Ars is running a lengthy piece on modularized Windows and why it would suck. There are quite a few reasons why I do not believe this to be true, mostly looking at the increased control modularized Windows would bring to all categories of users.

I am still holding to my belief that Windows would be better off as a split product. I believe I understand the technical difficulties no doubt present in such an undertaking and while in no way ‘business educated’, I consider it to be a more sensible approach to answering the very different needs of home and corporate users. Modularized Windows is almost what I had in mind, though not quite as drastic but much easier to attain.

So far, Microsoft has tried to blur the line between the separate versions of Windows¹. While we do have different editions for each version², they are still one product: Microsoft Windows XP. In this, Microsoft tries to make Windows feel the same, whether you are using it for work or at home, but it is quite clear the semblance stops here.

Modularized Windows is certain to please IT personnel aiming to control what users do with their machines. We can only speculate on how granular the compartmentalization of features would be, though one would assume such a thing as a Media Package would exist³ and it would be kept out of most enterprises.

In the Ars article, a claim is made that, as a developer of third party software, one likes to rely on certain components ‘being there’ in Windows. That, when selling your product, you wouldn’t want to tell your customers they also need to buy component X.Y.Z for Windows before they can use it. I do not see this as an issue. First, targeted applications are sold to people that need them. Chances are, they have already bought additional components for various other things and they understand the idea behind this model. Certainly, if you want to sell me productivity software, I might not see the need for the Enhanced Security Package, so I will look at alternatives⁴. But not being bullied into buying a much more expensive edition of Windows just for the sake of one feature is set to please far more users.

I agree I would hate to see a subscription model for features. They may work fine in terms of online applications (photo galleries and the likes) but I wouldn’t want my hard drive encryption to expire at some point. That would make my data a bit too well protected.

The downside for Microsoft is that, beyond core features⁵, they open themselves up to competition. Sure, maybe it helps out with anti-trust regulations, and I’m all for that. From a business perspective — and Microsoft is a business, trying to make money —letting users choose to pay for a DVD player, being the right thing, is the ‘not going to make us money’ thing. It allows for VLC, the Mozilla Suite and Open Office to swoop in and garner some serious market share, even more so than now. The de-coupling will make those other components independent applications⁶ and more people will slowly understand they do have an option in what they use on their computers. So while the EU and the USA anti-trust commissions will suddenly find their work days become dull, not having to fine Microsoft for anything, so might Microsoft find their application market share quickly declining with the home users.

But that is about it. Home users don’t really care sufficiently for these things, most of their time is spent online. The slow realization that the platform is becoming irrelevant for that 80% of the time is one of the reasons we hear of so many ’switchers’. Forcing the user to buy a more expensive edition of Windows just so that they can watch a DVD is a downright silly idea; giving the user a really cheap starting point — commodity hardware and bargain software — might still keep users around.

The business, on the other side, relies on Microsoft to fix problems when they don’t have the time themselves to do so. Enterprises want communication suites, backwards-compatibility layers and seamless evolution between versions of Windows. They would pay more to secure them so long as the cost of ‘the alternative’ remains unreasonable⁷. They will also enjoy the control they get over what exactly they buy — and everyone loves control.

I do not think a modularized Windows will hurt the end-users as much as some think. It will not even be such a huge step from the present-day scenario of the ridiculously ‘diverse’ editions of Vista, though quite possibly it will be far, far easier for the consumer to understand, free of complex table diagrams with checkboxes. It will also offer business more control over what they install, not only to increase security but also productivity. The right Windows edition for a job may not be the ‘fullest’ edition available, which is why the ‘pick your own’ approach could prove to be quite successful.

1. Throughout the article, a Windows version is a major release, such as Vista; an edition relates to the various flavours of a version, such as Vista Home Basic or Ultimate [↩]
2. e.g. XP Home and XP Professional [↩]
3. Not unlike Apple’s iLife and iWork suites. While iWork may be seen as an alternative to productivity software similar to Office, iLife stands on its own as one of the core marketing points of OS X [↩]
4. Really, this de-coupling can be done quite poorly if care is not taken. It would not be an easy design feat for software architects to split features equitably, as it would not be easy for developers to know what subsets they need [↩]
5. Which, really, are the operating system [↩]
6. Where this applies, such as media playback or browser integration [↩]
7. This bloody compromise that plagues the software world, where old applications are forced to live beyond any kind of original expectancy. No source code, no migration, no developers to be found. The plague of the accounting system one of my previous jobs had, running in VMS, completely unnatural for anyone that has only started using a computer in the late 90’s [↩]