did say so in citable print. The fact that they didn't even say so anywhere we can find makes it worthy of little more than a "Some proponents claim..." footnote to the article, I'm afraid. --Nigelj 18:50, 6 December 2006 (UTC)
You're making a common mistake. The word "standard" in the real, actual English language in use today (as opposed to the jargon we invent on the Internet) means, amongst other things, "Something used as a measure for comparative evaluations"; perhaps you've heard the phrase "the standard by which"? When people talk about Internet Explorer being a "standard", that's how they are using the word. This has nothing at all to do with the W3C's specifications and "recommendations"... IE is deemed a standard simply because it's in common use. It is, of course, quite accurate to say that IE doesn't follow W3C recommendations for web standards, but that doesn't make IE -not- a standard. -/- Warren 19:09, 6 December 2006 (UTC)
Scraping the barrel a bit there, if I may say. The problems with a 'standard' like that include that it exists for the benefit of just one group of shareholders, and that it keeps changing. MS have just brought out IE7, which no doubt will introduce a whole new set of such 'standards', 'features' and bugs. That's why I used the phrase "published, internationally agreed standard" above. That wasn't a 'common mistake'.
But what we need here is a citation to a source in which Microsoft explain that they put this "non-standard" box model into IE's code on purpose, i.e. that it wasn't a bug - that they really did read the W3C spec and decided to code up something different. Until then, all we have is a blogger who likes it and thinks he has can justify it, an ex-Microsoft employee who published a workaround for it, and a bunch of published sources that actually do refer to it as a 'bug'. If that's all we have then the article is seriously skewed toward Koch's article, and needs an overhaul. --Nigelj 19:45, 6 December 2006 (UTC)

The W3C CSS3 Working Draft cited in the article says:

The property 'box-sizing' was first proposed to provide an upgrade path for certain browsers that interpreted 'width' the wrong way [...] However, newer versions of those browsers have already fixed the bug and it is not clear that these properties are really needed anymore.

(Emphasis added) --Nigelj 20:00, 6 December 2006 (UTC)

Microsoft's documentation is verifiable proof that their box model implementation was intentional and the W3C's own CSS1 Recommendation affirms that Microsoft's box measurements are standard CSS1 and therefore 'A-Okay'[edit]

Measuring Element Dimension and Location at MSDN

    • This diagram proves nothing in and of itself but will be helpful as a general reference to understand the components of a box. These components, fundamentally, include the box's size (which is comprised of its margin, border, padding, and contents), and the box's position related to its container. I do hope we can reach consensus on at least this point?
    • The diagram at this page shows detailed information for positioning and measuring a box within an HTML document. Many of these dimensions are scriptable objects only (such as offsetHeight and clientWidth) but this diagram also includes the following important style attributes:
      • margin
      • padding
      • border
    • This is then an almost complete diagram of Microsoft's box model. What is missing however from the diagram is the "width" attribute and how the three properties above relate to the width measurement (this of course is the pillar of this dispute)...so read on to the next section where I'll explain further...
    • It is also interesting to note that in Microsoft's implementation of the box model, all block level elements behave exactly as their body element. The body element you'll note (is a block-level element) and the margin, the border, the padding, and its contents all fit within the browser window.
Tangential argument: By W3C's logic the contents of the body should fit in the browser window while the padding, border, and margin should extend outside the bounds of the window and should by the same logic produce vertical and horizontal scroll. Does anybody want to say, "that's dumb"? I do.
By the W3C's CSS2 logic, my body includes only the contents of my interior anatomy but my skin and hair (the "border" perhaps?) shouldn't be measured as part of my overall width. However, (all W3C logic aside) I know that the waistband on my pants must be large enough to envelope my interior anatomy (analagous to W3C's 'contents') + the layers of my skin (analagous to the "padding" and "border" perhaps?) plus my body hair and some breathing room (analagous to the "margin" perhaps?). I know that the overall width of my body includes all of those components. Does yours?
Therefore, (and for a host of other reasons already scattered throughout this article and talk page) I'd say that the W3C and Internet Explorer got it right with CSS1, and the box model "flaw" was introduced by W3C in CSS2 and further propogated by Firefox and other browsers who were so quick to implement the latest so-called "standards". (I know that will make some folks frustrated, but this is the talk page and I think it's a relevant argument to introduce.)

