b++ To support or not?
Internet Explorer 6 is an outdated browser. It’s hard to develop advanced interfaces for it. Its quirky support for CSS and proprietary JavaScript is hard to grasp. However, many people still use it.
On the other hand, web developers and web designers are seasoned computer users, which are most of the time after the latest technology — software-wise, hardware-wise, interface-wise. This breed often has several browsers for different contexts.
Ordinary people (who are not dealing with web for money) find it hard to remember browser history stack, don’t need to learn keyboard shortcuts, they don’t know the difference between Comic Sans and Cooper Black, and they don’t know the difference between Internet Explorer and Firefox.
Regular Internet users don’t need Firefox extensions, or care if rounded boxes are done with pure CSS or with background-images…
That is our job — we need to know those differences, and provide the best possible experience.
That’s why CSS people are called experts. They can do miracles, when everything else fails. Any fool can code for the latest browser, but experts take care of every browser with significant share.
Consider this Wikipedia article: Usage Share of Web Browsers.
In the end, let me share some quick numbers from a few random high profile web sites. All of these are in the company of six- and seven-figures visitors/mth.
| Internet Explorer | 62% |
|---|---|
| — IE 6.0 | 53% 32.86% |
| — IE 7.0 | 46% 28.52% |
| Firefox | 34% |
| Opera | 3% |
| Internet Explorer | 48.90% |
|---|---|
| — IE 6.0 | 55.25% 27.01% |
| — IE 7.0 | 44.45% 21.73% |
| Firefox | 44.96% |
| Opera | 4.95% |
| Internet Explorer | 59% |
|---|---|
| — IE 6.0 | 53% 31.27% |
| — IE 7.0 | 47% 27.73% |
| Firefox | 37% |
| Opera | 3% |
| Firefox | 59.32% |
|---|---|
| Internet Explorer | 31.29% |
| — IE 6.0 | 51.51% 16.11% |
| — IE 7.0 | 47.85% 14.97% |
| Opera | 4.94% |
| Firefox | 52.72% |
|---|---|
| Safari | 32.73% |
| Internet Explorer | 13.02% |
| — IE 7.0 | 62.81% 8.15% |
| — IE 6.0 | 37.09% 4.82% |
001—2008.09.03.13:14
nice article, that I was trying to say in my article (http://3kolone.org/index.php?id=47 –> sorry only in croatian) which was also the topic of discussion on mi3dot.org (http://www.mi3dot.org/forum/viewtopic.php?t=13678&highlight=korisnici+manje+vrijedni)… someone who call him self a webmaster or web developer can’t say … IE users fu… off, or I don’t care get yourself a new web browser …
002—2008.09.03.13:59
So IE 7 users haven’t even tipped over IE 6 users among the general population in Croatia … good to know.
As a reference, you could update the article to state at what percentage of IE 5.x users your company decided to stop supporting that browser in the past.
003—2008.09.03.14:41
Since IE is still widely used by “regular” users (best shown on Coolinarika stats) i would say you should still do miracles for it through progressive enhancement.
But that largely depends on what you are building. Does your local intranet some kind of cool administration page has to have IE6 support? Probably not. but it’s still a good perk.
004—2008.09.03.14:56
I see your point about being an expert, but dumbing our techniques (and as a result the end-user experience) down to IE6 has really limited what we can do on the web. Miracles aside, front-end development would be much easier if we could exclude IE6 from the design process. …and if we don’t start pushing it out of common use, who will? Does that mean I’m going to start throwing “Sorry! This page doesn’t support IE6″ errors on client projects? Heck no. But it might be worth consideration on sites with smaller IE6 shares like QBN.
005—2008.09.03.15:39
I’m all for the progression, but there is a difference between progressive enhancement and quitting support.
Of course it’s easier to abandon something, but I say let’s provide access to everyone and offer enhanced experience for users with better browsers.
Quit support, and the site will stink in IE. Those users will never come back. Add enhancements and announce it — and some of them might switch and even thank you for opening their eyes.
006—2008.09.03.17:34
@Marko: Couldn’t agree more.
At the end of the day if 10 - 30% of users of a site use IE6 then really thats something one has to cater for and consider as it still represents a lot of traffic to a site. If not then happy days. However, if IE6 must be supported then it doesn’t mean one should raise a white flag and surrender.
My view: educate users, help users to make an informed decision and let them see what they are missing out on. Open a dialogue and communicate. Telling people do this or else from my experience never goes down too well. Obvious I know and as you say Marko, in the end, some might see the light, others might not or maybe not straight away. But hell thats life!
We are getting there with IE6 but just not quite yet.
007—2008.09.03.18:08
“That is our job — we need to know those differences, and provide the best possible experience.”
- precisely!
008—2008.09.03.20:18
Javascript that alerts the user they are using an outdated browser and should update.
Do this enough and people will get sick of seeing it.
Anthony
009—2008.09.03.20:49
I don’t comment very much, but this is a very interesting topic for me. Many people ask me if I code for IE6, and my answer is always - yes. It is still a major browser on many sites (and in my experience this depends on the people who visits the site - if it is a site aiming for developers and designers, you will see much more visitors with Firefox or Safari).
