Lessons from LinuxCon North America 2011


Every conference, at least the good ones, has a theme to tease out and lessons to learn. LinuxCon North America 2011 was one of the best, and having the good fortune to be in Vancouver, BC last week for LinuxCon, I learned quite a bit.

20 Years of LinuxNow, I’m not just talking about technical stuff. There were plenty of great technical sessions and case studies. For instance, I learned quite a bit about Oregon State University’s Open Source Lab (OSUOSL) during Lance Albertson’s talk Friday. But what I’m talking about are the larger lessons and themes that come out of a conference. When you get a bunch of people in an industry together, you often find some larger trends worth thinking about. LinuxCon was definitely no exception.

It’s the Enablement, Stupid

One running theme at LinuxCon was the idea that Linux is not important. Yes, you read that right. More accurately, Linux in and of itself is not that important — it’s what Linux enables that’s important.

In its first two decades, Linux went from a hobby OS to world domination. Sure, Linux doesn’t run every computer — but when you add up all the computers that Linux does run on, it’s pretty easy to argue that we’ve reached the point of world domination. If not domination, certainly Linux has achieved ubiquity. From mobile phones and countless embedded devices to running the bulk of the world’s super computers, Linux is everywhere.

Why is that? It’s not (just) because Linux is or was technically superior to the alternatives. It’s because the licensing allowed Linux to be molded and pressed into service on the smallest and largest computers we use — and everything in-between. It’s because the community around Linux has welcomed and encouraged adoption. It’s because companies were able to collaborate on Linux while competing with each other. It’s because the changes that benefit one company or developer benefit everybody.

As Marten Mickos noted during his keynote, the past decade or so for Linux has been all about disruption. Linux has been a disruptive force and computing, and enabled many more. Would MySQL have taken off without Linux underneath? Doubtful. Would PHP, Perl, Ruby, Python (and others) have been as popular without Linux? I don’t think so.

Now, as Mickos said last week, Linux is enabling innovation in cloud computing, database technologies, and so on. People are not as excited by Linux these days as the technologies that it’s enabling — and there’s nothing wrong with that at all.

Think Differently About Collaboration

It’s no surprise that Clay Shirky delivered a fantastic talk. Shirky’s keynote, “Good Collaboration in 2 Words: Structured Fighting.”

When people talk about collaboration in the FOSS community, it’s often cast as a harmonious activity. It’s not. At its best, collaboration is (to use Shirky’s terminology) “structured fighting” where opposing views and approaches are pitted against one another and the best ideas and technologies are adopted. The process may be contentious, but when done right it works.

Shirky also talked about the fact that most large collaborative communities are really small collaborative communities, where a (relative) few integrate work from a larger body of people. What’s the difference? It makes a huge difference when thinking about the structure of a community, how it should be governed, what is and isn’t healthy, and how much of the load is being carried and by whom.

The whole keynote was fantastic, of course, but the idea of “structured fighting” is something that’s particularly interesting.

Vision and Commitment Equal Success

During Dr. Irving Wladawsky-Berger’s keynote, I was struck by just how right IBM got it when it decided to back Linux. Thanks to hindsight, IBM’s decision looks perfectly ordinary and obvious now — but it really wasn’t obvious to many people when IBM jumped on the Linux train.

IBM’s bets on Linux have paid off for IBM and the rest of the community as well. Why? Because IBM understood what Linux meant early on, its leadership had a vision for Linux, and they committed to Linux and followed through to success. Lots of companies involved with Linux failed or didn’t do as well as they could have, due to lack of vision or a failure to commit over the long haul. Other companies lacked vision and tried to keep Linux down — and largely failed there too. (Or perhaps they did have a vision of Linux achieving world domination, and thought they could derail it. Fat chance.)

IBM isn’t the only company that’s profited from betting big on Linux, of course. Look at Red Hat, which is poised to become the first billion-dollar Linux company by the end of this year. A number of other companies have done well with Linux, and the best results consistently come from those companies that have a vision and stick to it.

Conversely, the SCO strategy of turning on the rest of the community when the going got rough? (Remember, the SCO that sued IBM was not the original SCO — it was the remains of Caldera that took the name of SCO after buying some of the SCO Group’s assets.) Pure unfiltered fail.

Don’t Get Cocky

Linux has succeeded wildly, but that doesn’t mean that eternal success is guaranteed. Eben Moglen talked a bit about the patent scariness during the 20 years of Linux panel with Jon “Maddog” Hall, Dan Frye, and Jim Zemlin. Moglen, eloquently as usual, talked about the next push against Linux and free software — and said that there will be a very fierce effort to push free software out of the marketplace over the next decade. I do believe he’s right, so while we’re celebrating the amazing successes Linux has had over the past 20 years, it’s not sufficient to say that we’ve succeeded. There’s much, much more to do.

Mickos also noted that we need to work on keeping the cloud as open as Linux. It’s not going to happen without effort, and if we do nothing, the trend is not towards openness, even when technologies are being built on Linux and open source. So, despite success so far, there’s no room to get cocky just yet.

Do What You Love

Linus Torvalds’ Q&A with Greg Kroah-Hartman was a great deal of fun. Though Torvalds is famously not fond of being in front of a large audience, he’s a fine public speaker and quite fun to listen to.

It wasn’t Torvalds’ answers that really sparked an “ah-ha!” moment for me, though. It was just the realization that all of the good things that have come from Linux really stem from Torvalds (and others) following their passions. Torvalds joked about finding a new career, but it seems pretty clear that he’s enjoying what he’s doing now.

Nobody Knows What’s Next

Looking back on 20 years of Linux, it’s natural to want to say “OK, now what?” What will the next 20 years hold?

The IT industry is not known for having a shortage of opinions and predictions, yet few people are willing to make any big bets on what the next 20 years will hold for Linux. While some people and organizations bet heavily on Linux a decade or more ago, few imagined the world exactly as it is now. When Red Hat’s Jim Whitehurst took the stage to talk about it, he had a hard time nailing down the next 20 years.

Whitehurst wasn’t alone. Few folks at LinuxCon were looking in the crystal ball much past the next five years. It’s easy to predict that Linux will continue to be a major player in cloud computing. It’s equally easy to joke about the next year being “the year of the Linux desktop,” but it’s hard to nail down what the future holds over the long haul.

At least we know that there will be another LinuxCon North America in 2012, and that’s going to be in San Diego. It will be really tough to top this year’s LinuxCon, but it will no doubt be an excellent event. I’ve learned my lesson — and I don’t plan to miss it.