'Width' in Microsoft's CSS Attributes Reference

    • This information describes how the width of a box is measured. But the most important thing to note is at the very bottom of the page where Microsoft states that "This property...is defined in Cascading Style Sheets (CSS), Level 1 (CSS1)."
    • So, the CSS Level 1 Recommendation defines the formatting model exactly as Microsoft implemented it. You'll note in their diagram that the "box width" is equal to the content + padding + border + margin. This is exactly as the box model in Internet Explorer behaves by default.
    • I believe therefore it's evident that Internet Explorer's box model does in fact follow W3C recommendation (their CSS1 recommendation) and therefore any argument suggesting that the box model doesn't adhere to any "standard" is ludicrous. Microsoft didn't make a uni-lateral decision on this; they followed the spec. Perhaps everybody's just mad that they didn't adopt CSS2 fast enough? It doesn't adhere to the CSS2 recommendation, but are we really arguing about 'which' standard should have been followed? No...from all previous arguments presented here it appears that most ignoramusus are claiming that they: (1) didn't follow the 'standard'; and (2) therefore it's flawed. Would you all like to clarify your argument and now begin saying: (1) they didn't follow the 'CSS2' Standard (specifically) and (2) therefore it's flawed.
    • I believe it's also evident that Microsoft did all of this quite intentionally.

So...flame away! --Regards. 03:27, 7 December 2006 (UTC)

Sorry, I won't flame you. I will merely point out that presenting arguments for or against various conclusions is original research that is disallowed in Wikipedia articles. What we need is not an argument that Internet Explorer's box model was coded as intended, but a source that explicitly states that. All we can say in Wikipedia articles is what the cited source says. We cannot do any analysis or synthesis to reach a conclusion that is not directly stated by the source. Go ahead and read the no original research policy which will explain the situation in more detail. -- Schapel 03:39, 7 December 2006 (UTC)
This is strange logic you have. Whatever happend to "innocent until proven guilty"? Let me put it this way:
  1. 20 people are looking up at the sky.
  2. 1 person says, "the sky is falling".
  3. the other 18 people argue about whether or not it's falling. Some agree, some don't.
  4. 1 person says "the sky is not falling".
  5. nobody gets hit on the head by the sky -- but person in #2 wants proof that the sky is not falling (and demands a citation from a verifiable source. Yet, the only sensible conclusion is that person #2 is the only person who is wrong and it's unfortunate that they were so influential upon the rest of the community.)
Another example:
  1. A bunch of people live in the world.
  2. George Bush says, "Iraq has weapons of mass destruction".
  3. Everybody argues about whether or not this is true. Some agree, some don't.
  4. The U.N. says, "Iraq does not have weapons of mass destruction".
  5. Nobody finds weapons of mass destruction -- but George Bush wants proof that they don't have weapons of mass destruction (and demands a citation from a verifiable source [as though Iraq and the U.N. weren't verifiable sources]. Yet, the only sensible conclusion is that George Bush is the only person who is wrong and it's unfortunate that he was so influential upon the rest of the community.)
That's what is happening in this talk page:
  1. a bunch of people are using MSIE.
  2. 1 person says, "the box model is buggy". Actually, he said it is a "bug".
  3. the others argue about whether or not it's a bug. Some agree, some don't.
  4. 1 person (me in this situation) says "the box model is not a bug".
  5. nobody experiences a bug -- but person in #2 wants proof that the box model is not a bug (and demands a citation from a verifiable source [as if Microsoft's own documentation and the W3C CSS1 specification aren't reliable sources]. Yet, the only sensible conclusion is that person #2 is the only person who is wrong and it's unfortunate that they were so influential upon the rest of the community.)
Can't we just omit the word "bug" from the page title and get on with it? Without that word, this would be a great page to document the box model's behaviour in MSIE; but with that word is just misinformation.

