They Shoot Browsers, Don’t They?
Issue № 253

They Shoot Browsers, Don’t They?

Proprietary innovations by browser vendors are nothing new. Internet Explorer alone has given us XMLHttpRequest, innerHTML, and colored scrollbars. In each instance, we were free to use or ignore these non-standard extensions. Now Internet Explorer is introducing a new proprietary technology in the shape of version targeting. But this time, the only way to opt out of using the technology is, perversely, to use it.

Article Continues Below

Ball of confusion#section2

When I first read about version targeting here in the hallowed pages of A List Apart, one point confused me. At the end of Eric’s heartfelt article detailing his reaction to the proposal, the final section seemed to suggest that IE8 would, by default, behave identically to IE7. “That can’t be right,” I thought. Surely I was misreading Eric’s words. To clarify the situation, I asked Chris Wilson what would happen if IE8 were to encounter a valid, well-formed document with a strict DOCTYPE. My worst fears were realized when he confirmed that the browser would behave exactly as if it were its predecessor.

This is gobsmackingly audacious. Imagine a new version of Word that behaves exactly like the old version of Word unless the document it is processing contains a hidden instruction to unlock any new features. That’s what Microsoft is demanding that web developers implement. Unless you explicitly say otherwise, IE8 (and IE9 and IE10, ad infinitum) will behave exactly like IE7.

My incredulity couldn’t be assuaged by the obvious explanations for this behavior—that Microsoft was being stupid or “evil.” The Internet Explorer team is made up of good standards-savvy developers. They must have a good reason for proposing a solution which, on the face of it, appears so crazy.

Destroying the web to save it#section3

Microsoft’s proposal was triggered by a traumatic event: the upgrade from IE6 to IE7. Internet Explorer 6 languished in the doldrums of non-development for many years. Eventually, spurred on by the encroaching market share of rival browsers, Microsoft released Internet Explorer 7 sporting far better CSS support than the previous version.

Because IE6 stagnated for so many years and because it remained the market leader, a whole generation of websites had emerged that were coded to the quirky but predictable vagaries of that browser. These websites appeared to “work.” That is, they looked fine in the most popular browser on the market. But when IE7 was released, these websites were inevitably rendered differently. IE7, with its improved support for Web Standards, rendered these sites in much the same way as any other standards-compliant browser. Despite a concerted campaign to encourage developers to use conditional comments instead of browser-specific hacks, Microsoft received a barrage of complaints from website owners upset at the way that IE7 had changed the game. This is what the Internet Explorer team are referring to when they talk about “breaking the web.”

That’s a loaded phrase that doesn’t stand up to closer scrutiny. Firstly, what’s at issue here is not “the web” but “some websites”. Secondly, rather than “breaking”, it’s more accurate to say “displaying differently.” Finally, it’s important to remember that we are talking about how websites are displayed in one browser: when the IE team talk of “breaking the web,” what they really mean is that their browser will display documents in much the same way as other modern browsers do. Would that really be such a bad thing?

Won’t somebody think of the children?#section4

On the face of it, being the market leader is something to aspire to. But think of how much responsibility that entails. Would you really want to innovate and push the boundaries when even the smallest changes could cause disruption for thousands of your customers? This is exactly the kind of paralysis that Microsoft is trying to break out of. The version targeting proposal is a good solution to this deadlock. With the addition of one meta element, websites can specify exactly how they should be rendered (in one browser).

Furthermore, had Microsoft implemented the X-UA-compatible instruction in IE7, they could have saved themselves a whole mess of trouble. Instead of requiring developers to revisit their style sheets and strip out their browser-specific hacks, they could have instead told website owners to simply add one line to the head of their documents. While it’s hard to imagine that the move from IE7 to IE8 will cause the same upheaval, it’s reassuring to know that Microsoft has thought ahead. Version targeting allows site owners to freeze rendering (for one browser) to a specified browser version. That’s a good thing. While it probably won’t affect standards-savvy developers like you or me, it offers a simple solution for site owners who don’t want to worry about the future. Better still, the fact that the X-UA-compatible instruction can be sent as a header means that this issue can be taken care of by sysadmins with one small tweak to their server configurations.

But even that is asking too much, according to Microsoft. Instead of asking that developers who want to opt out of future improvements do so with the addition of a meta element or header, Internet Explorer expects standards-savvy developers to actively opt out of version targeting… by using version targeting.

The reasoning here is that less savvy developers shouldn’t have to worry their little heads about adding one extra line to their documents. Instead, they should be encouraged to continue to write to the quirks of one specific browser version from the market leader. That their documents will “break” in other browsers is not Microsoft’s problem. The counterpoint to this condescending worldview is that standards-aware developers are the ones best placed to add a single line of markup to their documents—though, for some unexplained reason, the instruction for up-to-date rendering (IE=edge) is strongly discouraged.

This strategy is doomed to failure. Standards-aware developers, by their very nature, will object to adding a line of unnecessary markup to their documents just to get one single browser to behave as it should by default.

Fear of drowning#section5

While most of the web development community saw the release of IE7 as a welcome return to form, within the corridors of Redmond it was viewed as a failure. Microsoft simply cannot afford a repeat of the IE7 upgrade. Version targeting is a technology born of fear. A fear of “breaking the web”—which really means “rendering some websites differently in one browser”—has prompted the draconian default behavior.

Whether this fear is well-founded or not depends on just how drastically IE8 is going to “break” existing websites. Personally, I’m rather puzzled: what exactly are they planning to add in the next version of their browser to make the web asplode? If IE8 is going to differentiate itself from its predecessor by having better standards support, then surely we can assess how it will render websites by simply viewing those websites in a standards-compliant browser like, say, Firefox, Safari or Opera.

Lonely at the top#section6

There was a time when Friendster was the biggest social networking site on the web—MySpace and Facebook were little more than distance glimmers on the horizon. There was a time when Netscape Navigator was the undisputed king of browsers and Internet Explorer was laughable challenger playing catch-up. On the World Wide Web, the status quo is a mutable, shifting thing. The proposed default behavior for version targeting is predicated on events that took place during a short span of years when Microsoft, having emerged as the top dog, pulled the plug on its own browser. There is an unspoken assumption that the only meaningful way the web is experienced is through one browser: Internet Explorer.

We are being told that the default version targeting behavior is necessary because without it, the web will turn into a messy crime scene of breakage (in one browser). If Microsoft are to be believed, the self-crippling default behavior of IE8+ is necessary to save the web (in one browser). Whether you agree or disagree with the default behavior comes down to faith: faith in Microsoft accurately foretelling the impact that IE8 will have.

I would much rather base my judgement on facts. There is an easy way for Microsoft to prove the necessity of mandatory version targeting: release a beta version of IE8 with version targeting disabled by default. Then we can see just how badly the web breaks some websites render differently in one browser.

I’ve listened to and understood all of the arguments in favor of the proposed default behavior: all of them assume that without self-crippling, IE8 will make a mess of a significant portion of the web. If that fear is borne out by an uncrippled beta release of the browser, I will back the proposed default behavior. Until then, I ask that Microsoft honor their promise from many years ago and allow their browser to render a valid, well-formed document with a current DOCTYPE to the best of its abilities.

