Hacker Read top | best | new | newcomments | leaders | about | bookmarklet login
IE6 is dead (www.ie6death.com) similar stories update story
248 points by Eduard | karma 2879 | avg karma 3.9 2014-04-08 05:45:26 | hide | past | favorite | 185 comments



view as:

Reports of IE6's death are greatly exaggerated.

We run a SAAS app. We recently saw IE4 in our logs. Not a single one of our customers has upgraded past IE8.

IE6 is alive and well in many large companies.


Consider the possibility that someone was playing with old browsers.

I joined a new project where IE 8 is the lowest required supported browser.

This is nothing unusual when you do SaaS aimed at business customers.

We have to support IE7 on some of our software, only dropping IE6 a year ago or so ... Other software minimum requirements is IE8, hoping to drop it by the end of this year.


Or their user agent header...

Windows 2000, maybe?

Anything that runs IE4 by default is well past end-of-life and is probably riddled with viruses.


it was unconscious for quite a while, so why did it die today?

Because it is off XP life support now.

XP has life support turned off, however, it'll still survive for a few years despite that.

The end of MS support doesn't seem to have caused any mass migrations off of windows XP, and it will be a long time until it's less popular than windows 8.


Yes, I'm afraid reports of IE's death have been greatly exaggerated.

Third time's the charm?

The exaggeration of IE6's death has been greatly exaggerated.

Well, now this thread got bigger. At the time of my posting perhaps quarter of the comments said this.

The Dutch government recently signed a contact to have MS keep supporting XP with security updates, so it's not dead at all.

Yes, so I'm actually paying for the life support :(

the second badge at the bottom (the small one) has its own countdown timer. seems there's an off-by-one-month bug.

(i always wondered why some libraries treat jan as 0 but the first day of the month as 1)


Total guess, maybe 1 is thought of subconsciously in the implementers minds as a name, while number for months are so unlike the month name that they're thought of more like an index.

"Never passed the ACID test" :) Made my day!

That is actually unclear to me though. ACID1 was even included as an easter egg in IE5 for the Mac [0][1]

[0] http://en.wikipedia.org/wiki/Acid1 [1] http://en.wikipedia.org/wiki/File:About_Tasman_IE5_Mac_OS_X....



meanwhile IE 5 is alive in Wichita Kansas. Error message received today: HTTP_USER_AGENT': 'Mozilla/4.0 (compatible; MSIE 5.0; Windows NT; DigExt)

because the old thing doesn't like URLs without a protocol:

//ajax.googleapis.com/ajax/libs/jquery/1.8.2/jquery.min.js

<WSGIRequest path:/ajax.googleapis.com/ajax/libs/jquery/1.8.2/jquery.min.js/


This month alone, my site hat 3 visits each by IE 2 and IE 3, 14 by Firefox 1.5 and 11 by Netscape 4.0...

That's odd. what type of content does your site have ? Because it seems like those browsers would only be run by some tech savvy person doing it as an experiment.

I miss Netscape 3. I used to make lots of javascript dhtml/art on it and it had a very ghostly way of twitching.


what's the benefit of excluding the http:? I'd never seen that prior to a few years ago when the jQuery links started doing it.

It's a protocol relative link. So if the page is served over http, it'll be an http link, and if the page is served over https, it'll be an https link. See http://en.wikipedia.org/wiki/Protocol_relative_URL#Protocol-...

And it breaks unexpectedly (and sometimes hilariously) when people use it in html email, it turns into a file:// "link".

You can't have external css or js in an html email. [1] Are you talking about links? Or image tags?

[1] Well you can, but no one will fetch them.


browser chooses http or https for you. Same as current page was requested.

it will reuse the protocol of the current page, think http or https or spdy or whatever comes in the future.

The browser will load the resource from whatever protocol the current page is loaded from. So if the page is on https, the URL will be accessed as https, otherwise as http.

dagga, FYI, you are hellbanned. I am not sure why, your comments look reasonable. I would email a staff member or create a new account.

You sure it's actually IE5 and not a bot? Lots of bots reuse old UAs to avoid detection

Reports of its death are, from my experience, greatly exaggerated. It's still fairly widely used in businesses; as they say "if it's not broke, don't fix it." (Despite the fact that it is broken on many sites.)

For anyone that wants to use a modern CSS framework that supports IE5.5+, check out Min (http://minfwk.com). Disclaimer: I made it.


Yes, my company precisely does work in running IE6 in newer operating systems such as Windows 7. A case study of why and how we helped Symantec in this area is here http://blog.nektra.com/main/2014/02/12/nektra-escalation-sup...

I know this is a bit pedantic but it's definitely not 'fairly widely' or 'quite a lot' or 'some', there's a tiny, tiny percentage of businesses using it.

