2009, the year we drop IE6?

5 Jan 2009 | Jon Hobbs-Smith
Fact: IE6 is a pain in the ass.

Speak to any web developer and they'll tell you the same. IE6 has more problems than I've got time to mention, but here are just a few examples.

  • It's got loads of rendering bugs, including the peekaboo and guillotine bugs.
  • PNG transparency isn't supported.
  • position:fixed isn't supported.
  • The box model calculates widths and heights of elements incorrectly.
  • min/max-width/height aren't supported.

IE6 has also been getting a lot of bad press recently...


The question is, when is it right to stop supporting an old browser and what are the risks involved?

If it was up to most web developers then Microsoft would have forced users to upgrade to IE7 a long time ago, but that clearly isn't going to happen. Microsoft seem perfectly happy to let users hang on to it, and they have been, loads of them.

Despite IE6 being released in august 2001 (click here for a great list of things younger than IE6) and being superseded by IE7 in 2006, usage still stands at 20%. That may be down from a peak of 60% in January 2006 but it's still 1 in 5 users.


Who is still using IE6?



So, the question is, when there are plenty of better alternatives around, why do people still insist on using it?

Here are some possible reasons...

  • They are on a work computer and they have no choice.
  • They don't know any better or don't know how to upgrade.
  • They have just bought a computer with XP installed (an increasing number of netbooks for example) and it comes as standard.
  • They like it. I know, I know, but some people like familiarity.

So, will we be dropping support for IE6 in 2009?



The short answer - No!

2009 will probably be the year that IE6 usage drops below 10%, but even that's too much to warrant dropping support on consumer driven sites. None of our clients would be impressed if we told them 1 in 10 of their visitors wouldn't be able to use their site. So, in short we're going to have to wait a bit longer.

The long answer - Kind of.

Let's look at the options available to web developers when it comes to dropping support for IE6

1) Don't support it, don't test in it, ever.
2) Reduce functionality for IE6 users.
3) Make sure that everything works in IE6, even if it doesn't look quite right.
4) Make sure everything works and looks perfect.

... and a possible 5th, make sure everything works in IE6 but ask your users politely to upgrade.

I think that in 2009 we'll be looking more towards option 3 than we have in the past. We still intend to make sure that every user is able to use every site we build (wherever possible), but as user numbers dwindle we'll probably be less inclined to spend hours making sure everything looks pixel perfect in IE6

There is one exception. We'll probably drop support for IE6 in our Content Management System and Ecommerce platforms this year, for the simple reason that they are only used by our clients, who are all in a position to make sure they are using the supported browsers that we specify.


What factors should you base the decision on?


The choice for you might not be as clear cut as it is for us. There are a number of factors that might make it more acceptable to drop IE6 support.

  • What browsers are popular amongst your users? If you run a tech blog then your numbers might be under 1%. Check your logs/analytics to see.
  • Do you have any say in getting your users to upgrade? If you're developing an intranet or similar then you might be able to specify supported browsers.
  • Who loses out if the site doesn't work? If you're going to lose sales then I'd suggest it's not a good idea.

There may be other good reasons not to bother supporting IE6, but for us, it's something we're going to have to live with for a while yet.

So, that's where we stand, what about you?




Tags:

comments

Jamp Mark @ 9 Jan 2009 9:24 AM

1 reason could be that some people are having problem upgrading to ie7 due Window's Genuine Advantage issues.
here is a comment that says so
http://www.jampmark.com/web-scripting/dynamic-text-gloss-effect.html#JOSC_TOP

supporting ie6 can sometimes mean hacks which takes time to find and thus impractical.

Rob Yates @ 22 Jan 2009 6:17 PM

I have a site with mostly visitors from businesses and IE6 still represents over 60% of our overall IE traffic. Unfortunately, I don't have a machine or VM laying around with IE6 anymore so I have to rely on emulators and browsershots.org to do testing. Uggg.....

Jon Davis @ 22 Jan 2009 6:28 PM

I've been on #3 (make sure IE6 works, even if it doesn't look quite right) for some time. Frankly, where the chicken-or-egg scenario comes into play of who should upgrade first, the user or the developer's baseline, I think that the web developers HAVE to take action first. Users will have no incentive to switch if all their sites work just fine.

