Hacker News new | past | comments | ask | show | jobs | submit login
Microsoft decides to automatically update Internet Explorer for everyone (geek.com)
626 points by ukdm on Dec 15, 2011 | hide | past | web | favorite | 171 comments

The article points out that MS will still provide blocking tools for companies. Corporations are the major source of IE6 browsers and I'm not sure this will have any impact on them. The best we can hope for is that high consumer adoption rates will force many more sites to drop IE6 support which might spur companies to finally test and upgrade.

The is a common claim, but has anyone actually researched this? Is there any evidence that this is still true? It's my understanding that most of the IE6 market share is in Asia, where many people are running pirated versions of XP.

Pirated versions of XP are perfectly capable of running IE7/8.

The problem is that a few years ago Microsoft pushed Windows Genuine Advantage updates automatically via Windows Update (they might even have done this twice), with the result that all those pirated versions of XP turned off Windows Update forever and never installing anything from microsoft.com.


I partake in my fair share of Microsoft bashing but am I understanding your point here.

You're criticizing MS for not pushing updates to pirated versions of their software?

They forced an externality on to the rest of us.

The marginal cost of sending updates to pirated machines is probably negligible, and pirates adapted to WGA within a couple weeks.

So – they managed to convert a few Western users who didn't know they were running a pirated copy. I would place a good bet that the total volume of pirated licenses remained the same, though.

More recent experiments have shown that piracy is really more about customer service than straight up cost [0]. As a result, we have large hordes of unpatched machines that are easy to convert into zombies and running outdated software the rest of us have to support.

[0] http://www.geekwire.com/2011/experiments-video-game-economic...

>running outdated software the rest of us have to support.

Do we really, though? If the boxes running IE6 are in Asia and elsewhere in the developing world, how many of us really need to develop for them? I can't think of many websites that need to cater to Chinese pirates.

> running outdated software the rest of us have to support.

Why would you support outdated software only used by pirates? Just don't support it to force the pirates to update.

I am from a developing country (Bangladesh) and use a pirated version of Windows (because you cannot buy the genuine one at stores). If I wasn't a web developer I would probably be using Windows XP and IE6 (easier to crack).

I regularly pay for stuff on the internet (domains, hosting, invoicing apps, ios apps, etc. etc.) as long as Paypal is supported (we cannot use credit cards to make international payments).

You need to support as many people as possible with your webapp, you never know who your customer is.

Just because they knowingly or not pirated a copy of Windows does not mean they won't pay for your web app, if at least not in the form of eyeballs for your ads.

That seems like a fair criticism. Those pirated copies of XP will now never be updated again. This will both contribute to the persistence of IE 6 as well as allow security flaws in XP and IE 6 to go unpatched which will increase the potential for these machines to become part of a bot not or spread viruses. At the same time Microsoft's actions didn't prevent any of these pirated copies from continuing to be used, so the overall outcome is a net negative for everyone.

Then they'll get viruses and have to reformat, or call Microsoft support and find out that it's pirated (if they didn't already know) and eventually end up upgrading. If nothing else, the fact that hardware vendors no longer make Windows XP drivers is about to kill off IE 6 once and for all anyway.

Aside from the Windows Genuine Advantage issue, I've been told that a lot of these XP installs have a bunch of social networking stuff hacked in. Simply updating would break them, because they aren't running stock dlls and exes anymore.

I do some limited IT support for a large (140,000 employees), multi-national defense company. Everything is IE6, and IE6-only. If I use IE9 to access our internal apps, it complains I have an unsupported browser.

I know our competition works the same way, as does pretty much the whole government services sector. That's a few million employees in the U.S. alone.

It's great to have a washing machine that works for 10-20 years, but why couldn't the government and corporations pick a _good_ washing machine. Why IE6 instead of some sane browser?

Anyway, there's enough non-IE6 work to do that I generally refuse to write any IE6 code (and when I do, I don't spend more than a minute or two copy-pasting some conditional comment found via Google.)

> Why IE6 instead of some sane browser?

Why a Motorola Razr instead of an iPhone 4S? Because the latter didn't exist. When IE6 came out more than a decade ago, it was the absolute best browser available. In 2004 when XP SP2 came out, it cleaned up a lot of reliability and security issues. You just have to look at IE6 with a 2001-based lens and realize what the alternatives were at the time.

IE6 was good enough to become the standard for slow-moving corporate types, and they're sticking with it. The every-few-weeks upgrades of Chrome and (now) Firefox are way too fast for them.

A lot of these big companies will stay with IE6 till the bitter end in 2014, when free XP support (security patches) goes away. I wouldn't put it past some of them to pay for XP support rather than switch to another OS, or to negotiate extended XP support into their Microsoft upgrade contracts.

http://www.ie6countdown.com/ has a good breakdown of the percentage of IE6 users around the world. Indeed, China leads the pack in the largest percentage of IE6 users.

This survey from 2007 (I know it's old now) showed 55.2% of corporations used IE6. I'm pretty sure it is one of the sources of the "big companies use IE6" claim. http://www.dailytech.com/Firefox+Makes+Big+Gains+In+Business...

And here's a newer poll showing 29% of respondents use IE6 at work with "no end in sight" http://www.readwriteweb.com/enterprise/2011/03/it-poll-do-yo...

I worked for a Mainland Chinese manufacturing company for a while, and they run all their desktops on Windows 2000. They have several applications that targets IE only, and will not run (properly) on other modern browsers such as Firefox or Chrome.

What I have found is that the IT guys in China compile many "recovery disk" with a pirated copy of Windows that people use to reinstall their OS. A lot of these disks are XP-based, but more people are running Window 7 now.

Lastly, several Chinese companies have developed their own version of browsers. I haven't really looked into what kind of rendering engine they run, but they appear to be some kind of rebranded IE. Does anyone have the scoop on this?

I think there's a very strong "If it ain't broke, don't fix it" attitude in China, and most of the people I've met here don't consider IE 6 to be broke, so they see no reason to fix it. After all, every mainland Chinese website still supports it (circular, I know) ...