Our UK website, not aimed at tech savvy, had 0.13% IE6 over the last 2 months compared to 0.32% the year before and when I last looked a significant proportion of that is bots and scrapers pretending to be browsers anyway.

Half our traffic comes during the business day (we actually see a significant switch from desktop to tablet/mobile at around 6).

For all intents and purposes, certainly for consumer websites, it's dead.

For those interested, our IE7 is at 1.72% (we stopped testing IE7 a couple of months ago as we don't have the resource and it was already converting badly any way).


there's a tiny, tiny percentage of businesses using it.

Not really. Probably many fortune 500 companies (I know several) continue to use IE6 and pay for virtualizing IE6. That's a significant amount of money in licenses and support. That's why companies such as VMware or Symantec have this feature in ThinApp and SWV.


If you're virtualizing then you're probably using it for one specific purpose and not general use, so imo it might as well be dead if it's not being used for general tasks.

Does your traffic come primarily from the UK / Europe? I imagine the situation is much worse in the US.

Why would you imagine that?

http://www.modern.ie/en-us/ie6countdown

UK, 0.5%, is higher than the US, 0.2%.


Thank you. Bookmarked for sure.

It was replaced by IE 8.0, long live the king! :(

With the death of XP I would expect to see a usage drop, not only in IE6 but in IE7 and IE8 too.

IE8 is the latest version of IE that can be installed on XP, I would assume that most people using IE8 would be XP users.

If anyone has data on IE8 OS versions I would be interested.


For our app I've been monitoring IE8 usage for the last month. It is almost exactly one-for-one with XP users (YES!) and dropping a little at a time every day. When I first started looking the rolling 30 day average was roughly 15%. It has since dropped below 12%. I would imagine by the end of the year and a wonderful rash of XP viruses and worms, a lot of our customers will have had to drop both of them.

Except it is the browser that comes by default in Vista and Windows 7 installations at the corporate level, so XP has little influence on it.

It would be better if IE, period, was dead. Otherwise we're just switching one terribly obsolete IE for another terribly obsolete IE.

We are writing enterprise web application that uses WebGL for displaying complex 3D models and we think that IE11 is damn good. Thus, I've no idea why are you calling it terribly obsolete.

I had moaning about how backwards IE11 was earlier this week, but last week I was moaning about Safari and the week before it was Firefox.

We've reached the point where I'm spending about equal time dealing with bugs in different browsers instead of IE being the source of 95% of all pain.

This is a good thing, but I sometimes miss being able to just hate on Microsoft :)


Give it a year.

Yes, then there will be IE12.

We'll have IE12 by that time.

MS has made updating the browser a lot easier than it used to be in recent versions. So I don't think we're going to run into any major issues where we have to wait for a new version of Windows to gain popularity before we see a browser version die off.

At my employer, they finally pushed a full blown enterprise wide upgrade to IE10 with IE11 just around the corner (August).

Unfortunately in parallel, they could not get enough people to submit sites that needed compatibility mode, so to prevent massive help desk problems they decided to push USE COMPATIBILITY MODE FOR ALL SITES. This basically downgraded everyone to IE7 mode permanently. So much for progress...


My Employer does this also, all sites are our intranet switch IE into IE7 compatibility mode. Does matter if the need compatibility mode or not. We recently released a version of the web app I am building and I wish you couldve seen the horror on our faces when we saw the app in IE 7 mode.

Well, unless they actually change it in group policy, this is the default action for IE when it connects to anything on what it considers to be Intranet.

You can override this by setting X-UA-Compatible to IE=edge as a HTTP header on your server. The equivalent meta tag is ignored in this situation.

Or you could ask IT to change group policy, but I have no idea where that setting might live.


>X-UA-Compatible to IE=edge

I believe we added this as a meta tag on the HTML pages, we didnt think of changing the HTTP header on the server.

I doubt they will be willing to change the IT group policy because the app we built has a smaller userbase than other apps. That and they are still supporting web apps from the late 90s - early 00s designed for IE 5/6.


You cannot override "Use compatibility mode for all sites". It's a master switch. No X-UA-Compatible header, No meta tag, and no DOCTYPE will save you from this scourge.

Just checked the analytics on a non-IT website:

IE represents 10.45%

Out of that:

    IE6      .28%
    IE7     2.03%
    IE8    19.81%
    IE9    16.57%
    IE10   16.15%
    IE11   45.13%

Everyone needs to do their own math, though. I run a network of news & information sites for business professionals in different industries and the numbers vary a bit, but are almost all much "worse" than your numbers.

On utilitydive.com (covering the utility/energy industry) it's 35% IE and of that 40% is IE8. IE6 and IE7 are much less, but they also don't really work with our latest redesign so it's a bit hard to tell.


