I started programming in 1984, at the age of 6. Back in those days, everything you needed to be a “programmer” (the title of Software Engineer or Developer didn’t exist yet) was contained in a single computer—in my case, an Apple II+ with 16 kilobytes of RAM.
I was able to learn to write software, primarily because the gap between writing software, and running software, was so small. Often, I didn’t even bother to save my small, Apple BASIC programs to a 5-¼” floppy disk at all—I simply left them in RAM, where they naturally sat when I typed them in.
Fast forward 30 years, and I now have two daughters, both of whom are trying to “learn to code”. And while the intervening decades have made computers vastly more capable, they have also made them more complex. What was once possible with a single machine (the one sitting in your living room or kitchen, no less), now requires the use of “the cloud”, and an arcane set of tasks with a poorly defined mental model called…deployment.
This was the central issue that I was recruited into NASA to address in 2008—how to improve both security and efficiency by unifying NASA’s application development into a common platform. Originally called NASA.net, this project quickly ran into a then-common roadblock—the lack of agile and programmatic infrastructure to support this platform.
The past 5 years have been a detour to address this lack of agile infrastructure—a detour that could now be easily termed the “OpenStack” years. (Although when the five of us sat down at HouseKu for a long weekend to write Nova, the precursor to OpenStack, we never expected it to take more than a few weeks.)
I’m delighted that OpenStack has achieved visible world dominance—there isn’t a single major cloud software vendor that hasn’t elected to use OpenStack as the core of their product offering, and it’s rapidly achieving critical mass in the public cloud space as well. Piston, the company that I founded to commercialize a hardened OpenStack software product, now boasts a who’s-who of Fortune 100 customers and scaled-out production deployments on a global scale.
Of course, the rest of the IT industry hasn’t been sitting around while we were busy with OpenStack. Cloud Foundry has attacked the very heart of the “deployment” challenge, using the same weapons that we used to fight hardware abstraction—open source licensing, a broad community that embraces both users and commercial vendors, and a rapid and iterative approach to product development.
With OpenStack and Piston both on firm course, I’m taking this opportunity to climb “up the stack” as it were, and join Pivotal as Field CTO for Cloud Foundry. This position will give me an opportunity to work directly with the Fortune 100, on the front lines of 3rd platform adoption. I will also be working with the Cloud Foundry product team, translating these real-world challenges into user stories in the backlog.
By joining Pivotal, I have the lucky privilege of jumping to the “front of the line” in platform innovation, and taking advantage of five years of parallel effort by James Watters, the broader Pivotal engineering team, and a global open source community. I am delighted to return to “Mission One”, serving the needs of both developers and operations staff by embodying in systems software, the accumulated wisdom of this last decade’s “Internet Unicorns”. With any luck, Pivotal CF will be in every high school in North America—before my children arrive.
About the Author