Talk:Quirks mode/Archive 1

.txt
Mention firefox uses qirks mode for .txt files too... --Jidanni 2006-04-16

Quirks mode as "legacy"
WOOO-HOOOO!!! yeah, your jealous. uh-huh. KL is sooo kooler than u. ah ha! admit it.

(I don't know who added the above lines, but it certainly wasn't me) Ubernostrum 05:33, 4 March 2007 (UTC)

Pages which render as intended only because of quirks-mode behavior are indeed "legacy" pages; they are designed and created using techniques and practices which are outdated, deprecated, and actively discouraged by the relevant standards body, and continue to function only because of complex backwards-compatibility features implemented in major browsers. Something which functions according to outdated or deprecated behavior and is supported only through explicit backwards-compatibility features is, by definition, "legacy". —The preceding unsigned comment was added by Ubernostrum (talk • contribs).


 * No, that's not the definition of legacy. See Wiktionary or reference.com.  The term is being mis-used!  People can create pages that don't use "standards" mode today.  In fact, people do this all time, perhaps far more often than you realise.  New things which are created now absolutely can not be defined as "legacy".  More importantly, note that none of the external links used for this article use the word legacy anywhere; this is simple editorialization on the part of Wikipedia, and it has no place here.  Warrens 01:49, 16 April 2006 (UTC)


 * Yes it is. Quoth your Wiktionary link: "(computing) of a computer system that has been in service for many years and that a business still relies upon, even though it is becoming expensive to maintain". Pages which are built for quirks mode rely on (check) syntax and behaviors which have existed for many years (check), even though it is becoming expensive (to browser developers) to maintain (check). Seems to fit. As for references, if you'd really like me to I can happily cite Mozilla and Opera calling this "legacy behavior". I will be re-inserting the term and adding those links, and if you feel that it is still inappropriate despite my arguments, please feel free to begin any of Wikipedia's formal dispute processes regarding its use in this article.


 * One more while I'm at it: noted CSS expert Eric Meyer uses the terms "legacy bugs", "legacy layout" and "legacy authoring" while talking about quirks mode: http://www.ericmeyeroncss.com/bonus/render-mode.html —The preceding unsigned comment was added by Ubernostrum (talk • contribs).


 * Great, so now we're citing sources with a clear point of view on the subject (pro-web-standards) as authoritative? Just what Wikipedia needs more of, POV-pushing.  Nothing you've said addressed the key point of the fact that new pages are being created using quirks mode.  Yes, there are plenty of older documents which certainly qualify as "legacy" as they were created long before the modern standards became widely used.  That, however, only tells part of the story:  Newly-created pages aren't automatically "legacy" by virtue of leaving out a single statement at the beginning of the page, but they do render in quirks mode.  Don't believe me?  Go look at a Myspace profile, CNN.com, or BBC World News, all of which render in quirks mode.  Nobody in their right mind would argue that these are "legacy" pages; they're some of the most-visited pages on the Internet!


 * This article describes the existence of one of the two valid methods by which a page can be rendered. Is that clear?  It's VALID.  I know this is an extremely difficult concept for standards afficionados to accept, especially since their stance is a POV propogated by well-spoken people like Eric Meyer, and they expect others to accept this as The Truth.  It is, of course, in their own selfish self-interest to push their preferred web standards and to refer to the other option as "legacy".


 * Wikipedia's overriding philosophy is still one of neutrality, however, and every topic, including this one, must be presented as such. To this end, I've re-written the overview to balance out the issue of "legacy" pages vs. "new" pages which both use quirks mode rendering.  Please, instead of simply reverting these changes to try and push the Eric Meyer-approved view of things, support a consensus-building approach to getting all the correct and relevant information into this article.  That's what we're here at Wikipedia to do.


 * p.s. don't forget to sign your talk page comments with four tildes. Warrens 07:35, 16 April 2006 (UTC)