Here's bloomberg.com:

IE represents 25.00%

    IE6    0.11%
    IE7    3.20%
    IE8   34.39%
    IE9   32.94%
    IE10  16.48%
    IE11  12.87%

That's really interesting, how did you get these statistics (do you work for Bloomberg, or do they have this publicly available somewhere)?

I work for Bloomberg. (And yes, I verified that web guys didn't mind sharing the numbers :))

An apparel client in the US over the last 90 days:

    IE6    0.07%
    IE7    1.33%
    IE8   19.44%
    IE9   18.67%
    IE10  16.24%
    IE11  44.25%

On a general consumer website since the start of 2014:

IE: 32% - 11: 42% - 10: 16% - 9: 18.4% - 8: 20.2% - 7: 2.4% - 6: 0.07%


The zombie certainly still continues to live on shitton of computers across the globe running unauthorized copies of Windows XP for customers who couldn't care less about official support from Microsoft as they never had it in the first place. Yeah, unfortunately.

As a Web Developer I do not count those people as a potential target group. Or even a subgroup for that matter.

Great, now we just need to kill IE7... and 8, 9, 10, 11, ad infinitum. :-)

I'd be happy with just 7 and 8 for now. 9+ has usable support of CSS3 at least.

10+ has WebSocket and strict mode. I'm happy!

As others have pointed out, IE6 isn't dead yet. Any person or business that is stubborn enough to use IE6 nowadays is probably also stubborn enough to keep using Windows XP for another year or ten.

And I'm afraid a lot of us have been complicit in the unusually long life of this dreadful monster, as well as of its just-as-dreadful offspring, IE8.

By "us" I mean everyone who has ever developed a website in the last few years and consciously tweaked the code to make it work in an old version of IE. Or anyone who claimed to discontinue support for IE[6-8] but quietly kept fixing IE[6-8]-related bugs in their spare time.

Because every time a website tweaks its code for compatibility with an old version of IE[6-8], IE[6-8] gets another drop of unicorn blood to sustain its life. Since the majority of websites even today work relatively well in IE[6-8], the inconvenience of occasional breakage does not outweigh the inertia of IE[6-8] users. Even among people who call themselves programmers, I know of plenty who just don't bother to install updates. They won't act until their pants are on fire.

If we really want to kill IE[6-8], we need to set fire to their pants. We can't just sit back and count the days until somebody else (Microsoft) mercifully kills IE[6-8] for us. How about we all agree upon a certain date on which we intentionally break our websites and services for IE6 users? We should bring as many websites on board as possible, just like what happened with the SOPA blackout. Any IE6 user who tries to visit any marginally popular website on or after that date will be refused service and told to 1) upgrade to Windows 7+ and IE 10+, 2) install another browser, or 3) nag their bosses to let them do so. That's right, no Google, no Facebook, no Twitter, no Wikipedia, no reddit, no webmail, no lolcats, no porn (!) anymore for you. Within a month, IE6 will be wiped off the face of the planet forever, and whatever remains will be safely confined to intranets. Just like smallpox. Three months later, we repeat the blackout for IE7. Six months later, we boycott IE8 as well.

Don't we, the HN crowd, have the power to pull off a boycott like that?

Will it be an ethical thing to do? Will it even be legal?

I don't know. It does sound rather perverse to refuse service to IE6 while keeping the website fully accessible for Lynx.

But I do look forward to the day when website developers collectively decide to stop being a slave to browser market shares and take the business into our own hands once and for all. Because IE[6-8] deserves it. It has hurt us all a thousand times more than the outgoing CEO of Mozilla has ever hurt gay couples. We can't just sit and wait for it to die. We need to point our keyboards at it and shout Avada Kedavra with extreme prejudice.

/daydreaming


> If we really want to kill IE[6-8], we need to do something more drastic, and do so with extreme prejudice. How about we all agree upon a certain date on which we intentionally break our websites and services for IE6 users?

Maybe for a blog, but I'm not telling 10%+ of my customers to piss off. If the cost of supporting them outweighs the value of their business then I won't support them any more. But I would never intentionally break a site for them.


These people might be using Microsoft software, but it would be a stretch to call them Microsoft customers. When was the last time they actually coughed up any cash? Besides, this is a bed Microsoft made, so they need to lie in it. The longer people stay with XP / IE 6-8, the worse the Internet gets for everyone. M$ is the only company that can do anything about it, and it's in their best interest to, so they just need to quit dilly-dallying and just do it already.

And is 10%+ of the Windows install base really XP???