Still, new consumer machines often have Windows 7 installed, from what I've seen, so things are moving forward. I don't have any numbers though.

Maxthon is quite popular. They use Internet Explorer's Trident engine, newer versions also ship with Webkit and switch engines based on the site being rendered.

I've seen first hand that even a large "fortune top 10 places to work" silicon-valley company forces IE6 on their staff. You don't have to look very far.

I did see Chrome installed as well so that might be their solution :)

My experience has been that the US government is a big IE6 consumer as well.

Do you have any research to backup your claim?

Actually, yes.


IE 6 holds a 1% market share in the US. How many corporations in the US are mandating IE 6 if it's only got a 1% market share?

You'll also note that IE 6 has a nearly 28% market share in China. Now, this doesn't say that these are pirated installs, but they are obviously XP or older, and this is clearly where the bulk of IE 6 market share is.

Keep in mind where browser usage statistics come from, in this case, NetApplications. If your not visiting some chosen list of sites you are not counted. Thus IE6 installs used mostly internally or stats from sites that don't really support IE 6 won't end up getting counted properly.

On the other hand, if you're developing a website for the general public, you probably don't care much about people who only use IE6 for internal sites either.

If you rarely or never go on the actual internet, the fact that you're using an old browser doesn't matter any more than if you're using an old native program to do the same things.

Of course, there is still the issue of people who do go on the internet but don't visit any of the sites that actually track these statistics; however, I don't know how much of an issue this is because I don't know which sites do the tracking.

Agreed however recall the comment I was replaying to, specifically this:

>IE 6 holds a 1% market share in the US. How many corporations in the US are mandating IE 6 if it's only got a 1% market share?

While the points you've made are valid they are irreverent to the discussion at hand.

When you start your sentence with "It is my understanding" and careful questioning there is no need for giving him the third degree.

Furthermore, apparently "future versions of IE will provide an option in the product for consumers to opt out of automatic upgrading".

Source: http://windowsteamblog.com/ie/b/ie/archive/2011/12/15/ie-to-...

If Redmond insist to have this option, please make it super hard for users to do it :D

I don't think that it will be a big issue since the major reason for end users to still have IE6 around is inertia.

The major reason for end users to still have IE6 around is that they use proprietary applications which were written with IE6 in mind, and they see no good reason to spend money to upgrade these applications (if the vendor is even still interested in doing so) if it still works.

So, as long as there is a way to opt out, I imagine there are a lot of corporate folks who will do so.

I think these are best addressed with the X-UA-Compatible header, and various other ways to trigger the use of an older IE engine. Otherwise, users are offered a choice between broken legacy internal apps, or breakage on most of the rest of the web.

If that requires a change to the legacy application then it's never going to work.

There's a reason companies can't upgrade from IE6 without a significant cost, and it is usually one of the following:

1) The company has lost the sourcecode, and the original engineers have left.

2) The company doesn't even have the source code, and has no rights to modify the application.

3) The application is vendor-supplied, and said vendor requires a large cash payment to make the 'custom' change.

4) The application is vendor-supplied and the vendor refuses to support IE6 in the version of their application that the company have deployed. Their new, super expensive version does support more browsers.

5) The company has little control over their IT. Their business depends on a thousand tiny applications built by countless different parties of varying ability. Simply auditing the whole mess is too costly.

So lots of companies take the easy (and cheap) route out and decide to stick with an ancient browser.

Adding a single header can be done as simply as using nginx as a 'transparent' proxy in front of the app. No need to make application changes.

or can be added to policy/registry that open http://legacyapp:6583 in ie6 frame.

Modern versions of IE do not contain a compatibility mode for IE6, hence the problem. (The official solution is virtualization.)

Sounds reasonable and as something only power users would be aware of.

Updating using Windows Update is the biggest pain in the ass, and it's great to see a much more convenient solution.

The update will still be delivered via Windows Update, it just won't prompt you like it previously did.

Big corporate is such a weird space.

On one hand, you're right: big corporate will block this, and that will limit the impact, and thus I think a lot of the hallelujah in this thread is over-optimistic. But MS would never force such consequential upgrades on those customers; that's part of why MS owned and still owns the "enterprise" market (decreasingly lucrative though it may be).

On the other hand, I doubt that consumer adoption driving sites to drop support for IE6 will have any substantial impact on what megacorps decide to deploy.

One of the things that really blew my mind this year was a large ($40MM) software development project I became familiar with (a worldwide internal system for a multinational corporation) that concluded -- in 2011 -- and required MSIE 6. MSIE 6! Doesn't even run on MSIE 7, much less any modern browser.

While I personally think that's insane -- if you are that specific (not to mention antiquated) with your browser requirements, why don't you just code a native app? -- I've also never developed software with a team larger than five, and certainly don't know the nitty-gritty details about spreading the work over a dozen countries and hundreds of developers, the vast majority being low-cost Chinese and Indian coders. So I'm not judging (or at least I'm trying not to).

But my point is that Big Corporate just wants their freaky "web-based" apps to run predictably for the projected 6-year deployment timeframe and does not give one flying fuck about whether their staff can access the new hip and way-superior version of <ANY OTHER WEB APP ON EARTH>. Unless said had real business value to large enterprise, but then, if it did... it would probably support MSIE6.

So while this news is welcome, it's still unlikely that the computing capabilities of the typical large-company locked-down employee PC are going to become awesome, or even good. Even MSIE8, while way better than 6, still isn't something you jump for joy having to develop for.

This matters less and less, though, as the percentage of people who have a way more capable browser in in their pocket increases. So I think that it is starting to make a lot more sense to give up on corporate machines and dead-end legacy browsers, and instead spend that effort on scaling your app's interface to smaller form factors.

