Chris Wilson, Internet Explorer 7 and CSS 2.1 compliance

"The main problem is that IE does things wrong at such a fundamental level that there are bugs stacked on top of bugs. Many CSS properties have multiple bugs in them" - David Hammond, August 10th 2006

"I have been trying the new IE 7 beta series, and I have to say, from a developer's point of view; big disappointment. There are a few CSS changes; mostly long overdue bug fixes, and a few token extra parts of the CSS standard, just to try to make it look like they are not too far out of the game. The upgrade concentrates almost entirely on the user interface (...) Keep up the good work Dean Edwards, looks like we are still going to be needing you." - Mark Wilton-Jones, August 2006

"there are still annoying bugs that haven't been fixed, new bugs that have been introduced, and important features that are missing. In other words, it is still a pain to make things work properly in Internet Explorer. (...) It's incredibly frustrating to find important parts of CSS 2.1 missing (...) IE 7 will not even catch up to where other browsers were a couple of years ago" - Roger Johansson, September 28th 2006

Microsoft's commitment to implement and to comply with W3C web standards

Microsoft's commitment over the years regarding implementing and complying with W3C web standards in its IE browser has been weak, erratic, incoherent, inconsequent, unpredictable, unreliable and not trustworthy.

More than 5 years ago, Microsoft used to have a webpage defining its commitment regarding W3C web standards which was saying:

"Microsoft is committed to working with the World Wide Web Consortium (W3C) to implement W3C-approved HTML standards, and has confirmed its pledge to work through W3C and other standards bodies on enhancements to HTML and other key Web technologies."

... but, very quickly, that webpage (originally located at http://www.microsoft.com/standards/intro.asp) was removed from microsoft.com website and never reappeared again. Of course, such web-standards-commitment-from-Microsoft webpage would be entirely unneeded, pointless if a majority of webpages at microsoft.com website were all using valid markup code and valid CSS code to begin with, would pass validation testings at W3C, despite an imperfect IE browser. At other browser vendors' (mozilla.org, opera.com, konqueror.org, icab.de, etc.) websites, it is completely the opposite that is happening: a very wide majority of their website webpages use valid markup code, valid CSS code, trigger standards compliant rendering mode, will pass validation tests (otherwise will not generate hundreds of validation errors per page) and they are constantly and indisputably improving their browsers' conformance to W3C web standards.

Regarding W3C web standards compliance, Microsoft either does not listen or does not understand or just does not care.

"CSS 2 is over six years old, and Microsoft have had employees in the CSS working group the whole time. (...) Given that Microsoft are the world's biggest software company, it looks very much like standards are deliberately being sabotaged. Is it any wonder web developers are starting to get royally pissed off with Microsoft?" Jim, July 21st 2004, IE Blog

Microsoft is a big, slow and awkward company when it comes to understand and to meet appropriately, in a timely manner (or even proactively), the needs and demands of its customers. Webstandards.org and many other web standards advocacy groups and websites have during years (many years before June 2004) demanded better support, better compliancy and increased conformance of web standards in IE.

If Microsoft really and seriously cared about W3C web standards compliance, then it would have fixed this simple testpage:

http://archive.webstandards.org/css/winie/inline.html

from webstandards.org, reported by webstandards.org to Microsoft, created by webstandards.org for Microsoft some 9 years ago! Today, December 15th 2006, MSIE 7 final release still fails to render that webpage accordingly.

Some of you may think that I am harsch with Microsoft here but W3C web standards is much more than just CSS. W3C web standards includes also HTML 4.01 specification, DOM 2 interfaces specification, User Agent Accessibility Guidelines (UAAG) 1.0 specification and other key specifications: when comparing with other browsers (Opera 9.02, Firefox 2.0, Safari 2.04), the latest MSIE 7 is behind - way behind - other browsers in all these specifications.