But, there is also another thing. Nevertheless I sometimes really hate IE, it is in some way a challenge for me. And, as Jason says, we cannot do much under IE6, I find IE6 a real challenge and I’m going to miss it some day ;)
010—2008.09.03.21:18
I think that dropping or not support solely depends on market for which site is targeted.
37signals will drop ie6 support on October 1st. AFAIK they (and Apple) are the first ones to go that road. Hope many will follow. :)
011—2008.09.03.22:59
Ignorance is never an excuse. Haven’t we already excluded older Mozilla and Mac IE users, for example, by using current, valid XHTML and CSS specs? They can ‘access’ the content without the experience a current browser offers.
As a developer I wouldn’t exclude 100% of IE 6 users on a client’s corporate site, but I might deny them access to some features or enhancements with a notification that if they were using a more advanced browser they would have access to those features. It’s a nicer way of giving them the ‘get with it’ message without giving them the finger.
Having said that, we decided on our own corporate to exclude IE 6 users completely and show them a nice notice that they need to download (insert list of other ‘good’ browsers here).
We did give them the finger, and that’s our choice because it’s our site. But we wouldn’t do that on a client’s site unless they requested it.
012—2008.09.03.23:03
Normaly I use Firefox on all plattforms (Linux, Apple and Windows). But if a site demands IE, i will stick to IE6, since this ist the latest version running on W2K and I refuse to use XP or Vista.
013—2008.09.04.02:50
developing for ie 6 really isn’t that difficult
014—2008.09.04.03:45
well, working on the biggest GLBT social network system, brings excat the same issue…support or not…
i say damn this thing, and by jan2009 the support should be dropped by at least 50%….
015—2008.09.04.09:47
I think it is all about progressive enhancement. Do not forget the IE6 crowd, but it is up to you whether you give them the same experience or not.
Me personally and what I am trying to push more than ever at my company is using forward thinking CSS for the next generation of browsers and providing a less engaging view for IE6 browsers and below. As long as they can access the same set of information nobody is losing out.
016—2008.09.04.10:29
@Mislav: Lower pie-chart share is not the only argument for dropping browser support.
For instance, users in corporative environments with Win2K were able to upgrade from IE 5 to IE 6, but that is not the case with upgrading from IE 6 to IE 7 without additional costs for software licensing.
Next, IE 6 is the default on Windows XP and regular Jane doesn’t bother upgrading.
There is still significant number of home users with pirated Windows. Since Vista is somewhat harder to crack, I really don’t expect those visitors to upgrade soon.
017—2008.09.04.11:15
Developing for IE6 or not clearly depends on the target audience. Your stats above reinforce that. It’s the same as with people who use modems or other slow connections (as ISDN in Germany).
If you can afford to scare away 1/5 or more of your audience you can ignore them. ;-)
018—2008.09.05.16:51
The people using IE6 won’t upgrade unless you give them a reason to upgrade.
019—2008.09.11.09:12
My stance as of late (some 4-6 months back) has been ‘make it work in IE6, but don’t bother with the details.’ Meaning exactly what you pointed out–CSS rounded corners or image based ones–IE users don’t care, the list goes on.
More on this from my POV in this blog post: “Optimizing Designs For Different Types of Browsers” (read the comments as well.)
020—2008.09.16.16:12
i think that we have to stop supporting IE6… :)
if not, there will always be someone who is using it…
021—2008.09.18.16:48
Yes. No. No more support for IE.
OF COURSE I’m not going to just kick them off my site (much) but I’m not going to stress out trying to get element x to stop from visibly overflowing its container. Nor am I going to add extra markup to take care of floats. Nor am I going to add a single extra line of javascript for any version of Internet Explorer.
I know exactly(ish) how to make pages work in those browsers, and the disgust I’ve felt in having to learn how to treat IE differently has put me off using that knowledge.
If there’s money involved then I guess you should write code that is compatible with bad browsers (IE6, IE7, IE8) on a very basic level - BUT for personal sites just test in Opera or Safari or Firefox. Don’t even bother opening Internet Explorer. If it works, well, that was lucky. If not, tough, the IE users will have to upgrade.
I know it sounds really bad. I know it’s insensitive. I know that the average user may not have understood five years ago that there are more than two browsers, but now things are different.
After all, the user went out and bought a computer in the first place (Prrrrobably).. Many of them have Photoshop or Winamp or Microsoft Office installed. These things are not (always) OEM software — they have to be hunted out and _got_.
So if you’re going to go out and buy fancy graphics software, then you may as well take two minutes to upgrade your browser also. It’s better for the internet, and if you choose wisely it will be better for you too.
Should we support ANY version of Internet Explorer? No, not until the next modern version is released (IE X or IE XV) .. We shouldn’t kick them off, but we shouldn’t go out of our way to make our code work in those archaic, legacy browsers either.
If they switch, they will be doing themselves a favour. So we should encourage them - gently - and aggressively.
lol @ “if you are submitting a comment” et cetera. Ahhh I had a great chuckle!