Breaking the web?

By crisp on Saturday 7 April 2007 00:54
Category: Browsers, Views: 3.724

Chris Wilson, as Chair of the new W3C HTML WG, stated:
[...] We can't, for example, change the behavior of how we support CSS floats in IE7 without requiring an opt-in, since we would change layout significantly for half the web.* When we break the web, it's our fault, even when we're breaking it to improve standards compliance.

If the way IE7 handles floats is incorrect according to specification (and yes, there are still many issues there and on numerous other points), and other browsers deal with them in a correct matter, how can this be 'breaking the web' when apparently half the web is already broken to all users of non-IE browsers? Would for instance fixing the float-model in IE really break half the web? I sincerely doubt it because I would have noticed it if half of the websites I visit with my 'more up to standards' browser would appear broken to me now...

C'mon Chris, lay off your IE hat and your double agenda. When a couple of pages 'break' when IE becomes standards-compliant than Microsoft is to blame for ignoring existing standards for so many years and thus allowing it to come to the point where pages 'break' many years after dato, and MS should just take that blame. This is just playing chicken and I honestly don't feel there is much need to it since most respectful websites are already dealing with these issues in a good manner. Like you said: *49% of the top 200 US sites are in "strict mode" - no matter if that means they're actually fully according to the specifications, they're obviously aware of them.

Besides, I can't really think of anything that would cause major breakage in todays' websites if MS were to fix them in IE-Next's HTML(4) or CSS(1 and 2) area. Most problems in those areas simply cannot be mistaken with intended behavior - the problems with the CSS float model for instance are clearly bugs that just lead to unexpected behavior - so these problems don't require a "really really" standards compliant mode but should be fixed for all modes.

Volgende: HTML5: Microsoft and the opt-in catch 04-'07 HTML5: Microsoft and the opt-in catch
Volgende: Is 100.000 times match() enough for you? 03-'07 Is 100.000 times match() enough for you?


Comments are closed