Future imperfect#section7

Version targeting is not a bad idea. The choice of delivery mechanisms—meta element or server header—is inspired. As an optional feature, this could prove to be a real lifesaver in some development environments. As a mandatory millstone however, it strikes a blow against progressive enhancement. [1]

The proposed default behavior for version targeting in Internet Explorer solves the problem of “breaking the web” in much the same way that decapitation solves the problem of headaches. In its current state, version targeting is a cure that will kill the patient. Version targeting could have been an opportunity for Microsoft to demonstrate innovation. Instead, the proposed default behavior demonstrates a fundamental misunderstanding of the World Wide Web, a place that according to its creator, Sir Tim Berners-Lee, will always be “a little bit broken.”


[1] On the plus side, as-yet unpopular DOCTYPEs such as HTML5 can be used to trigger up-to-date rendering from future versions of Internet Explorer. That’s reassuring for the future but HTML5 is not ready for use today—any DOCTYPE that still includes the FONT element still has some issues that need to be worked out. Besides, once HTML5 is widely deployed, Internet Explorer will probably freeze its rendering for those documents too.

91 Reader Comments

  1. I feel your pain on principle, but the reality of the matter is that I’m already using conditional comments to target for both IE 6 and IE7. You say the team is staffed by standards savvy people, but their product doesn’t reflect that. Z-indexing on positioned elements, float issues, filters, and non-css stuff like dynamically adding options to select elements via innerHTML, the list goes on. How about adding support for xhtml mime type? Heh. What’s one more inclusion of IE browser version targeting?

  2. As there are thousands of websites out there tweaked to look in a specific way with a specific browser the default behaviour of IE8 saves as a big time saver for all the website authors. Those who want to implement their websites for IE8 will have to do some changes and testing anyway. So I don’t think it’s too bad to add this extra line.

    The only problem I see that this solution is not “beautyful” in the eyes of a techie. Well, but it’s pragmatic.

  3. The idea that IE may ‘freeze the web’ without developer intervention, from version 8 onwards (IE8, IE9, IE10, etc), is ugly and a massive disappointment. Particularly when a similar solution would be quite elegant (opting _into_ backwards compatibility, as opposed to opting _out_ of it).

    I don’t understand why some people seem happily resigned to deal with specific versions of a browser for the rest of their careers. Just because that’s the case now (with conditional comments, etc), does not mean it should be the case moving forward.

    I see the irony: it’s one small line — what’s the harm? However, this one small line challenges the evolution of the web and promotes web stagnation (in one browser).


  4. As much as I agree with your conclusions, I’m not sure about one assertion:

    bq. If IE8 is going to differentiate itself from its predecessor by having better standards support, then surely we can assess how it will render websites by simply viewing those websites in a standards-compliant browser like, say, Firefox, Safari or Opera.

    Surely most sites that will break are those that have ‘if IE…’ logic trying to cater for broken behaviour in IE 6 and 7? Those sites will still render correctly in a standards compliant browser, but (presumably) not in IE8, unless they’re specifically targetting older versions separately, with no catch-all ‘other IE’ case.

  5. bq. Imagine a new version of Word that behaves exactly like the old version of Word unless the document it is processing contains a hidden instruction to unlock any new features.

    Is this really the case? Surely it would be more akin to a new version of Word displaying the old document how it was supposed to look; which seems to me precisely the behaviour you would want. If every time I opened a Word ’97 file in Office 2008 it looked completely different to how it originally looked I’d get a bit annoyed.

  6. Having had a month, pretty much, to think about this issue, I’m still no closer to coming to a firm conclusion – but my gut says that the proposed default is the best possible option.

    1. It allows intranets, mission-critical sites such as online banking, un-maintained Geocities sites, and cat diaries to continue to render as correctly as they do now, by simulating the IE7 rendering engine. (Or IE6, if that is their baseline and they choose to add the meta tag.)

    2. It allows standards-aware developers who fixed their sites when IE7 was released to _do nothing_ and be confident that their clients’ sites will not break. If you sold your services as being futureproof, this is a big win.

    3. It allows standards-aware developers to use edge targeting (or the HTML5 DOCTYPE, or some other setting) to be able to create sites that take advantage of the advances present in IE8 (and FF3 et al), further setting ourselves apart from the hacks.

    That all makes sense to me. Where I have doubts are around the accessibility issues (as documented by Bruce Lawson); JavaScript compatibility issues, including across frames, as pointed out by several people; and, most importantly, Microsoft’s ability to release pitch-perfect backwards-compatible rendering engines that don’t introduce new errors on top of the ones they are trying to reproduce.

  7. I don’t see the purpose of new web browsers, if people can decide to display their sites the way former versions do.
    I don’t doubt, that standards savvy people are working for Microsoft. That’s why I suspect another reason behind Microsoft’s plan.
    C’mon Billy boy, do it like the automobile industry! Buy your engine from a competitor and paint it microsoft.

  8. Those ‘Microsoft’ engine control units (ECUs) in Formula 1 racing cars ? That’s what they did. Bought McLaren electronics, changed the badge 🙂
    I simply do not believe that MS can recreate all of IE6s bugs, from now on until the end of IE, without getting it wrong a few times. So the switch wont do what it says.
    FireFox is a lot better, and in Europe has 25% of the browser market, and they feel very happy to actually fix bugs as opposed to codifying them and promising to have them forever, unless the web page asks nicely.

  9. Jeremy, I’ve been following this debate with interest and definitely support you in your view. It’s good to see ALA responsibly allowing you the opportunity to put the counter-point to Eric’s original article.

    My thoughts have consistently echoed yours – as in what on earth do they think IE8 is going to do to jeopardise the rendering of so many websites, given that a lot of pain has been cleared by the switch to IE7? But then it started to dawn on me that actually all those site owners who now have reasonably coherrent rendering in IE7 are not the point.

    The point is the thousands of companies out there who have intranets and web apps poorly developed by short sighted business owners and developers that will only render using IE6. These businesses have actively put off upgrading, both to IE7 and by extension Vista and this is the real reason MS are so keen to be able to guarantee predictable behaviour going forward. So they can say to their customers ‘we promise that what you create with our new products will continue to work in the future’.

    Unfortunately we face a divergence. Not between one browser maker and another. But between the established corporate world and how it uses the web and the real innovators. Not only are the corporates heading down an evolutionary blind alley but they’re being lead there by their ‘trusted supplier’.

  10. Rather than saying which version of IE the site is built for”¦ can’t we say when the site was built?

    Then IE just needs to say “this site was built before X, but after 2006-10-18, therefore this website must have been built for IE7”?.

  11. Einstein said, “Insanity: doing the same thing over and over again and expecting different results.”

    This is exactly what Micrsoft is asking us to do. “Keep doing the same thing, we’ll change, really. Just wait till the next version, we’ll get it right then.”

    We already have to use conditional comments to target IE6 and IE7, now we have this new switch.

    Do people seriously believe things will change in the future?

  12. bq. Can’t we say when the site was built?

    I don’t see how that’s any easier than just putting in the current version of IE. Plus it gives you less control over the behaviour.

    bq. Unfortunately we face a divergence. Not between one browser maker and another. But between the established corporate world and how it uses the web and the real innovators.

    Does the divergence not already exist? Is that not the problem: they don’t want to break the ones that have diverged down the non-standards route.

    It does seem like deep down this is all about Microsoft making more money. If it was really about the intranets and such then those businesses could just use older versions of IE. But of course then they can’t upgrade to Vista. So if you bring out a version of IE that supports the older intranets you can then get them to upgrade to Vista – it’s not like they desperately need tabs and an MSN Live search bar. The best way to make sure intranets don’t break is simply not to upgrade from IE.

    Having said all this, i still don’t see Version Targeting as a huge issue. Sure it’ll make Microsoft’s lives harder. But as long as you use purely standards code you should be ok for the future if you use targeting.

  13. bq. Then IE just needs to say “this site was built before X, but after 2006-10-18, therefore this website must have been built for IE7”?.

    That assumes that people always write sites to the latest standards. Which they clearly don’t. There are new sites being launched every day that rely on IE’s bugs to render properly. There are probably new and revised pages being uploaded to sites every day that still require IE6 bugs to render properly.

    All you’re doing is making the targeting even more complicated and unnecessary than it needs to be.

  14. _This is pulled from my response on Eric Meyer’s blog article._

    I agree with the direction IE8 is taking. The problem is that people are confusing standards for requirements. According to the W3C the standards are recommendations and not requirements. The W3C provides standards that are relegated to mere guidance while the browser vendors provide the requirements. This is not my desire to favor browser vendors, but the reality in which the internet exists.

    *Identifying the problem – faulty reasoning*
    The problem is confusion between those “R” words. If the standards are not requirements then the idealism for standards lust based arguments looses idealistic appeal. The rational result is to ensure the recommendations become requirements, or at least enforceable options.

    *Identifying the solution – a migratory path for browser vendors*
    Unfortunately the standards are written as doctypes and not schema, so validation of code is not capable of being enforced. Browser vendors are more capable of enforcing the standards that are otherwise unenforceable merely through field use. The solution to the noted problem is help browser vendors carry the torch of standards enforcement. Notice that I said *_standards enforcement_* and not *_standards compliance_*. In order to achieve this most desirable solution without immediate alienation of backwards conformance browser vendors must adhere to a three part plan:

    1) – *_Abandon slackware:_* Browser vendors must implement an optional standards mode that fails on non-compliant pages. This will allow designers to create future compatible content that is nothing less than well formed.

    In the case of IE8 the IE8 processing engine must be optional if it actually wishes to be strict to the standards. Internet users must be aware that IE8 is not requiring conformance to the standards of IE8, but it will. This allows content owners to get their assets up to par before they are exposed as flawed, archaic, or simply incompetent. This would also satisfy the requirement of point 3 – timely notice.

    2) – *_Version control:_* Browser vendors must implement a version control system to warn web content owners that the browser will significantly limit backwards compatibility in future versions. This will allow time necessary for content owners to conform to the standards before their pages begin to break.

    3) – *_Timely notice:_* Browser vendors must make public what a browser version will no longer support and what new features it will support 9-18 months in advance to allow the market to prepare for those changes. Without timely notice the previous two points are irrelevant. The point of timely notice is not to benefit designers or asset owners, but to protect the browser vendors. Browser vendors must view themselves as always legally liable and take steps accordingly to continually provide innovation while ending support for older versions that impair such innovation.

    *Intended result – standards become requirements*
    The goal is not to improve code. The goal is to make content asset owners aware that archaic content and markup methods will expire and that it is their liability to stay up to date. Without liability there are no requirements.

    *The only alternative for standards achievement – abandoning HTML*
    If standards are important they would be enforceable. If browser conformance to those standards were important then they would be given the liability of enforcement. If the mentioned steps are too much to ask for then HTML creation will never be standards conforming. Requirements are extreme in their strictness and lack of compromise or they are not requirements. If standards are not entirely followed then they are not standards.

    If the mentioned steps are not appetizing then simply abandon HTML. Create a new markup language that is more semantic and defined by schema so that validation is a requirement.

  15. bq. The reasoning here is that less savvy developers shouldn’t have to worry their little heads about adding one extra line to their documents.

    I don’t think the argument is that they *shouldn’t*, rather that they won’t, however much we want them to.

    bq. This strategy is doomed to failure. Standards-aware developers, by their very nature, will object to adding a line of unnecessary markup to their documents just to get one single browser to behave as it should by default.

    I dunno. I add a heck of a lot of code to get IE 6 to (almost) behave how it should. And even if we don’t add the line, that just means our sites will be rendered by the IE 7 engine for ever more. I don”˜t think that means the strategy has failed. If anything, Microsoft will get fewer calls complaining about web sites that look different in IE 42. So, from their point of view, they win.

  16. This article touches on the real reason for Microsoft version targeting: “… a technology born of fear”.

    This has nothing to do with “helping build a better web”, “saving clueless web devs from themselves”, “helping standartistas build better sites”, or any of the other nonsense coming from the Zeldman camp.

    This is all about protecting Microsoft their market share, which is being paired away every day by secure, standards-compliant browsers.

    If they can create a de facto IE7 standard, which this proposal will bring about, all the fully standards-compliant browsers are suddenly broken in terms of the new ‘IE7 Standard’. Bingo! MS has the only non-broken browser!

    It’s absolutely shameful the handful of ‘big hitters’ who are arguing for it, and worrying that so many others are falling for their bogus arguments.

    Anyway, what’s the debate about? Microsoft will do what suits Microsoft profit margins, and fuck the rest of them. Nothing new there.

  17. I can’t help but feel this decision has already by made. So spending a lot of time debating the pros and cons seems like wasted time. There’s a good and a bad side to this. MS are the big boys in town and until that changes we will always have to accomodate their quirks. Just use “edge” and that should do for most of us me thinks.

    My own feelings on this can be summed up like this:

    Can’t Microsoft just release an intranet version of their browser for people to use when IE8 doesn’t seem to work?

    I guess not, but for me the best compromise is to actually let the user decide using a “Does this site look broken” button which users can click on to switch to IE7 mode. I know that’s probably far too much to ask of users, but wouldn’t that be nice?

    Finally, I still don’t believe that IE42 will ship with support for IE41 all the way down to IE7. MS will pull the plug on this eventually, once the “business case” has been made for it.

  18. Surely IE7 and conditional statements already cleared a path for IE8?

    How can IE8 ‘break the web’ if there are already conditional statements in place to ‘fix’ the site for previous versions?

    Won’t it just look the same as compliant browsers?

  19. Jeremy:

    bq. …less savvy developers … should be encouraged to continue to write to the quirks of one specific browser version from the market leader. That their documents will “break”? in other browsers is not Microsoft’s problem. The counterpoint to this condescending worldview is that standards-aware developers are the ones best placed to add a single line of markup to their documents…

    This is an excellent summary of what’s wrong with the proposal — it’s predicated on an eternal continuation of an out-of-date status quo.

    Thanks for writing this article — it adds some clarity to the anti-targeting argument (and I agree 100%).

  20. Surely MS should stick with the method of version targeting they have already started rather than introducing a new META tag.

    Why not just add a new conditional comment, something like



    At least then there’s a modicum of consistency.

  21. The only issue with this that I haven’t seen addressed is how Microsoft is going to handle all this backwards-compatibility in IE47. It’s going to be one bloated browser. This is obviously a doomed plan unless they’re only going to use it for a few iterations and then drop off support for it at a later time. If that’s the case, why introduce it in the first place?

    I’m not saying that I have a better solution for Microsoft’s problem, but I don’t understand what happened to the web standards movement. Instead of promoting standards, we’ve moved on to accepting ignorant developers and even catering to them with this new idea of versioning. So, some websites don’t work in an antiquated browser? It’s not like they can’t be fixed (if they’re even worth fixing), and it points out who the competent developers really are. The advantage of the Internet is that it’s flexible; it can be re-made. You can’t “break the web”.