No, I mean they are my customers! Right now, today, I have 10%+ users on IE6 through IE8. If I can afford to dedicate the resources to support them, I will. I don't believe in punishing users and I certainly don't believe in turning people away if I don't have to.

I'm not sure about XP specifically, but I get way more than 10% of my traffic from IE6-8. IE8 alone is over 10% on some of my business news sites. Don't forget IE8 shipped with Windows 7.


I'd be curious to know what that IE6 support involves nowadays.

There's a difference between having certain simpler layouts (although that probably affects conversion too), and IE6 actually messing up a purchase flow on an e-commerce site.


I actually dropped support for IE6 already, and IE7 is only "best effort" and is currently pretty wonky since our last redesign. We don't release updates that don't work in IE8, though. Too many users. Luckily for us it's mostly just display issues and I have no problem removing nonessential features or sections that are too hard to make work.

But if it's breaking the flow of your site, it might well be more trouble than it's worth.


It's sounds like you are doing what everyone else is. If the cost of support < cost of retaining those users than you fix it. You've realized IE6 is not worth it at this point and sounds like IE7 is getting to that point. For some people supporting anything <IE9 doesn't make any sense. I'm curious about IE8 users though. It seems that after IE7 it would be a much faster life cycle. Doesn't IE9 auto update? So our last hope is to get IE8 user to upgrade 1 time and then we're good?

I work with some enterprise customers and they seem to be mostly IE8 or IE9. We have one dreadful customer who still wants IE6 support but we've officially dropped IE6 & IE7 support. We still do direct hacks as part of consulting for that customer.


Sure, blindly supporting everything regardless of cost is just as foolish as intentionally breaking support to punish old IE users (IMHO).

A disproportionate number of our most important users are running IE8. I don't know if it's because they're still on XP (and thus can't upgrade past 8) or some other reason.

I'm not an expert, but I think it's so much that IE9 auto-updates as Win7+ does a much better job of pushing updates that include new browsers.


Unfortunately IE10 is the first browser to start auto updating silently by default.

I don't have numbers, but I'd be willing to guess that a fair number of IE6, 7, and 8 users were on corporate PCs which prohibited installations of newer versions of IE or other browsers. Microsoft can't do anything about those people! They have been try to end support for Windows XP and IE 6 for years, and newer versions of IE auto-update unless the user specifically tells the OS not to.

Actually Windows XP is 29% (http://www.networkworld.com/community/blog/end-windows-xp-ha...) though it is steadily dropping.

But still, making your app/service compatible with IE6 is a hassle for you and preventing you from doing a better job in other areas which harms your profits.

Perhaps we can take the hit for you. We could cause your customers to stop using IE[6-8] by specifically not fixing the sites to be backwards compatible. Your users will note that although your site still works in IE6, every other site won't. Then if we can drive down the 10% IE6 users of your site to 1%, then will you stop spending time making things compatible for it?


Sure, it'd be great if everyone used a modern browser. I don't think it's ever a good idea to intentionally punish users, though. I'm pretty sure anyone using IE6 (or 7 or 8) is very aware they are on a crap browser. Tons of sites have told them or just appeared broken. They likely aren't in a position to be able to do anything about it.

That's exactly why any hypothetical boycott of IE[6-8] will have to be an industry-wide affair like the SOPA blackout. If everyone boycotts IE[6-8] at the same time, nobody will suffer any competitive disadvantage because of their participation in the boycott.

Well, for one thing, my competition isn't just other websites. And the risk isn't just that people will find a non-blacked out site.

Anyway, I'm against anything that's intentionally user-hostile as a matter of principle.


An IE6 boycott would be user-hostile in the same way that the SOPA blackout was user-hostile.

In the SOPA blackout, millions of innocent people were intentionally blocked from all those sites for 24 hours. Their convenience was sacrificed in order to apply pressure to people who actually were in a position to make a change.

The purpose of an IE6 boycott will be very similar. The purpose is not to hurt innocent workers who have no control over their computer, but to apply pressure to managers who actually are in a position to change things for the better. Even if the only change they make is to allow employees to install Chrome Frame.


The purpose of SOPA was to call attention to a bill that had enormous implications for the world at large.

The purpose of deliberately breaking IE6 support would be. . . what? To tell customers that we don't like old clunky technology? Seems just a teense less noble, don't you think? Especially in response to a situation that the free market can easily handle unmolested.

If you want to not have to worry about IE6 anymore, all you have to do is stop worrying about IE6. You don't need my help to do it. Though, if the only thing stopping you is fears about alienating customers I'd certainly be glad to help there - just send 'em my way. Can't say I even know how to support IE6 anymore, but if they've been able to cling to it this long I imagine there's a good chance their money's green enough to make it worth my time to re-learn.


