Someone from usability company Etre wrote in to inform me they that conducted a study on many corporate websites to see how the new Internet Explorer 7 displayed them.
Turns out that many of these websites are busted. I haven’t done enough research myself to determine if IE7’s new rendering engine really adopts standards as their blog claims. To quote:
The CSS standard does not provide a way to target specific browser versions and as a result the Web developer community has developed CSS filters (also called “CSS hacks”)… As we fix these bugs and improve CSS support, some CSS filters will stop working.
My hopes were raised when I read further that Microsoft is trying to strike a balance between adopting standards and also making sure that not too many websites break. Of course, ultimately every website should move towards standards.
For now, we have to make do with a partial adoption. I’ll leave it to experts like Zeldman/Alistapart, Meyer and Tantek (yes, the box model hack needs to be updated for IE7). Big thanks to the Web Standards Project for facilitating the reporting of browser bugs to Microsoft.
We can expect to see new tutorials on CSS hacks for IE7 soon. And business will boom for web standards consultancies. Designers who code the old-fashioned way (that’s a LOT of Singaporean designers IMHO) will have to upgrade or lose out. If everyone’s IE on Windows PCs gets automatically updated in time, people will be scrambling to make sure their corporate websites look fine on IE7.
Good on ya, Microsoft.
Technorati Tags: IE7, internetexplorer, IE, Microsoft, webstandards
Update: Zeldman’s post on IE7. Also see IE7 bugs and fixes, part 1.
I think IE7’s purpose is to make the web standards developers’ jobs easier and not convert non-standard developers since IE7 _will_ as much as possible render table-layout just fine like in IE6.
So I beg to differ with this statement: “Designers who code the old-fashioned way (that’s a LOT of Singaporean designers IMHO) will have to upgrade or lose out”. We’ll need real XHTML to address this issue since XHTML is not (or should not) forgiving when rendered in UAs.