 * Regarding your points:
 * Web standards as a "point of view": HTML and CSS are open specifications, maintained by a a vendor-neutral international standards body with the widespread consensus of major players in the industry. This is roughly as POV as would be referencing building codes in articles about structural engineering. You seem to have a personal grudge of some sort regarding this issue, however, and I'd appreciate it if you could set that aside while discussing this article.
 * Regarding the authority of the link to Mr. Meyer's comments: regardless of your personal feelings, Eric Meyer is a noted expert on the subject of CSS, and the link to his work was not inserted into the article; I mentioned it here on the talk page, and only here on the talk page, as further support for the use of the term "legacy".
 * On the use of "legacy": you claim that it "propagates" a so-called "Eric Meyer point of view". However, the two links I actually added to the article are both from browser vendors: one from the Mozilla Foundation and one from Opera Software. They both characterize quirks mode behaviors using the term "legacy", hence the term "legacy" is appropriate. And again, pages which rely on quirks mode are relying on outdated and, often, non-standards-compliant behavior and impose a significant compatibility burden on developers of web browsing software. This is, as I've reiterated, pretty much the definition of "legacy" in computing. The fact that pages continue to be created using these techniques does not make the techniques any less outdated, does not make the features they rely upon any less non-standard, and does not change the burden they create on browser developers.
 * I'm going to re-insert the "legacy" terminology; if you still feel that this is a serious issue which affects the quality or accuracy of the article, please say so here and we can go through the appropriate channels to resolve it. Ubernostrum 08:33, 16 April 2006 (UTC)


 * I've edited the article a bit to ensure that it only references documentation from browser vendors in describing quirks mode vs. standards mode behaviors. I also made a minor edit to the "legacy pages" section to try to make its language more accurate and involve less insinuation about developers not following standards. Please let me know what you think. Ubernostrum 09:02, 16 April 2006 (UTC)


 * You're completely ignoring the thrust of what I'm saying, and you've reverted every change I've made to the article -- in spite of a request to not do so -- in favour of prose you've written yourself. What do you expect me to think, other than that you don't want to work collaboratively on this, and that you're here to push a specific POV?


 * The phrase "legacy pages" didn't appear in the article before you rewrote pretty much the entire article and added that characterization in. Where are you getting this phrase from? Eric Meyer doesn't talk about legacy pages -- he talks about legacy browser behaviour.  Microsoft doesn't use the term "legacy pages" -- they talk about "previous versions".  Mozilla doesn't use the term "legacy pages" -- they talk about legacy browsers.  Opera doesn't use the term "legacy pages" -- they talk about legacy browsers, too.  Who is talking about "legacy pages", other than you?  It's not in any of your references.


 * Don't you see the issue here? The problem isn't described by anybody noteworthy as being "legacy pages".  Again I point to CNN, BBC World News, and other major web sites which render in quirks mode as notable samples of web sites which are most certainly not "legacy pages", yet are subject to the issue described in this article!


 * I'm going to edit the article, once again, to remove this characterization, because it's not accurate, and is not described as such by any of the article's references. This is in accordance with WP:Verifiability and WP:No original research.  I'm also going to reinstate the Quirksmode.org link, which has been in this article since day one, and is considered a fairly authoritative page on the subject.  You've removed it for no particularily good or justifiable reason; it really should be here.


 * Once again, I request that you NOT SIMPLY REVERT my changes. And please, lay off with this "formal dispute process" hand-waving... they're not there to go running to every time someone makes an edit you don't like.  If you don't want your writing to be edited mercilessly, do not submit it.  Warrens 10:38, 16 April 2006 (UTC)


 * You should also review WP:External links to learn how to create them properly. Mediawiki will create numbers for you; you don't have to fudge it yourself. Warrens 10:43, 16 April 2006 (UTC)


 * The term "legacy" did not appear in the page until I edited it. However, the article was also marked as a stub and marked as in need of cleanup to "conform to a higher standard of quality". I had some free time and I was familiar with the subject, so I edited and expanded the article; I'd hope you'll agree with me that it's now more thorough and useful as a result.


 * And again, I cite the definition of the term "legacy" which you provided from Wiktionary: "(computing) of a computer system that has been in service for many years and that a business still relies upon, even though it is becoming expensive to maintain". The techniques used by quirks mode pages have been in service for many years. Check. Businesses still rely upon them. Check. It is increasingly expensive to maintain compatibility with them. Check. Opera and Mozilla do not directly refer to these pages as "legacy pages", but their documentation is not concerned with pages -- it is concerned with browser behavior. They refer to quirks mode as "legacy behavior"; logically, would a page which is not a "legacy page" be reliant upon "legacy behavior"? Bottom line: calling them "legacy pages", regardless of when they were created, is accurate, because they rely on expensive and difficult backwards-compatibility routines in browsers to render their outdated and/or deprecated code. At this point I'm unaware of how to more clearly articulate this fact. Ubernostrum 22:51, 16 April 2006 (UTC)


 * As to your edits being reverted out, I do apologize for that. I was using the most direct method available of getting back to the versions you had changed, which was simply to copy/paste and save. Besides, I figured I could rely upon you to be by shortly after to re-insert anything you felt needed to be placed back in the article.