Unless you have an unlimited developer budget (and who does?) spending time and money on IE support is also user-hostile. It's just hostile toward a different set of users.

That money could be spent on improving the user experience for people with good browsers, developing new services, lowering prices, or any number of other things that would directly benefit the 90% of your users who have actually entered the 21st Century.


gp said "intentionally"

Deciding to allocate the resources to IE support happens by accident, you're saying?

part of the meaning of the word "intentionally" suggests "for that reason" not plainly "on purpose".

While it's certainly a good general policy to not abandon significant segments of your user base, this is not a conflict of user ignorance vs developer laziness. The real question is whether preserving IE6-8 functionality is actually a good way to serve your customers. The resources you dedicate to retaining that 10% might be better used to improve your service for the other 90%. Focusing on modern browsers will also help to attract new customers who won't lock you deeper into these technical shackles. While this doesn't have to mean directly antagonizing people, it may involve making the site unusable and 'ruining' the experience for some.

As ever, it totally depends on the client base.

I'd suggest that most organisations that only had to support 10% IE6 use in the past have already actively made the decision not to support it any longer (I know that happened a couple of years ago for all the orgs I was working with).

For other companies, their customers are 90% on IE6. I know that's a hole of their own making, but in some cases it's a massive undertaking to dig out of. There's money to be made in supporting those guys :)


Just curious, any idea why you have so many users (10%) still on IE8 or under?

I just checked my usage, and I only have 8% using IE period -- all versions. Of those, only 8% are on IE8 or under -- which is only 0.6% of all users.


Lots of reasons for that - including your stuff doesn't work on IEx. Observers paradox.

We have a network of B2B news and information sites for various industries. I think this is probably pretty standard for sites targeting enterprise users at non-IT companies.

It's actually kinda interesting to see the variations between industries like Healthcare vs Construction vs Utilities. If I get time maybe I'll write up a blog post.


Please do. It does sounds interesting.

I agree with JoeAltmaier, but to add to that it depends on your product. eg. If you run a non-profit crm, you will have higher percentage of IE users. As opposed to a git hosting platform, you would have negligible use of IE. Put simply, it depends where on the tech scale your users are and products tend to appeal to some section of the scale.

We work in healthcare software, and IE7/IE8 support is mandatory, otherwise we'd be ignoring potential customers and losing current ones. Sure, they have iPads now, but healthcare IT moves really slow and we have to support that pace.

It's really not as simple as stated above. It's not up to us to kill it off.


I also work in healthcare software, and we've decided on skipping IE7 and IE8 software. It's all about what you can convince your client to let go of. It makes sense to run modern and less bug ridden browsers.

If they attempt to access using older browsers, we bring up an explanation page taking them to an ever green browser of their choice.


We suspect that most of our winxp/IE<8 traffic comes from people stuck in corporations that haven't/can't upgrade with no privileges on their computer.

Have you run metrics to see what the ROI on your IE6 support is?

Do you know what percentage of your revenue these customers generate?

Maybe it's 10% of your user base, but 0% of your profits. Or 50%. You won't know until you run the numbers.


You might also like to run metrics as to how those same IE6 users evangelize your service. They might be really good at word of mouth, or at recommending other enterprise-level customers in via the golf course.

Or - more realistically - you could actually consider 10% of your customers a large, valuable chunk whatever the ROI and accommodate them.

Your choice.


Well, they're not valuable if the ROI is negative.

I think this is talking about immediate ROI, that is customers that actively purchase products, and indirect ROI, or customers that serve some other valuable purpose.

As most sites follow some kind of 80/20 rule, be careful to dismiss a small percentage of users if they fall in the critical 20%.

Then again, if you have data that shows IE6 users are just parasites, get rid of them.


Good advice! Unfortunately, our IE8 users are disportionately more valuable than average.

I keep trying to imagine developers deliberately breaking their source code to fail in Bill Joy's vi to force users to upgrade to vim.

IE6 is certainly nowhere near as capable as modern browser but if you don't need fancy HTML and whamodyne JS then maybe it's not such a handicap after all?


It barely even supports CSS. IE6 was released more than 12 years ago and it was behind even then! I'm amazed people are still able to screw with their code enough to make anything work on it.

I agree with you. If you build your site with Progressive Enhancement then there really should be very little you need to do. We never support IE6, but generally will support IE8 and occasionally depending on the client need to support IE7 too. Sometimes the css is screwed up with IE6 (if we even bother to test it) and it's just a case of a small tweek to a style but we wont go out of our way to fix it (or make any thorough testing)

It's not that easy. I worked supporting Virtual IE6 users and they use it in the context of a Java based internal web application or a web application which requires some specific ActiveX that only work with IE6. In general, to migrate these applications are a nightmare and the IE6 is only connected to an internal site, it never connects to internet.

