RFC-1 was published 50 years ago. Infrastructure management emerged as a discrete field two decades later, and devops a decade after that. As we work to realize the full potential of networked computing and the decentralized organizations it enables, what will the next 50 years bring, and what can we do to hurry things along?
In this talk, we’ll touch on some of the roots of current infrastructure management tools and the motivation and computing science theory behind them. We’ll attempt to reconcile some long-lasting disconnects by referring to current thinking in distributed consensus systems. Finally, we’ll (carefully) extrapolate forward to see what the community likely will be working on between now and mid-century, and we’ll announce some new initiatives in that direction.
Steve Traugott’s 1998 USENIX paper “Bootstrapping an Infrastructure” helped kick-start our community.
He’s been coding since the 1970’s, on platforms ranging from embedded systems to supercomputers.
He helped port the Mach kernel to IBM mainframes, and UNIX System V to PC’s.
He is former Vice President of trading floor infrastructure engineering for Chase Manhattan Bank, former NASA, and a U.S. Air Force veteran.
Today, Steve and his family operate an innovation space in Livermore, California.
Their community of companies conducts engineering and manufacturing operations in support of aerospace, physics, energy, and global security efforts.