In a fascinating recent book called The Tipping Point, an American journalist, Malcolm Gladwell, speculates on the ways in which social change can spread like an epidemic, quickly and often unexpectedly. It can be argued that Linux in particular, and open source in general, has likewise reached a tipping point in mind and market share within the IT industry.
This is likely because several things are happening at the same time. The first is that achieving value for money, never far from New Zealand IT budget minders’ minds, has become more important due to a kind of post-dot-com, post-September 11 malaise (regardless of the actual shape of the national economy). It's also become even more possible. This is because, setting aside the still large questions of support and in-house expertise, it is now possible to create a fully functioning client-server or web-based infrastructure out of mature open source technology. A third is the taking of steps by incumbents to corner more of your budget, such as Microsoft’s Software Assurance licensing scheme. But more of that later.
We’re still quite some way from widespread Linux and open source applications on the desktops of big companies. It's about existing investments, it's about the cost of changeover, it's about a shift in management mentality. But if there is the desire and the will, it's going to happen. The open source server market is healthier. But even if one in four servers ships with Linux, as the box-counters tell us, it’s still a few flakes in a snowstorm. Local figures are hard to nail down, but Linux server sales worldwide totalled about $US400 million for the first quarter of this year, according to Dataquest, while Unix servers accounted for about $US4.5 billion and the mostly Windows Intel server market was about $US4 billion. Moreover, Linux is probably just now eating into the Unix market more than Windows.
At bottom, it’s a world of choice more than it has been in a long while. And that goes both ways. Just as we often hear of a Linux box beavering away on some dedicated task within a Windows-only site, developers tell me an NT box wrapped in web services standards could just as easily be a fifth columnist in an otherwise Unix or mainframe shop. Choice also means that, on the development side, Java and .Net will have to get a lot closer yet. But this is inevitable, at least according to one guru I heard, a Big Fiver who's paid a lot more than me to prognosticate on technology.
Microsoft, which has got into a commanding position on so many fronts, doubtless feels beleaguered because of all the tech and mainstream press that open source is garnering. But it could well be time for it to disregard all the Register and Slashdot sneers and view what's happening with something of the spin of the pro-life/pro-choice brigades. People aren't anti-Microsoft; they're pro (their own) choice. To its great credit, Microsoft is showing some interest in developing a more open source-like support community and is apparently to show at a LinuxWorld Expo in August. Time will tell.
Regardless of what its detractors say, Microsoft didn’t become a fabulously successful software company just through marketing. XP, for instance, which I recently installed, so far seems a stable, sensible OS. But for a company that’s done so extraordinarily well, Microsoft appears to find it hard to really listen to its users. Product activation, for example. XP's not that good that it warrants a loss to users' privacy. Do users really want bloatware, requiring greater system resources and delivering functionality almost no one will use? Passport, anyone? Software Assurance? Palladium?
There’s increasing hard evidence that IT execs in New Zealand are unhappy about Software Assurance. They feel they're having a more expensive system of licensing their software foist on to them. The irony is, there’s also some feeling that lower yearly dues might have been enough to quieten much of the anger. If this is true of a majority, it suggests it’s not the concept of an annual fee that’s at fault – it’s coming, like it or not, they realise – but the cost of that imposition.
Regardless of what happens with Microsoft's US antitrust trial, or the Commerce Commission’s fingering through the entrails of software licensing here, change elsewhere will push all that aside. The customers of Microsoft -- and any other vendor, for that matter -- are going to mix-and-match even more in future to suit themselves and their companies’ cheque books.
Sure, it’s probably making for harder administration, it's demanding a rethink of support structures, of business relationships, of attitudes, but the days of forced homogeneity, of being locked in to an IT straightjacket, are gone.
Your thoughts, with Software Assurance in the subject line so I don't delete it as spam, to Broatch.
Broatch is deputy editor of Computerworld. Send letters for publication to Computerworld Letters.