I agree that I can't drop support for IE 8 or risk losing customers. On the other hand, if facebook/youtube/wikipedia were to do it, and they were the bad guys instead of me, I'd be very happy.

Oh I'll fix you're IE bugs all right. But I make sure to show you you're gonna have to pay me twice for the privilege.

>I don't know. It does sound rather perverse to refuse service to IE6 while keeping the website fully accessible for Lynx.

But it's not. By keeping your website useable in Lynx, you make it a great deal easier for handicapped people to use your site. And to search engines. And to whatever more use cases there are (low bandwidth,..).

These concerns are just not valid for IE6. At least I can't think of any.

Also, you can always redirect the IE6'es to the same version as for Lynx, so you're not really refusing service...shrug


> Also, you can always redirect the IE6'es to the same version as for Lynx

That's an interesting idea. Just wrap all the CSS and JS references in a pair of <!--[if !IE 6]> --> <!-- <![endif]--> conditional comments ;)


Any person or business that is stubborn enough to use IE6 nowadays is probably also stubborn enough to keep using Windows XP for another year or ten.

No, two things:

i) It's not about being stubborn. Companies have invested a lot of money and time to run critical applications. Even if they have the money the risk of migration is high.

ii) They have already migrated to newer browsers! they just run IE6 side by side with modern browsers using application virtualization.


i) and ii) are both excellent excuses to keep using IE6 on the intranet.

But there is no excuse for browsing the public web with IE6 anymore. Especially if there's a modern browser installed side by side with IE6.

Ideally, a firewall should be set up so that IE6 can't access anything outside of 10/8, 172.16/12, 192.168/16, and localhost.


it would be interesting to analyze the demographics of the people who are using IE6 these days. Probably IE6 is working for them and they don't even know about all the technical and security aspects discussed here.

They all work for a very large risk-adverse organisation. Any private person who ignored the repeated calls to upgrade will long since have drowned in malware.

But grandmothers and great grandmothers are included in those organizations? ;-)

If that is true, then your typical consumer web site or even web app probably does not need to support IE6. If they are so risk-averse they probably will not be allowed to visit your site anyhow.

A large part of the reason why IE6 even works anywhere outside of corporate intranets in 2014 is that we, web developers, have put in countless hours of painful and completely unnecessary work to make it work.

This is the job you signed up for. It's a site easier situation now than 15 years ago. You think supporting browsers is hard now? (I can't believe I get to be the crotchety old man in this thread). Perhaps you don't remember the NN4 vs IE4 days. NN4 didn't even have a DIV tag, it had a LAYER tag, and IE4 didn't have the LAYER tag, it had DIV. The only way to support both with all of the DHTML junk people wanted (like they want the damn-near equivalent malarky out of HTML5 today) was to emit separate markup for each browser from the server. You got to maintain two different versions of the site for desktop users.

I just really don't have any sympathy for you. It's not anything approaching hard these days. These days, you just write everything in a compiles-to-javascript language it figures out the browser compatibilities for you. That shit didn't EXIST. You were considered an ASSHOLE if you did anything with JavaScript.


A situation being better than it used to be is a poor argument against making further progress.

Hopefully, one day I will be the crotchety old man complaining about how the new kids never had to worry about IE8 issues.


If I remember correctly, <div> and <layer> could be nested.

Just because I signed up for a job doesn't mean that I have no right to try and make that job less unpleasant, using the force of the collective if necessary.

After all, that's the whole point of progress. How dared our ancestors complain about working 12 hours a day in dangerous coal mines when their own ancestors worked 16 hours a day under even more dangerous conditions? IE4 vs. NN4 was trivial by comparison with that, but the general principle holds.


Why can't we write a super hack-ass extension for Chrome that emulates IE6, warts and all? Then just src it.


What you're really saying is that we in the software development and hardware engineering industries have completely failed to provide a compelling, positively reinforcing reason to upgrade. In response, you suggest negative reinforcement, take something away from these "people" [1] rather than give them something.

What did I do with my Pentium II 266MHz machine with less hard drive space than I have RAM now? I (played games) | (listened to music + (wrote code | wrote papers)). What do I do now on my quad-core supercomputer? I (play games) | (listen to music + (write code | write papers)).

What code did I write back then? Low level graphics operations in C. What do I write now? Low level graphics operations in... JavaScript. Now, instead of using an integrated debugger to suss out memory leaks, I get to use printf statements to suss out... memory leaks. [2]

We've made computers smaller, we've made them prettier, we've made them take less energy, but we haven't made them particularly more useful. At some point, the PC became Good Enough. And unit sales took a well studied drop.