--Regards. 04:05, 7 December 2006 (UTC)

There's no "strange logic I have." I'm not presenting any argument at all. I'm simply pointing out Wikipedia's official policies, which we all must abide by. If you have a problem with those policies, I suggest you debate them in the proper place. We have cited reliable sources that call the behavior a bug, and therefore get to say it's a bug in the article. If you want to say something different in the article, you must cite sources that say the same thing you want to say. -- Schapel 05:13, 7 December 2006 (UTC)
None of the sources you have cited have said why it is a bug. The word has been used loosely to describe an authors general opinion of the behaviour, but there are no sources who have said clearly that "it is a bug because...". Fine, you can say that some people call it a bug. But the title of this article?!?! A page called "Internet Explorer box model bug" should rightfully be a page which describes in detail the bug. But this article is about the "Internet Explorer box model" and therefore it should be called "Internet Explorer box model" -- and the notion that some people refer to it as buggy should be a mere footnote in the topic!
What logic do you have to support the notion that the word "bug" should be in the page title? Which "bug" exactly is documented in this article? —The preceding unsigned comment was added by Formation (talkcontribs) 05:27, 7 December 2006 (UTC).
The bug is what many sources refer to as the Internet Explorer box model bug. Why not read some of the sources we cite? They will all explain to you exactly what the bug is. From the source I just gave, the bug is that "Internet Explorer incorrectly includes the padding and borders within the specified width, resulting in a narrower box when displayed." I would justify having the term bug in the title by noticing that it is the term used by many to describe the problem, and the term unambiguously communicates that Internet Explorer's box model is wrong, incorrect, flawed, a problem, a defect, or any other synonym of the term bug. It is not merely a different, equally valid, box model. -- Schapel 15:01, 7 December 2006 (UTC)
Actually, it is just a different, equally valid box model. It's exactly the box model described in the W3C CSS Level 1 Recommendation. Do you need me to repeat that a few more times?
Get it yet?
I've read your citations -- and all of them are clearly biased, uninformed opinions suggesting that Internet Explorer's box model *should* be somehow different (i.e. more like Firefox).
Your citation that "the bug is that 'Internet Explorer incorrectly includes the padding and borders within the specified width, resulting in a narrower box when displayed.' (which is the basis of your entire argument) is simply not true. What did the author mean by "incorrectly"? It includes the padding and borders within the specified width exactly as the W3C recommended in CSS1. The only thing verifiable about your citation is that the author is misinformed and/or purposely biased.
Would you like proof of that? Got it, how's this: MSIE's box model behaves according to the CSS1 recommendation and this proved by the observation of hundreds of millions of users, hundreds of millions of times per day. Every time a page is rendered in an MSIE browser (and the box model behaves exactly as it should according to the CSS1 recommendation), my argument is reaffirmed and yours becomes more ridiculous.
So, again I'll ask the question: What's the flaw? What's the bug? --Regards. 01:47, 8 December 2006 (UTC)

And in case there still some confusion...the W3C CSS1 Recommendation also says this:

"The horizontal position and size of a non-floating, block-level element is determined by seven properties: 'margin-left', 'border-left', 'padding-left', 'width', 'padding-right', 'border-right' and 'margin-right'. The sum of these seven is always equal to the 'width' of the parent element."

Do you want me to repeat that too?

"The sum of these seven is always equal to the 'width' of the parent element."
"The sum of these seven is always equal to the 'width' of the parent element."
"The sum of these seven is always equal to the 'width' of the parent element."

Is there still any doubt that MSIE's box model behaves exactly this way, and is therefore not flawed as you're arguing?--Regards. 01:56, 8 December 2006 (UTC)

I'm not arguing anything at all. I'm merely pointing out what reliable sources say. Even if you prove IE's box model is correct, we still cannot include that statement in the article without a reliable source that says it because the threshold for inclusion in Wikipedia is verifiability, not truth. Get it yet? -- Schapel 02:11, 8 December 2006 (UTC)

From Microsoft's "CSS Enhancements in Internet Explorer 6": Earlier versions of Internet Explorer calculate the width and height properties in a way that does not comply with the CSS1 box model.

