The Future Ain't What It Used To BeThe Future Ain't What It Used To Be

China will officially roll out the unwelcome mat for Microsoft. In some ways, this process is already underway: The Chinese government has implemented a "buy local" IT policy, which it recently enforced by ordering the Beijing municipal government to break off a planned business relationship with the company.

Besides the Chinese public sector rejecting Microsoft and embracing Chinese-made (and largely open-source) software, I also predict the government will apply heavy pressure on the country's private sector to follow suit. As a result, Microsoft's problems with pirated software in China today will look like child's play compared to the mess Redmond will face a year from now.

Enterprise compliance initiatives, such as those addressing Sarbanes-Oxley, will pose a major challenge for the open-source community. It's already clear that auditors are digging deep into firms' IT organizations, looking for weaknesses that could compromise the accuracy and integrity of their financial data. This won't stop the growth of open-source in the enterprise, but it will force firms to work only with vendors that can vouch for the integrity of their open-source products. In the end, this type of accountability will only benefit open-source vendors, giving them a much more solid footing as they compete for enterprise market share. But in the short term, it could cause a lot of worry and more than a few misguided doomsday predictions--and plenty of FUD from proprietary software vendors eager to get in their licks against open-source competitors.

Given the first two predictions, it might seem like a safe bet that Microsoft will finally see the writing on the wall and make a major effort to embrace the open-source model. Don't take that bet: I expect Microsoft to file a patent infringement lawsuit against at least one, and possibly more than one, major open-source development organization. Worse yet, it will quickly become clear that Microsoft has a solid case against its chosen target, given the sheer depth of the intellectual property war-chest the company has built over the past few years. Mitch says: "This ties in with my prediction, above, about the GPL taking a pounding in court. Matt, I think you agree with me, and I also think you've correctly identified Microsoft as the guy who's going to deliver the beating."

The SCO lawsuit won't exactly go away, but it will gradually slide into complete irrelevance. First, the Microsoft lawsuit I mention above will completely overshadow the SCO-IBM case. Second, it will become even more clear over time that SCO doesn't have a case, but the company simply won't have any choice but to continue fighting. Third, SCO's legal team, which is now working under a $30 million fee cap, won't be able to drop or settle the case without their client's consent, but they also won't devote any more resources to this lost cause than they have to. Mitch says: "See, this is the prediction I was making about the SCO lawsuit last year. I agree with this prediction of yours, Matt, but I think it's already happened."

Sun is going have a serious public relations problem with the Linux and open-source communities by the end of 2005. Several events will conspire to make this true: Sun won't come any closer to open-sourcing Java, and in particular J2SE, than it is today. I personally think there are good reasons for Sun's position, but I also think the company is going to do a lousy job defending its decision. I also foresee serious dissention within the Java Community Process, which may be on the verge of falling apart by the end of the year. Sun won't follow through on its plans to open-source Solaris. (Yes, I'm going out on a limb here: The Open Source Initiative just gave Sun's proposed open-source license its blessing, and Sun plans to begin the process as soon as this month.) At the same time, Solaris 10 won't live up to the company's Linux-killer expectations. This will get ugly: Heads will roll, and they're likely to be very highly placed heads. Although Sun insists that its feud with Red Hat doesn't extend to the greater Linux community, the company is likely to do something--or several things--that suggest otherwise. Finally, Sun won't join Eclipse, even though it's becoming more obvious by the day that refusing to do so is a short-sighted, self-destructive position.
Naturally, I have to include a Firefox prediction, so here it is: By the end of 2005, the open-source browser will have a solid 10-12 percent share of the market. The problem is, this will happen during the first few months of the year; after that, Firefox is going to hit the wall. I don't like this prediction; in fact, I hate it. But as my colleague Scot Finnie recently pointed out, Firefox's fate rests largely in the hands of corporate IT departments that have standardized on Internet Explorer. There are plenty of good reasons why enterprises should get off their butts and do what it takes to make Firefox work for them. Unfortunately, I don't think any of these reasons will sway the vast majority of CIOs and IT directors who are simply too risk-averse and short-sighted to make a clean break. Mitch says: "I expect they'll have that market share by the first quarter. And I won't predict that they'll hit the wall--I'm so surprised to see Firefox get as much market share as it has that I refuse to predict any limit to its growth."

The desktop Linux market this year will be plagued by a series of false starts, dashed hopes, and extravagantly well-funded Microsoft FUD campaigns. In spite of it all, by the end of the year Linux will have a small but significant--and steadily growing--share of the desktop market.

Several readers predicted that Microsoft will go beyond simply patching Internet Explorer and release a major upgrade this year. A lot of folks in Redmond will have to put in some serious overtime to push out a major IE update before the company releases Longhorn, and most Microsoft-watchers don't think this is a serious possibility.

The old Microsoft--the one that reduced IBM executives to a bunch of blubbering babies in the 1980s--would take the Firefox challenge seriously enough to sit a few dozen developers in front of their workstations, lock the doors, and tell them they can leave when Internet Explorer 7 is ready to go. The new Microsoft is so complacent about its dominating market position, especially among those enterprise IT executives I mentioned before, that it wouldn't even think about deviating from its game plan. If this prediction is going to pan out, the old Microsoft will have to come out of hiding, and fast.

One reader had a good take on what is necessary for desktop Linux to gain any real traction. "BillK" writes: "My prediction for Linux for 2005: More Windows users will switch to Linux. Not much, eh? I had better explain. Windows users come from an insulated, GUI environment where (almost) everything is done for them. They will expect no less from Linux. They expect to click on an icon to install a package. Every tweak, mod, table change, etc. is done via a GUI picture. If a Linux distribution expects them to open a console, use apt-get, or even think about dependencies, then it will be dumped immediately. Distributions like Xandros, Ubuntu, Linspire, etc., which attempt to mollycoddle the user, will sweep the board for ex-Windows users. Face it, Linux experts who build their own distributions will become a tiny hobbyist clique. The vast majority of future Linux users will not have a clue about how it all actually works. They will just use it. That's what becoming a mass-market OS means." Amen. I hope the Linux distro vendors are listening, because their success depends on whether they grasp this deceptively difficult concept.