Open Source Windows?

I had to laugh at Steven J. Vaughan-Nichols tongue-in-cheek suggestion that Microsoft could solve its Vista uptake and quality problems by open sourcing the OS.

Hiding behind his suggestion is the reality that, despite putting thousands of people and billions of dollars into it, Vista doesn’t seem to deliver high quality — this despite several years of Microsoft’s commitment to quality (in fact, at one point, they used the quality initiative as a reason for one of the many delays to the product).

Of course, despite its lack of quality, the OS doesn’t really deliver much in terms of features, either, as I described in an earlier post. I mean, after five years of trying, transparent icons of your open windows and widgets on the desktop?

What I think is really interesting about Steven’s blog posting is his quite rational diagnosis of Microsoft’s problems and prescription to address them: let open source coders loose on the product so that the benefits of open source development can be applied to solve the problems.

Open source development methodologies have shown that they fix bugs quickly, rather than letting them fester for years. Open source also allows end users to solve problems that those who control the product don’t understand or care to fix themselves.

In the abstract, Steven’s ideas have merit, but they founder on one central reality: they conflict with Microsoft’s business model, and when user needs and business models clash, business models win every time.

After all, what end user demanded Vista’s byzantine hardware security mechanism? The answer is no-one. Hollywood content providers influenced Microsoft, which was willing to listen to them rather than the millions of end users that are inconvenienced by the feature. On the other hand, Microsoft is looking to form lots of business relationships with those providers for other products and services: Zune, Xbox, the inevitable DRM-friendly response to YouTube. In other words, Microsoft’s business model trumps end user needs in the core software offerings.

This isn’t just a rag on Microsoft rant. I mean, what else would you expect any rational commercial entity to do?

More to the point is where this need to protect its business model will take Microsoft.

By releasing a complex, expensive, light-featured, user-unfriendly feature intrusive, quality-compromised product, Microsoft is failing to meet the needs of its core user

base. And this is where trouble looms.

People who want less expensive, good enough alternatives are not being addressed by Microsoft’s products. And this makes alternatives like a Linux desktop more acceptable: perhaps not quite as good, but a lot less expensive and unencumbered by all of the stuff thrown in to help Microsoft’s other business initiatives.

Clayton Christensen addresses this situation extensively in his books. One example that stands out is the rise of the inexpensive steel minimills when big integrated steel companies ruled the roost. The steel those minimill companies turned out wasn’t that good (i.e., it was the Linux desktop of steel), but it was a lot cheaper than the alternative offered by the integrated companies. And because the minimills had a different cost structure and business model, they could be happily profitable in low-price parts of the steel business.

The minimills started out with rebar, a low-quality product that the integrated companies didn’t make much profit on. In fact — and this is crucial — the integrated companies were glad to lose the rebar business because it improved their profitability. In other words, you drop low-margin revenue and your remaining high-margin revenue makes your overall profit margin increase; everybody’s happy.

But the minimills kept improving their  products and repeatedly took away markets from the integrated companies. The minimills could make money in those markets, while the integrated companies were happy to drop them because they were lower-margin than their most profitable product offerings. Every time the integrated company dropped a market, their margins improved.

You don’t have to be a fortune teller to know where the steel story ends. As the integrated companies retreated to the highest-margin tip of the pyramid, eventually the minimills took everything away from them.

How does this tale apply to Microsoft? Well, first off, there are tremendous numbers of people who aren’t well-served by its products. I read somewhere where the company gets 2/3 of its business from North America. One can speculate that a lot of the rest comes from other parts of the developed world. For all the other countries in the world, the products are too expensive, and they’re likely to turn to less expensive alternatives.

But there’s another aspect of this situation that needs to be looked at, and I think it’s more troubling for Microsoft long-term. Those steel minimills

were able to offer lower prices because they had lower costs and found those markets profitable at prices the integrated companies found unpalatable. In other words, the minimill’s cost structure enabled them to repeatedly take markets away from the established players. And this is Microsoft’s Achilles Heel.

Open source development and distribution is inherently less expensive than Microsoft’s proprietary approach, and can therefore thrive where Microsoft cannot. And Microsoft is locked into a business model where it has thousands and thousands of people (i.e., committed costs) dedicated to its development and distribution methodology.

Moreover, beyond the fact that all those thousands of people are dedicated to the Windows products, necessitating a high cost structure, is the reality that all of those other Microsoft initiatives like Microsoft Live, etc., etc., etc., are funded by Windows and Office. Microsoft requires the profits of those products to pursue its business model in other areas.

There’s a tremendous arrogance in high technology that sneers at the losers of other industries. The unspoken thought is “we’re a lot smarter than those guys in the steel and auto industries, we’d never make the same kind of stupid mistakes they did.”

Guess what? They were the geniuses of their time, visionaries of the highest technology of the age. They foundered on their success, not failed because of their stupidity. As markets evolve, the exact same factors that made you successful in the past become shackles in the future. Microsoft requires the profitability of its core markets to fund the losses of everything else it does. And when the profitability of the core market deteriorates, all the costs of all the loss-making initiatives compounds the misery. Eventually, Microsoft’s core profits will be overwhelmed by a bloated cost structure, swollen by the company attempting to play in every profitable technology market — even those that it can’t be profitable in and that directly clash with its core market offerings.

I’m a lot more impressed today by the potential for the Linux desktop than I have been in the past. It’s moving from the utterly inadequate to the tolerably acceptable — and that’s enough for an important segment of the current market as well as the currently unaddressed market. I’ll be attending the next Linux Foundation Desktop Architect Meeting in June. The first DAM gave rise to the Portland project, which has

made great strides in fostering Linux desktop applications. I’ll give a readout on my thoughts after that meeting.