That might be one of the biggest ironies of the monumental deluge of software patents these days: we live in an era that hasn't invented anything novel in 10, maybe even 20 years!

[1] You do seem to talk about them like they are animals

[2] Progress indeed.


> [1] You do seem to talk about them like they are animals

Few people choose to use IE6 of their own free will. Most IE6 users are stuck with IE6 because a large bureaucracy told them to do so. And when talking about large bureaucracies, I find it entirely appropriate to talk in terms of positive and negative incentives and reinforcements because that's the language they understand the best.

And I agree with you that, in some sense, we really haven't changed much in the last 10-20 years. There's been massive quantitative improvement, but that doesn't necessarily translate to qualitative changes.


Some may say you're a dreamer

But you're not the only one


To be fair, IE8 is not nearly as bad as IE6. Most of what IE8 is missing degrades gracefully or can be shimmed, and to be honest my biggest complaints with it are the sorry debugging tools and its JS speed, both of which were within an order of magnitude of the state of the art at the time of its release.

To be even fairer, IE6 was a stable and innovative browser when it was released. The painful part didn't happen until years after MS stopped maintaining it and blocked IE upgrades past 6 on Win2K.

To me the biggest issue is the lack of CSS3 selectors, which makes developing even with preprocessors a hassle that entails lots of unnecessary markup classes and other stuff.

at a company I worked at they only had IE6 available on their cargo ship. There was no other choice or question. I found it interesting that the software had to be degraded to work on outdated hardware and software, simply because the people paying for everything said so.

IE6 will keep living in the enterprise space, and many parts of the world. I hate it but it's really never going away.


> As others have pointed out, IE6 isn't dead yet.

> we need to set fire to their pants.

I remember going to a friend's house for dinner. He had spent all day upgrading the version of Linux on his laptop, and wrangling with a large number of thorny technical issues. After hours, he boasted to me of his success.

"Look! When you type ls it now displays the results in color!"

That was the best he had to show for a whole day of kludging around. He seemed to feel he had done something virtuous -- the equivalent of walking ten miles to school, perhaps, uphill both ways, in the snow. It seemed to me that he had merely wasted his time.

Damn me as lazy if you wish.


IE10 is too aggressive a goal. Can't require OS upgrades...

Chrome and Firefox work just fine in Windows XP.

It's not dead the same way as Mosaic isn't dead. Meaning there is absolutely no reason to care whether anyone is using it or not when you develop some site. If they open that site in old IE - it's their problem.


There are at list other 3 version over the 6 that should be erased from the Earth surface. Just assuming someone would keep the last 2. Not me.

Actually, XP has a shorter life-span time here in Belgium (online banking activities are blocked for Windows XP), so people actually buy a newer PC or upgrade their PC for a newer Windows.

Alternatively they could change the user agent?

They could do so, but non technical people don't even know what that is :)

> in Beligium ... online banking activities are blocked for Windows XP

Do all the banks agree to shut out XP on their sites? Legal requirement?


They say it's because the updates make the OS less secure. So they are all going to disable Win XP because of vulnerabilities.

PS. They also don't support linux


Tell that to the 23% of people still using XP.

Quite a few of those are likely to be IE6 users, primarily due to governments and bureaucratic corporations.


We have a large client with over 2,000 'zombie' XP/IE6 machines out there. Ugh.

I have only two words to say: if only.

On a positive note, now that Microsoft has officially withdrawn support for Windows XP, IE6 may meet its long overdue death faster.


Huh, I just went to the acid3 test in Firefox 28 and it failed. (Animation wasn't smooth, final rendering did not match expected image at all.) Is this surprising? I thought acid3 was a fairly old test.

I ran the acid3 test on my Firefox 28 on OS X and it passed 100/100.

Besides, the acid tests are more a worse case scenario and may not be indicative of real-world cases.


I just went to acid3 with Firefox 28 and got 100%...?

I got "100%" too, but there's multiple criteria for passing:

   - 100%
   - Animation is smooth, no stuttering
   - The final rendering matches the provided image

On Chrome (Version 32.0.1700.77) it went to 69/100, then a moment later continued to 99/100. I reloaded the page and that time it went all the way to 100 smoothly.

It was smooth for me and matches the reference rendering. Maybe your machine is under powered (Netbook or something?)?

Xeon E3 1240 v3 (Haswell) ;-).

Intel i7 950 @ 3.07GHz, discrete nVida graphics, Ubuntu

Maybe you are using integrated graphics or your graphics driver in on the FF blacklist? Platform?

There has to be some reason why--


100% on IE11, 97% on Chrome 34. Interesting.


I do mostly contract work in the U.S. for state and local governments. The RFP's (Request for Proposals) always say something about using HTML5, responsiveness and the latest web standards.

