2007-03-11

Who writes Linux? Been obvious to me all along ...

Who writes Linux?

That's a broad question that is extremely difficult to quantify or even qualify. But a Linux Weekly News article by corbet tries to quantify the contributions -- from many different metrics and angles -- who works actively on the kernel.

Ironically enough, even though IBM maintains almost entire platform ports and architectures on Linux, along with Freescale, Intel and select others -- and that accounts for anywhere from 2-10% in various metrics -- it's not the #1 maintainer in any major metric. It's easy to get a lot of lines of code, patchsets, etc... when you maintain the hardware for your platform. Even HP gets far less credit for this, or it comes bundled with Intel contributions.

In fact, the overwhelming contributor in virtually all metrics -- excluding unknown (which is significant in several metrics) -- is the same company I've been trying to educate people about for years. They have no other interest but Linux and Linux alone. They have no other products. They have no proprietary solutions at all. They sell a 100% GPL-centric product line, giving away everything to all its competitors, hording not product line (let alone its code) from any other competitor (let alone the community).

I have said this over the last 10 years, and a majority of Linux advocates (especially newer) have just looked at me like I'm dumb -- along with the other, brave few that joined me -- because such an overwhelming majority couldn't be wrong in their eyes. The self-fulling prophecy. Sigh, even now people are saying this is a flawed and inaccurate set of metrics because of the results.

The only bias in this article is that it is biased towards vendors who produce their own hardware, and need to donate that knowledge to the kernel in the form of code so it is well supported. Take the 390 and other ports out and you'd see IBM's share drop tremendously. Same deal why you see vendors like Freescale and Intel, although one could argue in these latter two cases that most of their contributions are affect products others than their own.

The list is dominated by the various, anonymous contributors, select organizations and institutions and the -- at the top -- the one, sole commercial entity that does not sell one single, proprietary product. Makes you think about who is really and truly about 100% Linux, and those entities who have other interests than just Linux. That's not to demonize these other companies at all, and I'm happy to see their involvement. In fact, it's a testament to their wish to involve the community with the support of their products, because they realize the value of the community.

But it is something to keep in mind the next time you want to rant about Red Hat. Especially since Red Hat doesn't have a piece of hardware they only sell or design, or a select subsystem or component that they alone are the only one that uses (even if some would argue otherwise).

People have complained for years that Red Hat's "influence" has changed the path of Linux -- from the readoption of GLibC 2 in Red Hat Linux 5 to the push to GCC 3 in Red Hat Linux 7 / Enterprise Linux 2 to the implementation of Native POSIX Threaded Libraries (NPTL) in Red Hat Linux 9 / Enterprise Linux 3 to SELinux Mandatory Access Controls / Role Based Access Controls (MAC/RBAC) in Fedora Core 3 / Enterprise Linux 4. Now you know how Red Hat is able to "influence" Linux, especially the kernel.

Just like the sign at the animal shelter that reads ...
"If you don't like what you see here ... Volunteer!"

True community involvement, by individual choice, in a free-market, capitalist society -- the ideal form of "public good," instead of forcing everyone which quick becomes differing views of "what is right." But what do I know? I mean, the majority of Linux advocates know far more than me, they've been telling me, among other Linux developers and consultants, that we're just wrong. After all, Red Hat is the Microsoft of Linux -- although now more and more think Novell is just as evil too, possibly more so.

Even I, a man who has built 90% of his career over the last 10 years on Linux solutions with Red Hat, don't have anything negative to say about Ubuntu, KDE, etc... (not at all), unlike the seemingly majority of "advocates" who blast each other back and forth with commercial-like marketing sewage. In fact, leadership in the Linux community (note, I didn't say "leader" or even "leaders") isn't about saying anything negative about your competitors, or even preventing them from integrating with your solutions. That's something we -- the Linux community as a whole (putting many advocates aside) -- left the commercial software world for, so why would we drag it here?

I thank God everyday that Linux is a meritocracy and not a Democracy. Funny, the Fedora Project tends to have the same organization too, and not everyone gets a vote. And they get things done, not always as everyone wants them, but in the way that moves Linux forward. Because while you can't make everyone happy, many of us who have been implementing Linux as an enterprise solution (not just for web services) for the past decade are getting what our clients say they need.

If you disagree, remember, Linux can use volunteers just as much as the animal shelter. That's how things get done, and how you get your "influence."

Related Blog Entry:
The Five Types of 'Linux' Corporations

1 comments:

wow power leveling said...

Americans everywhere humor A detention wow gold notice was written like this: a wow power leveling police car with stones, to win wow gold the detention center for seven wow power leveling days all-inclusive accommodation replica rolex Tour Value; hit send 2 a beautiful bracelet, wow power level fashionsuit, police transport; more more surprises , the former can enjoy free shaved 10; before the 100 can play with power leveling the dogs, the guests were presented massage sticks, electric shocks to CHEAPEST power leveling the dead skin beauty care services.