Original:http://www.alistapart.com/articles/fromswitchestotargets
. Com
Author: Eric Meyer
Web Teaching Network
When I read Aaron Gustafson's beyond Doctype:web standards, Forward compatibility, and IE8, my first reaction was to profoundly negate this view. Aaron's description of the version-targeting mechanism is completely wrong, completely backwards, and is exactly the opposite of what we should do. I've been in the field of web development for more than 10 years and I'm against every nerve. . Com
Why do I have to resist so much? Part of the reason is that the target converter is like the Revenge of the "browser sniffing" technology. Really, the sniffer is a necessary way to deal with incompatibilities between them, but it doesn't work in the end, before many browsers are right to support the standard. Not long after you uploaded your script, a new version of the browser appears and destroys it. Browser sniffing technology fragile, self-defeating nature is the resistance behind the ideals that bring standards to our browsers. If you legalize it from the browser's code layer, then he will be a force for undermining the standard path.
First, I am bothered by the target converter because it runs counter to the development of forward compatibility. This used to be the best way for our industry over the years to survive the difficult discovery of the browser wars. We focus on future development, mostly using pervasive and stable functionality, and then using "cutting-edge technology" that doesn't affect the normal use of our site--which is gradually becoming "progressive". An example of this approach is the technology described in "Go to print", which allows progressive browsers to display linked URLs on pages that need to be printed, but does not break the print effects of browsers that do not have this feature.
For the target converter, what is planned for the future, and what forward-looking, is almost wiped out. Browsers promise to always be backwards compatible. For browsers, the goal converter is like a time machine, and the idea is that when you load IE 7 with IE 10来, ie 10 makes it work like IE 7, no matter what happens over the years. Web Teaching Network
Thus, as a developer, there is no need to pursue hyper-reality browsers. I can even assume that browsers will always support what I do, even those short-sighted, browser-specific, and in any case require standards. As for the direction that browsers expect to support: CSS or JavaScript, or HTML5 ... And who cares? Web Teaching Network
Realistic Survey
So, who cares? A list of Apart readers, indeed, and most of us. However, after investigation and analysis, most of the content of Web pages did not attach much importance to the principle of standards-based, forward compatibility.
Yes, we have made a great progress. The education of developers has also produced some fruit. Nevertheless, we must treat these correctly. We don't have (the standard) to reach all the people, maybe never. Some sites are developed based on what the current browser can do, regardless of whether the control specification is wrong or if the behavior is correct in other browsers. . Com
This leaves the browser vendors in a dilemma when facing their flaws: fix it or keep it? One of the most classic examples is "Internet Explorer's original width and height", which is an incorrect execution of the CSS specification. The IE team became aware of the problem shortly after releasing IE 3 ... But until IE 6 was really repaired, such delays slowed the application of CSS and triggered all the JavaScript sniffing and CSS Hacks.
DOCTYPE's conversion did save them, allowing IE 6 to keep the old (wrong) behavior in "quirks mode", to do the correct parsing under "Standards mode"-a mechanism introduced by the Mac version of IE5, which was soon adopted by other browsers.
Let's think about that, through DOCTYPE conversions, browsers effectively recognize two states: old and right. This is one of the newest and greatest methods of the day before the DOCTYPE conversion.
[1] [2] [3] Next page