 * For now I'm done with this; I've asked some knowledgeable people to help me find further citations which will support the use of the term "legacy", as apparently conforming to a definition of the term published at Wiktionary is insufficient grounds for being able to use it on Wikipedia. Hopefully that will be able to resolve this issue once and for all.


 * Here's a nice one: http://www.motive.co.nz/glossary/legacy-content.php. "Quirks mode enables legacy HTML documents to still ‘work’, and should be triggered when the underlying code is known to fail technical standards (and if there is no intention/budget to revise legacy content)."


 * And while I imagine Ian Hickson will be struck down as a "standards aficionado", he's been a developer for both Mozilla and for Opera, is a member of the WHAT-WG, and he consistently refers to quirks mode itself as "legacy mode". See, for example, this document: http://www.hixie.ch/specs/alttext-1.2 Ubernostrum 06:50, 17 April 2006 (UTC)


 * In further support for the use of the term "legacy" or legacy pages", I submit a quote from Devindra S. Chainani, the Group Program Manager for Expression Web Designer: "Prior to IE6, browsers built in non-standard rendering, without conforming to W3C recommendations With wider acceptance of standards, browsers began to operate in two modes – a ‘quirky’ mode called Quirks mode and a ‘standards’ mode. The former was used to support legacy pages." http://blogs.msdn.com/devi/archive/2005/10/14/481227.aspx Mcwilson 13:24, 17 April 2006 (UTC)


 * Microsoft on the use of DOCTYPE to switch rendering modes: "DOCTYPE Switching was introduced to enable browsers to render both standards-compliant and legacy Web sites correctly." Source: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnaspp/html/aspnetusstan.asp


 * Warren, in your opinion is the use of the term still "not accurate"? Ubernostrum 14:44, 17 April 2006 (UTC)


 * This link is dead. But, yes, I would say that using a selfserving definition from MS is NOT an accurate definition of "legacy".


 * Tyrerj (talk) 08:41, 8 March 2008 (UTC)


 * Are the definitions from Mozilla and Opera also intended to serve Microsoft, in your opinion? Ubernostrum (talk) 22:04, 8 March 2008 (UTC)

Microsoft polluted HTML as "legacy"
The above discussion is absurd. Many web pages are currently being written using syntax that doesn't conform to any version of HTML. There are two causes of this:


 * 1. Writing the code to support a browser rather than the standards.


 * 2. Using Microsoft web development tools that deliberately produce code that does not conform to W3C standards -- polluted HTML -- or web development tools from MS compromised vendors.

To call this legacy is simply to follow the Microsoft party line. The Orwellian PR that comes from MS is worse than NewSpeak (1984).

The impartiality of this article should be seriously questioned as it appears to use language from MS news releases.

Presuming that web standards should be followed is not a POV.

Tyrerj (talk) 10:37, 5 March 2008 (UTC)


 * If you feel there's something in the article (remember, this page here is the discussion page for development of the article, not the article itself) which is unfairly biased, please point it out specifically so it can be reviewed. Ubernostrum (talk) 04:45, 6 March 2008 (UTC)

I thought that I was clear. This is not about "legacy" code or backwards compatibility. The use of "Quirks" mode is to try to correctly interpert the HTML like web page language that MS has promulgated. If not for this issue, there would be no "Quirks" mode but rather the browser would read the DTD and base the (Standard mode) interpretation on that. The purpose of Quirks mode is not backward compatibility but rather to try to interpert HTML like code that does not conform to any version of HTML.

There aren't a few things in the article that need changing, it is this whole premise that needs to be changed and the whole article rewritten based on the unfortunate reality of the situation rather than appearing to be based on MS press releases.

This lack of impartiality in this artice becomes a much larger issue now that MS is using it to justify their 'MSpeak' statements about IE8:

http://blogs.msdn.com/ie/archive/2008/03/03/microsoft-s-interoperability-principles-and-ie8.aspx

Perhaps a greater understanding of the problem would be gained by running web sites built with MS web development tools through the W3C validator. It should then become obvious to anyone that MS has added a lot of what could charitably be called extensions to HTML.

The problem that browsers face is that they must somehow correctly interpret these "extensions" without having a standard to follow -- this is what Quirks mode is for. But, the problem has now become worse because some web pages with a valid DTD in the header still contain these non-standard extensions to the HTML or XHTML version used.

