Don’t give me this “We Used to Do Big Things” Crap. We do.


ImageI was having a discussion yesterday with someone who decried our country’s lack a manned space program. I agree, the fact that the US doesn’t have a manned spaceflight program is a step back. But then he said, “our generation hasn’t done anything like land someone on the Moon… we used to do big things.”   I looked at him funny because…

…he’s just plain wrong.  The Linux kernel trumps the moonshot both in terms of engineering effort and societal impact by a few orders of magnitude.  The kernel is the largest, most complex collaborative effort in the history of the species. That may sound somewhat grandiose, but it’s very much true. The Linux kernel is over 17 million lines of code and is growing at an average rate of 3,500 lines per day. Nearly 1,300 developers contribute to Linux with versions like 2.6.25 generating more than 12,000 patches. The Linux kernel powers over 93% of the TOP500 Supercomputers. The kernel is at the heart of Android which has a nearly 60% share of the mobile operating system market with 1.5 million device activations a day. The kernel also powers millions of servers across companies that have transformed the way we consume information and communicate with one another such as Yahoo, Google, Facebook, and Twitter.

This is a silly game show to play, but if I had to choose between a Saturn V with three dudes and a lunar landing module strapped to the top of it and a DVD with the Linux kernel, I’d point to the DVD as having a bigger impact. I’d also argue that the benefits of today’s innovations are more equitably distributed and are not driven by global nuclear conflict.  That’s a good thing. Instead of generating dusty artifacts for museums in DC we’re busy creating software that people can use. I’ve singled out the kernel because I’m convinced it’s at the center of several transformative shifts.  

It is also the operating system that will send us back to the Moon (on a private, SpaceX rocket).  Here’s a quote from Robert Rose at this year’s Embedded Linux Conference:

Linux is used for everything at SpaceX. The Falcon, Dragon, and Grasshopper vehicles use it for flight control, the ground stations run Linux, as do the developers’ desktops. SpaceX is “Linux, Linux, Linux”, he said.

We still do big things.  So keep your sappy 60s sentimentalism to yourself and submit a patch.

12 thoughts on “Don’t give me this “We Used to Do Big Things” Crap. We do.

  1. Yes there is a large tribe of people who are constantly harking back to a “golden age” whether that is the renaissance, the 1800s, the world wars, the 1960s etc. If anything we’re in the golden age *now*. Never has science, technology and overall human development progressed faster.

    As you explain very cogently, the Linux Kernel is better than a moonshot on so many parameters. Another thing, its a shared resource — not owned by a government or a company. Its owned by humanity — thanks GPL.

  2. It’s not about work invested (certainly not in LoC) it’s about ambition. The Linux kernel can capture 100% of all markets and grow to 42 million LoC and it wouldn’t change the fact that it’s just an operating system.

    The point stands that while we’ve visited the moon of this planet, we’ve yet to physically venture further in over four decades.

  3. I mostly agree, but the moonshot didn’t only benefit a few people. The technology developed from it has a huge impact, and is still important. For example, that DVD with the Linux kernel on it uses material science (designer plastics) developed during the space race.

  4. “This is a silly game show to play, but if I had to choose between a Saturn V with three dudes and a lunar landing module strapped to the top of it and a DVD with the Linux kernel, I’d point to the DVD as having a bigger impact.”

    One of the reasons for doing so is also the simple fact that Saturn V was almost a makeshift contraption. It’s not something you’d want to build in large numbers on modern machines, in fact, even if we had complete blueprints for the whole thing including the full manufacturing steps for the individual components (which is, what I believe, we’re missing), you wouln’t be able to convince a CNC factory that this is a good thing to build. Desinged for manufacture, stressing the *manu* part. All by hand.

  5. Clearly there’s a “What’s come first? The chicken or the egg?” problem here. One could only say these things as a thought experiment in reaction to an off-hand remark. I think there was ultimately a huge benefit of the moonshot, one of which was the acceleration of solid-state transistors that ultimately enabled the development of the internet.

  6. I’d have though they’d want to go with a real time operating system, rather than Linux. Or can you modify Linux in a way which gives it real time operating system properties?

  7. The collective WE still do huge things but I think the context people usually levy this criticism in is that the United States government doesn’t do huge things anymore, which is a different story than the kernel.

  8. Wrong wrong wrong, wrong and wrong.
    This whole thing is wrong.

    The federal government cannot hold intellectual property. Every single bit of technology developed for Apollo is available to the public, patent free, and the result has benefited society greatly.

    Take for example the very first self contained computer built using ICs. Apollo set the stage for the entire information age. Imagine if IBM had been the first to develop such a thing, and secured a patent on it. It would have set computer development back for entire decades.

    We are *still* studying what we found in just a few short lunar missions. The benefit to science is gigantic. The lunar laser ranging experiment alone has helped physicists to prove parts of general relativity. Were talking fundamentally increasing humanity’s understanding of the universe here – but that’s not as important as a kernel?

    Linux just isn’t that great. Sure its a great thing, but theres nothing dramatically different or special about the Linux kernel when compared to other, similar kernels. There’s some great things about it, but it isn’t some unique thing – its a unix-like kernel with some good user spaces available and a nice phone OS built around it that a lot of people use. That’s it.

    1,300 developers contribute to linux? over 400,000 people worked on Apollo. They invented new materials. Global tracking systems. A mission control organization who’s structure and procedures are regarded as *the* example for all real-time engineering and troubleshooting groups to ever come after.

    Hell, the Apollo guidance computer could restart itself over and over without ever loosing state – half the time Linux cant even close a laptop and wake it up from sleep without loosing its damn mind.

    Your seriously going to say that a good, but ultimately replaceable unix-like kernel is more important to society then almost half a million people working together to build the most powerful machine in the history of the human race, successfully allowing us to travel away from our own planet for the first time?

    That’s dreaming really freakin small. And that’s what’s wrong with America.

  9. I agree about the importance and impact of the kernal… don’t forget that we probably don’t get to the kernal w/ out the manned space program, however.

Comments are closed.