"People go to the Web for sites, not the browser, much as they go to their PC for apps, not Windows."
That's what Microsoft Corporate Vice President Dean Hachamovitch told attendees of Microsoft's Internet Explorer (IE) 9 beta launch in San Francisco this week, where Microsoft made available for download a public Beta 1 of IE 9. Hachamovitch's comment -- and Microsoft's showcasing of a number of big-name partners customizing their sites and apps for IE 9 -- got me thinking about what happens when the line between apps and sites are not so clear-cut.
Here was my thinking: If IE 9 is designed to blur the lines between Web sites and Web apps, does that imply that Microsoft is counting on content developers to do with IE what the company did with Windows -- namely, to build apps that work better in Microsoft's environment than anywhere else? In other words, is Microsoft looking to use the "Windows first/Windows best" strategy that worked for operating systems with the Web? (Looks like Technologizer's Harry McCracken's been wondering about some of these same things, as he noted in his post this week entitled "The Unwelcome Return of 'Best Viewed with Internet Explorer")
A hypothetical example: Will the Rough Guides travel site/app in IE 9 look and work the same as it will in Chrome, Firefox, Opera and Safari? Will it require the developers of that site to maintain completely different versions of their site for IE 9 than other HTML5-compatible browsers?
Ryan Gavin, a Senior Director on the Internet Explorer team, provided answers to my questions on this. Here's a transcript of our e-mail exchange:
MJF: The idea of IE-9-optimized Web sites seems to me to be contrary to Microsoft's idea of “same markup” everywhere. If some sites will look/work better if optimized to support certain features, isn’t this still creating a situation where devs can’t write once and browse with anything?
Gavin: There are two things here.The first is that IE9 has HTML5 and other web standards at its center: this support along with our work with the W3C and the creation of comprehensive standards test cases are helping to fulfill that goal of ‘same markup’ for developers. Once you have that, you can unleash the full potential of HTML5 through things like fully hardware accelerating the browser through Windows, like what did with IE9. This allows developers to use the full power of the PC, and create interoperable HTML5 experiences that are far richer and more immersive than what we know today.
The second thing is that developers should be able to bring their sites out of the browser box. By taking advantage of pinning, Jump Lists and notifications in Windows 7, sites like Facebook, Twitter, Amazon, WSJ, Discovery, and the Killers can make their sites behave more like native applications with just a few short lines of code. Of course, these same sites viewed using another browser will still work, but customers will remain inside the browser box.
MJF: What happens to sites that don’t customize for IE 9 and just create “standard” HTML/CSS sites? Do their sites look/work less well than the IE 9 optimized sites?
Gavin: Absolutely not, is the short answer. As I said, the goal of ‘same markup’ is grounded in interoperability of the web. HTML5 / CSS sites will work great in IE9. And since IE9 includes compatibility mode that falls back to IE7 or IE8, users will find their sites become more beautiful across the board when visited using IE9. You can go to http://beautyoftheweb.com and play with a number of new HTML5 experiences that are being built by sites like USA Today, Quicksilver, MySpace and check out how they work in IE9.
MJF: Once a content producer customizes for IE 9, will that site still work well in other HTML/CSS browsers, like Chrome, etc.? Or will devs have to make tweaks for them to work/render correctly with other browsers?
Gavin: Our goal is that developers never have to make code tweaks, and if they do, it’s at the feature level not the browser level. To give you some more context, today web developers end up writing whole pages of code for “browser detection” – if browser = x, run y page. That’s what we want to fix. Instead, developers should be able to write a page of code and if they want to add something they write a feature level detection instead.
Video is a good example of this: a developer writes their whole page and wants to include a video so they write a feature level detection for just that one part. So if someone is using an older browser version and wants to view a video, run a plugin; if a customer is using IE9, run H.264. You can imagine how similar settings could be inserted for other browsers, too.
And the big thing here is when developers write feature level detection, they can get a return on that code (like return on investment). You can imagine a developer recognizing the resources they saved from code being more interoperable, and investing those resources for to create additional richness in a different area (like inserting code for other cool features on the site).
MJF: The idea of Web sites looking/behaving more like apps also makes me wonder about Microsoft’s (former?) stance that local Windows apps are the reason users want Windows PCs with Windows apps running locally. Doesn’t the work you are doing with content providers to make sites more “app like” make the Microsoft “local apps are better than Web apps” position moot?
Gavin: Customers tell us the experiences they have with their content, favorite sites and applications is what matters. Increasingly so, it’s going to be impossible to tell where “running locally” stops and “running in the web” begins. It’s just one end-to-end experience. We think the best experiences will need to use the whole PC to create the level of richness that tomorrow’s web will demand. Today, what you get from many of your native applications on Windows will start to appear in web applications as they take advantage of GPU powered HTML5 in IE9, and you are in fact using the power of the whole PC through Windows to enable that. The reality is that web sites get better as we continue to help them tap into Windows software and PC they run on.
MJF: I feel like MS is going with IE9 the same way it has gone with Bing: It is using telemetry data to customize for the greatest-use cases, but leaving folks who are “outliers” -- like me -- outside the equation. The reason I don’t use Bing as my default search engine is it is optimized for things like shopping, travel and celebrity gossip. But often Bing can’t find a story I wrote that I’m looking for. I am wondering if this will happen to me with IE 9, as well. Is going with the “majority experience rules” going to drive the 20% who are outside the “norm” to Firefox , Chrome, etc…?
Gavin: We improved IE9 based on data from our customers, similar to our approach with Windows 7. The keyboard shortcuts and features users are familiar with, favorites for instance, are still there – we recognize our customers pick Windows because they want choice.
That said, it’s been interesting to look at the usage stats of things like the task bar on Windows 7 versus the favorites bar on the web. Some sort of favorites bar has existed for 15 years and only 4% of people have added a site to their favorites bar. Whereas, less than a year after Windows 7’s availability over 33% of people have pinned at least one of their favorites apps, and 87% of our customers launch applications directly from the task bar. Based on data like that, we built IE9 to make things work the way our customers now find familiar. You shouldn’t have to learn new ways of doing things to browse the web. Browsing the web is the number one activity on Windows, so why not just make it work like you’d expect. Aero Snap, Jump Lists, pinning, and even our new UI are all reflections of our focus on making your sites shine and work they way you’d expect in Windows.
Anyone have any additional observations or questions about customizing sites and apps for IE 9?