  22. Microsoft broke the web when they decided to include the MARQUEE and BLINK tag, and decided not to follow standards.
    Microsoft broke the web when they realized that Javascript was not in their control and created Jscript to fill their corporate needs.
    Microsoft’s been broking the when since I have memory of it, including it’s own non-standard technologies (like ActiveX), ignoring the direction the WWW was taking and creating their corporate bubble, an ill fated environment where rules are different that those of the real world.
    And now they want *us* not to broke _the_ _web_?
    Which web, Microsoft’s web or the W3C web?
    Frankly speaking, I’m tired of MS. I’m tired of always playing by their rules. I’m tired of always have to take care of their problems, of their lousy browsers, of they not playing by the rules and adapt all I do to fit their _peculiar_ view of the web.
    It’s like one of those bad relationships that you just can’t get out of.
    That I’ll have to modify my code every time a version of IE comes out to use the last engine of the browser? Yeah, sure…
    Either they create an _Intranet_ _Explorer_ or something like that with the same engine of IE6 to satisfy their corporate customers or they get out of the browser business at all.
    This time, I’m not going to help them fix _their_ _web_.

  23. We have all been hearing the “broken web” argument. I think that’s a red herring which hides other reasons. Knowing those reasons might help us make better decisions about what we want to do.

    For example, David One said:

    This has nothing to do with “helping build a better web”?, “saving clueless web devs from themselves”?, “helping standartistas build better sites”?, or any of the other nonsense coming from the Zeldman camp.

    This is all about protecting Microsoft their market share, which is being paired away every day by secure, standards-compliant browsers.

    I have yet another that I have not seen mentioned in the various places discussing the proposal. Yes, we hear “intranets” as a factor, but let’s explore that a bit further. I know that one very large corporation (over 300,000 licenses) said “No Thanks” to Vista last year. Others have too. The reason isn’t exclusively about IE breaking intranet applications, but IE is definitely a factor. Other reasons include the need for more memory, faster machines, and long arduous testing cycles before upgrading.

    When a firm has 50,000, 100,000, or 380,000, seats, saying “No Thanks” to Vista is a gigantic revenue loss for MS. MS is used to nearly automatic upgrades from these large firms and now they’re hearing far to many “No Thanks” from them.

    The biggest thing this proposal does is enable those large firms to upgrade with no IE changes at all. From what I see, this is definitely not about breaking the “everyman’s website,” but about Vista revenues.

    So, should we agree to freeze the web at the IE7 level so that MS can regain their lost Vista revenues?

  24. Just had to point out my favorite line of the article, very well said my friend 🙂

    “The proposed default behavior for version targeting in Internet Explorer solves the problem of “breaking the web”? in much the same way that decapitation solves the problem of headaches.”

  25. I’m a bit amazed that people still defend the move of Microsoft by saying “but otherwise intranets that are based on IE6 will break” or “old unmaintained sites will still function”.
    The brilliance is the inclusion of the HTTP header, intranets could be fixed on IE8 by adding one single entry to the webserver fo the whole company. Cheap, easy to maintain, applauded by every manager garding a budget.

    What I don’t understand is why people insist that old sites won’t be broken. What they mean to say is that IE6 rendering won’t be broken, but if mom and dad replace their old Dell PC by a shiny iMac they will very much get a broken website. Microsofts domination will end, that is for sure. The question is when. Galbraith has shown in his studies that not a single monopoly is sustainable.

    The whole point of this article is the argument that Microsoft made this decision driven by fear, but what they fear will happen anyway at some point in time and will only work in their disadvantage, not in their advantage.

    I’m not implying that the decision has been made by the IE team alone, but that they had to negotiate this as a viable option to move forward.

  26. Jeremy: thanks for sticking to the argument against the default behavior; I was sorely afraid that this issue of A List Apart was going to be two arguments in favor of the version targeting without any balance.

    The only part of the article that puzzled me was the second paragraph of “Fear of drowning” (and it was only puzzling in light of the announcements that the DOM has been overhauled in IE8). If the DOM is really that much improved in IE8, then I can understand the concern about breaking the behavior and functionality of web sites tailored to lte IE7.

    However, I’m still not convinced that this is the only way of implementing version targeting that reaches the goal of standards compliance and backwards compatibility.

    I would presume that most scripts sniff for MSIE in the UA string. If IE8’s UA string is changed to exclude MSIE (and perhaps include Gecko and/or keep Mozilla), then scripts will act like IE8 is a Gecko/Mozilla variant and thus give it code that’s compatible with the improved engine.

    This solution admittedly doesn’t solve the intranet problem. However, I would think that making IE6 available in Vista would cost *less* than including previous rendering engines with each new version of IE. They could even lower the cost of producing IE by utilizing a standards-compliant engine with their “new” UA.

    There are lots of solutions to this problem; I think it’s just a matter of analyzing data. If the IE team has done that, it would be great to see some numbers to put (at least some of) us at ease. While they’re at it, seeing a game plan (such as always keeping in latest-standards mode, or how they are going to decide when to drop an old rendering engine from a new version) wouldn’t hurt either.

  27. Jeremy,

    Thanks for the article. I still agree that this is a boneheaded move. Not so much for Microsoft, but for the rest of the web standardistas.

    I’m sorry, Jeffrey, but I don’t think that freezing rendering at IE 7’s level is the right move, whether that be for CSS support, Javascript support or otherwise. I don’t understand why it is that you support this so fervently.

    Why are you for non-advancement of web standards implementation in a market leader? Because that is what will happen. Especially if you take into account all of the clueless web developers. Breaking their sites is a way of educating them. That education could be in the form of showing them how to fix their website the easy way by using the version targeting or the better way by actually implementing the standards. Why do you think we need to protect them? Were you not the one that hung banners on websites proclaiming “Down with bad browsers”?

    I don’t mean to attack or insult, I am just perplexed. This solution does not make sense, yet one of the people involved in bringing awareness of standards is wholeheartedly supporting it.

    I am not against version targeting. I am against default behaviour. I will not create a better web. At best, it will confuse standards newbies, at worst it will discourage the unwashed masses from learning standards ever.

    To whomever said that W3C standards are recommendations and not requirements, that may technically be so, but consider the alternative. If you start telling people that they don’t need to use standards for web development, what is going to happen?

    It’s too scary to think about that.

  28. bq. I don’t understand why it is that you support this so fervently.

    Fair question. I don’t support it fervently. I accept it as an inevitability. It is what Microsoft intends to do. And it’s the only way we’ll get proper DOM support in IE.

    Netscape won the scripting wars (its JavaScript became standard ECMAScript) but lost the browser wars.

    IE “won” the browser wars but lost the scripting wars. Its JScript is nonstandard.

    If IE hadn’t “won” the browser wars, it and its nonstandard JScript would be an amusing memory of failed technologies gone by, like Netscape’s JavaScript Styles, or Apple’s Hot Sauce. Remember those? Oh, what fun we had.

    But millions of people wrote JScript before the web standards movement could take hold. And many still write it today.

    Opera, Safari, and Firefox don’t have to worry about that, because they support standard ECMAScript. Rock on!

    Microsoft, in IE8, will support the DOM correctly, or so we have reason to believe. But if people targeting IE write JScript, their sites won’t operate correctly in IE8, unless Microsoft makes improved compliance an opt-in.

    Under the circumstances, putting a gate in front of proper DOM support seems like something Microsoft must do, whether we like it or not.