And sometime shortly after (and sometimes before) project kickoff we always see that a large percentage of their employees are using IE8 and we have to support it.

I make a point of describing in detail what kind of resources will be required to do both a modern website/app using the latest standards and one that will support IE8. I mention that, for the most part, they are mutually exclusive and I will effectively have to build the site (at least the "modern" parts of it) twice and this will cut back on the features I'm able to implement within the project budget and timeframe.

This usually makes the customer take a closer look at their actually IE8/XP install base and I've had several customers decide they would speed up migration from IE8/XP and remove the support requirement for that browser.

Just doing my small part to help kill IE8 - one small government agency at a time.


I would think the myriad security holes ie8 presents would be compelling enough a reason.

Especially now that there are no more security patches. But if you've never worked for a government agency then you might be surprised at:

1. How little management cares about technology as long as it's working.

2. How competent/paranoid some of the sysadmins are and how severely locked down some of these networks are because of it.


You're fighting the good fight, sir or madam.

While I commend you for finding a way to communicate to your clients in a way that allows you to not develop for IE8, you're not really being truthful. You can use "HTML5" with IE8 as part of html5shiv.js of course and IE8 doesn't need to be responsive - its a desktop browser. In most cases you can almost certainly deliver an IE8 solution without a lot of extra work.

I admit that I may overemphasize but it's never as simple as just using a shiv. I always run into additional debugging and support issues when building for IE8.

There is most definitely additional work required to support IE8.


HTML5 is a lot more than the new semantic elements.

http://html5test.com/compare/browser/mybrowser/ie-6.html


As a dev I have been trying to convince the project managers of the following theory of people who use IE6/7/8.

They must fall into one the following categories

1. Windows XP, which does not allow them to upgrade to IE9 AND do not know that other non-sucky browsers exist

2. They work for a company that requires IE8

3. They are masochists who love pain)

4. They are sadists who enjoy dealing out pain to Web developers.

Proposed solutions were:

1. Tell them that other browsers have better experiences.

2. Point out that they can usually get around this by using a Chrome frame. Also point out that IE8 is riddled with insecurities and give them a form letter for their IT department asking for an update.

3. They love pain, there is nothing we can do about this. They probably enjoy having incomprehensible boxes instead of icon fonts on half of the sites they visit. Grids be dammed, they love misaligned columns.

4. They probably don’t even use IE8, they (or their botnet) just use a User Agent String that mimics it. These people are asshats, should be ignored at all costs.

I felt this was a quite pragmatic approach, needless to say we still support these users and their shoddy experience.


In regards to your first solution, I keep going back and forth on whether or not it's professional to add a conditional warning that detects an old browser and explains the upgrade process. I know there are some scripts you can throw on your sites that automatically do this, but I'm not sure if that's a good idea.

googleapps, facebook, youtube, nike, among many other gigantic brands seem to think it is just fine to encourage their users to have a better experience.

I didn't realize there was actually an end-of-support date for IE6.

Is there one for IE7?


Today is the end of paid support for IE6.

Today is also the end of unpaid support for XP (which I assume includes IE7 - IE8).


I really can't say I'll be in mourning over this. I've lost count of how many times I've had to convince clients to upgrade even to IE7 because IE6 isn't compatible with some aspects of WordPress.

Having to support IE8 Compatibility mode for a bank is worse than IE6 - at least IE6 issues and fixes are documented. IE8 Compatibility mode is a world of pain, add in some proxy servers and latency sensitive JS and you have a fun fronted time.

It's not dead. I work for a large health organization (100,000+ employees) and IE6 is still the default browser. Users can manually update.

Needless to say, but I'm still using XP.


I'm currently working in a very heavy enterprise space, most fortune 100 banking/insurance, etc. companies use our software. Our requirements are only IE9+. I'd be scared of the legacy companies that need older support

I never understood why IE updating lags so far behind. It is free. I understand why updating the OS is a bigger deal. It costs money and is a huge change that involves a lot of work for some.

I suppose Microsoft is partially to blame. You can install any version of Chrome and Firefox on XP, but only up to IE8.


Is it though?

If you're maintaining sites that are difficult to upgrade, I'd recommend looking at IE11 Enterprise Mode. It lets you run with pretty good IE8 compatibility (e.g. ActiveX) in IE11, taking advantage of improved speed, security, and base OS improvements.

20 slide deck: http://view.officeapps.live.com/op/view.aspx?src=http%3a%2f%...

docs: http://technet.microsoft.com/library/dn640687.aspx

good blog post: http://love2dev.com/#!article/The-New-Internet-Explorer-Ente...


The browser we love to hate. The version we love to hate.

Interesting (see footer) how many of those site were out there.


Legal | privacy