Can we move on now? AlistairMcMillan 03:17, 8 December 2006 (UTC)

BTW You should try reading that last thing you repeated. It doesn't say what you think it says. "The sum of the seven is always equal to the width of the parent element." Not that actual element in question but the parent element. Note that one of the seven properties is width. As in you add the width of an element and its margins, borders and padding and you get the width of the parent element. AlistairMcMillan 03:26, 8 December 2006 (UTC)

Published author Paul "I prefer the IE box model" Peter Koch: "Although Explorer 5 Windows mended quite a lot of Explorer 4 bugs, it perpetuated other glitches in CSS (mainly the box model).'[5] AlistairMcMillan 03:31, 8 December 2006 (UTC)

Microsoft browser developer Tantek Celik: "...the "incorrect" box model dimensions for browsers like IE5.0 and IE5.5 for Windows which got the CSS box model wrong" [6] AlistairMcMillan 03:31, 8 December 2006 (UTC)

Author of books on CSS, Jeffrey Zeldman: "...the incorrect CSS implementation in versions of Internet Explorer for Windows prior to IE6." [7] AlistairMcMillan 03:34, 8 December 2006 (UTC)

Seriously are we done yet? AlistairMcMillan 03:40, 8 December 2006 (UTC)

I don't think we have to waste too much more time on just one person who floods the talk page with five-line headings and identical text repeated over and over again - that is not rational debate. I think we should just get on and fix the article - it's still gives some undue weight a few non-notable or bloggers' opinions. --Nigelj 23:18, 8 December 2006 (UTC)

I think we should revert it back to the pre-"pro-IE" July 6, 2006 version [8] and start improving from there. AlistairMcMillan 00:49, 9 December 2006 (UTC)

I was looking at older versions of the article yesterday, and I found them to be much more informative and useful than the current version. The July 6, 2006 version you point to looks as informative as anything I saw, and even has a diagram to completely clarify what the bug is. Perhaps we should add information about Paul Peter Koch (and other web developers, if we can state that verifiably and without weasel words) preferring the IE box model, even though he considers Microsoft's implementation a bug, and also add the proper, non-buggy CSS3 way of requesting the IE box model (the new "box-sizing" attribute). We should also explain why CSS3 supports the IE box model in a standard way — is it because some web devlopers prefer it, or is it an easy way to make pages that rely on the bug standards compliant? -- Schapel 03:52, 9 December 2006 (UTC)
Re your last point, remember that CSS3 is only a draft at the moment. That reference to it I quoted above has a reviewer's comment that clearly says, "it is not clear that these properties are really needed anymore" regarding the IE-box-compatibility features, so it might well be that that suggested support does not survive into future CSS3 drafts. As for the re-write being based on that older copy, I agree with you and Alistair above. --Nigelj 18:15, 9 December 2006 (UTC)
Really, after reading all this, I think it's pretty clear that Microsoft's continued support of their box model over W3C's demonstrates that it is absolutely not a bug, but an intentional design decision that they hope (in their hubris) to make the standard. I don't know if CSS3's implementation of the box-sizing property is a capitulation to IE's majority share of the browser market or if it's just the fact that having such a property makes a lot of sense. In any case, I'm tending to agree with Regards's point that use of "bug" (an unnecessary pejorative) in the title diminishes the article's ability to maintain NPOV. The results for "Internet Explorer box model" forward to here, when they should more ideally (from my perspective) forward from here to there. I think it's fine to have reference to designer dissatisfaction with IE's model as well as reference to sources that describe their intentional implementation as a bug. It just seems out of place in the entry's title. Either that or change the entry to be about the perception of IE's box model as being a bug. --The Dane 17:48, 15 March 2007 (UTC)

You do realize that Microsoft support both box models since version 6, right? The fact that they continue to support their flawed box model is not unusual. Maintaining backward compatibility is a big issue for Microsoft. They always strive to maintain backward compatibility. Raymond Chen writes about this all the time. AlistairMcMillan 18:12, 15 March 2007 (UTC)


A lot of you are missing the point. Microsoft and the W3C released separate, documented standards for defining the visual properties of an element. Just because one became more accepted as a standard does not immediately place the other one as "bug" in the traditional sense. That's like saying that a Corel Wordperfect document has a bug because it does not open as expected in Microsoft Word. Or that HD-DVD is a bug in the high definition disc, because it doesn't work as expected in my Blu-Ray player. It isn't a bug. It's a difference in spec. I may have personal opinions on the matter, about whether or not I agree that if I stuff a 5 1/2-foot refrigerator box with 2 1/2 feet of padding on all sides, that the height of my refrigerator box is now 6 inches. Really? Because it kinda still looks 5 1/2 feet tall. But that's the standard, and yes, it's good for everyone if we all agree that the height of the box is 6 inches. The Standards for Refrigerator Measurement disagreed with G.E. on how to measure the box, and we've sided with the SRM. Great, move on. But G.E. was not wrong for suggesting originally that the box is 5 1/2 feet tall because, um, it is. --66.119.170.242 (talk) 20:52, 26 January 2010 (UTC)