    So I’m looking on the bright side. IE will finally have proper DOM support, and will continue to improve its CSS support. That is good. Whether you dislike the opt-in or not, better web standards is a good thing.

    What is the alternative? Microsoft can’t commit seppuku for having sucked in the past. It can’t put itself out of business. If it had to choose between the masses that think IE=web and the smart minority that use web standards, it wouldn’t choose standards. If Microsoft _stopped_ supporting standards, we’d have real web balkanization again. We’ve come too far to contemplate splitting into IE and standards camps. The only way to go forward is in synch — with everyone supporting standards. Microsoft intends to do so, but this is the price.

    Would it be better for you and me if Microsoft could be as good as Firefox without requiring us to opt in via a meta element or http header? Yeah, it would be better for you and me. But it isn’t going to happen.

    And there are some benefits for you and me to what Microsoft intends to do, as I outlined in the end of my article.

    The glass is more than half full.

  29. I see the point of Jeffrye Zeldman, and I understand and respect his point.
    But my question still remains: why us? Why put the burden on us, who cared for standards, who learned how to use it, who promoted them?
    I still think it should be Microsoft who takes care of it. Fix it at the server side, put a checkbox on IIS that says “force IE6 compatibility”, release an _Intranet Explorer_, train your clients on how to make things work for them or something. But please, don’t put the burden of your errors on the guys who care about standards. The burden’s been _always_ on our side, and it looks like Microsoft wants it to keep things that way…

  30. First of all, Jeffrey, thanks for a fair and detailed answer. I do now see the difference in opinion. You have accepted it as an inevitability. I am not prepared to do so currently and as long as we’re discussing it and not actually using it.

    You’ve asked what would I have Microsoft do? In my fair opinion, what they SHOULD do is take the considerable time and effort it will take in recreating IE 7 and IE 6 bugs in multiple rendering engines for IE 8 and put it towards educating the ones who are not aware of standards. I don’t personally care for the people who wilfully ignore standards. They do so on their own peril. However, the ones who are simply not aware of standards-based web design could and can be helped. Hell, I will be willing to personally jump in and help out with the education. I have been doing so at my current position and I have made changes.

    I understand that Microsoft is wary of sites not working in IE 8. Who can blame them? They got unfairly blamed for IE 7’s improved support. It is very hard to trust them, though. How can one do so after the browser wars, JScript, etc.? Freezing standards support on the market leader to what it is now is not a good move. I don’t even think it’s a good compromise.

    I don’t want them to kill off IE, I don’t want IE vs standards separation. But I also don’t want stagnation and, while IE 8’s DOM and CSS support may be terrific, if we don’t educate the uneducated, our field will stagnate.

    Do you feel otherwise?

  31. Sorry, first paragraph should read:

    First of all, Jeffrey, thanks for a fair and detailed answer. I do now see the difference in opinion. You have accepted it as an inevitability. I am not prepared to do so currently and as long as we’re discussing it and not actually using it, IT WILL STAY THAT WAY.


  32. If you are Microsoft, the proposed default behavior makes perfect sense. *If you are Microsoft*… is the key here. The “fear” mentioned in this article is not the fear of “braking the web,” it’s the fear of loosing market share. That’s what this is all about… This default behavior is a *business* decision. If you side with Microsoft decision, then you agree that Microsoft market share is somehow linked with the future of the web. Pretty shady if you ask me…

    Thankfully, at this point, besides Zeldman and Microsoft, I can’t seem to find anyone who still argues that this default makes sense.

    At the end, it won’t matter though… Microsoft will do what it does best: more money.

  33. I think the proposal for a beta version is a great idea, we don’t even need v.targeting turned off, we can use “edge” if we’re curious for testing and see how it goes.

    And honestly, I think it could be an eye-opening experience for a lot of developers to see their sites render unpredictably. From my experiences, if you’re writing clean, semantic markup with solid CSS behind it your websites should work just fine cross browser/platform without any CSS hacks, browser sniffing or conditional comments.

    I welcome the challenge, but I do hope Microsoft listens to the public about the default behavior in v.targeting and also provides a good lab for beta testing.

  34. this is because Microsoft are self serving… there’s loads of legacy apps all using the old IE engine and instead of fixing these MS has chosen to simply maintain compatibility with their existing systems and give web developers the choice to opt out of quirks to standards… this simply is not right!

  35. I can’t believe some people are seriously entertaining the idea, I do agree with a couple of comments asking why we (the ones who have worked so hard to make the web better) have to carry this burden?

    Why us, this opt-in proposal goes against everything we have been working for, why do we have to opt-in to standards compliance? Considering we have already opted in by doing things right, those that chose to do things wrong should be the ones that have to deal with their decision to “code badly” they should be given the option to opt-out of standards compliance, not the other way around.

    I feel like we have been digging out way out of this hole for a decade, and just when it seems we have reached the surface we cop a fresh load of dirt. Well it’s easy to see where the next decade is leading… A lot more digging.

    Shame MS shame, and shame anyone who is willing to lay down a take it up the proverbial.

    My 2c

  36. [quote]That assumes that people always write sites to the latest standards. Which they clearly don’t.[/quote]

    But they should. And now they don’t have to. EVER.

  37. Whether you dislike the opt-in or not, better web standards is a good thing.

    Better web standards for who? We just allowed crappy ‘developers’ to NEVER use web standards, ever, for all eternity. Did I say we? I meant MS…

  38. bq. If they can create a de facto IE7 standard, which this proposal will bring about, all the fully standards-compliant browsers are suddenly broken in terms of the new “˜IE7 Standard’. Bingo! MS has the only non-broken browser!

    And what’s so different from now? Microsoft has created an ‘IE6 standard’ that leads clueless authors to write pages that are broken in fully standards-compliant browsers.

    At least this method will allow authoring technologies, including (if we’re unbelievably lucky) “Save as HTML” in future versions of Orifice, to generate standards-compliant code _with_ the necessary tag to invoke IE8 mode.

  39. I do wish MS would just bite the bullet, finish the transition to full standards compliance the way they should have done with IE7 (and the pains of transition would *already* be a thing of the past). All of this would be a non-issue. I do understand and even sympathize with MS, what with upsetting its hoard, but standards are important and non-compliant sites really need to be brought up to date. It’s for everyone’s own good in the long-term. It seems to me that this whole issue also presents a new set of problems as it concerns the “adoption of new technologies”:

    I see this move as nothing more than buying some time. The inevitable is inevitable, and eventually everyone will have to be on the same page.

  40. I think what microsoft is opting to do with ie8 may have made sense with their release of ie7, since ie7 rendering engine is so vastly different from ie6. not sure why Microsoft would want to freeze browser rendering from ie8 going forward.

    I’m not sure that their intent is to inhibit browser development looking to the future, but rather to preserve how legacy sites render. But as stated in this editorial, this is precisely how their choices can play out. Thanks again for another thoughtful piece of analysis.

  41. In reading the previous articles on this subject I was unaware that the default would be set as basically the IE 7 version. That seems ridiculous. It makes more since to me to have the default on the “edge version” which sounds like the way it is now on every browser. So that when new advances come out people are more likely to use them. I’d also say that making the people that don’t want to change there site add the one line of code. I think they could handle that, and the beta idea sounds like a good “safe” first step for testing if a site will break.

