There’s reason to hope Windows 9 will be better

Terry Myerson is a tough guy to pin down. Where his predecessor, Steve Sinofsky, would post pages-long blogs explaining in raptuous telemetric detail many of the components that made Windows 7 and 8 (and by extension 8.1) tick, Myerson has a reputation for letting his products speak for themselves. That’s going to make it hard to second-guess where the next version of Windows is headed — but all indications at this point are encouraging.

Let’s start with a few points that, while debatable to some, appear to my long-suffering Microsoft-observer eyes as self-evident:

• Windows 8 (and 8.1) is one of the largest disasters in Microsoft’s history. I don’t fault Sinofsky for trying, but the Jekyll-and-Hyde two-headed monster didn’t and doesn’t work, and Microsoft would be foolish to throw more money that way. Tablets and phones have rightfully eclipsed old-fashioned desktops and laptops on many fronts, but Microsoft’s actions unnecessarily accelerated the traditional desktop Windows’ decline. Windows 8 threw most of Microsoft’s 1.4 billion Windows customers to the wolves — and they aren’t buying it. (Touch tiles on Windows Server? Puh-lease.) On the bright side: Win8 has cemented Windows 7 as the operating system of choice for companies and individuals locked into the Windows architecture. That’s not a bad outcome.
• Microsoft was crazy to bifurcate Windows RT — the “modern” programming API set — into a phone version (called Windows Phone Runtime) and a tablet/desktop/everything-but-phone version (WinRT). I wrote about the sorry state of WinRT apps on Windows Phone 8 a little over a year ago, in response to Gartner’s press release that said, “[Windows 8] provides a common interface and programming API set from phones to servers.” That was and is pure marketing drivel, and it led to one of the most basic flaws in the Windows 8 approach. At the risk of re-stating the obvious: Microsoft should’ve grown Metro “up” from phones to tablets, instead of “down” from the desktop. Sinofsky and crew thought they would be clever and move the old-fashioned desktop user base to Metro. Few customers bought into it. In fact, they ended up nearly killing the aging golden goose — and alienating a lot of corporate and consumer customers in the process.
• There’s still a huge demand for desktops and laptops running a non-touch-centric (but possibly touch-enabled) version of Windows. That demand’s declining, admittedly, but it isn’t going away any time soon, and demand for the old desktop may actually improve with voice, gesture, and other non-touch technologies that are emerging. Microsoft’s ignoring — alienating — a huge market by trying to graft a toaster onto a refrigerator, er, an upstart touch interface onto a venerable environment that plays by mouse rules.

Here’s where the optimism comes in. Although it’s very hard to read between the lines with Myerson — and his views only rarely reach the press — as best I can tell, he gets all three points. He’s uniquely qualified to pull both the old-fashioned desktop and Metro out of the fire. It remains to be seen how he’ll respond to the challenges or if he’ll get the corporate backing to pull it off.

Consider Myerson’s history. Prior to December 2008 — just before Sinofsky took over the ill-fated Windows Vista project and started working on Windows 7 — Myerson was in charge of Exchange. When he took over the Windows Phone engineering effort, his first significant decision involved scrapping the Windows Mobile operating system and replacing it with Windows Phone. In July 2011, Andy Lees, then the widely respected president of the Windows Phone division, was pushing for Windows Phone and Windows to grow together — an effort said to be supported whole-heartedly by Myerson.

At the same time, Sinofsky was pushing for Windows Phone and Windows to grow together as well, but in a slightly different way: He wanted to absorb Windows Phone into his Windows empire. If you look at the corporate politics from a software point of view, both Lees and Sinofsky were working on WinRT (called Windows Phone Runtime on the phone side), with Lees working from the phone up, and Sinofsky from the desktop down. Push came to shove, Lees was moved aside in December 2011, and Myerson took over the entire Windows Phone effort.

The two approaches to WinRT didn’t grow together, they grew separately. Sinofsky grafted one version onto the desktop in Windows 8. Myerson built a different version for the new Windows Phone 7 platform, changed it significantly for Windows Phone 8, and ne’er the twain shall meet. That said, there’s some progress. Windows 8.1’s WinRT has a few concessions to smaller devices. Windows Phone Blue (8.1? 8.5? who knows what it’ll be called) is said to have some serious scaling capabilities. But the two flavors of WinRT are still very different. Even though the desktop-down WinRT and the phone-up Windows Phone Runtime are rapidly converging, one is from Venus and the other from Mars.

The existence of two Windows Stores reflects the underlying problem: There’s a Windows Store for the desktop-down version of WinRT, and a Windows Phone Store for the phone-up Windows Phone Runtime. They’re as different as night and day — and that’s a big part of the problem for Windows developers looking for mobile markets.

In May 2009, Ray Ozzie talked about “three screens and a cloud,” with the three screens being the size of a phone, a PC, and a TV and with solutions “delivered to us in some coherent way” to each of the different screen sizes. That’s the goal here, but from the developer’s side — to make the development tools scalable to all screen sizes and amenable to cloud glue.

Maybe I’m being overly optimistic, but I think it’s possible we’ll soon see a merging of all the touch-centric pieces under a single umbrella programming paradigm, which you could call WinRT. That was Lees’ vision more than two years ago. Now, with Myerson in charge of the whole shooting match, it may come to reality.

Myerson has been talking about the approach, although in typical Myerson fashion the public words are few and far between. At the Microsoft financial analyst’s meeting last week, he said:

We should have one set of developer APIs on all of our devices. And all of the apps we bring to end-users should be available on all of our devices… Windows RT was our first ARM tablet and as phones extend into tablets, expect us to see many more ARM tablets, Windows ARM tablets in the future

The legacy desktop Windows Win32 apps aren’t going to migrate from mousing Intel to tapping ARM anytime soon, but if you interpret Myerson’s statement as referring to new apps, we have a glimmer of hope. Perhaps there’s a retro niche that can be carved out for the old-fashioned desktop, a separate legacy island that can be showered with benign neglect, while the breakneck development cycle focuses on WinRT running on all platforms. 

With one foot in Exchange and the other in Windows Phone, Myerson may be able to figure out how to get a Windows 9 — several Windows 9s — out the door in a way that respects the proclivities of both mousers and tappers. One can only hope.

The future is mobile. No doubt about it. But the past — and the present — needs better desktop support.

This story, “There’s reason to hope Windows 9 will be better,” was originally published at InfoWorld.com. Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest developments in business technology news, follow InfoWorld.com on Twitter.