Tyrerj (talk) 08:31, 8 March 2008 (UTC)


 * Er, no. Plenty of tools besides Microsoft's have produced and continue to produce nonstandard HTML (see: Dreamwaver, Homesite, InDesign, browser-based WYSIWYG builders...); the fact that there were a large number of these tools, and a large number of people who'd never learned a standardized methodology or who learned HTML prior to the push for standardization, was the original impetus for quirks mode. Trying to pin this all on one vendor is simply incorrect, then. Also, I find it somewhat amusing that when I first added the term "legacy" to this article, based on descriptions by non-Microsoft vendors, it was condemned as being too much in favor of the standards POV, and now apparently you feel it's too much in favor of a Microsoft POV. There's a contradiction there, I think. Ubernostrum (talk) 22:02, 8 March 2008 (UTC)


 * I listed 2 reasons. Writing to the browser rather than writing to the standard is part of the problem.  I said that.  But if that were the only issue then Quirks would only apply to HTML-3.2 and earlier.  But that is not the case.


 * The current situation is that non-standard code is being written today using MS tools and other tools designed to conform to IE rather than the W3C standards. To call this "legacy code" is absurd.  To make things worse, this polluted code often has a DOCTYPE header that identifies it as a specific version of HTML when it is NOT.  This means that browsers need to be able to always run in Quirks mode and to start in Quirks mode if a URL for the DTD isn't listed.


 * Yes, it is propper to blame this all on one vendor. MS understands the advantage of exploiting bugs and non-standard code "extensions".  The problem persists till today because IE continues to be designed for non-standard code.

Tyrerj (talk) 11:51, 25 March 2008 (UTC)


 * HTML 3.2, though old, does not in and of itself require special-case workarounds; it's a documented, open standard and anyone can implement it if they choose, alongside other documented, open standards for other versions of HTML. The problem quirks mode addresses, as always, is documents which do not conform to any documented, open standard, or which rely on misinterpretations or misapplications of such standards in order to render as intended. This is, as I've stated, not unique to Microsoft or to documents targeted toward Microsoft products; browsers have implemented and will continue to implement non-standards-based rendering modes for a variety of reasons (see, for example, Mozilla's "almost standards" mode, which has nothing whatsoever to do with people targeting Internet Explorer), and as a result stating that these alternate rendering modes -- including quirks mode -- are entirely due to Microsoft would be factually incorrect and wilfully dishonest. Ubernostrum (talk) 20:27, 31 March 2008 (UTC)

Konqueror 3.2
Whoa... Konqueror 3.2 is *really* old. Maybe someone with knowledge of doctypes should update this table to at least Konqueror 3.5. Bramschoenmakers 18:32, 25 June 2007 (UTC)
 * If I'm not mistaken, Konq 3.5 handles doctypes identically to Safari and Gecko 1.0.1+. I could be wrong, though, and can't truly be certain without real testing -- but I'm lazy. ♠ SG →Talk 13:49, 27 June 2007 (UTC)

IBM dtd
Should indicate what mode browsers are put in for this: <!DOCTYPE html SYSTEM "http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd">

puts some versions of mozilla into almost-standards mode, others into quirks mode. —Random832 17:16, 21 September 2007 (UTC)
 * Why given that there are dozens of doctypes in use out there, why should this one be given specific encyclopedic focus? (OK, it is a bit special, but still)Hsivonen (talk) 21:50, 30 March 2008 (UTC)


 * OK, I see the point: I was surfing that night at http://developer.mozilla.org/en/docs/Mozilla's_DOCTYPE_sniffing and found many mores. I think you're right and we can delet this one Mabdul (talk) 22:56, 30 March 2008 (UTC)

Spelling: "prolog" vs. "prologue"
Just a quick note to explain my reversion of the spelling: section 2.8 of the W3C XML specification designates the opening declaration of xml version and encoding to be the "prolog", not the "prologue", and the spelling "prolog" is included in the formal grammar of XML. As such, Wikipedia should use "prolog" rather than "prologue" when referring to this feature of XML. Ubernostrum (talk) 20:08, 5 December 2007 (UTC)
 * That was my mistake. Nice fix. Ronark (talk) 03:04, 9 December 2007 (UTC)

HTML5
The article says "While this DOCTYPE (which does not contain either the version of HTML in use, or the URL of an HTML Document Type Definition) would trigger quirks mode: <!DOCTYPE html PUBLIC>". Doesn't it contradict the last (HTML5) row in the table ? --Fenring (talk) 10:26, 26 December 2007 (UTC)
 * The HTML5 doctype is <!DOCTYPE html> the table data is correct for that particular doctype, which is different from <!DOCTYPE html PUBLIC>.Hsivonen (talk) 21:55, 30 March 2008 (UTC)