  42. Suppose someone decides to keep their site at a certain lower IE version. . . a couple years down the road that version that looks fine on an older IE version will probably not look so fine in other web standards browsers IE Firefox, Safari, Opera. At that point if they want their site to not break on other browsers they have to play catch-up no matter what. That is unless they only care that it works on IE browsers, which seems doubtful.

  43. For any company who has to maintain many web sites over time, anything that we can do that keeps the sites from breaking is good news. I would have preferred that they default to the standards approach, but understand the “don’t break the web” mentality. I hope a default meta tag can be set at the server for all pages, which would allow us to set our own default. But we will definitely be using this feature, and

  44. bq. Standards-aware developers, by their very nature, will object to adding a line of unnecessary markup to their documents just to get one single browser to behave as it should by default.

    Oh, I don’t know about that. I’m pretty standards-aware but it’s not my religion. I’m also deeply practical. I’m going to do what gets the job done. Sure I want to do it as efficiently and elegantly as possible. But I’m certainly not going to eschew practicality in favor of an ideal. I’m going to weight the benefits of my actions and figure out a happy compromise.

  45. Zeldman: “If Microsoft stopped supporting standards, we’d have real web balkanization again. We’ve come too far to contemplate splitting into IE and standards camps. The only way to go forward is in synch—with everyone supporting standards. Microsoft intends to do so, but this is the price.”

    From the information that’s been publicly available, I see no basis for your statement that this is the “only way to go forward in synch – with everyone supporting standards”.

    With the default rendering of future IE versions being IE7. I see no roadmap from Microsoft that indicates when IE7 rendering will disappear. I see no declaration from Microsoft of when IE7 rendering will be removed or end-of lined.

    That means, with the information we have publicly available right now, web standards developers will have to support IE7 permanently. That alone doesn’t sound like a way of “everyone supporting standards” – that sound like “everyone supports IE7”.

    So that makes the meta tag switch a complete waste of time. And its clear Internet Explorer has no future or practical roadmap. The permanent existence of IE7 is the balkanisation of the web – a permanent unhealable scar on the web.

  46. Does anyone else see this as MS taking the cowardly way out?

    “We’re afraid to ‘break the web’, so we’re going to freeze time and display websites in our current (but still outdated) web browser. Forever!”

  47. missing from Keith’s post is the manical supreme evil overload laugh… “Mwahahahhaha, MWAHAHAHAHAHA!”

    Then it would be complete and accurate.

    I’m yet to be convinced that Microsoft’s proposal(? – plan) is for the benefit of the web.

  48. Thanks Jeremy for a great read, after Aaron Gustafson’s article last month I was starting to think MS has another organisation in its pocket.

    @Zeldman, how is this the only way forward?

    I understand microsofts fear of losing marketshare. Hence, their wish to support their previous browsers. But how is standard compliance by default not a better way forward?

    “My take on the whole IE8 meta malarky”:

  49. You know, I think about this controversy, about this IE6 vs. standards, Jscript vs. Javascript and what immediately comes to my mind is the OOXML vs. ODF battle. Why contribute to a established standard? Let’s do things in the Microsoft way…

  50. IE isn’t only for displaying websites. Lots of Windows specific software uses it. Software firms had to rush out patches and re-patches following IE7s release. We have some clients who cannot upgrade from IE6 because their software suppliers cannot/won’t patch their product. Hence the need for consistency between versions.

  51. I’d be more convinced that this is actually a step in the right direction instead of them attempting to freeze the to preserve their market share if there was actually a clear road map to full standards compliance that MS has said they will hold themselves to with legal penalties happening if they fail to keep their promise.

    As is with how closed up IE development is and all the problems both past and present, I remain unconvinced that this is a step towards standards compliance or is good for anyone but MS.

    Anyone can say anything, but actions speak louder than words and the actions of the past do not match what they are saying now. The current IE team may very well care for standards support, but even if they do, someone over them cares more about keeping the monopoly going and likely is only letting the IE team move to standards if it involves methods that don’t harm or improve the status of the Windows monopoly.

    Without an agreeable road map to standards implementation I don’t believe any of this is actually about improving standards compliance.

  52. @Ault,

    While that is a good point, I do have a comment on it. Who’s to say that there will be consistency between versions? Unless Microsoft intends on keeping the IE6 rendering engine, the IE7 rendering engine AND the new more standards-compliant IE8 rendering engine all in one browser, how can anyone guarantee that the bugs of previous versions will be perfectly replicated in the new ones?

    Let’s move that up to IE version 15. Now what? Will the customers be willing to put up with a bloated browser full of bugs and security issues associated with all of the versions from 6 up to 15?

    The way I see it, the only way this can be resolved is to support standards as much as it is practical. If that means that applications need to be patched, because they stupidly relied on IE6 or IE7 bugs, so be it. Education is the key.

  53. This is not the answer. Although, it’s good to know our good friends like and will be sticking around. Who needs progress in terms of technology or more specifically, the web? My good ol’ reliable Windows 98 running IE6 is working just fine. Crazy Old Grandpa George keeps telling me to upgrade because standards this and standards that, future this and future that…blah blah blah. Am I going to listen to some old kook? No thank you. I’ll stay on this train and learn as little as I have to for as long as I can. God bless America and God bless Microsoft!

  54. I just want to say what a relief reading this article was after reading the articles by Eric Meyer and Jeffrey Zeldman. Here, here, Mr Keith!

  55. @Stephen Down: “And what’s so different from now? Microsoft has created an “˜IE6 standard’ that leads clueless authors to write pages that are broken in fully standards-compliant browsers.”

    The big difference now is that MS, Zeldman and Meyers (is there anyone else?!) want to introduce a broken rendering mode as standard. The ‘IE6 standard’ has been rapidly eroded over the last few years as more of us code for standards and then put a sticking plaster on MSIE via conditional comments. This proposal introduces a legitimised ‘standard’ that benefits one company at the expense of its competitors. It’s an M$ wet dream.

    Microsoft will steam roller on and do what they want – as always, but it’s infuriating that they have advocacy from people who have previously been on the side of web standards. Microsoft will continue with their usual “embrace,extend and extinguish”:, but it’s evident that the vast majority of web devs are seeing straight through this attempt to subvert open standards.

  56. The amount of gas being expended on “The Great IE Versioning Debate” of 2008 continues to confound me.
    Viewing it all from a distance, you would think that people were concerned about really, really bad things happening. But what does it all come down to really? A bunch of CSS selectors, many of which no-one will ever use? Is form triumphant over substance here?
    Look, as with any software, it all comes down to the installed user base. If average users are slow to upgrade, whose fault is that?
    Moving the web forward, standards-wise, is a waiting game. Forcing authors to re-write their content to accomodate a new browser release that will take many years before it is adopted by the many, frankly, seems somewhat arrogant, if not immoral to me.
    I, for one, hope that Safari, Opera, and FF follow suit with the versioning meta tag.
    Please, please tell me what awfully major harm will be done?

  57. Microsoft doesn’t care about ethics. With great power comes great responsibility. Now that’s just awful silly! With great power comes a lot of money. Yes, that makes a lot more sense.

    Sigh. Too bad my happy little world of computing had to be perverted by them.

  58. header(“X-UA-Compatible: firefox”);
    See it here:

    As I’m running Ubuntu I will have to wait until tomorrow to test in IE and check that it still renders the site. Once I confirm, then I’ll add it to my other sites, about 15 in all. If I must do UA targeting, then I’ll target the right UA.