"For the last couple of years, we developers have been struggling with IE incompatibilities while creating and testing our sites. (...) IE7, however, was supposed to solve all these bugs, and add those all missing features. Nothing could be farther from the truth. (...) Although they've successfully fixed many bugs, it's still quite evident that their browser is still far from perfect." - Guillermo Rauch, March 18th 2007

"Solid test cases we can access and bug reporting would help" - Chris Wilson, August 10th 2006

This is exactly what many web developers, web standards advocates have been doing during years (via webstandards.org or otherwise/elsewhere) and, in particular, when they finally heard in June 2004 that Microsoft and its IE dev. team was listening, then the provocative I Love This Browser at IE blog site. Hundreds of web authors overwhelmingly wanted Microsoft to address the issues of conformance and correctness of web standards implementation in its IE 6 browser, a particular wish which became official in a post from Chris Wilson on March 9th 2005. At that time, there was already a lot of websites with articles (which included testcases, screenshots, detailed analysis and explained workarounds) demonstrating IE 6 bugs.

Hundreds of users, web authors reported in good faith, with good intentions all kinds of bugs, spec. violations, useful links, relevant quotes, constructive feedback criticisms and demands, during months at

Still today, the latest version of MSIE 7 (MSIE 7 final release, build 5730.11, released on October 24th 2006) still fails many CSS 2.1 tests. In less than 2 hours, I was able to identify 25 testpages coming from Ian "Hixie" Hickson "evil" CSS website:

25 Ian "Hixie" Hickson CSS 2.1 testpages which fail in MSIE 7
link to actual test pageIE 7Firefox 1.5Opera 9Other browsers which
pass the test page
link to actual test pageIE 7Firefox 1.5Opera 9Other browsers which
pass the test page
Shorthand border propertyFailsPassesPassesNS 6.2, NS 7.0, Amaya 9.53
Character escapeFailsPassesPassesNS 6.2, NS 7.0
Character and CaseFailsPassesPasses
At-rulesFailsPassesPassesNS 6.2, NS 7.0
Forward Compatible parsingFailsPassesPassesNS 6.2, NS 7.0
Malformed declaration missing valueFailsPassesPassesNS 6.2, NS 7.0, Amaya 9.5
RGB color syntax error handlingFailsPassesFails
:visited linkFailsPassesPassesNS 6.2, NS 7.0
Dynamic pseudo-class :focusFailsPasses?Passes?
DOM mutation on :hover effectFailsPassesFails
Explicit inheritanceFailsPassesPassesNS 6.2, NS 7.0
Margin-rightFailsPassesPassesNS 6.2, NS 7.0
link to actual test pageIE 7Firefox 1.5Opera 9Other browsers which
pass the test page
Margin-rightFailsPassesPassesNS 6.2, NS 7.0
MarginFailsPassesPassesNS 6.2, NS 7.0
border-width-rightFailsPassesPassesNS 6.2, NS 7.0
thin bottom borderFailsPassesPassesNS 6.2, NS 7.0, Amaya 9.54
thin left borderFailsPassesPassesNS 6.2, NS 7.0, Amaya 9.54
border-widthFailsPassesPassesNS 6.2, NS 7.0, Amaya 9.5
border-colorFailsPassesPassesNS 6.2, NS 7.0
border-topFailsPassesPassesNS 6.2, NS 7.0, Amaya 9.54
Floating elementsFailsPassesPassesNS 7.0
Floating imagesFailsPassesPassesNS 6.2, NS 7.0, Amaya 9.5
(Amaya 9.54 fails)
border conflict resolutionFailsPassesPassesNS 7.0
border conflict resolutionFailsPassesPassesNS 7.0

"I'd also like to turn comments like 'IE 6 and IE 7 failed to render even basic columnar layouts in CSS' in to solid work items, but I don't understand the problem - much of the web uses columnar layouts that work in IE." - Chris Wilson, August 10th 2006

It's simple to understand, actually. Microsoft's repeated failures to implement correctly many CSS 2.x properties - when and where it particularly mattered for columnar layouts (abs. pos., rel. pos., float, clear, etc.) - lead web authors to search for, to create and to use all kinds of non-compliant workarounds, ugly fixes, invalid code, dirty hacks regarding IE... and they did this during years with newbies copying their column templates (or via looking at their webpage source code) without even understanding their hacks on top of that.