Frameset Doctypes
Someone just added placeholders for frameset doctypes. What's the point? They seem like a distraction in the table. When they are used for establishing framesets, quirks/standards doesn't really matter for the frameset itself. OTOH, they shouldn't be used for content that actually appears in a frame.Hsivonen (talk) 21:53, 30 March 2008 (UTC)


 * I was the crazy mn who reformated this whole table ;)
 * The point is: There are diferent behaviours in the different browsers (e.g.: in the HTML4.0 ) Why we shouldn't show this? Mabdul (talk) 22:56, 30 March 2008 (UTC)

IE and almost-standards mode
The section on almost-standards mode defines it entirely as something exclusive to Gecko-based browsers, yet the charts show IE as entering almost-standards mode in a variety of circumstances, so either the charts are wrong or the almost-standards section needs to be filled out by someone who knows more details. &mdash;Largo Plazo (talk) 18:03, 7 July 2008 (UTC)


 * This is kind of tricky, because while IE and IE's developers refer to a "standards mode", that mode is functionally closer to the "almost standards" mode of Gecko, and IE does not currently have a stricter compliance mode more in line with the "standards mode" of other engines. A note explaining that would probably be a good idea. Ubernostrum (talk) 04:20, 8 July 2008 (UTC)

Time to remove deprecated browsers columns?
NS6, Gecko pre-1.0.1, Opera 7.5 and IE Mac are really deprecated and unmaintained browsers. Is it time to remove these columns or do you plan to keep this not useful information forever? —Preceding unsigned comment added by 195.235.227.10 (talk) 15:42, 19 January 2009 (UTC)
 * Why not? I mean quirksmode is constructed for old browsers behaviour... why not including old browsers?  mabdul 0=* 18:59, 19 January 2009 (UTC)
 * Because no one should care about these deprecated, unmaintained and unused browsers. Just take a look at the table, NS6 rendered all XHTML as standards mode but all current browsers render XHTML transitional and frameset as almost standards. What I mean is that the columns about old browsers are not useful to anybody nowadays as nobody is using them and any current designer does not look at them for browser compatibility. —Preceding unsigned comment added by 195.235.227.10 (talk) 10:45, 20 January 2009 (UTC)

I agree, only current browsers or layout engines supporting quirks/(almost)standards modes should appear on the list. VanisherX (talk) 10:35, 21 January 2009 (UTC)
 * o.O the list shows only browsers/lacout engines that support quirks. something other doesn't make sense!  mabdul 0=* 14:48, 21 January 2009 (UTC)

By the way, IE 8.0 and Opera 10.x are not mentioned in the table, which are important browsers for advanced and general net users. Maybe just add IE 8.0 and not Opera 10.x (Opera 10.x is still in beta stage......).--RekishiEJ (talk) 11:58, 10 June 2009 (UTC)

I also think older, little used browsers should be removed from the table. Also, why were the Firefox and Mozilla entries in the table changed to Gecko? All the other headings list browsers. Even more confusing is that Netscape 6 is also a Gecko browser, but it's listed by the browser name instead of Gecko. Let's strive to be consistent, understandable, and useful. -- Schapel (talk) 16:48, 9 May 2010 (UTC)


 * As I described in the cleanup-tag: there war a really good table that got crippled and I didn't found the time to change it back. why using gecko? since gecko has a really wide range of browsers that are based on. We could change all to browser engines. (but I don't think that we find a consents for that) mabdul 17:55, 9 May 2010 (UTC)


 * It's a needlessly complex table for people who want to quickly look up information in 2010. Let's make all the column headings list the most popular browsers instead of engines. This will make it far easier to use for readers who don't know what Gecko is. Let's remove the years-old, little-used browsers, such as Netscape 6, Mozilla 1.0, Opera 7, and Konqueror 3.2, which nearly no one cares about in 2010. Let's focus on making the table useful for the average reader. -- Schapel (talk) 14:41, 10 May 2010 (UTC)


 * Wikipedia should be a encyclopedia: it doesn't matter what in 2010 is standard. or let us create a second table for the "historic" browsers. mabdul 19:23, 10 May 2010 (UTC)