Criticism[edit]

Among the many problems in this new section are:

  • weasel words (e.g. "several web designers")
  • No sources cited
  • Speculation on why CSS3 has the box-model property. It may be to appease the "several web designers" that prefer the box model, or it may be simply to make it easy for sites designed according to the box model bug to be made standards compliant and work in all browsers. We need a reliable source for this information rather than a guess.
  • The CSS3 specification is currently just a working draft, but this section makes it sound like the box-model property will definitely be included in the final specification. The fact is that is may be removed at any time.

I don't have a problem with the content as long as these problems are addressed ASAP. -- Schapel 22:19, 5 February 2007 (UTC)

Schapel, I was not even finished with editing and putting in a selection of sources when you wrote that comment. Now that I discovered it and the lengthy discussion here, I have incorporated some of your other suggestions. However, the weasel will be hard to avoid since its impossible to determine what fraction of designers is supporting the one or the other model, and whose opinion counts how much. However, the points made in favor of the IE model and against the W3C specification themselves are valid and cited in a number of sources, and therefore belong to the article as an existing alternative view. Bluebird47 22:55, 5 February 2007 (UTC)
There probably isn't any way to "determine what fraction of designers is supporting the one or the other model." The best you'll be able to do is to cite sources and attribute opinions. I completely agree that the points made are valid, and even if they're invalid they belong in the article. It's not our place to judge whether points are valid or not. We simply report what those points are, attributing those points by citing reliable sources. -- Schapel 23:20, 5 February 2007 (UTC)

Bug in the title[edit]

Should we have the "bug" in the title? I know it has been discusse before but things have changed since then. While it is definitely a bug as far as compliance to standards (circa IE6) is concerned, things have changed now. Its a part of the CSS3 working draft. And IE now exposes the border box model as a CSS feature, to be opted in using the -ms-box-sizing attribute. That definitely is not a bug.--soum talk 16:50, 25 March 2008 (UTC)

I agree. Although this question has been debated for a very long time (just look above in this page), I think that there are no strong arguments, if any, to say that IE pre-IE6 box model was bugged. Indeed, there is enough evidence to proof that this is NOT a bug! It should be called a non-W3C-compliant behavior. In my opinion, the article could explain that some sources states this behavior as a bug, when it’s indeed just a non-W3C-compliant behavior. Insisting that this is a “bug” is a strong non-neutral point of view. --Wizard IT (talk) 18:51, 23 August 2008 (UTC)
In some future CSS3 article, we may discuss the detailed features of CSS3, and maybe mention their history, if someone publishes the W3C's discussions and thinking processes in a way we can cite. This article is about the 'Internet Explorer box model bug': It was a bug (as stated by Soumyasch above), it was called a bug by those who wrote about it at the time (including those who found and published the workarounds), it never got fixed, and it caused lots of web pages to display inconsistently until developers learned the CSS hacks to work around it. This article is about the bug where some MS developers either didn't read the spec before writing their code, or read it and decided they could do better than the W3C spec. Everyone called it this at the time, and that's why this is the right name for the article. The -ms-box-sizing feature and the stuff in CSS3 presumably aren't bugs, but this is not about them - this is about the 'Internet Explorer box model bug'. How much clearer can that be? --Nigelj (talk) 18:31, 24 August 2008 (UTC)
It clearly isn't a bug. A bug is an error, not a difference of opinion. The major problem is that the CSS for reasons best known to themselves chose a nonsensical "box" model which is counter-intuitive (as demonstrated by the metaphor of a real, physical box and its contents) and which has no merits of its own. It is a very great pity that they could not recognise that they had boobed, and changed the official spec to the superior IE model. At least we have the grudging "box-size" inclusion to work with. But I find it hard to believe that anyone would think the W3C spec is superior to the IE spec on merit alone. So anyway, it clearly isn't a bug, by any normal definition of the term "bug".82.71.30.178 (talk) 15:20, 29 July 2012 (UTC)

