Hacker News Reader

> 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.

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.

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.

The opposite was actually tried by Google for a while: https://www.google.com/chromeframe?prefersystemlevel=true

You've also got https://chrome.google.com/webstore/detail/ie-tab/hehijbfgiek...

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.

>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   conditional comments ;)

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.

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.

> 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.

Some may say you're a dreamer

But you're not the only one

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.