(Disclaimer: that's what I think, but not what I actually do. I work on web-based software for medium-sized businesses, so I don't have to worry much about what browser users happen to have. In this context, you can have requirements, and I tend to just say, 'In order to decrease development costs, the system requires the Google Chrome browser. But don't worry: it's free, secure, cross-platform, and has become the dominant web technology standard.' That last part is a bit of customer-reassurance hyperbole, but so far, no problems. (And of course the same code typically works fine on Safari if there's an iOS requirement.))

>One of the things that really blew my mind this year was a large ($40MM) software development project I became familiar with (a worldwide internal system for a multinational corporation) that concluded -- in 2011 -- and required MSIE 6. MSIE 6! Doesn't even run on MSIE 7, much less any modern browser.

This happens all the time at the hospital I work at. I'm constantly flummoxed by how behind the times software vendors consistently are. We're still stuck on XP for the foreseeable future because some of our vendors are still pushing updates that are totally incompatible with NT6 with no timeline for when it might work.

They've had half a decade to fix their software, and they just aren't doing it. I understand that Enterprise moves slowly, but you eventually reach a point of absurdity, as well.

One problem is that as a vendor, we have to work to the lowest common denominator. It's still generally a requirement that you support IE6, despite the department of health encouraging hospitals to upgrade, because they have legacy apps where the vendor has disappeared or the upgrade is too expensive / painful for the hospital to bear.

We actually ship customers a custom build of Chromium which can only access our application (hospital policy generally dictates they must use IE to access the internet for "security" reasons) - we would love them to just use modern browsers so we can ditch our custom client but it doesn't look like it's going to happen any time soon :-(

I worked at a hospital in 2008, where, mid-summer, the IT department had just approved the use of Windows XP Service Pack 2 (which was released in August 2004).

Note that this doesn't mean they were actually going to start deploying it. It just meant that they were open to the idea of doing so. I'd be surprised if they'd actually upgraded any machines yet.

The law firm my wife works for has just switched to the new version of office, because Word Perfect is no longer supported by their billing software. I don't even think their version was new either.

It's kind of sad that you'll only modernize when you'll no longer get paid. You'd think improving efficiency would be motive for modernization, but I guess not.

Probably why so many hospitals are choosing iPad over PCs.

Our software is so tied to PCs and IE that that's not even an option.

Just curious, what makes you think the "enterprise" market is less lucrative than it used to be?

No growth. Everyone who could possibly use a Windows license has a Windows license and drags their feet about upgrading.

You're either busy growing or you're busy dying. As a potential shareholder, I care more about growth than straight up profitability.

"Entreprise" is not limited to Windows. Far, far, far from that. I don't know your background, but in software development (let us exclude plain Website just publishing information), most of it is written for enterprise customers, not private consumers. Excel had around 10 developers years long. I understand cloud is making the difference more blury, but there are just so different needs across different countries, industries, etc - work is in the enterprise. Consumer tools scale, you don't need much people to maintain and develop these.

You're either busy growing or you're busy dying.

This is extremely simplistic MBA talk.

By this logic, as soon as you saturate the market, you are dying?

Yes. You need to find/create more markets.

Look at Apple. iPod -> iPhone -> iPad.

If I invest 100 dollars in you, I want to receive a percentage back after a certain while. You need to be able to give me 103 dollars back the next year just to beat inflation, i.e. for this to not have been a complete waste of my time.

In real terms, this means you need to grow your business by at least 3% in order to give me that return.

I'm not the right person to talk about this, as my knowledge is thin and my metaphors are poor, but even if you're the sole owner you start to lose money if you don't grow at a reliable rate.

You're right in that most businesses need to grow to increase their value (generally because growth is built into their current valuation) but the narrative you're using isn't on target.

It's entirely possible for a business to give investors good returns even if their business or market isn't growing. Profitability and cash returns from a business can grow as revenues decline if the company either increases operating efficiency, or reduces in investment in recognition of the declining opportunity. MBAs would call this a "cash cow" business - one that should be milked.

We dont tend to see this much in tech because growth is such a fundamental part of our valuations that no company is willing to admit it's in such a position. And you'd probably lose all your engineers over time. But when you look at say, Microsoft, and strip away their non-performing businesses, that's what it looks like: a company that has some assets that produce a lot of profits, but aren't really growing any more, yet if managed properly can continue to throw off gobs of cash for years to come.

Please correct me if I'm wrong!


>Profitability and cash returns from a business can grow as revenues decline if the company either increases operating efficiency, or reduces in investment in recognition of the declining opportunity.

Sure. When I say growth, I mean "make more money this year than you did the year before". This doesn't directly translate into "make more widgets", of course – that's just the (in an ideal world) easiest and least complicated way of achieving that.

>yet if managed properly can continue to throw off gobs of cash for years to come.

Until! It dies ;0.

>Sure. When I say growth, I mean "make more money this year than you did the year before".

If I understand tomkarlo's point correctly, he's saying that you don't have to make more money year to year in order to have a viable business. Rather, you simply need to keep expenses low enough to manage a respectable profit for your investors. That profit need not necessarily grow year to year for investors to find it worth holding.

Of course, valuations will be made completely differently for such a company, but that doesn't mean they can't exist.

It's the fallacy of the excluded middle that convinces people that stability has no value.

It's this statement that is incorrect: "In real terms, this means you need to grow your business by at least 3% in order to give me that return."

In almost any definition "growing a business" means increasing unit sales, or increasing revenues, or both.

Neither is necessary to provide acceptable returns to an investor - there are great investments around in businesses that have long since peaked in terms of unit volume and revenue. Much of the private equity world focuses on these kinds of investments.


Duly noted. I was using it as a catch all for "make more profit than the year previous".

> If I invest 100 dollars in you, I want to receive a percentage back after a certain while. You need to be able to give me 103 dollars back the next year just to beat inflation, i.e. for this to not have been a complete waste of my time.


> In real terms, this means you need to grow your business by at least 3% in order to give me that return.


Suppose you want 3%.

You give me $100 and I buy $100 worth of goods which I then sell for $106. I give you $3 (there's your 3%), keep $3. That leave me with $100 to buy goods to do it all again.

