« RSS Plainly Explained | Main | Happy Halloween! (Athens, Ohio) »


After finishing the major work for a new web page, I discovered the page wasn’t properly rendering on Internet Explorer. I expected some discrepancies because the page layout used some moderately advanced CSS. What I didn’t expect was that the page would be comlpetely unreadable because IE rendered the 89% opaque div framing my content completely transparent.

When one of Pam’s Presidio colleagues/classmates, explained that the "tesellated" background made the text difficult to read, I assumed IE was rendering the background more transparent than Safari or Firefox. I had no idea Pam’s colleague was being diplomatic given the atrocity which presented itself to him in his browser.

The next morning, I had a chance to see the screenshot he had sent, and Pam and I made an emergency trip to Alden library. I was fairly upset at having to spend time debugging for a web browser whose limitations seem part of a (partially abandoned) strategy to maintain browser market share. I decided to redirect IE browsers to a page with a different CSS stylesheet, one that rendered the underlying div completely white, and I placed a notice at the top of the page stating

You have been redirected to this page from [WEB PAGE], which requires a higher level of CSS compliance than the browser you are using.

Mozilla-based browsers including FireFox render the page as it was originally designed.

Thank you for visiting.

in a salmon-colored font at 180% normal size.

I asked Pam what she thought, and she said it was unnecessarily confrontational. I argued that such confrontation was necessary.

I was unhappy that I’d spent so much time making the page fancy—slightly reducing the opacity, debugging the layout of the content divs, choosing aesthetically-pleasing colors, and getting insufficient sleep for five days—that the idea of debugging for a browser developed by a company known to undermine the adoption of open standards placed me beside myself. Using confrontational digital print, I wanted to bully visitors into using a standards-compatible browser.

Pam calmly suggested that doing so might alienate visitors and that waging this battle so visibly might offend my intended audience. I’ve been fighting this battle for years and, through the smoke of open standards weapons fire, the possibility of damaging my relationship with the less-technically advanced members of my audience seemed necessary collateral damage. I compromised by reducing the font size, muting the color, and demphasizing the contrast between the announcement and the main text.

In the between and following a full night’s sleep, I decided that I needed to be able to run Internet Explorer from a computer which I controlled and so began the process of installing Parallels and Windows XP on nitwit (my MacBook Air). The 2-day process involved a 3-hour round trip to the Apple Store in Easton, Ohio (fake malll city), and buying licenses through my university from third-party software vendors such as Microsoft and Parallels.

The setup process was relatively easy, especially when I forwent customizing the virtual machine and selected the recommended “typical” setup. I set up Keyboard Maestro to open a virtual machine with an installation of Windows XP and in the last dozen hours or so, I updated that machine with Internet Explorer 7 (XP SP 2 comes with IE 6 which contains an unfortunate bug due to the reegineering of how IE 6 handles ActiveX requests in response to Microsoft’s settling Eolas’s suit) and updated XP so that it would properly handle DST.

Now, I am able to test against IE 7, which is as good as it gets in terms of CSS support for mainstream users of Windows XP and is the apparent default for academic institutions.

Following my installing Parallels on nitwit, I re-compared the page to the IE-redirected page and it took a fraction of one second for my less sleep-deprived eyes to see how much cleaner the fully opaque version was compared to the fancier 89%-opaque version I had built for bleeding edge CSS-compatible browsers. The styling I’d implemented using advanced CSS code produced an inferior product, a less readable page.

Surely, part of the problem is I’m not such a great a sucky designer, but another part of the problem is that usability is founded upon simplicity and it’s easy to forget such a fundamental point when working with information technology.

But most importantly, everyone should have access to so thoughtful and skillful a consultant as I do. Thanks, Pam, for the great advice. end of article



You're very welcome, Johnnie. If you can raise awareness about the benefits of open-source software with out alienating your target audience, I will always recommend that approach. If you can raise awareness and bring your readers on board with you, better yet!


I agree with pam on this one, even though I arrived on a different train.

An engineer's job (yes, that is what I think is going on) is to produce a product that meets intended goals while capitalizing on the strengths and transcending the constraints and limitations seemingly imposed by available resources. In the end it is our challenge to be creative when we don't have perfection in resources.

I think that the idesigner has to deal with whatever browsers the intended audience uses, it is just part of the territory.


Part of the frustration can be the inefficiency but another factor can be when it eats into your ability to earn a living. We all understand the famous pie chart. I've started letting the customer decide by offering options that work in IE7 (the only version Microsoft supports) Safari, Firefox, Opera, and Konqueror and another option that's 40% more expensive that also works in IE6. I bring in a bugs-whiz to handle the delta. The answer seems to be delineated along financial means - for a high-revenue site, being able to capture those Win2K/IE6 users from BigCorp is worthwhile. But I don't get any acid reflux from it anymore.

BTW, nice blog!