The Linux Extremist

I used to work in possibly the most blinkered development atmosphere in the United Kingdom. The owner of that company lived, ate and slept Microsoft; a devotee of Visual FoxPro, it was difficult for him to imagine anything better than that. To give some idea of the difficultes that this creates, Foxpro is a single threading application; it can handle only one transaction at a time. When he demanded that we develop our e-commerce solutions using Foxpro as the backend database, chaos ensued.

I remember once giving him a copy of Ubuntu Linux, in the hopes that might shake his point of view, an attempt which was bound to be as successful as trying to convince Tom Cruise that a squirty little midget like him shouldn’t play an action hero, but I made the effort nonetheless. As the owner accepted the CD ROM, he floored me with the following question: “How does one develop solutions for Open Source servers using a Windows PC?” I had to ask him to repeat himself to ensure that my ears weren’t playing tricks on me. They weren’t.

It dawned on me that Microsoft is like that for some people. They contaminate the horizons, and managers and professionals get locked into a “comfort zone” which is not easily broken; they get used to Microsoft providing all the solutions to every problem, and regard anything outside the Microsoft Zone with suspicion. That is precisely how Microsoft would like it: they would like to provide solutions up and down the line, from enterprise computing to your MP3 player; unfortunately for them, this has the negative consequence of them being a Jack of all trades, but a Master of few.

This idea of a one-stop shop for all your IT needs has its attractions; at least one can ensure interoperability between systems. For example, it should be child’s play to send a spreadsheet from Pocket Excel to Desktop Excel. Similarly, it should be easy to develop solutions for Pocket Explorer as well as regular Internet Explorer.

However, trying to be all things to all people is simply not a viable strategy; what tends to get lost in the attempt is quality. It is very rarely stated that MSN Search is superior to Google, or that Windows XP is superior to Linux or Macintosh OS X (at least in terms of stability and performance). The Playstation and Xbox are roughly comparable, and Windows for mobile devices is not seen as infinitely superior to the alternatives.

Indeed, I do not recall Microsoft ever being considered “superior”, rather, they were just more available. I remember a conversation I had in 1997, when I was working for a publishing company. My boss asked me to lay out on the line what the difference was going to be using Windows NT versus Caldera Linux; I had problems finding reasonably priced contractors to build solutions for the latter at that time, a situation which was not helped by the fact that my company was based in a rural area.I told him: “Using Caldera is like getting a custom built car…it’s stable, it’s elegant, it’s beautiful. However the initial costs will be higher and when it does break, which will be rare, getting people to fix it will be more expensive.”

“And Windows NT?” my boss asked.

“It’s like getting a Ford Escort. It’s adequate, nothing more, and when it breaks down there are loads of people around to fix it.”

Less than 10 years later, the shoe is on the other foot - the availability of Open Source developers is on a par or greater than that of Microsoft-only programmers; the costs of using Open Source are unquestionably less, yet the solutions are still beautiful.

Thanks to this change in fortune, Microsoft’s retreat has begun in earnest, but rather like a spoiled child, when told, “What part of No don’t you understand?” they still reply, “All of it.”

Rather than see Microsoft do a full re-think, we are witnessing an expansion of their all-encompassing philosophy: not only is there MSN Search, there is also the proposed MP3 player, and of course, Windows Vista, which will incorporate an even larger bundle of functions in a single OS, trying to be ever more to the user.

Previous experience should have indicated that this is only a recipe for mediocrity. However, Microsoft is apparently immune to learning. They should be taking hints from the Linux model, whereby tools are small and specialised. The kernel is not “everything”, and development is carried out by many individuals and organisations, rather than under the umbrella of one, all-encompassing corporation. Linux users are very happy to have a kernel from one project, a browser from another, and an office suite from yet another. Open standards ensure that there is consistency between projects, and the results are the conglomeration of the work from the Masters of Few Trades, rather than the thin gruel from a less-than-optimal Jack of all.

Microsoft can continue to rely on those who refuse to be saved by the Linux Revolution. However this group, in my experience, is shrinking; when the bill for licensing costs comes in, ever more managers are saying, “Surely there must be a better way.” There is. Sure, there will be people like the owner devoted to Visual Foxpro, but they are a dying breed. If Microsoft doesn’t want to become a dinosaur themselves, they’ll have to learn to narrow their focus, specialise, and most of all, learn from their rivals and their mistakes. If so, they have a future. If not, the horizons they clutter are going to be ever fewer.

Trackback URI | Comments RSS