I've thought about this a wee bit and I don't like it. For many of the reasons that the codepoet hates it, and more.
I'm just a user for the most part when it comes HTML. If you are looking at my web stuff for examples of how to do it; I'm the poster boy for the new anti-pattern. Look in the dictionary under wrong and you'll find my picture.
Why is this? Partly because I'm lazy and partly because my tools suck and well, goto 1.
All I want to do when I put up a weblog entry is to write some stuff, twiddle it a little to deal with any code or other 'do not touch' type data and post it. I spend too much time every day fighting with all the little in's and out's of Internet related arcanum, I don't want to have to do it here. Writing for the web should be easy (and it's not).
To make this happen, I need tools that make few compromises (or which allow me to dynamically make them if perhaps I should), and a display engine that makes no compromises (if possible). If the system engine is as correct as possible, then the Safari folks will need some sort of fork, otherwise, the HTML display engine will be fragmented. The power of IE in the Windows world is based not only on the installed base of IE, but all the other pieces which base themselves on the system libraries. As an engineer, I recognize that inevitably we'll need to deal with problems in WebCore (KHTML) and Safari. Do we want to do so now, when everything is so new and unsettled, or would it be best to wait until we have the majority of the problems settled? And I know, everyone wants their site to work in Safari three months ago, but is that reasonable? I'd prefer patience if possible.