Note that this biz doesn't grow at all but it can throw off 3% as long as I can keep buying at $100 and selling at $106.

What about inflation you say? Multiply all the numbers after the initial $100 by your inflation factor the first year and use the updated numbers (and your inflation factor) each subsequent year.

It shows just how isolated the startup community can be from traditional business. Stable businesses can make plenty of money; growth is simply the way you increase your profits even more.

Really, your sentiment is reminiscent of the dot-com refrain of 'get big or go home'. Those were the true champions of growth-as-profit (and it sure turned out well for them)

"""Yes. You need to find/create more markets."""

Actually you don't "need" anything. There are tons of examples of profitable companies that do one thing / product, or several things / products and do it for a century or more, with thousands of employees that do not look to expand on other markets. Even more so for companies with less employees.

"""If I invest 100 dollars in you, I want to receive a percentage back after a certain while. You need to be able to give me 103 dollars back the next year just to beat inflation, i.e. for this to not have been a complete waste of my time."""

Yeah, that's why the stock market promotes cancerous growth of companies.

I don't mean to be adversarial, but could you provide examples?

>Yeah, that's why the stock market promotes cancerous growth of companies.

I have troubles with this sometimes. I lie awake at night and think about the little I know about finance and economics and I wonder - is this all a zero sum interaction? Abstractly, the earth is a closed system, right?

Of course, value and money are also abstractions and so aren't directly pegged to any physical constants, so maybe in real terms it doesn't matter either.

Anyhow, my understanding is that the narrative I used above works for stock markets and single shareholder companies.

I have a fixed amount of hours/year to invest in my company/enterprise. Let's suppose it's profitable and yields oh 250k in profit/direct wages to you, the single shareholder.

If your total amount of profit remains constant, you're losing money just via diminished purchasing power. Your employees will demand raises, which means you have to raise prices and/or squeeze more productivity out of your company, yadda yadda yadda. If you don't keep growing… you end up finding yourself in a position where you will begin to contract, and it's easier to go down than to go up.

"I wonder - is this all a zero sum interaction? Abstractly, the earth is a closed system, right?"

No, the amount of value in the world is not fixed [eg, 1].

[1] http://www.paulgraham.com/wealth.html ("The Pie Fallacy")

My suggestion is stop laying and wondering and start reading. Economics and finance is a fascinating topic if you want to truly understand the history and theory of money and finance.

I don't mean boning up on company reports (though the ability to read and understand them is a good life skill) but rather truly trying to understand what money is, what it means, why so many apparent paradoxes exist.

To wit; wealth is not a zero sum game. Hard numbers like valuations come about just from an agreement at a point in time, which are just the opinions of two people at a flash in history. Even the value of a dollar in your wallet is a massively volatile thing in terms of what you can purchase with it.

A good jumping off point is Niall Fergusons 'The History of Money' (book or video). Get through that and then decide which of the sub-themes you want to delve into further.

I've been reading Sacred Economics: http://www.realitysandwich.com/homepage_sacred_economics

It's the clearest explanation of this I've heard in a long time. However it's a bit long and deep, if you want an essay length version of the argument, it is here:


"""I don't mean to be adversarial, but could you provide examples?"""

Well, clothing and accessory companies, restaurant chains, publishing houses, film studios, etc, even software houses (has EA made anything else than games --anything substantial to their bottom line, that is?).

But don't think brand names. Think of the tons of companies we don't necessarily know by name, but exist and serve their markets: construction companies, package delivery companies, etc.

OK, I interpreted your original comment as saying that big corporations in the aggregate spend less money on software and its associated services than they used to, and I couldn't imagine how that could be true.

Well, it's still pretty lucrative, but for Microsoft, it doesn't go anywhere anymore.

I am persuaded by some of the recent arguments (John Siracusa's maybe?) that both the innovation and the money in general-purpose computing industry have moved over to the consumer side of the equation, and that this change has put MS in a worse position than they've traditionally been in.

Dominating corporate computing (for soooo fucking loooong) made MS billions and billions, but it also held back their two products, Windows and Office.

They couldn't ditch compatibility, they couldn't make disruptive moves, they needed to pre-announce their roadmap... and also, I'd guess that they probably just didn't see the business-to-consumer weight shift coming, just as they didn't see the Internet coming until it ran them over. (Neither did any of their peers, though -- Apple still had their own branded AOL-clone at a late stage, what was it, eWorld? The Internet ran over everybody as the 80s turned into the 90s. And MS is itself one of those hidebound big corporations; it's natural for those organisms to be slow.)

Anyway my point is that megacorps actually want a pretty limited computing product, compared to consumers. Barring esoteric business needs, they don't need, or even want, their employees to be able to, say, shoot HD video, make unlimited anonymous video calls to anybody in the world, track themselves jogging with accelerometers and GPS, play Angry Birds, and so forth. In general, megacorps want cheap, reliable, slow-moving tools that match their slow-moving planning and frankly limited needs. And did I mention cheap. Cheap.

Back when a PC cost as much as a used car, business purchasing drove the industry. But the human spirit and Moore's law went marching on, and in recent years we have a variety of desktop, laptop,and pocket phone/computers that not only outperform those used-car PCs of yesteryear, but may also feature wireless and nationwide magical Internet connections, ones much faster than the 128K ISDN-line that cost me 40,000 yen per month in the 90s.

So now you can buy a pretty kickass computer--one that fits in your pocket, or handbag, or backpack, whatever you prefer--for the price of like, two steak dinners with a date. (Maybe three or four for the macbook air.)

Cost-conscious companies with big headcounts don't need any of that good shit. So they end up with, say, a fleet of Dell Vostro 620S desktop PCs with 4GB RAM, archaic spinning-platter HDDs, and Windows 7 Pro (or XP still, if they are on a volume license (which they totally are)). So we're talking about $700 with a shitty Dell monitor. 3-4 year deployment life. MS'll make whatever percentage of that, plus a bit on the Office they'll probably use, but that's it for that duration.

