I've been dolling out advice to Microsoft since 1994, when I got on the Windows 95 Beta (then called Windows 4.0) in order to co-write my third book, which was about that OS. I'm sure most of that advice was unwelcome at the time, since by its nature you only provide this kind of information when there's a problem, real or imagined.

Other times, Microsoft has reached out to me. During the Windows Vista debacle, for example, I was asked what I would do to improve consumer perceptions of the OS, which wasn't as bad as the pundits and tech bloggers claimed. Frankly, I didn't have many specifics, but I cautioned the company that it would need a steady series of highly-regarded releases before it could ever possibly recover. I pointed out that consumer distrust of Microsoft in the wake of Vista was like the distrust that US citizens developed in their government in the wake of various scandals in the 1960's and early 1970s. It was possible things would never get better, I thought.

With Windows 8, more than ever before, Microsoft's core business is standing on the edge of a precipice. I firmly believe that Microsoft needs to get this one right, that if they don't, if they achieve only a Windows Vista level of "success"--that is, that Windows keeps selling simply because it is Windows, but many users and businesses quietly downgrade to some previous release--then it's the beginning of the end: Customers will simply turn en masse to simpler computing devices such as those based on Apple iOS and Google Android, and Windows will slowly become passé. It will continue to be the most popular OS on traditional PCs, yes, but that market will shrink compared to the larger market for mobile devices. Irrelevancy follows.

If, on the other hand, Windows 8 is wildly successful--and by this I mean a Windows 7 style of success but on both traditional PCs and new, iPad-like tablet devices--then Microsoft will pull off an amazing transition that, I believe, is unprecedented in the admittedly short history of the personal computing era. That is, it will have made the transition from being dominant in the PC market to being relevant in the consumer electronics market. It will establish Windows as a credible alternative to iOS and Android, and will bring along hundreds of millions of customers--perhaps billions--as a result.

Is this possible?

Of course it is. But since we're in a weird quiet time between the incomplete Developer Preview version of Windows 8 from September and the feature-complete Beta version of the OS that Microsoft will deliver in late February 2012, and because the software giant is painfully silent about several key details of its strategy, especially around ARM-based versions of the system ... we just don't know. We don't know what the company is going to do, how  it intends to establish a foothold in this new and one day humongous new market, and most certainly why it has engaged in a veil of silence to date. It's all a big question mark.

What a wonderful time to speculate.

Of course, given these conditions, I've been speculating about Windows 8 for a long time. Last year, around this time, I wrote two articles about this very topic, How Microsoft Can Fix Windows 8, Part 1: User State Virtualization and How Microsoft Can Fix Windows 8, Part 2: Virtualize Compatibility. Both, notably, involve virtualization of some kind, and for a reason: I really wanted (still do) to see a split between the backwards-compatible but messy and hard to update underpinnings of the past and a new, presumably more modern and sustainable future. Virtualizing some of Windows' key features seemed like a good approach. Well, to me.

I also wrote an article last year called How Microsoft Can Fix Microsoft (in 2011). In it, I argued that maybe it would be better for the company to "start over from scratch" with a smaller, componentized, and simplified version of Windows, rather than try to meld its top-heavy current product to new form factors like tablets.

Microsoft, of course, has split the difference. We now know that Windows 8 will include a new runtime engine, the Windows Runtime or WinRT, that will provide a new environment for Windows Phone-like, Metro-style apps and that this new runtime will be the default shell. But the old shell, the Explorer desktop, will still be present, and not virtualized, giving access to all of the applications and functionality of the past. Microsoft calls this "the best of both worlds," but for many users the experience of moving between the two environments is jarring. That may become less of an issue once we have more Metro-style apps to replace our current, old-school Explorer applications. But it's controversial, no doubt about it.

After Apple in February released a developer preview version of its now-current Mac OS X offering, Lion, I wrote an article called What Microsoft Can Learn From Mac OS X Lion. This article can form the basis for any discussion about "fixing" Windows, since Apple and Microsoft are clearly on complementary trajectories when it comes to combining their respective desktop and mobile OSes. And what I recommended there applies, largely, to the discussion today. Key suggestions include:

One SKU. Microsoft currently markets what appears to be about 117 different versions of Windows 7. Apple, meanwhile, has exactly one desktop OS called Mac OS X Lion, and if you want to run it on a server, you pay $50 and download an add-on. That's the way it should be done. Microsoft is doing wrong by its customers by nickel-and-diming them over which features are available in which product versions. It's shameful.

Better resolution agnosticism. The classic Windows GUI is built on resolution-dependent bitmap graphics, so onscreen elements are whatever size they are, and while some offer a few different versions, the built-in controls for enlarging UI controls has always been hokey at best. Mac OS X has always done a better job with graphical and textual fidelity, despite also being built on bitmaps. But I've got good news here: While the Metro-style apps of the near future are not technically resolution independent, they will utilize Silverlight-style transforms to provide something pretty darn close. That is, an app scaled to run on a 1920 x 1080 HD display will still look great, even though it was developed for a 1366 x 768 screen.

Get the Recycle Bin off the desktop. This is sort of irrelevant now since Microsoft is moving beyond the desktop interfaces of the past 15 years.

Simpler app launching. Lion includes an iOS-like app launcher called Launchpad, but now we know that Windows 8's Start screen will go many steps further with a customizable interface of live app tiles. So this one is already in the can.

Simpler app discovery, installation and maintenance. Mac OS X Lion includes an integrated App Store, but now we know that Windows 8 will as well. So done and done.

These suggestions are what they are, but with many months having elapsed since I wrote that article, during which time we received a working if incomplete version of Windows 8 called the Developer Preview--which I'm using as my full-time desktop OS, as you may recall--some more suggestions come to mind. Are they enough to put Windows 8 over the top? Who can say? But this is what I've come to after a few months of being firmly immersed in the Windows 8 Developer Preview.

ARM is for devices, x86 is for PCs. Steve Jobs' legendary "cars and trucks" discussion (where he said that iPads were like cars and PCs were like trucks) has really resonated with me this year, and it's quite clear that an iPad or iPad-like device is exactly the right level of functionality (pretty high) and complexity (pretty low) for many, many people. But of course traditional PCs aren't going anywhere, and even if PC sales in 2012 are "flat" as predicted, that means that almost 800 million PCs will be sold between 2011 and 2012, an amazing number that can't be ignored. So Microsoft needs to address both the "car" and the "truck" market.

ARM compatibility is, I believe, the key to making the former happen. I've made this argument before, many times, so I'll just summarize it here. ARM-based versions of Windows should be treated as devices, with no classic Windows desktop or ability to run Explorer-type applications at all. They should offer only the new WinRT, Start Screen, and Metro-style app compatibility, and can take on the iPad and Android tablets toe-to-toe.

Meanwhile, "Intel compatible" (x86-x64) versions of Windows 8 should be treated as PCs, and can come in any number of form factors, including slate-style tablets of course. But because these versions of Windows 8 will offer more functionality, including backwards compatibility with today's applications, they will be the more premium (but also complex) systems. 

How Microsoft can/will market the difference between the two is unclear, though the company has publicly stated that it has a plan for differentiating between ARM- and x86-based variants. That should be interesting.

Multitasking. Some people take exception to me using the word "fix" in the titles of these articles--well, not "some" people, they're almost always Apple enthusiasts--but in this case at least, the word is apt. The current multitasking model in Windows 8, where you can only run two apps side-by-side in very limited ways, is most definitely broken.

There's a lot to be said here, but the big thing for me is that Windows 8 should work like a car when it is a car; that is, this simpler multitasking model is fine for a keyboard-less slate tablet device. But when you run Windows 8 on a 30-inch screen, or on multiple monitors, with a desktop machine, two apps side-by-side is unacceptable. Completely unacceptable. Speaking of which....

Full screen doesn't work on a real PC. Ditto for the full screen apps requirement in WinRT. This works fine on a tablet, but is completely unacceptable on a real computer. Don't believe me? Try to spend a day just using different browser tabs in true full screen mode (hit F11 to enable). It's not just impossible, it's infuriating. It does not work.

Microsoft needs to enable a "truck" mode where the Start screen and its apps can be accessed via a separately floating windows on large screen displays. It's that simple. 

Windows Phone. As with ARM, it's unclear what Microsoft plans to do with Windows Phone and whether the next, or some next, major version of the OS will be based on Windows 8. My advice is to go for it, and to allow existing and new Windows Phone apps to run like widgets/gadgets on top of a PC desktop, in little 480 x 800 windows. How neat would that be? 

(On the flipside, my early forays into Windows 8 development have been less than positive. This is odd to me since Windows Phone development is excellent, and easy. Maybe I'm missing something, but I'll investigate this more in the new year.)

I wrote separately about how Microsoft can fix Windows Phone in 2012 if you're curious.

Really integrate with the cloud. There are signs that Microsoft intends to integrate SkyDrive, Windows Live ID,  and other cloud services deeply with Windows 8. This can't go far enough, in my opinion. Everything Windows 8 does should occur in the cloud and be synced with the PC you're using. Your user account and settings, your documents, your media, your apps, everything. You should be able to seamlessly and securely log on to any Windows 8-based PC anywhere and get access to your stuff. And then you should be able to move on, because nothing is left behind. This will change the dynamics of work-related travel in significant ways, especially if Windows 8-based PCs can be built into hotel rooms, airplane seatbacks, and the like. You could watch part of a movie on the plane, then finish it in the hotel, all without actually carrying around a computer. The potential is limitless.

Strip it down. In keeping with the earlier advice about "one SKU"--which, let's face it, is a non-starter--Microsoft should strip as much cruft out of the OS as it can. And I mean be aggressive about this: Remove junk like WordPad, Paint, and whatever else, and put all that in the Windows Store, for free, along with what used to be called Windows Live Essentials (Mail, Messenger, Mesh, Photo Gallery, Movie Maker, and so). Make Windows as small and light as possible and let users find and grab the stuff they want. Microsoft can put "app starter packs" in the store, or promote their stuff however they see fit. But by removing this stuff from Windows, they will also remove the inability to upgrade and update these apps regularly. And then they can be improved steadily, which they'll need to be, since they'll all be new, if you will, in Windows 8. (Since they will be Metro-style apps now. Right?)

That's what I've got for now. But something tells me this list is going to grow or at least change once we learn more about what Microsoft's planning for Windows 8, first with the Beta release and Windows Store preview in late February. I can't wait.