That's the whole point of web standards. To document this `whatever` minus the extra features offered by some browsers plus the features the IE team were too incompetent to write.
The problem is that the real world differs so markedly from the standard that really they're just an "in theory..." reference, or a "hey that will be cool in 5 years if it gets built" reference.
HTML5 features are a wonderful example. If I'm targeting the iPhone, the new location API is a goldmine, but otherwise it's meaningless (for now) because 50%+ of my audience (IE/Opera) doesn't support it. I know I need to build in a good fallback anyway, but the point is that the standard only matters insomuch as it's actually implemented.
One look at the different local storage/web storage implementations bears this out.
I don't really care if W3C argues itself into oblivion if the browsers leapfrog them and implement something useful (see: canvas/iecanvas).
If the browsers are outrunning the standards by a significant margin, rendering the standards irrelevant, then the process is broken.
Agreed. My point is simple- throwing web standards out of the window is no solution. It serves as a guideline (or "documentation") for both programmers writing browsers, and authors writing web pages. I'm not asking for strict adherence, just a guideline. If you try to implement a TCP/IP stack strictly as per the RFC, you'll find that it can't communicate with any real-world operating system- that doesn't mean that you throw RFCs out of the window. Things have always been, and will always be, engineered to "work".