Meanwhile, a guy and his wife will drop eight times that on smartphones over the same duration. iPhone, Nexus, Galaxy, whatever. (But yeah, mostly iPhone.) Or a tablet of some kind.

Because they want all that good new shit, and it's no longer so expensive that they can't have it.

So consumers, wow what a shitty word, okay I mean 'people doing their own things and living their lives' -- let's call them PEOPLE for short -- can now afford computers. Waitresses, stadium hot-dog boys, tax drivers, high-school students. That was not so much the case when MS started to achieve hegemony.

Back then, not only could PEOPLE much less frequently afford computers, but also computers weren't nearly as capable of doing the cool shit that PEOPLE are actually interested in. PEOPLE want to do cool shit, not just do some boring-ass work. Video mashups, GPS flash mobs, music buying, music bootlegging, angry birds, porn, cold fusion, etc.

Therefore, the companies aimed at PEOPLE are the ones who are, and have been for a while, making the more ground-breaking and innovative products. Lighter, simpler, more sensors, more stable, brand new re-architected OS software. The PEOPLE are the ones recently pushing the limits in computing. Demanding new, ground-breaking products. Rewarding innovation.

Computers were prohibitively expensive. But now they're not.

Computers used to not be able to do cool shit. Now they can.

So even if the ENTERPRISE market is still bigger overall in terms of units, ENTERPRISE wants shittier products, updated less frequently, for less money. That's what MS has provided over the past decade (with the exception of the less money part, har har).

Microsoft's products met those needs, but at the cost of tying MS to that legacy, and thwarting their ability to compete for the faster-evolving PEOPLE market, arguably now more important.

They are now trying to reinvent themselves. With Windows Phone surely, and it seems they are trying to make a break with the past even with Windows 8, though they are understandably halting in their efforts, loathe to disturb the waning cash cow. But serving the PEOPLE and serving the ENTERPRISE increasingly seem to be mutually exclusive. PEOPLE want new, awesome, fast, again! ENTERPRISE has different priorities. So MS has its internal schizophrenia. And no success on that front so far.

I didn't mean that the enterprise market was becoming less lucrative for everybody, although relative to the consumer market, it probably is. I meant that for Microsoft specifically their dominance of the enterprise over the past 15+ years has been milked for everything it is worth and has now become a hindrance to them, keeping them stuck in a narrowing rut.

You say that the enterprise is “lucrative” and then you say that “megacorps want cheap” software.

I don’t dispute either of these statements. I share the observation that price is the number one concern of those who don’t understand technology and that there’s a lot of money to be made in that space. But what to you think enables both of those statements to be true?

Is it that developers are able to con technophobes into contracts and processes that appear cheap but ultimately end up costing them more? Or, perhaps technophobes actively demand things that lead to that situation e.g. waterfall. Or some other possibility?

It turns out that "cheap" software at the scale where you're rolling it out for thousands of employees, plus maintenance contracts, plus consulting fees, plus servers needed to manage all those PCs, plus... still add up to be a remarkably lucrative market.

Hey veidr, Your ideas are intriguing to me and I wish to subscribe to your newsletter (cold fusion... made me snort baba ganoush out my nose) - update your profile with further information, dagnammit!

While I personally think that's insane -- if you are that specific (not to mention antiquated) with your browser requirements, why don't you just code a native app?

They pretty much have. They are probably using VBScript for event handling and depending on various IE6 behaviors that are no longer supported in current versions of IE and never were available on other browsers. They've written a "native" app for the IE 5/6 platform.

>why don't you just code a native app?

It sounds like that's exactly what they did,

They just hosted most of the code on a server. It's kind of like the "native" iOS apps that just use a webview component to display a web app. Sticking with IE6 just means you have all the hassle of a native app (like needing to make sure IE6 gets installed and stays installed on every machine) with none of the benefits.

At the time, IE6 WAS installed on every machine and using it as a platform for internal apps was VASTLY easier than deploying client/server apps that needed a powerbuilder or some such runtime on each desktop.

Almost as easy as CICS. ;-)

Well, almost all of the hassle, anyway. They probably don't need to maintain and distribute an installer for this particular enterprise app.

Educational institutions using certain testing software from a textbook publishers will probably block updates also. The problem is that, like a lot of enterprise software, the testing sites are fairly fragile and don't react well to changes in browsers. After getting into a situation where students couldn't take tests do to an update, most system admins curse at the sky and then block updates until given the go ahead. It sucks, but the book publishers make poor software.

// ok - just to head off all the switch testing, books, etc. comments - none of those are options. certification testing sites are kinda non-negotiable

There are also things like JAWS screen readers and other add-ons that educational institutions MUST use to comply with ADA regulations which, surprise, generally depend on specific versions of IE to function (though personally the ones we use at my work are at least compatible with IE 8).

There's not really a very good alternative to this-- Firefox has a history of breaking it's extension model with every major version, and Chrome is still in relative infancy (though I personally use it every day), so I think it's understandable that "enterprise" software companies haven't rushed to ditch IE for them.

The best we can hope for is that high consumer adoption rates will force many more sites to drop IE6 support which might spur companies to finally test and upgrade.

It won't, if my experience is anything to go by.

I believe most of the 'IE6 only' apps are not 'on the internet' but in-house only applications.

Ours certainly are.

Yeah we use IE6 (on four year old laptops no less!) because there are applications that we Techs use that the company won't replace and the applications can only be used in IE6 apparently.

Waiting for five or ten minutes for a failed Adobe Flash update to try and load (fails since I am not a desktop admin) each time I reboot or power on and then a full two minutes for Outlook to start is not fun thing to do every day.

No thanks to our outsourced IT to a three letter company.

It's a great change not because of IE6 (which, most likely, won't be affected at all) but because of future versions of Internet Explorer. These are now supposed to be released annually, which could be a curse for developers if there was no automatic upgrade mechanism (see: http://paulirish.com/2011/browser-market-pollution-iex-is-th...). Today's decision of Microsoft means it will be a blessing. Basically, all of client side web-technologies will iterate a lot faster and in 2-3 years the vast majority of users will sport the newest version of a browser by default.