"It is very important to fix CSS/HTML/DOM bugs as they appear, otherwise with time, developers find ugly hacks to work around them and then, much later, Microsoft can not fix these bugs because nobody asks to fix these because fixing these would break their webpages. A very miserable and stupid situation IMO where web design[ers and webpages] become prisonners of bugs and financially hostages of bugs and incorrect implementations. We [web designers and browser vendors] eventually all waste/lose time, money, respect, etc. anyway. (...) Considering the efforts it takes to develop webpages and considering the difficulties, time, efforts, budget/finances involved to develop and to later update and maintain a website, it is imperative for Microsoft to fix in a timely manner, to correct ASAP, the CSS/HTML/DOM bugs that are known and reproducible." - Gérard Talbot, December 6th 2004

"Until IE is reasonably standards-compliant you can expect no quarter from us. We've spent too much time designing web sites that will work for CSS and then work for IE. We've destroyed budgets, pissed off our bosses, irritated clients and made development more expensive because of your little (Rosemary's) baby.

We web developers [are] fully aware that when your sweaty, monkey-boy of a boss was leaping around onstage chanting 'developers! developers! developers!', he most certainly wasn't thinking about us." - Chris, August 10th 2006, Albatross! (Chris Wilson) Blog

IE bugs and IE problems affecting basic columnar layout fit perfectly the explanations above. You will not find basic (or semi-complex or complex) columnar layouts in CSS on the web which does not use/have/rely on hacks - many hacks - targeting specifically IE bugs and IE flaws. How could it be and how would it be otherwise for web designers when IE 6 has been and still is the most popular browser in use on the web and when CSS bugs in IE 6 have been lingering on and on during the past 5 years?

A simple and good example of basic columnar layout that fails in IE 6:
Toby Inkster's 3 column layout at
http://examples.tobyinkster.co.uk/3col

We can easily read the following in the code:

"If you thought Netscape 4 was a pain, now we come to Internet Explorer.(...) we feed IE 5, 5.5 and 6 a bogus 'left:-10em' value. Otherwise the left column will overlap the centre one."

A very typical hack. The web is full of webpages using/relying on CSS hacks of this sort to overcome IE 6 bugs/limitations.

Another good example on this hard and widespread reality: in the .css files of this (Skidoo) document
Skidoo: 2 or 3 Column Layout
http://webhost.bridgew.edu/etribou/layouts/skidoo/index.html

there are at least 15 different kinds of CSS hacks aimed at IE 5+ /* in the comments */ of the Main stylesheets and Extra stylesheets. In those .css files, there are extraordinary efforts to finely control layout bugs in IE 5+.

There are many other examples (pages, demonstrations) showing that MSIE 7 still has problems with columnar layouts: it's just a matter of searching the web, e.g. CSS Layouts at incutio.com has dozens of columnar layouts based on CSS.

Bottom line: a non-conformant browser will require non-compliant code to achieve a particular webpage design. A non-compliant browser will often work "accordingly" with non-forward-compatible tricks or with ugly hacks when it comes to semi-complex layout or to a demanding/precise webpage design.

"bug reporting would help" - Chris Wilson, August 10th 2006

Of course it would help, but ...

"how the hell are we suppose to file bugs [at Microsoft's public IE bugtracking system] if you just close them? I could understand if you changed the status from Active to Future or something like that to indicate that the problem can't be addressed in time for the IE7 release, but marking real bugs as Won't fix just makes me delete the 'bugtracking' system from my bookmarks. Why bother reproducing a bug, creating a simplified test case, redacting everything if you are going to dismiss it right away?

If you want us to take you seriously, then you have to start behaving seriously." - Alphonso, August 10th 2006

"I have already submitted a number of bug reports to the Internet Explorer feedback system with simple and straight-forward test cases. I have been disappointed with the way the feedback system has been run, specifically how so many reports are marked as 'By design' or 'Won't fix' when, according to the explanations in the comments, they should have been filed as 'Postponed' or simply switched from a 'Bug' to 'Suggestion'. I'm trying to be as helpful as I can." - David Hammond, August 10th 2006

Internet Explorer feedback system a disaster: "the [Internet Explorer bug reporting] system has failed." David Hammond, July 30th 2006

Conclusion: MSIE 7 still fails to deliver the CSS compliance that the web of 2006 and web authors deserve in 2006.
Microsoft is still disconnected from web authors' demands, needs and perspective.
Web authors of 2006 deserve a higher degree of CSS 2.1 conformance (and a more complete CSS 2.1 support) from an IE browser and certainly deserve a higher level of commitment into supporting and complying with W3C web standards.
Gérard Talbot, September 6th 2006 (updated on December 15th 2006 and February 2nd 2007)

Valid HTML 4.01 strict! CSS compliant Web standards project