Windows 7 is about to come out, so why in the world would I want to write about the next generation of Windows? One, because we already know much of what we need to know about Windows 7. It is better than Windows Vista and Windows XP, with the advancements of the former, and the footprint and lack of aggravation (for the most part) of the latter. Most of the reviews have been glowing (here is a balanced review by Harry McCracken), and I've been using the product on all but a couple of machines myself (which are waiting for RTM code) since January. Most of you will wait until Office 10 ships or RC1 comes out (whichever is last) before adopting (watch out for XP mode) but there is no real mystery in the product. Windows 8, on the other hand, is largely a blank slate and by starting to talk about it now, maybe we can help it become what it needs to be.
So this isn't what Windows 8 will be, but more along the lines of what it should be if Microsoft wants to strengthen its franchise rather than bleed share to Apple and Google.
Windows 8: OSOD
Operating System on Demand. I know it's kind of scary when I create yet another acronym. Windows has become synonymous with bloat; it is full of stuff you don't want or need, yet you have to get all of it. The EU, and likely the U.S. again, under the current anti-monopoly administration, will have problems with the number of things bundled in the platform. These bundled products are not only either not complete (Media Center) or not competitive (Media Player), some are even obsolete (Telnet-who uses Telnet?).
The problem for Microsoft is that there is an increasingly large amount of code in Windows that shouldn't be, for a lot of reasons. And, for a lot of reasons, this code likely should only be added to Windows on an as-needed basis, not defaulted into the product. Even RDP (how many people actually use RDP?), yet this remote control offering is defaulted into the product and it is on a list of things that often have to be patched because, even though you aren't using it, a hacker has found a way to misuse it and take control of your system.
This suggests that Windows 8 should be vastly leaner and that parts should only be added as you need them. Some, like Active Directory Support, already come with an extra charge (the difference between Windows Premium and Windows Professional). This way, the OEMs could load a lower-cost OS (which is what they want to do) and if the buyer needs Active Directory support, they could buy it along with their implementation of Active Directory and better match the cost with the solution rather than have it spread across products. It is interesting to note that Active Directory support, or Domain Join, comes native in the MacOS and is an extra charge for Windows, which has always seemed both foolish and counterintuitive in a competitive environment. I mean, why would you actually put your product at a competitive disadvantage intentionally?
So, I think Windows 8 should be vastly leaner, likely closer to Windows Starter Edition but just Windows, as the default, with the ability to load both free and enhancements that come with a fee on top of the base OS, likely through a Microsoft Application store, which enterprises would intercept and block after specifying the specific load they wanted on their hardware.
Connected by Default
Windows was created in a world that was occasionally connected. Windows 8 will launch is a world that is virtually always connected even when laptop hardware is involved because WiMax and LTE will make such connections readily available and affordable in much of the developed and undeveloped world.
As a result and as a competitive hedge against Google, much of what is delivered to Windows 8 will be centrally hosted, which suggests stronger security links to back-end services to ensure users are not compromised, and a higher reliance on these back-end services by default. This suggests aggressive, performance optimized caching because even though being network-connected will be more common, network loading will likely make bandwidth relatively unreliable and tie perceived performance to how effectively the OS caches to create the appearance of a high-speed network even when one is not available.
This is expected to be a key advantage to the Google Chrome OS and one that Microsoft will need to match to avoid competitive displacement. Just like on the browser today, speed in the future will be king and Microsoft will need to make sure its solution is both fast and secure enough to protect its base.
With ARM growing into the smartbook/netbook space, if Microsoft doesn't embrace the platform, it is likely to bleed market share badly to Google if Google is successful with the Chrome OS. Betting a competitor will fail is always risky. This was best demonstrated when IBM bet against Microsoft in the early 90s. But, to do ARM would mean merging the Windows Mobile 7 and Windows code bases to an even greater degree, and this may be a bridge that Microsoft is unwilling to cross. Still, since we are talking about what Windows 8 should be, not what it will be, I would add this to the list of things that need to be addressed as a competitive hedge against Google and Apple, both of which are likely to grow their ARM-based offerings (with Apple, the possibility of an Apple-sourced ARM chip it might take into the PC space continues to grow).
This would clearly separate the Windows desktop from the Windows server code base, something it has needed to do for some time anyway for the good of both products, and the result could be better focused products in both areas. As the dominant vendor, I don't think Microsoft can afford to be x86 only, given the ramp and popularity of products like the iPhone because, increasingly, they will be moving into the PC segment and, if Microsoft doesn't play this right, they will move Microsoft out.
The right path is to bridge the hardware architectures with one product, either growing Windows Mobile 8 into the desktop space, or driving Windows 8 down into smartbooks/phones. Second would be to leave the products as they are but more closely relate them to each other so that users would see them as more similar in capability than different.
Clearly the screen sizes and widening uses of this future OS will force a rethinking of the user interface. As we learned with Windows Mobile, the same user interface on a big screen does not translate well to a smaller screen, and touch will be common in this time frame. Different vendors will want different solutions depending on the device and it is also likely that some IT organizations may want either their own UI for certain classes of workers or to use one off a list of options that best matches the needs of their user base.
This ability to transform the user experience already exists in HP TouchSmart PCs and HTC Windows Mobile and Android smartphones and will be a big part of Windows 7 and Windows Mobile 7. It should be even more pronounced in Windows Mobile 8.
To be competitive in a vastly changed and connected world, Windows 8 and Windows Mobile 8 will have to go through massive changes. Strangely enough, Windows Mobile 7 will be closer to where the platform needs to go, in terms of content, services and pricing than Windows 7. This suggests some interesting dynamics between those two groups over the next few years.
Microsoft is at a break point. It will either step up and embrace the future again (and it is amazing how many times Microsoft has accepted massive changes, from IE's birth, to .NET, to open source) or lock down in the past and watch Google, Apple or some other yet to be identified (IBM?) competitor pass it by. Wouldn't it be ironic if it was IBM? We'll see what the future brings and whether Microsoft rides this wave or is pounded into the sand.