Perhaps I'm being overly optimistic, but this has just made my christmas and new year. If this has a genuine impact, and means people are running IE8 in the worst case, then I will be a very a happy dev.

Patiently waits to see browser usage trends once this rolls out...

One thing I've noticed is another niche that doesn't quite fit into the "corporation blocking updates" bucket. VMs. I've used a few services at work that run on auto-started VM images and IE8 is installed on those images. It isn't that someone is blocking updates necessarily, it is just that it takes a human to actually go and update the main VM image and update the browser on it. This seems to make it persist longer than it should because VMs are either updated on some long time horizon or there is not a push to proactively change it if the current situation "just works".

I see your point, but wouldn't that concern vanish the moment IE updates silently, in the background, without the user even noticing? If that would be the case, they would start out with an outdated IE, but be automatically updated to the most recent version as soon as they connect to the internet.

Well usually the VM images are completely locked down because they are throwaway. Every time one starts up, it starts up from the master and the master has the original IE8. So even if the VM did allow updates, it would silently create a new problem.. a potentially massive amount of waste as all the random VM images are always downloading/updating IE whenever they start up.

Finally! I have been waiting for this news for a very long time. I wish Microsoft would also push something like Chrome Tab to those users who opt out of the update, so a website can set some special http header/meta tag and then the website gets rendered with lasted version of the rendering engine.

That sounds bee-yoo-tee-ful.

Good news, but XP users still won't be able to upgrade past IE8.

Yes they can: Chrome, Opera, Firefox, Safari ...

Neat, how do those handle legacy activex sites.

They come with a convenient "minimize" button that the user can click to reveal the IE icon on their desktop.

Maybe Microsoft could enable an auto update to XP to a limited Win 7 OS :)

I honestly can't see the reasoning behind this.

IE9 implements too much of its modern HTML5 and security features using Windows 7. DirectX 10, which it uses for hardware accelerated graphics, text and video does not run on XP. XP is eons behind the times in security features. It's not just a marketing decision, at some point if they wanted to do these things they had to say "this version of IE will not run on our 10 year old OS."

Hmmmm.... I wonder how other browser vendors are able to pull it off then?

This answer doesn't seem to carry much weight if Firefox and Chrome can do these things and still remain secure.

They use OpenGL instead of DirectX for hardware acceleration. Microsoft's obviously not going to do that; using DX plays to their strengths in every possible way, from what their developers and toolchain is already optimized for, to what's already rendering the entire desktop in WinVista/Win7... so it could even mean IE10 on Win8 tablets will be better on battery life than FF/Chrome. I'm not an expert on this, but DX is likely the objectively better choice for hardware acceleration on Windows -- the reason FF/Chrome use OpenGL is because they have to support other platforms IE doesn't, and doing everything twice in OpenGL/DirectX to get the best result on Windows while still supporting Mac/Linux has obvious downsides.

The latest version of Firefox does not run on PowerPC (my parents are both still on PowerPC Macs). Mozilla has no financial interest in people buying new Macs.

They want people to buy a newer version of Windows.

And more subtly, the "support windows, support office" mentality that forces each team to link arm-in-arm with other team. Everyone slow is sped up, everyone fast is slowed down.

Btw, running that forward, everyone fast learns to slow down, and the slow ones aren't pulled ahead as much.

Or more accurately, it's not financially viable for a company to support a 10 year old OS.

And yet Firefox and Chrome have no problem releasing modern browsers that work on XP.

I applaud Microsoft for finally taking the step, but I wonder how much this will actually change the stats around browser market share. My understanding was that most IE6 users, even in the developing world, were admin-imposed.

Either way, it's good to finally see them moving forward.

Right, I had to support software at one point that required IE6, when IE8 had been released for some time. Maybe this will push those companies that continue to sell IE6~ only software to notice their time has run out.

My experience at present is that only a very small number of corporates are still on IE6. IE7 is the new lowest denominator.

Unfortunately our hospital EMR software, which I access from my office, does not work on IE9. I had to downgrade to IE8 to get it working.

Out of professional curiosity, I'd like to ask: do you work in the UK?

No, the US

This is great news: It means IE8 won't be the next IE6. Of course there are a lot of users still restricted to Windows XP, but their market share is dropping fairly quickly. Of course there will always be large corporate environments where updates are much slower, but I would say these are under 10%, perhaps?

IE8 is the max for XP so until XP goes away it will be the new IE6, which is at 1% in the US. So, yes IE8 will have another decade.

Not a decade for XP and IE6, but certainly more than two years. Close to a decade for IE8 potentially though.

XPsp3 drops off extended support in April 2014 (see http://support.microsoft.com/lifecycle/?ln=en-gb&c2=1173 for detail) at which point it will get no new security updates unless you are a company that can afford to pay for them to be done (and if you did need to try that, it would no doubt be very expensive).

It is highly unlikely that the large companies (who are the main bodies holding on to XP and IE6) will allow themselves to still be using XP or IE6 beyond that point.

Unfortunately IE8 will work on Vista and 7, so I can see some corporates upgrading to IE8 and no further for now. The lifecycle lookup chart at MS (http://support.microsoft.com/gp/lifeselectindex) doesn't list dates for IE7 or IE8. Assuming that IE8's support window is tied to Windows 7 the same way IE6 and XPsp3 were tied for support purposes (as W7 was delivered with IE8 installed), they'll be able to use IE8 under Windows 7 until 2020 (http://support.microsoft.com/lifecycle/?p1=14482) if they are stupid enough to try.

AFAIK even IE 5.01 was supported on Win2000 SP4 until 2010!

Is IE8 really that bad with regards to standards ? That's an honest question.

In 5 years it will be compared to the other browsers, just like the others are much further ahead of IE6 now. IE6 used to be a good browser, too, when it launched.

The problem isn't good/bad. Problem is that IE6 is nonstandard.

