IE 9; why the 'same markup' mantra?

Better support for standards and ‘same markup’ are more significant than improved Acid 3 results or even the faster results on the Sunspider JavaScript benchmark in the IE 9 Platform Preview 2.

It can seem ironic when Microsoft calls for developers to demand more from browsers and repeats its ‘same markup’ mantra with the fervour of a convert, but Microsoft rarely implements a standard it doesn’t want to give a thorough workout. Putting in a standard simply because it’s a standard doesn’t seem to be very interesting to them (especially before the standard is – well, actually standardised); Microsoft puts in a standard because it can see something clever that it – or other developers – can do with it. That’s frustrating if Microsoft doesn’t care about a standard you’re interested in and it can seem a cavalier attitude to standards, but as I’ve said before, the best word to describe Microsoft is pragmatic. And when Microsoft sees a standard as useful, it gets involved in the standards process because it wants a standard that enables the clever ideas it has - so while the Acid 3 test covers 100 different elements, with the79 new HTML 5 test cases Platform Preview 2 adds, the test suite Microsoft is donating to the W3C is already up to 183 tests - and it evangelises the standard.

There’s also an understandable desire to set the record straight. For years developers have complained about having to create different markup for IE. But by implementing features planned for HTML 5 and CSS 3 - like border styles and rounded corners - earlier, the so-called ‘modern browsers’ have added some unique syntax of their own. IE general manager Dean Hachamovitch pointed this out to us when we asked what the focus of the second preview was (it’s ‘same markup’); “You expect this one line of code to work and it doesn’t. You need to have this one line for WebKit and then you have to duplicate this line for Mozilla but you need to change the order of words since they did it a little bit differently - and that's before IE shows up. There's something to call out to the community there.”

Although he never says it, my impression is that Hachamovitch wants the Web developer community to hold all the browser companies to the same standard (no pun intended) . What he does say is: “This is the window in time for the industry to go ‘we really need to do a better job at the same markup thing’. Developers have this amazing power to make the Web great. The details of how the browsers work to help them or hold them back are key. The theme of same markup is that developers have to raise their expectations from browsers.”

But if all the browsers deliver same markup, what is going to distinguish IE 9? Until the rest of the browser shows up (what we’re getting now is just the rendering engine) Microsoft can’t talk about security (an area where IE has gone from terrible to ‘if you don’t do anything stupid you're pretty safe’) though we expect it. In another turnaround for IE 9, one answer is going to be performance according to Hachamovitch. “It's the same markup [in IE 9] and it’s the same markup run better and it's run better because of the hardware acceleration and the performance advantage. These aren't necessarily distinct; they support each other and complement each other.”

GPU acceleration alone won’t be enough because the other browser companies can do that too; (multi-threading and the just-in-time use-the-second-core JavaScript compiler also contribute to the impressive speed of a rendering engine that isn’t in beta yet). Hachamovitch (whose team is of course part of the Windows division) acknowledges that; “I think about [hardware acceleration] more as an advantage for Windows than anything that has to do with IE. Browsers on Windows can offer users amazing hardware acceleration; all the browsers on windows can do this. To me it’s an advantage to Windows customer. Frankly it’s always hard to be the first person to do something; you get to work through all the wrinkles and issues. But Windows is a well-documented platform – you will see other browser vendors pursing this and some not and consumers will see an amazing advantage.”

Talking of browser vendors who do and don’t pursue hardware acceleration, we also asked Hachamovitch why the speed comparisons they list on the IE blog are against shipping and beta versions of the other browsers but don’t include versions like the Firefox nightly builds that had GPU acceleration in? “The thing that’s interesting about using a nightly is there are never any promises. If you grab one and performance is bad, they say you should never have used that nightly, if you grab one and performance is good… You never know which one to take.” IE 9’s previews are far further apart than nightly builds (seven weeks this time and always eight weeks or less) so developers know which one they should be testing – and that’s who Microsoft is reaching out to with the previews. -Mary