Question

The question is quite theoretical. Let’s say you’re implementing a normal (or small or complex) theme for a website and IE is an important requirement. The question is: how frequent you would do the design fixes for IE (or any other browser that makes you headache)?

Just a short of my thoughts to raise the debate:
If you’re doing it frequently, you probably do unnecessary fixes that probably will be changed. Although if you’re doing it rarely, you would have to fix too many things.

I’m interested in the efficiency, time and headache factors.

Answer 1

I suggest developing in a single browser, then testing your layout in multiple browsers at the end. My workflow usually goes like this:

1. Develop using Safari.
2. Run finished site code through http://validator.w3.org and correct errors.
3. Test in IE6, 7, and other browsers.
4. Apply any amends to the layout using conditional IE comments & stylesheets.

I’ve found this approach to be much faster than constantly hopping between browsers during development. Writing valid code and checking for obvious errors (step 2) seems to solve a good proportion of early IE layout issues. The rest can be handled with browser-specific rules applied in conditional style sheets.

There are some who’d say that you should develop the site using whichever browser its core audience will be using. The approach has merit, but I tend to work faster in a browser backed by good developer tools that gets great Acid3 test scores.

Others favour solutions like IE7-JS, which uses JavaScript to make early IE versions behave more like modern browsers. This can be the right approach, but any solution depending on JavaScript won’t be ideal for all scenarios.