I'm not sure there was a browser that did standards better at the time of IE6's initial release. Maybe Opera.

And in my opinion, the bigger problem with IE6 was that it was buggy -- both where it purportedly followed standards and where it had its own way of doing things, behavior often didn't match spec or had unexpected side effects. It took devs years of trying to figure out where all the bodies were buried before everybody understood how it worked.

Yea, many people forget that IE6 actually improved standard compliance over IE 5.5.

There's another side to it too that isn't popular, but IE was in and of itself a standard. When you own 90% of the market, whatever you do is a de facto standard. And I was pretty okay with that up until they decided to stop doing anything innovative with the browser.

With regards to the HTML5 and CSS3 level standards that Firefox, Chrome and IE10 are going to be supporting by the end of 2012?


Compared to these, the IE8 JavaScript engine is quite slow too.

it's worth looking a few years ahead if IE8 will have another decade of popularity, which isn't impossible. Even without that longevity, the hindrance of the installed base of IE7 and 8 will grow over the next couple of years.

We know what IE10 will have, and roughly when, since it's part of the win8 plans.

> the IE8 JavaScript engine is quite slow too

I would go with very slow. Faster than IE6, but not by nearly as much as more modern browsers are faster than it. DOM manipulation is very slow compared to more modern browsers (recent versions of Chrome, FF, Opera, ..., even IE9) too. DOM changes being slow is what you'll notice more with most code.

As an aside: make sure you set X-UA-Compatible (see http://msdn.microsoft.com/en-us/library/cc288325%28VS.85%29....) otherwise you'll have no idea if people looking at your stuff in IE8 are using compatability mode or not.

> I would go with _very_ slow.

Yup. Here's a chart to put IE6 v IE8 v IE10 javascript engine into perspective: http://goo.gl/3ACqo

I'm surprised to see that much difference between IE6 and IE8.

Though IIRC the sunspider test is all javascipt processing and no DOM manipulation, so perhaps the fact I pecieve IE8 as not being that much faster may mean that its DOM layer is the bottleneck I'm experiencing. Not unlikely as I don't currently do large amounts of nmuber-crunching and text processing in anything I have to test under IE.

It's not great. A large amount of the nice fancy stuff that makes designing a website easier and more efficient is not available.

But it is a huge step up from IE6 so if a large chunk of those people suddenly find themselves at IE8 then the world will be better for it.

I wish MS would go back and try to update the IE8 render engine so that it can use some of the newer standards of IE9 or even IE10.

Yes. That's an honest answer.

What I don't understand is why doesn't Microsoft provide an upgrade that actually allows IE6 (and possibly other versions) to run as they did, on a site specific basis. This way corporations that need some IE6 only app can run them forever, but their employees can still use the rest of the web unhampered by this. Having it possible to run multiple "virtual browsers" within a single browser would also thrill web developers who want to test their sites on all the browsers without having to have multiple machines.

Yes it would be a bit bloated, but the default install would probably be the one that just ran the latest version.

I would just like to see a standalone version of IE6, everything wrapped up in one binary, so that the silly, completely unnecessary one-or-the-other choice wouldn't exist.

True, that would help too, but it makes it challenging in a setting where people need to be able to email url's (possibly to the internal application) and the like. If it was specific to a domain or page, it would be a bit more seamless.

Such neatness is contrary to Microsoft philosophy, but hey, I see microsoft making wise thought through moves more and more often. Maybe they will slowly become an interesting company again. They're still enjoying the aftermath of the great brainwash of the 90s. But to be fair, there have always been interesting things coming from microsoft. So we'll see.

It also means all IE9 users will get IE10. So by this time next year, 80-85% of users will be on a very modern html5 browser.

I'm not a fan, and I don't understand the gushing. As developers, I suppose folks are glad to reduce their support matrix?

But as users, creators of highly customized workstations and rabid fans of particular development environments, doesn't it bother anybody but me, that the browser choice has been hijacked?

Sure, its just stupid Windows users, they don't care. Is that it?

Every IE UI is different, and they seem to be spiralling down is usability. I'm particular about optimizing my own time and changing UI to suit Microsoft's agenda is definitely going to piss me off.

Sure, its just stupid Windows users, they don't care. Is that it?

Chrome follows the same model, which is what I use, and I prefer it. So, I think it's a good idea, and not because Windows users are "stupid."

If you run Windows, you accept what you get. The same if you run Mac.

If you want to customize it you run Linux, simple as that.

There are prices and trade-of for all platforms. I run on them all I accept that Windows is a walled garden. I should not expect my things to stay as they are.

What? I'm calling you out. Sure, you can customize your OS if you run Linux. You can also customize your OS if you run Windows or Mac, and it's not more difficult. More to the point, you can install an alternate browser or even an alternate version of IE if you so choose.

>> As developers, I suppose folks are glad to reduce their support matrix?

Not just the number of elements that we have to support, but the sheer amount of time spent in trying to emulate and workaround severe limitations of previous versions, which easily added up to a significant portion of development time.

>> that the browser choice has been hijacked?

To me the browser choice should be between brands not builds.

When you choose to keep an out-of-date version around, you are actively halting the innovation pace for the web.

You can opt out of it, but it will be enabled by default. I'm sure it's not meant to target technical users, but simply those who do not know any better.

Old browser versions have been a detriment to web development for years. This is definitely a step in the right direction.

Pushing updates is on par for chrome and firefox as well. What specific ways is IE's usability "spiralling down"? I quite like the improvements made from 6 up to 9.

And here I thought that all non-updated IE versions were installed in environments where Sysadmins blocked the updates (and will in the future). It would be interesting to see a number of how many installations this change could really target. Like, how many IE 6 to 8 are actually in the wild. Can't be that much, imho.

The great majority of IE6 installations today are in China [1], and I presume that is a region where _personal_ installs of IE6 still are quite prevalent. Of course, these people could have updated if they wanted, and this change won't affect IE6, so it's still a great challenge to get people to modernize.