  59. I know plenty of people have posted their own critiques and solutions, but I would appreciate your opinion on “my idea”: which to sumarize is version targeting where the default is one version behind the current. This would give developers a padding of a few years to update their code, while still encouraging standards to grow and improve.

  60. @Richard,

    To address your question. The only people that would be forced to re-write their websites are the people who have neglected the standards or, even worse, added the proper DOCTYPE switch and then proceeded to write IE6-specific code. Anyone who has up to this point developed to web standards would not have to dramatically change what they have written.

    This is why it makes sense for IE 8 to default to standards mode. All we would have to do by then is remove IE 7 and lesser version’s conditional comments.

    You said that it will take years for IE 8 to be fully adopted. Why is that? For that matter, why is it so for IE 7? IE 6 is full of bugs, display ones as well as security ones. Why didn’t MS push the upgrade the way they’re pushing the Vista upgrade? The conventional thinking is that MS doesn’t care about IE. MS hasn’t done much to correct that thinking. Now they’re saying that all versions of IE from 8 on up will always default to IE 7 rendering. What does that tell you about MS’s plans for IE?

    Because they’re the market leader, they need to listen to the market. Other than a few members of WaSP and a few people beyond it, everyone has denounced this idea as a bad one. Why do we still continue to clamor about it? It’s because nothing has been done to assuage our fears, nor is there a feeling that we’re being listened to. As opposed to the supporters saying “Well, what else are they to do?”, they could initiate a proper discussion of the drawbacks and advantages of this decision. However, I don’t think that rationality and logic have anything to do with it.

    Sorry if I come off ranting, but this is something I feel is important. To summarize for you, it’s a bad idea because it will effectively freeze the standards movement, or at least badly slow it down. It’s an extremely bad idea for the other browsers to follow suit. In fact, all three browser makers you’ve mentioned have publicly come out and said that there is no way they’re implementing this. And, why should they? They’ve been implementing their rendering engines as close to the standards specifications as possible. IE 8 is supposed to be close to, if not at, that level. It makes no sense not to take advantage of it.

    Hope I answered your question.

  61. This raises the question for me: Will this stop at ‘full versions’ of a browser? In the future will we be selecting through meta tags individual elements of a release version. Give me the overall version 7 browser but with the opacity qualities of IE9 and the base font set of IE13?

  62. @Richard Fink

    Amount of gas expended? Great way to make a point by starting with an insult to everyone on both sides because they care enough about a subject to debate about it, though from your later points it seems it’s only really targeted at those who don’t agree with version targeting the MS way. Maybe I’m reading you wrong, but your closing statement really does make it seem like an attack on everyone who doesn’t agree with your side. Those of us who do not agree that version targeting is the right way to handle this really do believe that bad things could come about on the web from this move by MS. Your CSS statement is a little flawed, how about I fix it for you, try “A bunch of CSS selectors that are already finalized and ready to be used in every browser except IE, and are already in use with fall backs by the more aware developers.” People already use CSS that IE can’t understand then protect IE from itself using multiple techniques that have been around for years or just let it gracefully degrade to plain text as it does with CSS it doesn’t understand. Please elaborate what you meant by the “form over substance” statement, as is it seems meaningless to me. The way I see it, with form being what’s made (a website) and substance being what it’s made of (underlying code), neither can be more important. It doesn’t matter what you make if it’s made out of the wrong materials, and the materials don’t matter if you built the wrong object. Both are equal and I see no relation to the current discussion, as I see no one saying what’s made is more important than what it’s made of.

    The slow to upgrade and advancement of web standards is important, so I’ll pay a bit more attention to it. With version targeting the way it is, users upgrading doesn’t matter any more for standards moving forward. We’ll get a bunch of half-rate web developers coding to IE7’s quirks and failings forever, and all the other browsers are going to have to keep those particular quirks in their browsers too or end up having problems rendering the web later. Your argument of it taking years for adoption is also not strong enough to be a point. Sure it takes MS over a year to get full adoption of a new browser, but no other browser maker seems to have a problem getting their users to update, maybe if MS unbound the browser from the OS and made it so everyone could update regardless of their OS they’d get an adoption rate of new versions like the others. Maybe if they added real value to each new release and didn’t make people jump through whatever new hoop they felt like adding in the adoption rate would also go up. “Forcing a re-write to accommodate new browser releases” is also not true, I’ve never had to re-write my code for anything except the release of IE7 and that was only a few minor changes to about 1/20th of all the sites I’ve done and manage. Never has a release of firefox, safari or opera ever made me re-write a site, it has always just worked.

    Luckily the other browser makers have already detailed why they won’t be using version targeting and the major stupidity that it is, especially for them to implement so we don’t have to worry about safari, opera, and firefox ever doing this to us. Version targeting is the first easy step towards MS ignoring standards and splitting the web again, even if they don’t mean for version targeting to do so, it doesn’t change the fact that version targeting makes it a lot easier for MS to do so. Perhaps this is the major harm you’re looking for, as I sure don’t want to deal with browser wars and a split web again, especially a war and split that comes at the expense of all the standards work that has been done.

    @Colin Steele

    I imagine it will only be possible to target a single version in total, so I’d guess no on being able to get the font set of 13, transparency of 9, and everything else from 7. Even if it were possible for MS to maintain every version of IE and slam it into a single browser, the ability to only use certain parts of multiple different versions of IE would multiply the work already done to an unimaginable level. Each individual piece of the browser would have to be compartmentalized into a stand alone form, and then a compiler would have to be included to grab each piece needed for every page and basically make a custom built browser engine on the fly. While I have to admit the idea of a compartmentalized browser that can do that seems cool enough that I’d like to see it some day just for fun, it’s highly unlikely that any group can put such a thing together and have it work well enough to be used.

  63. Even though IE6 still remains to be used by half of the world, the other half exponential grows.

    Browser are suppose to behave like a browser. You do it to look at website.

    IE, Firefox, Opera, ect all have the same function but are used differently.

    I prefer Firfox because of the tools it offers to web designer, and for how clean it is.

    IE on the other hand does not have that option.

    As for any continuing version of these browsers, they have will have new features in rendering or uses.

    But always, they will be a browser, and all of them browsers will have that in common.

  64. I got the impression from Kevin Yank’s blog over at SitePoint that this is, in fact, the case.

    Regardless, it seems like a reasonable compromise to me. Forward-looking standards savvy types shouldn’t really be affected if everything xhtml and up doesn’t require the new meta-tag to render standards-compliant in Explorer and devs can either cut and paste one more line of code or upgrade their skillsets to xhtml.

    Meanwhile, Explorer will continue to operate under a heavier load than necessary since it’s basically going to be multiple versions rolled into one. Microsoft may think it’s found a clever way to maintain the appearance of pushing standards compliance while continuing to keep its .net community in the stone age but they’re really just prolonging their own agony and reducing the quality of their own product.

    That is assuming this is only about html 4.x. If it’s across the board, I think it’s high time we got more active about marginalizing explorer by building fun pro bono sites full of the kinds of crap people like to link each other to at the office but refuse to render anything other than a statement why we’re not playing ball with IE this time around. If it’s “only one line of code” let the MS exclusivist .netTards deal with it.

  65. I think you nailed it. They ARE afraid, but I suspect that the Microsoft legal office is behind their fear.

    Think about it… If Microsoft releases IE8 and they decide instead to make it completely standards compliant, what would happen? For the most part everything would continue normally. But a small group of important sites around the world might break in such a way that transactions were no longer possible, or would be down for some unseen explanation. What Microsoft is afraid of is legal retribution from the companies that might be injured. They are afraid those companies will sues the be-Jesus out of them. To be fair… they are probably correct.

    But what I believe Microsoft doesn’t see is that the pendulum has already begun to swing the other way. They should consider that a non-standards compliant browser release may be a legal liability unto itself. What happens if they release IE8 and a group of sites that were designed to work according the standards don’t? Does this not open the door for legal liability?

    The only problem I have with all this as a Web designer and developer is that I don’t want to build websites and have to code for individual browsers. The ONLY logical way is to build sites to be standards compliant so we know they will work in every browser. PERIOD.

    Someday all browsers will be standards compliant, it is inevitable. But Microsoft is trying to play the game AND be the referee. They need to stop, reconsider their actions, and just play the game. Otherwise the legal community is going to call a foul and release the hounds.

    I have often considered sending Microsoft an invoice for all the extra work it requires to build a site that works properly in their various browsers. Perhaps after the release of IE8 the timing will be right?

  66. bq. That is assuming this is only about html 4.x

    Microsoft have said that it will affect HTML 4.x and XHTML 1.x – only doctypes that have not yet made it into common use, eg HTML5, XHTML2, will go into really _really_ standards mode by default.

  67. Jeremy, your article perfectly describes the dilemma many of us face with having to make exceptions for Microsoft’s Broswers, let alone different versions of their browsers ( I refuse to use that MSIE IF statement Malarkey … If I wanted If statements in my markup then I would playing around with oldschool QBasic. )

    Which brings me to the obviously massive news: – That IE 8 will be in standards mode by default.

    I love the last paragraph which says: “Long term, we believe this is the right thing for the web. Shorter term, leading up not just to IE8’s release but broader IE8 adoption, this choice creates a clear call to action to site developers to make sure their web content works well in IE. ”
    This is hard to believe but I’m willing to hold judgment till IE 8 ships. Not to mention how awesome it is that you and Zeldman posted your varying views on the topic last week and we are seeing this blog . . . now that’s clout! Anyways, I’ll close out by pondering if the classic pie chart will someday shift:

    See you guys at AEA’08:Boston!
    ~ Chris at

  68. @Srdjan Pejic: _You said that it will take years for IE 8 to be fully adopted. Why is that? For that matter, why is it so for IE 7? IE 6 is full of bugs, display ones as well as security ones. Why didn’t MS push the upgrade the way they’re pushing the Vista upgrade?_

    On the open web, IE is a “web browser”. On tens of thousands of intranets, IE is a “web browser” _and an application platform_.

    With Microsoft’s hearty encouragement, enterprise customers have spent billions of dollars developing applications for that platform. Because most of these applications were designed during the IE6 era, and were written for a captive audience that was guaranteed to have IE6, they frequently rely on behaviors specific to IE6.

    Along comes IE7. It can’t peacefully coexist with IE6 so enterprise customers have two options: stick with IE6, or test/fix every app that relies on IE to ensure that it works under IE7. And the latter path requires another choice: you either limit the fixes to stuff that works in _both_ IE6 and IE7, or you get to deploy all the application changes simultaneously, while deploying IE7 to the entire organization at the exact same time. Surprisingly, many companies chose to stick with IE6.

    Then along comes Vista. The outlay for a Vista migration is significant for everyone (new PCs where needed, testing/upgrading your business-critical Windows apps, replacing add-on hardware Vista doesn’t support, etc.), but worse for businesses still on IE6: they can’t deploy Vista to a single desktop until all their IE-based apps run under IE7, because IE6 _won’t even run_ on Vista.

    Now here comes IE8. You can’t run IE8 unless you’re on Vista, but you can’t run Vista until you’re off IE6. So if you want to move to IE8 you have to revise the apps that rely on IE to work under IE6 + IE7 *+ IE8*, then move the desktops from IE6 on XP to IE8 on Vista. You can’t take advantage of anything IE7 or IE8 specific during the application fixes, because that’ll break those apps for folks whose desktops are still on IE6 (i.e., everyone).

    As far as your company’s bottom line is concerned, all of the money spent on these migrations is bring tossed into a black hole — unless you want to argue that employees will be measurably more productive on Vista + IE7/8 than they are on XP + IE6. You’re spending that money because that’s the only option Microsoft is giving you. You either follow their upgrade path on their schedule, or you stick with what you have in hand and run critical business operations on unsupported software.

    If you were that customer, exactly how much “pushing” would you take from your Microsoft rep before you hauled off and punched him in the face?

  69. You know Jeremy,

    That word example you take is really an excellent one. Designers and hardcore computer geeks may hate the idea that a software is not completely streeamlined, customized and using its most recent, potentially performant and complex features, but the average user and mankind at large would love to not lose the ability to display and edit old documents.

    Really, I don’t see how one can stand for the idea that continued support of older versions is not something to be desired.

    If there was some mechanism of version targeting in word documents, it would be just great. I’ll never blame Microsoft for learning the lessons of the past.

    Aaron’s article on this very list has very good points about the fact that standards and DTDs alone are far from being enough.

  70. And here I thought we finally got through to Microsoft. Instead of backing up the web designers and developers who go through the time and effort of “doing it proper,” they’ll crumble under the complaints of those who are writing/authoring/whatever-you-want-to-call-it questionable code. And all because the number of those doing it “not quite right” are more than the ones who are.

    I wonder if this is how the “witches” of Salem felt when they were burned at the stake/pressed in the woods for using logic and scientific reason.

  71. I am already sick and tired of writing one website for every browser but IE, then writing for IE6’s quirks and compensating for IE7’s lack of progressive features (nobody can really demand a browser creator implement features from CSS3 before it’s released, but gee wiz Microsoft it’d sure be neat to see a border-radius).

    The idea that Microsoft not only thinks that’s okay but actually wants me to look forward to an future abundance of special implementations for each version of their quirky, obnoxious, sub-par browser deeply saddens me. I’m strongly tempted to start using their little conditional statements to redirect all traffic from IE users directly to

  72. All this talk of Microsoft protecting their market share really doesn’t matter – and what company wouldn’t. I do not care for one second which browser clients/visitors/grandmas use to access the markup I’ve created so long as all the browsers agree on how it should be rendered.
    I say, throw in IE=edge and move on. Keep using your plug-in friendly browsers, I know I will. At least we can stop wasting (or reduce the) countless hours hacking at the CSS.

  73. When I first heard of version targeting, I thought it had to be the worst possible solution. No doubt most of us keep track of what works in which browsers, but as time moves on, that list grows, requiring recoding of sites designed in the past. So, unfortunately, while I dislike the idea of version targeting, that’s the best case for its use: not having to worry as newer browsers appear next year and years from now. And while we can, hopefully, rely on the concept that our standards-based code will stand up over time, who really knows?

    I just keep thinking that version targeting is a little like those old “Best viewed in Netscape 2” buttons. While I’d rather that Microsoft had simply fully supported standards, they have a point: older sites will still have problems. So, with version targeting, at least we can get IE display nailed down. I guess.

  74. IE: the security-hole, opera: very fast and very secure, firefox: the browser with the most features.
    The rendering of these most popular browsers is good, but not perfect – so is the world!

  75. From what I can see on the IE 8 beta 1 website, version targeting has been abandoned in favour of a IE7 emulation mode selected by the user. Phew.

Got something to say?

We have turned off comments, but you can see what folks had to say before we did so.

More from ALA