Redmond Diary

By Andrew J. Brust

Blog archive

A Tale of Two Windows

As I write this post, Microsoft’s Windows 8-focused //build/ conference has just ended. The apprehension so many developers had around the show is now dissipated, the developer platform and tools have been detailed, and we have all been able to work with the operating system, and develop apps for it, on a touch device, for about 72 hours now. Most people, myself included, like what they saw. The OS is touch-friendly without being an iOS copycat; our developer skill set investments are nicely protected; a new generation of developers trained on the HTML/CSS/JavaScript stack of Web technologies can join the party, and Windows will continue to run on a greatly diverse set of machines.

In working with Windows 8, I have felt an odd combination of excitement and concern. Each time I feel anticipation for something in Windows 8 that looks really neat, I keep feeling a counterweight, an amorphous tug of caution, pulling me back. For about 2 days, I couldn’t put my finger on what was causing this Newton’s Third Law of Windows within me. And then I figured it out. Microsoft itself faces mutual, opposing market forces. And Windows 8 reflects the company’s admirable efforts at, and daunting task of, addressing them.

There’s touch, and there’s the keyboard and mouse; there are tablets, and there are laptops and desktops; there’s ARM and Intel; there’s Metro and the desktop, there’s WinRT and the .NET CLR. And of course, there’s the consumer and the Enterprise. Microsoft’s challenge is that it must appeal to new market segments and new trends, but it must also serve – and leverage – the Enterprise, and consumers who prefer classic Windows.

Booting Windows is almost an allegory for this attempt at mutual coexistence. Windows 8 takes users right to the new Metro-style Start screen but enables them to call up desktop mode applications directly. The Metro version of Internet Explorer can take you right to “desktop view.” The Metro version of Control Panel has a “More settings” option that takes you to the standard desktop version of that same applet. The operating system seems well at home on a tablet, but you can absolutely install it on a conventional laptop or desktop and use it as an upgrade. Or perhaps that would be better deemed a retrofit.

No other software company has to do this. Oracle, IBM and SAP focus on the Enterprise. Apple focuses on the consumer. Even hardware companies tend to divide into camps: Samsung and Sony and Acer look to the consumer and, I would argue, Dell looks mostly to the Enterprise. HP tried to serve both demographics, but it seems close to focusing on businesses more exclusively.

But Microsoft is in a special place. It can neither abandon the old nor deny the new. It must serve both masters. It must mediate, it must reconcile, it must negotiate coexistence. It must cross a chasm, and that is hard. But it then must bridge that divide, and that is harder still. To do so seems foolhardy, and yet it is necessary. It seems almost irrational, and yet it is oddly logical.

Enmeshing such different worlds might appear doomed to failure. And yet the sleeping giant that so many think Microsoft is, this company that started out making BASIC compilers for microcomputers in the 1970s, is taking on the challenge anyway. Because it knows it must.

This week at //build/, many started to believe that it can.

Posted by Andrew J. Brust on 09/19/2011 at 1:15 PM

comments powered by Disqus

Reader Comments:

Sat, Nov 12, 2011 Fallon Massey

Tablets are a sub $200 market, and by XMas 2012, possibly sub $100.

The WP7 OS was the only tablet OS they needed, Win8 will NEVER succeed as a tablet, it will find success in the ultra thin laptop(or convertable) market.

Microsoft didn't understand either the phone or tablet market sufficiently before leaping, but the enterprise will save the majority of their bacon on this one.

But if they were looking to pull an Apple style domination of the consumer market, well, all I can say is I knew Steve Jobs, and MS has no one even close.

Thu, Sep 22, 2011 Ed S Colorado

Actually, Apple has historically done an admirable job of balancing these same concerns. They were smart to split the OS's to optimize them separately for mobile vs. desktop. Now they are slowly bringing them back together. We'll see if that actually works. If it doesn't, they can just continue down 2 paths. If Microsoft's grand unified strategy doesn't work, what will they do?

Tue, Sep 20, 2011 Kirk Davis Bangkok, Thailand

@Brian Morris - I think you pretty much answered your own question; while I wasn't at //build/, I did download the developer preview and install it on a (virtual) partition over the weekend, and played with VS11, built a quick hello-world app, and explored some of the namespaces available. From what I can tell, the "metro" platform (discounting the HTML5/Javascript option) is a super-set of Silverlight 5, with additional namespaces (especially those in WinRT's "Windows" namespace) that enable apps to participate in metro "charms" like sharing, being on the list of apps to handle certain types of object (somewhat akin to how Android apps can handle intents), etc. So it's sort of in-between the Silverlight run-time and the full .Net clr, but closer to Silverlight. That's my take anyway. They can both also use VB.Net as well as C# as well. And during the keynote, somebody (I forget his name) took one of Scott Guthrie's Silverlight 2 projects from his blog, changed a few namespaces, swapped out a grid for a metro control, and voila - it was a metro app. So definitely I think you're right that they're very closely related, almost kissing cousins. This should make it much easier to build a Windows 8 metro app, and then build a Silverlight 5 version for use on Mac OSX and for Windows 7 users (eg, create an OOB version for non-Windows-8 users).

Tue, Sep 20, 2011 FredG

I had the exact same thought. The only anwer that come to mind is than MS can make money selling Windows 8 upgrade. Maybe the only problem with SL in MS view is that it does not bring enough money.

Mon, Sep 19, 2011 Bryan Morris

I've been watching with concern the MS de-emphasizing of Silverlight ever since Muglia shot his mouth off last year. Like everyone else, I've been confused, concerned, pissed off, hating the prospect of devolving to HTML/JS, and completely unable to understand Microsoft's motivations in giving the finger to its developer community. All the emphasis on HTML/JS was pretty worrisome, but recently a somewhat clearer, less bleak picture began to emerge that Build has brought into focus. During and after Build I did a lot of reading and thinking about what MS was saying and not saying. Initially, my impression was that it's not so bad. Today, I had a further thought that takes the form of a question I wanted to pose to this community:

How exactly, other than the deployment mechanism (app store vs. your own website) and perceived platform "reach" (no OSX, I guess), is developing a Windows 8 Metro app in C# different than creating a Silverlight out of browser app in C#?

The both use managed C#. They both use XAML. They both use WCF RIA services. They both make use of isolated storage. They both can access local devices. If anything, the Metro app can do more than the Silverlight OOB app. Am I wrong here? I'd really like to see what everyone has to say about this question.


Of course, why the use of the word "Silverlight" to describe their "new" app model by an MS employee is probably a firing offense is a whole other question.

Add Your Comments Now:

Your Name:(optional)
Your Email:(optional)
Your Location:(optional)
Please type the letters/numbers you see above

.NET Insight

Sign up for our newsletter.

I agree to this site's Privacy Policy.