[1] http://www.ie6countdown.com/

I thought so too, and it fits with what I've seen in large, charge-adverse companies. In fact, today I told a user that our site won't look its best or run quickly in IE7, which he's stuck with at $BigCo client site.

In my current experience, there are very few IE6 installations left in these corporates. There are some IE7, and a whole lot of IE8.

Should the headline read "Microsoft decides to automatically update Internet Explorer for everyone ... except most of them" ?

>> Sysadmins blocked the updates

There's something about this that really bothers me, maybe someone can better clarify.

Why are admins enforcing a policy that leaves the company target to just about every possible security hole that has been fixed in the past 20 years?

Is there a valid reason for this that I'm unaware of?

There are quite a lot of companies that have applications that rely on IE6. I, personally, know of a few financial institutions that can't update. This might not be necessarily related to their systems not being able to work with a newer version, though. Some office drones are just incapable of doing there job if the color of a button changes. I kid you not.

Other companies have a long(!) cycle when evaluating new software releases. I have heard of an insurance provider that rolled out IE7 just this year.

Define valid. The reason in the giant organization that I work in is broken internal applications. The number of internally deployed web apps that were targetted for IE6 and nothing else is truly astonishing. And the speed with which they get updated is astonishing as well. We finally deployed IE8 this year after almost 2 years of remediation effort. Which just kills me. (And of course because of thinking about security harkens back to the 1960s, we are a full year into exploring the deployment of Chrome or Firefox as a supplemental browser, but at present it is too risky)

I work for a local-government IT department in the DC area. We're on IE7, but we have automatic updates turned off.

At first, a few years ago, it was because there was one specific update that broke our Oracle software, so everything got turned off. Now, we keep them off because of inertia. We started pushing out selected security updates in the last few months because the security folks insisted.

Oddly, it's our policy to install the full set of updates (except IE8) to all machines that have issues. We also update our images with all the latest stuff from Microsoft Updates.

Large corporations run lots of browser-based applications. To upgrade these applications, or even to regression test them - even in virtual environments - costs money. Any money they spend on IT comes out of their profits, or is added onto their prices. Would you pay an extra 5 bucks to buy your car insurance from a company that has ditched IE6?

Legacy applications. When you have invested on applications that sadly work only on IE6, then you have no choice but to block the updates. It would be great if windows can run different versions of IE on the same machine, or even at the same time.

Microsoft once more does a fully automated update? What could POSSIBLY go wrong?!

On the same note, it literally pisses me off how companies decide to intervene with my software installation. Allowing an "opt-out" is worth as much as Google allowing me to opt out of mapping my access point by renaming it.

We are observing a notable shift where personal(!) computers and devices are being turned into consumer devices that we have no control over. Not to speak about the privacy related side effects.

Please, please, please let there soon be a pro-version of Linux on the desktop before the support cycle of Snow Leopard runs out.

"You’ll simply be bumped to the most current version available for your version of Windows (IE9 on Vista and Windows 7, IE8 on Windows XP)."

I was so hoping for IE9 on XP as part of this process...

Perhaps one of the reason why people are still capable of getting an acceptable Internet experience from IE 6 and 7 is because Flash 10 (and 11 for IE 7) supports them.

Thank goodness that there is a strong traction behind HTML5 stack, and the industry as a whole is less reliant on Flash to deliver good UX.

Without Flash, the capability of these older browsers will be reduced, and I'm sure they will get abandoned at an even higher rate.

No, people using ie6 are getting a crap experience. Seriously, fire up a vm and click around for a while.

The only people still using it are unable to upgrade for some reason besides laziness and inertia.

Indeed, Flash became popular especially during the IE6 stagnation.

Honestly, it is about time that they did this. I worked ont he IE team for a few years, and could never understand (aside from the enterprise argument) why users were not being auto-upgraded. "Hey here's a great new security model to replace the insecure previous version...but no rush on upgrading..."

The more I see Microsoft bend over backwards for big corps to retard the web, the more I think that maybe big corps should just stay off the web.

Perhaps staying on proprietary, native platforms that don't change as often is their best course in the future.

This is slightly making me think about adding a specific version of IE to my Crap Browser Notifier:


Maybe IE10 and up. Maybe.

Seems like a bit of a stretch to require jQuery for a snippet like that.

Finally. They should have done this a really long time ago.

So is this separate from the normal updates and installs regardless of your update preferences or is it part of the normal updates?

Interestingly (or perhaps ironically), this probably means that the IE chrome won't get updated lest it alarm the users.

Great! Now they just need to switch their entire browser codebase over to Webkit and sanity will be restored to the web.

Wrong. An all-Webkit web would be just as bad as an all-IE web. IE and Mozilla serve to keep WebKit honest. Plus, there's no sense in condemning them based on IE6 when IE9 has come so far.

In what way, since IE5.5 and XHR, has IE driven innovation in browsers? Serious question.

For one thing, they were the first major browser to properly separate the run-time for pages in different tabs. They were months ahead of Chrome, and Firefox and Opera still haven't caught up several years later.

They were the first to ship hardware acceleration, and they are still the only browser to ship hardware accelerated SVG.

They made a lot of mistakes.

Great that they abandoned the “intranet apps still depend on IE6” talking point. XP still won't move past IE8, however.

still seeing 20% IE6 usage in my Corporate clients in the UK. I have spoken with IT administrators and they give the reason of internal tools needing IE6 to run. Some kind of backwards compatibility might help but until hardware is upgraded and there is an OS upgrade from XP IE6 will be around for a while IMHO.

This is one hell of a bloody awesome news. I say the best ever for everyone on the Internet.

great news but a little bit too late in my opinion.

Web designers around the world rejoice.


Santa is real!

I wonder if it will keep replacing the shortcuts that I've removed.

<= Happy

Welcome to the 21st century.

No they don't.

You can still opt out.

Of course if you do that, we may end up pushing Chrome Frame on you, through an exploit (I would, if it wasn't illegal).

Fuck MS for forcing us to deal with their crap.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact