How I learned to stop worrying and love IE8

After an IE6-induced era of stagnation, the Web has adopted a steady pace of serious innovation and progress these last few years. Regular browser updates, collaboration on open standards, and some polyfill libraries make the Web a much more viable development platform than it used to be. Or does it? An exploration of the site paints a more depressing picture: Internet Explorer 8 is dragging us all behind with its market share of up to 20% (depending where you get your statistics). 20%! Are you going to leave 20% of your users behind? Should you still care that much? I’ll be arguing in this post that you shouldn’t, in many situations.

The first thing that I want to point out is that this seemingly huge market share is almost entirely on Windows XP, and is not going to go down anywhere as fast as you’d like. This is the hard core of technologically illiterate users who just won’t upgrade until their dusty old Dells finally explode. And China.

Windows XP isn’t supported by Microsoft anymore. There won’t be a new version of IE on it, and that is exactly why IE8 won’t go away any time soon. There also won’t be a new version of Office for XP, but XP users don’t care. Their browser is 60% toolbars, and it takes them fifteen minutes to boot because of all the malware they are running.  It will only get worse: such a large population, entirely unprotected, that nobody will ever come and rescue, is too beautiful and easy a target for the black hats. Seriously, are you going to spend precious resources optimizing the experience of your web site for those users?

Is this demographic really a significant part of your audience in the first place? If the site already exists, check your actual local stats rather than global stats. If you can’t do that, study your target audience, and compare it with the XP/IE8 crowd.

But the best argument may be that you don’t even need to do anything that you shouldn’t already be doing. Those users are already used to a severely degraded experience on their computers. If you are already applying principles of progressive enhancements, if your sites are designed to be usable with a screen reader, and only use the cool new stuff to provide a better experience to those who are able to experience it, you should be mostly fine, and that’s work you should already be doing (by law, in many countries). Those IE8 users will see some mildly weird stuff at worst, and otherwise will still see a usable site.

If your site or application relies on the cool new stuff for its basic behavior, chances are that IE8 users are not your audience.

In summary, design for accessibility, like you should, and give IE8 users the most basic experience. You’ll be fine, and the Web can move forward.


  • No custom fonts, no external scripts due to ie8 blocking bug, in fact no scripts at all. That also means no responsive - don't bother with respond - give them a basic experience. Then layer on progressive enhancement. Check out 'cutting the mustard' article on bbc.

  • 14 years since XP came out and we're still stuck with IE8. 8 years since Vista, so we're probably stuck with IE9 for the next 10-15 years. On the bright side, by 2035 everyone should be on evergreen IE... I realize the technical challenges here, but if the IE team could just spend the extra thousand hours porting Spartan back to XP and update XP and Vista (ignoring end-of-life), they'd be saving millions of software developers tens/hundreds of hours each. At that scale, it would literally have a positive global impact on stress levels, blood pressure, nicotine, caffeine, and alcohol consumption, headache medicine sales... Think about it :-)

  • That very probably won't happen, which is why I made the recommendations above.

Comments have been disabled for this content.