Requested move[edit]

The following discussion is an archived discussion of the . Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

nominator withdrawal, I (nominator) have decided to withdrawal this proposal.--Voidvector (talk) 21:57, 24 October 2008 (UTC) <result>

Internet Explorer box model bugInternet Explorer box model — The problem may historically be a bug, but due to its prevalence, it is no longer viewed by the web development community as a bug, but rather a different box model implemented by the browser. The different models are called "content-box" and "border-box" respectively in CSS. Major browsers now support proprietary tags allowing CSS developer to specify which model they wish to use [9]. — Voidvector (talk) 21:07, 20 October 2008 (UTC)

Survey[edit]

Feel free to state your position on the renaming proposal by beginning a new line in this section with *'''Support''' or *'''Oppose''', then sign your comment with ~~~~. Since polling is not a substitute for discussion, please explain your reasons, taking into account Wikipedia's naming conventions.
  • Support nominator. --Voidvector (talk) 21:14, 20 October 2008 (UTC)
  • Oppose - this is what it is commonly called all over the web and in textbooks. This has been brought up before - please see the archive. --Nigelj (talk) 19:56, 23 October 2008 (UTC)
[10], [11], [12]. Please try to be more civil and to assume good faith --Nigelj (talk) 20:40, 24 October 2008 (UTC)
"Please try to be more civil and to assume good faith." Same to you. --Voidvector (talk) 21:30, 24 October 2008 (UTC)

Discussion[edit]

Any additional comments:
The above discussion is preserved as an archive of the proposal. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

Support for Internet Explorer's box model[edit]

The current section with this heading begins with the point that the W3C considered this IE "feature/bug" as a possibility for inclusion in CSS3, and published these thoughts in their 2004 working draft. It goes on to say that the current working draft for the CSS3 'basic box model', dated 2007, does not include them, although one of the three properties considered in 2004, viz box-sizing, has been moved into another part of the overall draft, the 'advanced layout module', with a comment there that it may or may not be needed in the end.

The final paragraph mentions that there is now a proprietary CSS property that Firefox and its derivatives understand that reproduces this IE behaviour if used, but this carries a 'not recommended' comment on its developers' web page.

I think this covers the present situation in huge detail, and that some of these points are actually less than notable in the greater scheme of things. I would restrict this discussion to the part emboldened above and cut the rest on this basis.

What do other editors think? Could it be clearer? Are the latter details notable? --Nigelj (talk) 19:59, 23 October 2008 (UTC)

First of all, the article is not very long, including more detail gives user perspective of the current situation. Secondly, as far as I know, none of the modules is finalized, so citing one while ignoring another constitute favoritism/censorship. I don't even know why this article try to pin the problem on IE, since Netscape had a non-W3C box model. --Voidvector (talk) 20:13, 23 October 2008 (UTC)
Further proof that box model difference is not an IE-only problem: an old version of PPK's reference page. (press Cancel) --Voidvector (talk) 20:35, 23 October 2008 (UTC)
This article is about the Internet Explorer box model bug, that's why it focuses on IE and on this particular bug. Your new edits may be straying off that central subject into less relevant topics like recent W3C spec drafts, recent obscure Mozilla features etc. --Nigelj (talk) 20:48, 24 October 2008 (UTC)
You are missing the point, IE simply kept the non-W3C box model from pre-CSS era. To cover the the problem as "MS vs W3C" is misrepresentation. --Voidvector (talk) 21:20, 24 October 2008 (UTC)
No, the point is that IE keeps the non-W3C box model to this day, in various operating modes of IE6, 7 and 8. This can be seen as part of attempts to maintain a "best viewed in IE" web, that is different from the "worldwide" web. It costs a great deal of time, money and effort on the part of the web's designers and coders to try to prevent this becoming reality, despite these attempts.
There was no IE support for any CSS box model pre-CSS?! IE3 was the first IE to support any CSS and, "IE3 reliably supports most of the color, background, font and text properties, but does not implement much of the box model."[13] IE3 was released on August 13, 1996 and CHSS (later to become just CSS as it could be applied to more than just HTML) had been in existence since October 1994. When IE4 added more support for a CSS box model in Sept 1997, it was either willfully or erroneously contrary to the CSS1 spec that, by then, had been published by the W3C since December 1996. (see CSS#History and Internet Explorer#Version 3 and the citations leading from those pages) --Nigelj (talk) 22:05, 24 October 2008 (UTC)
So we both agree that there was a pre-W3C box model, correct? If so, instead of describing the issue as "standard vs legacy" which is what it is, why are you insisting on having the article centered around IE/Microsoft?
The non-box CSS is really of no interest to this conversation, i am not sure why you brought it up, but I would say Netscape 4 was worse in terms of CSS support. --Voidvector (talk)

