Tuesday, March 2, 2010

Why we should stop developing for IE6

This is not the first time I have said this but I really truly hope it's the last: developing for IE6 sucks.

Since Google announced its decision in January to stop supporting IE6 (as of yesterday), there has been much online discussion as to whether or not sites should continue to be developed for IE6. In developer communities this has been a hot topic for some time now.

Web developers hate IE6. Aside from the much-publicised security issues, IE6 is notoriously non-compliant when it comes to W3 web standards and has shocking CSS support. Developers spend a significant proportion of their time getting a site to work in IE6. Just to test the current website redesign entailed installing VMWare Player and running an old, tediously slow virtual machine. And woe betide the developer who starts off with IE6 and then tests in other browsers. In short it's a complete pain in the short and curlies.

So what's the big deal? Why not just stop developing for IE6? The problem lies in the fact that a significant proportion of website visitors still use it. Statistics range from 10% to 20% of the global browser market share, although, as the graph below (from StatCounter) shows, this is falling. These figures are not evenly dispersed. Only 6% of Australian web users browse with IE6, whereas 25% of Asian and African users do. In these regions it is the number one browser, presumably due to the higher concentration of developing nations and subsequent lack of infrastructure.

Of course, what is really important is that you understand the statistics for your target audience. My organisation's stats are similar to those above from StatCounter and we have very little traffic from Asia and Africa. In discussions with the Web Coordinator (who is also Project Manager for our web redesign) I have tried to persuade him to forget about IE6. My argument is that our target audience is largely young (generation Y school-leavers) and tech-savvy. As such, they are more likely to adopt modern browsers. Only 5% of our traffic comes from IE6 browsers: not a large proportion and one that will only decline.

However, he has made the call to at least attempt to support IE6, arguing that other universities still support it. This seems like a lame argument until you consider that a potential student using IE6 might end up at another university because they couldn't find the information they needed on our site.

I had a ponder about who on earth these people still using this outdated browser could possibly be. Due to Australia's vast size there are a few remote properties and townships that don't have good broadband connectivity. Downloading the later versions of IE would not be an option for this demographic (although they could ask Microsoft to send them a free CD if they thought of it and had the patience to find and load the web page).

Additionally, there are still workplaces which continue to use IE6 due to the reliance of internal applications. Not updating is often cheaper and easier than rewriting those old systems. And then there are those people, perhaps the older generation and those infrequent surfers, who simply don't know or care that they should update. However, I do wonder how many people in this category intentionally visit our site.

We try to develop a site that is accessible to all, including those with disabilities. We currently have no statistics on what percentage of our visitors are in this category. As such it seems fair that we shouldn't discriminate against IE6 users. (Although, it could be argued that they have a choice as to which browser to use.)

My personal feeling is that the difficulties surrounding the development of a website for IE6 far outweigh any perceived benefits. It is time-consuming, does not comply with web standards, is old and outdated, and its user-base is low and falling. I think developers should ensure that users of IE6 (who, it is worth noting, probably don't know how to turn off CSS in their browser) can get to all the information that any other user can access without a major visual assault on their eyes. That is, that information should be readable, accessible and navigable. Develop for IE7, IE8, and the latest versions of Firefox, Safari and Chrome and then create a separate stylesheet for IE6, stripping all but the most necessary styles. If you can get your valid, standards-compliant website to look cool and groovy in IE6 then fantastic. But if it doesn't, don't redesign your website. It's simply not worth it.

My current challenge is to get our site looking worthwhile (and as close to the original design as possible as per my colleague's request) without changing the HTML, which has already been sent off to the CMS consultants. I think ripping out my own hair strand-by-strand would be easier, quicker and less painful. I am considering the addition of a disclaimer stating that the site is optimised for IE7 and above. Although I have to do some work persuading the project manager to let me yet.

Google drops IE6:

Discussions and articles on supporting IE6:

IE6 security issues:

Web statistics:

1 comment: