Hail And Farewell, Part Two

For my last blog on open source, I've peered ahead -- inasmuch as anyone in this business can look ahead -- and made a few careful guesses as to what awaits us in both the open source and proprietary worlds.

The first thing is how both the "open" and "closed" worlds are going to continue to be inextricably bound up with each other. Software stacks that mix open and proprietary bits are going to become all the more prevalent and useful -- for both individual users (e.g., Android) and those building backends. There's too much that's genuinely powerful and important in both realms to shirk one for the sake of the other.

How this will happen doesn't seem to be unfolding according to any one plan, since every platform brings with it a different strategy. Linux, the by-default open platform, suffers from a lack of commercial-grade binary-only offerings. It's a great place to go if you're offering nothing but open source, but an entirely open source software market -- one where all the costs are in services or consulting -- simply isn't realistic right now. Windows is a closed and for-cost platform, but runs the vast majority of open source packages on top of it. And Solaris is both open and binary-compatible with multiple iterations of Linux, but hasn't made the kind of inroads that Linux has. (Not to mention there's a great deal of lingering worry about Sun's future in the hands of Oracle.) And so on.

But maybe this diversity of approaches is exactly what's needed. No one approach could possibly satisfy the whole market -- not as it is now, and not as it changes. The Apples and the Microsofts and the Red Hats and the Googles and the Suns/Oracles/MySQLs of the future will all continue to butt heads and stake out territories, defending them fiercely and turning an eye here and there to whatever new realms they can stake out. For such a landscape, you need to build a diversity of structures: a castle here, a bivouac there, a bridge across that divide over there. The landscape may never even itself out, but it can at least be traversed, and traversed with a variety of vehicles.

Out of such diversity has come many success stories. What's worth wondering about is how open source will fuel the future of software, one where applications and operating systems and platforms are all that much more fluid and interchangeable. Pop quiz: is a software appliance that runs in a VM a development platform, an application or an OS? All of the above? None of the above? Small wonder people are nervous about the big rising players all being folks developing web apps -- Google, Twitter, Facebook, the next big thing(s) after all those people. Most of what they create can only be consumed through them, and open source lovers are not fond of the idea of good, hard work on their part simply being used to make that many more walled gardens.

Thing is, even the garden walls can be vaulted over. In the past two years, some of open source's biggest adversaries -- both self- and other-appointed -- have faced up to open source being a fact of life. The most obvious is Microsoft, which has remained by and large a proprietary outfit but is gradually adding open source elements to its mix. The company's hesitation isn't hard to understand -- it has a legacy to uphold as a closed-source shop, and it has the wary eyes of stockholders on its balance sheets, not all of whom are sympathetic to the idea of "just giving things away." We may never see the vast majority of Microsoft's offerings produced in a way that will make open source advocates smile -- but maybe that's not the point. If nothing else, Microsoft has been scaling back the saber-rattling a bit (which, they've found, hasn't done much except make more enemies) and has started looking at smart ways to make open source strategically viable on their platform: e.g., IronPHP, which lets PHP apps run as native .NET binaries on Windows. The monolith has not fallen, but it has turned slightly to face a new direction -- no small feat in itself.

I know that on my own, I still plan on using a mix of things. I have an open source CMS and database for one site I run, and a proprietary stack on another. My desktop has Firefox and Outlook side-by-side; long may both of them wave. If by some miracle a future version of IE outdoes Firefox, or some iteration of Thunderbird gives me all the contact-management and calendaring features I've come to take for granted in Outlook, I won't hesitate to switch. But the proof has to be in the using, not just in the abstracts -- and I doubt anyone who uses software on a daily basis is going to seriously argue that point.

So what's next?

What comes next is always difficult to predict. Sometimes the future just ambushes you. To paraphrase film critic Roger Ebert in his discussion of the film Blade Runner, people of the Fifties accepted that the future (of, say, the 1980s or the 2000s) might have flying cars and world governments and food in pills -- but they scoffed at the idea that we'd still drink Coke and that rock 'n roll would still be on the radio. But we can still see a few things coming.

For one, there's no question that open source will fill most every container it can be poured into. There's room for it in just about every computing market, even if that position is not the dominant one. It may not have to be, either; all it has to do is exert a pull of its own. Android phones may never outsell the iPhone (and Android itself may earn jeers from open source purists, to boot), but the mere fact of its presence has alerted people to the possibility that a classy, powerful phone doesn't have to have a certain fruit logo on it.

The other -- and I can't emphasize this strongly enough -- is how, in time, the fact that something is (or isn't) open source may simply not matter all that much. Not in the sense that it would be a pointless distinction, but in the sense that it will simply be one choice among many -- and, in some cases, not even the most important one.

If open source's collective mission is to change the way software is designed and used, a goal like this would have more genuine benefit than almost anything else I can think of. The real dividing lines would be between what programs are worth using and which ones fall short -- not which ones hew to a specific licensing model or were built under the auspices of a given social contract. In many realms, this has already happened: most people don't care if a given CMS, for instance, is open source or not; they care if it fits their needs as a business. The next step will be to have the software makers themselves reach this stage -- where they can consider open vs. non-open without it being a political issue.

I don't know if that day will ever come, but it sounds like a pretty good goal to me.

I'll close with a few words from a fellow named Barrows Dunham, an American philosopher who had unabashed faith in humanity's ability to solve its problems and live both harmoniously and prosperously, without appeals to greed:

Here, then, is to moving Leviathan a little. And gaining a great deal from it.

Our "A New IT Manifesto" report looks at a variety of new approaches and technologies that let IT rebels take on a whole new role, enhancing their companies' competitiveness and engaging their entire organizations more intimately with customers. Download the report here (registration required).

Twitter: Me | InformationWeek
Facebook: InformationWeek