Having given you a couple of weeks to do what you like to this article, I have just tightened up the WP:LEAD. It is important to stick to the topic of the article at the very start and not begin the article with an argument disputing its very name, I feel. Having begun this discussion section asking the question as to whether or not a particular section was already too long, I am disappointed to see that it has been lengthened with a long preceding paragraph summarising what appear to be the opinions of a series of bloggers.

I don't have the time at the moment to read and appraise each of those references, but if they turn out to be blogs as I suspect, then I'm afraid they'll have to go. As I'm sure you know, WP does not generally regard personal blogs as WP:RELIABLE sources. If I get a chance, I'd like to go though and upgrade the whole article to the current recommended format for WP:CITE citations too - you know, using the format <ref>{{cite web | url=... | title=... | author=... etc}}</ref> rather than the current mish-mash. --Nigelj (talk) 20:50, 13 November 2008 (UTC)

Yes check.svg Done I updated many references. but the article needs more work anyway! mabdul 10:06, 27 July 2010 (UTC)

Requested move[edit]

The following discussion is an archived discussion of the proposal. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the proposal was No consensus Parsecboy (talk) 15:53, 6 December 2008 (UTC)


Internet Explorer box model bugCSS box model problem — The current title is POV. The box model problem is a problem that has plagued web design/development for years. It is not simply an Internet Explorer issue. Historically, this model was also used by other browsers, as stated and cited in the article. Even now, other browsers are using non-standard box definitions for some of the lesser known features (such as offsetX, offsetY).

The current title also gets in the way of neutral collaboration, as it may be used to label content/discussion on non-IE box model "bugs" as off-topic.

I withdrew my previous move proposal due to the fact that my earlier name did not fully capture essence the problem. — Voidvector (talk) 21:56, 30 November 2008 (UTC)

Survey[edit]

Feel free to state your position on the renaming proposal by beginning a new line in this section with *'''Support''' or *'''Oppose''', then sign your comment with ~~~~. Since polling is not a substitute for discussion, please explain your reasons, taking into account Wikipedia's naming conventions.
  • Support nominator. --Voidvector (talk) 21:57, 30 November 2008 (UTC)
  • Oppose; the assertion that this is a 'problem' that isn't limited to Internet Explorer needs to be put forth by a number of reliable sources. It's also called the "Internet Explorer box model" or the "traditional box model" by most experts. There's no justification for removing this name from the title of the article. If anything, the name of the article should simply be "Internet Explorer box model" -- the notion that it's a bug by no means a universal one. Warren -talk- 22:44, 30 November 2008 (UTC)
  • Support I will add the Netscape 4 boxmodel described in an article(link is somewhere above).(edit:oh and update the picture of course!)--mabdul 0=* 18:38, 1 December 2008 (UTC)
  • Oppose The article is supported by numerous citations to published and peer-reviewed university- and professional-level textbooks (as well as to out of date and current W3C recommendations and some private web sites) that refer to this bug and use the terminology in the current title. This terminology is what this article explains. It is clear from the proposal that the move would be to a newly-coined phrase, which some authors feel would provide them with a wider scope to write their own essays, and expound their own synthesis of wide-ranging and currently disparate ideas. Therefore, this article is based on material that can be verified in primary, secondary and tertiary sources. The proposal seems designed to open its door to original research and synthesis. --Nigelj (talk) 20:00, 1 December 2008 (UTC)
    • Comment The current article fails miserable in explaining the problem. It attempts to explain a structured problem, (i.e. "what is definition of 'width' and 'height' in terms of content/padding/border/margin?") by blaming it on the non-compliance of IE, and calling it a software bug. Also you are assuming bad faith by thinking that any new contribution or expansion of the article will be "original research". In addition, the current article title is a wiki coined term as none of the citations calling it a bug actually use the phrase "Internet Explorer box model bug". --Voidvector (talk) 22:53, 1 December 2008 (UTC)

