Linux: The Final Frontier?

Does a mature Linux mean there's nothing left to explore?

I have a degree in Physics. My particular affinity was in the field of optics, which is about as applied physics as you can get.

For the past few decades, there has been a serious challenge facing a lot of applied physicists--real ones, not amateurs like me: there's not a lot of new frontiers to explore. Sure, there are important nuances of the macroscopic universe still left to figure out, but these are the niggling details. For practical purposes, it's fair to say that we have this whole Newtonian universe of ours pretty much squared away.

[ And the best Linux desktop distro of all is... ]

The real excitement is perceived to be in the realm of quantum physics: figuring out the mechanics of the basic building blocks of the universe that behave in decidedly weird ways. Where's the particle? Here. No, wait, it's here. Or maybe its in both places. Ooooo...

If you hear a little cynicism in me, it's the conflicted nature of an applied physicist who loves his science, but secretly envies all the cool stuff those quantum folks are doing. Still, at least the applied team will always win in any interdepartmental sports competition: we know exactly where the ball is and how fast it's going. Period.

I bring all of this up because of two conversations I have had lately with members of the open source community about the maturity of Linux.

The first was a few weeks ago, part of an interview I was doing with Paula Hunter, the new executive director of the CodePlex Foundation, the non-profit organization sponsored by Microsoft that's intended to encourage and educate commercial software developers to start or improve their own open source projects.

Towards the end of the interview, I asked Hunter what attracted her to the CodePlex gig.

"The Linux marketplace has matured. The heavy lifting has been done," she said. "Working with the CodePlex Foundation is addressing the next generation of challenges."

For Hunter, the next big thing is to convince software developers on other platforms about the benefits of open source. Cynics will argue that this is just another Microsoft attempt to control that which it can't understand, but that's a valid argument I want to hold off for another time. Today I am more interested in Hunter's assertion about the maturity of the Linux marketplace.

In some respects, she makes a good point. In the enterprise and the datacenter, Linux is more than a match for any other platform, and triumphs easily. Resistance to Linux is usually based not in fact, but in FUD. The same situation is also true in the embedded space, though I think for now Apple's iPhone OS is a much stronger challenger to Linux-based devices compared to the strength of Windows Server and various UNIXes versus Linux in server-land.

On the desktop? Well, there the label of maturity also applies, but there's still some frontier left to explore. Yes, the Linux desktop is good, but it could be better.

Note that I am not trying to say there's no more room for innovation. There is, obviously, in all of these arenas. But there is strong evidence that the forward progress--the acceleration--of innovation is slowing down. In the "old days," developers had to work hard just to get an X desktop stable. Now the challenges are incrementally smaller: how does a GUI work with this new driver? How can its performance be enhanced?

Linux hasn't overcome every challenge, but it's succeeded over a majority of problems, and like any other mature product has begun to be more selective about the paths of innovation. One good example of this is the excitement generated by cloud computing and virtualization. Vendors see it as a way to make new money, and developers see it as the next new territory to explore and dominate. Cloud computing, until something new comes along, is the quantum physics of the Linux and open source community.

The second time this notion of new frontiers came up was in a conversation I was having with community members this week. I was spinning off the identity argument I made Monday, asserting that one of the barriers to creating an identity for a Linux distribution is that with the distros so close to each other in terms of technology, it will be tricky for different distros to distinctly identify themselves. It won't be impossible, but if Distro X wants to be "The Desktop Distro," then it will have a lot of competition for the same title from all the other distros in the alphabet. Better to stake out a unique identity claim that sets it well apart.

This happens all of the time in society. And can tell you (with sweeping generalization) that most Hoosiers are very different in temperament from Californians. And each group would be darn proud of that difference. But many people in both states will also identify ourselves as US citizens, which puts us in a larger group with broader shared views.

The frontiers for Linux may be diminishing, but there is still room to explore our own sense of identity and purpose within that community, both as individuals and as projects.

Ultimately, though, I wonder if the frontiers are indeed fading away. I concede that sometimes it seems that way, but the advances in cloud computing and embedded space also show us that there are still new worlds being opened up.

New worlds that open source software will be a perfect vehicle with which to explore.

ITWorld DealPost: The best in tech deals and discounts.
Shop Tech Products at Amazon