Now, if one site here or there does #1 and breaks IE outright, they are teh suxx0r, but if ALL OF US did #1 and break IE outright, then it would be a short-term tragedy for an otherwise immediate gain of total market overhaul.

So I think we as a web development community should set aside a bi-annual deprecation holiday--a day of the year, every other year, that we mark on our calendars that, on this day, we all agree to abandon a set of critically crippling legacy technologies. If the U.S. FCC can mandate this on over-the-air television (requiring all U.S. television stations to drop analog television signals by 2009, something that went into place about 5 years ago), we can surely invoke our own mandate for the Internet.

nav1 @ 11 Feb 2009 2:24 AM

I have to say that there is nothing worse than trying to make a site work in IE 6. It works fine in any browser, but the instant you try it in IE 6, you have a lot of work to do to make it work there. I just spent some time doing that today, but I decided that I will try to make it usable and put some notice for user to upgrade. I think that would work. They can access the site, but if the want to make it look the way it should they will have to upgrade.

Nicolas Mutis @ 17 Feb 2009 3:30 AM

I've found IE7.js (http://code.google.com/p/ie7-js/) to help quite it a bit when it comes to layouts and construction, but then it too will suffer from IE7 mishaps such as the horrible hasLayout issue. I understand that JavaScript shouldn't be used to control presentation, but this is an exception I'm willing to make.

Edward Caissie @ 27 May 2009 3:00 AM

I'm for option #3 (and #5). I've essentially spent more time on my current project making it work in IE6 than what I spent making sure the layout displayed to my satisfaction in IE7/8, FF3, Chrome, and Safari (Mac and PC).

I finally opted for a lesser graphic set and a theme appropriate suggestion to update.

M.S. Babaei @ 15 Jul 2009 7:01 AM

> 1 reason could be that some people are having problem upgrading to ie7 due Window's Genuine Advantage issues.
There's so many options for end-user: Firefox, Opera, Chrome, ....... They are free with tiny installer (less than 7/8 MB).

Daniel Rodriguez @ 12 Aug 2009 1:10 AM

Hey everybody, first things first, google IE Tester to find the most amazing tool since sliced bread. It allows you to test IE5.5,6,7,8 simultaneously!!!!!

Second, depending on the type of website, I choose whether to support IE 6 or prevent users from looking at the site entirely. Shopping carts are one such website. It may be naive to think that users who would shop online would also be willing to upgrade their browser, but that it the assumption I have made.

Social networking sites are another type that I prevent IE 6 all together. I would rather not spend days, weeks, or months trying to find workarounds for javascript and css limitations in IE6.

For informational websites, I put a simple message that provides links to download other browsers. google ie6 no more to find a website that provides a nice message to your IE6 users.

StrangeAttractor @ 31 Aug 2009 6:39 AM

Dropping support for IE6 doesn't have to be draconian.

Just a small notice that the user should update their browser, and that the site may not fully function properly in IE6 will be enough -- and supply links to upgrade to all the modern 2009 browsers.

This is what Facebook, YouTube, and other big players are doing.

It happened with Netscape 4, and it will happen with IE6 -- mere social pressure will be enough.

Even in corporate environments, it will eventually be embarassing to IT to hear complaints from executives (who until that moment never even knew what a browser was) that their software is out of date.

Don't ban IE6 users; don't support them; just let them know they should upgrade. We have two more recent IE versions now, so there's no reason to continue supporting IE6... as any professional web designer can tell you, IE6 is costing millions of dollars in unnecessarily man-hours.

So, I go for choice #1 -- don't support it, don't test in it. It will probably mostly work, but if it doesn't, the user has been duly warned. Eventually even really big corporate entities (banks, networks, etc.) will stop supporting it, and users (just like Netscape users in the past) will finally have to upgrade -- even the corporate IT departments that are actually the major culprit in keeping this decade-old browser alive.

(PS - And let's give IE some credit... aside from predatory monopoly practices, a major reason it killed Netscape was because it was the first browser to offer any real CSS compliance at all, even if it was buggy as all hell...)