IE6

IE6 is a browser that all web developers hate. Or at least that is the case if you tow the line.

IE6 used to be by far the most common browser and once was the most advanced. However, things have moved on since 2001. Web developers hate IE6 for its poor implementation of standards. Web developers hate the vagaries of the IE6 rendering engine and the performance of its JavaScript engine. These problems are well documented elsewhere.

However, the important people, the ones who matter, the ones who web developers do all their hard work for - the users - they don't seem to care nearly as much. This can be evidenced by the proportion of users who still use IE6. Even today, 8 years after its launch, over 15% of users typically visit a website with IE6. If I were Microsoft, I might just be congratulating myself for making a product so beloved of its users. Of course, where I to do that then I would naively be ignoring the massive proportion of users who have no idea what browser they are using, that there are other browers out there or how to switch.

We as web developers must also not ignore this set of users. There have been a lot of campaigns started recently to end IE6 support. Some have gone even further such as this one I just came across. I support the idea of getting our users using a more modern browser, but its up to them not us.

I don't get the big deal here. I stopped treating IE6 as a "fully supported browser" about 2 years ago. I think the problem here is this idea that a website must look identical in all supported browsers. What is wrong with a website looking subtly different in different browsers?

Of all the websites I have built from scratch in the last 2 years, not one has looked identical in different browsers unless it has been demanded from the client. They all worked in an identical way, but they looked subtly different - generally because browsers have different default rendering styles.

These differences have never been an issue. In IE6 you might get simpler button styles, you might get a more linear visual layout or you might find text looking slightly different. When I have shown them to a client and explained why they are different, the client does not care either - in fact they are thankful for the money and time saved!

I have only had problems with IE6 when I have been maintaining a site where the client has it in writing that browser support will include IE6 and that the site will be identical in all browsers.

In the many many demonstrations to clients that I have made in the last 2 years, I have never once been asked to use any browser other than the one I picked for the demo.

So, I just don't get the big deal with IE6 these days. Lets make a big deal about why we need to make websites look identical in different browsers on different platforms. Lets make a big deal about making websites more accessible. Lets make a big deal about building standards compliant websites with the simplest possible code.

If we spend our energies doing this, then we are more likely to have built-in zero cost support for many many browsers including mobile browsers, not just a narrow set often defined well before the release of a website and therefore out of date at its release.

IE8 - right, wrong or something else...?

For those that don't know I was recently struck down by a rather severe bout of bronchitis which has laid me low for 3 full weeks. I am delighted to say that I am almost fully recovered. Though just to be sure, I am still "taking the pills" - a form of prevention as much as cure.

Its amazing how similar that situation is when compared to Internet Explorer.

For many years, IE was suffering from an infection - non-standards compliance. There was a lack of anti-bodies to kill it off. But Microsoft finally took its sick child to the doctor who prescribed a welcome initial course of anti-biotics and so IE7 was born. But all was not fully well, so a second course was required.

As Microsoft now force-feeds its child the last dose of that medicine, we see IE8 emerge. Standards compliant, much improved performance and an all round more open approach to development.

But the virus remains, dormant for some but nonetheless very real and there. It has mutated into compatibility mode. Luckily there is a vaccination, a form of prevention, and that is a valid doctype and standards compliance.

Of course, you could stop reading the drivel I am writing and read about what Jon von Tetzchner from Opera thinks of IE8.

Whatever you think of Microsoft and IE, you have to acknowledge that they are trying. They cannot please everyone and certainly the remaining prevalence of IE6 does not please me, but I do believe they are trying. Now if only they could speed up their release schedules, I could finally breathe easy.

Browser detection

I am not a fan of user agent sniffing. I think it is a wholly inappropriate way to find out the capabilities of a user agent accessing your website.

However, it has its uses. Steve Souders built UA Profiler, for gathering browser performance characteristics. This is undoubtedly an important and intriguing project. In order to detect which browser you are using, he has written his own user agent sniffing code as he describes in his latest blog post.

He also offers to make this code available through a web service. This is an intriguing possibility. My first reaction was to think that this was an excellent move. A high quality user agent sniffer, improved upon by the community and its breadth of use that we could all access via a web service meaning we would not have to ask our users to download the code to do the detection.

Having calmed down slightly, I started to think more clearly. Overall, I am not convinced this is the best idea. Generally I feel that detecting a user agent from its user agent string is dangerous. This makes the assumption that no user agent spoofing is taking place, that your detection is granular and perfect enough and that browser features and support will not change. None of these are necessarily true.

I prefer to use feature detection. If I want to know whether the browser I am using is capable of applying a particular method, does that method exist?

The only reliable way I have found to detect a browser is for Internet Explorer only using conditional comments, both in HTML (useful for including CSS and JavaScript files for specific versions of IE) and in the JavaScript itself.

I notice that the latest version of jQuery (1.3) is deprecating its user agent sniffing. This decision is to be applauded.

I also realised that the next version of JSquared will actually add the ability to determine if the browser you are using is IE and if it is version 6 or another version. Whilst I might be adding a limited set of user agent support to my library, I am trying to do it in a fail-safe way.

IE8 - Version Switching

There has been much talk over the past few days about IE8 and its new concept of version targeting.

There is a lovely post by Robert Nyman where he considers an alternative. I agree with much of what he says.

My feelings on this matter do differ in some ways. I have a strong conviction that IE8 should render things as well as it can all the time. I wont object to the doctype switch - it has been around too long now - but there should be strong discouragement of its use.

I am well aware that for some the web will be "broken". But surely this is OK. It seems that browser vendors and especially Microsoft are helping those amateurs who think they can do web interface development. I have no problem with people building their own websites at all, but it is time to realise that writing good XHTML and CSS is not easy. Its time to put things right, have some short term pain but get that long term gain.

This combined with Robert's approach of standalone versions could make the world a better place. Tools for building websites "properly" will appear in time.

Just remember, XHTML and CSS are hard to do well, we want to encourage amateurs to have websites - everyone should be able to have a website - but its not easy.