Discussion[edit]

Any additional comments:
The above discussion is preserved as an archive of the proposal. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

CSS box model[edit]

I'm less interested in whether Bill's variation was a mistake than in how web designers can use the CSS box model. I'm even willing to help write an article about the box model, and I've started a disambiguation page. --Uncle Ed (talk) 17:03, 22 September 2009 (UTC)

The only web designers 'using' "Bill's variation" will be writing web pages for IE5.5 and earlier. This is a very small and diminishing market for them, I would say, in these days of online updates. The rest can refer to the main CSS article and the specs available from W3C. --Nigelj (talk) 19:55, 22 September 2009 (UTC)

Removal of referenced and sourced material[edit]

This series of edits by User:Ed Poor (also known as 'Uncle Ed' above) removed the following sourced statement and replaced it with a nicely deletable piece of weasel wording.

<removed>
The Internet Explorer box model bug is a software bug in the implementation of Cascading Style Sheets in earlier versions of Microsoft’s Internet Explorer web browser for Microsoft Windows.[1][2]
</removed>

<added>
Some software designers[who?] have branded this a "software bug".
</added>

No matter how long Ed has been editing WP, I propose that these edits are a clear reduction in the usefulness of this article. Where we once had two references to University-level textbooks on web technologies using an established piece of terminology and explaining it in the opening of this article, we now have mealy-mouthed weasel words that sound like a few designers were once a bit grumpy.

At the same time as he removed this sourced content from the opening, he said it was going into another article that he was writing. At the moment that other article consists of "The CSS box model is essentially a box that wraps around HTML elements, and it consists of: margins, borders, padding, and the actual content." [1] (The quotes are his and are there). Nothing except this one links to the new article.

Oh, and he 'moved' this article too. All this with no warning, no discussion and no consensus. I'm not going to get into a 'move war', but I am going to revert these edits until we find out what his motivation and plan is, and what other contributors think of this plan. --Nigelj (talk) 19:19, 22 September 2009 (UTC)

  1. ^ Shafer, Dan (2005). HTML Utopia: Designing Without Tables Using CSS. Melbourne: Sitepoint. pp. 124 & Appendix C. ISBN 0-9579218-2-9. 
  2. ^ Shea, David; Molly E. Holzschlag (2005). The Zen of CSS Design. Berkeley: Peachpit Press. ISBN 0321303474. 
Hey, speaking of grumpy ...
But seriously, folks, when I started out at Wikipedia, we had a rule called "Be bold".
In this case, it seems that you all have decided that the difference between Microsoft's implementation of the CSS box model and the W3C standard constitutes a software bug. I didn't know that, and from reading the article I felt that it was just a deliberate move on their part rather than, say, a coding error. --Uncle Ed (talk) 21:01, 22 September 2009 (UTC)

CSS box model[edit]

My motivation is to let others know all I have recently learned about CSS and its box model. I could care less whether IE has a "bug" or design flaw; or they just decided to be different to stifle their competition. All I care about is helping other web developers to write workable code.

Now if anyone wants to join me in some collaborative writing, then let's share our energy and explain what the box model is, how it works, and how to use it. I don't want to fight or argue. --Uncle Ed (talk) 02:09, 8 November 2009 (UTC)

If it is not a bug, but a deliberate choice to implement it that way, then it should not say "bug". Especially if the other major browser, Netscape, implemented it the same way. --Joshua Issac (talk) 10:55, 22 May 2010 (UTC)

Quirks mode triggering if "there are errors anywhere in the document"[edit]

This claim seems to be not true (kind of web design urban legend). MSDN and other relieble sources on browser modes (e.g. http://hsivonen.iki.fi/doctype/) don't confirm that any browser changes the rendering mode if the part of the document is already loaded and rendered. It's true that the errors in the document may be treated differently in Quirks mode and in Standards compliance mode, but these rules are defined only once — in the beginning of the document parsing, when Doctype is parsed. — Preceding unsigned comment added by 178.124.157.26 (talk) 11:26, 23 January 2013 (UTC)