 * Yes, Wikipedia is an encyclopedia. It is not a place for minute details that have lost their relevance. It certainly does matter what in 2010 is standard -- Wikipedia should first and foremost be useful today and in the future. I was thinking instead of columns, we could use footnotes to describe the behavior of the "historic" browsers. A separate table might be even cleaner. If we do so, we should make sure to add Opera 8, as it is mysteriously missing. Or we could decide to remove the information from the current version and allow readers to refer back to the old versions for the old information. Is there really anyone who will want to look up that information in the future? -- Schapel (talk) 14:51, 11 May 2010 (UTC)

Content-Type
It would be good if the Comparison table included which Content-Type HTTP header the page was served with, text/html or application/xhtml+xml. This is significant, as when XHTML is served as application/xhtml+xml, browsers can use their much better XML rendering engine. —Preceding unsigned comment added by Darxus (talk • contribs) 19:15, 17 July 2009 (UTC)


 * For switching rendering modes, the Content-Type doesn't matter (so far as I know), only the DOCTYPE declaration. And documenting the differences in how, e.g., CSS rules are applied in an XML DOM are probably out of scope for this article. Ubernostrum (talk) 23:58, 17 July 2009 (UTC)


 * You are incorrect. "How is the treatment of application/xhtml+xml documents different from the treatment of text/html documents?" "An XML parser (expat) is used instead of the tag soup parser." - Mozilla Web Developer FAQ.  And I apologize for the lack of a signature on my last comment.  Darxus (talk) 02:27, 18 July 2009 (UTC)


 * This article isn't about parsing, it's about rendering and, specifically, about the degree to which relevant standards are applied during the rendering process. In other words, not "what type of parsing engine will be used to produce a document tree from this input", but "will non-standard backwards-compatibility mechanisms be used during the visual rendering process". Ubernostrum (talk) 05:36, 18 July 2009 (UTC)


 * Using a different parser affects how a page is rendered.Darxus (talk) 21:42, 19 July 2009 (UTC)


 * Yes, but it does not affect the level of standards compliance used during rendering, which is the entire point of quirks mode and "almost standards" mode. An HTML 4.01 document with full DOCTYPE will get standards mode. An XHTML 1.0 document with full DOCTYPE will get standards mode. An XHTML 1.0 document with full DOCTYPE and an XHTML Content-Type will get... standards mode. In other words, the Content-Type does not change the fact that the browser will render using the highest level of standards compliance it can manage (it merely dictates which standard is used as the basis). Ubernostrum (talk) 01:10, 20 July 2009 (UTC)

ISO/IEC 15445
It's the only real standard (not recommendation) of HTML, so wouldn't it be useful to see what browsers do with it? —Preceding unsigned comment added by 80.99.3.247 (talk) 12:31, 11 January 2010 (UTC)

Table contains wrong data?
The notice above the table says that some of the information in the table is incorrect. What specific information in the table is incorrect? -- Schapel (talk) 18:41, 13 May 2010 (UTC)
 * compare this actual table with the table in the link I gave: the almost, quirks and standard mode are different in some browser as it is given here. I checked a few months ago the (old) table with the weblinks and it was correct or the weblinks were also incorrect. mabdul 19:08, 13 May 2010 (UTC)
 * I've taken a close look, and it's hard to compare because the order of the columns differs between the two tables, and the old table is so large I cannot always see the headers or footers, but I didn't see any differences. Could you point out a specific piece of incorrect information? -- Schapel (talk) 19:20, 13 May 2010 (UTC)
 * a long investigation I found: opera 7.5 is totally missing; with system identifier and without XML declaration, with system identifier and XML declaration, without a system identifier and without a XML declaration rows are partial removed although the layout engines handle them different... Now I don't think that there is data wrong, but there is data missing that leads us to the discussion above: remove layout engines and/or create a new table for historic browsers/engines. ('think you will know my opinion) mabdul 00:23, 14 May 2010 (UTC)
 * I added without a system identifier and without a XML declaration for XHTML 1.0. With this change, Opera 7.5 looks like it would be identical to Opera 9+. Perhaps we could change Opera 9+ to Opera 7.5+, and the table would be improved over the old one because it would then cover the mysteriously missing Opera 8. -- Schapel (talk) 15:06, 14 May 2010 (UTC)
 * I think you were right: since opera 7.5 and 9+ only are different in the frameset which isn't relevant anymore for most of the developers... mabdul 00:03, 15 May 2010 (UTC)
 * I made the change from Opera 9 to Opera 7.5. Is there any important information missing now? If not, I think we can remove the cleanup tag. -- Schapel (talk) 15:59, 18 May 2010 (UTC)