Now that there are real dollars beginning to finally accompany the many voices calling for government IT modernization, it’s perfect timing for Marianne Bellotti’s new book, “Kill It with Fire,” which cautions against the headlong approach the title connotes. She conveys this quite succinctly in the book’s pithy epigraph – with a quote from Ellen Ullman, a fellow author and computer programmer: “We build our computer systems the way we build our cities: over time, without a plan, on top of ruins.”
Bellotti is a software engineer and legacy-systems expert by trade, currently with Rebellion Defense. She previously served on a technical SWAT team in the U.S. Digital Service and built data infrastructure for the United Nations.
Bellotti joined us at our virtual studio at MeriTV last week to explain why she wrote the book and discuss its reception in the greater government IT community.
“I wrote it after spending a lot of time working on these very old government systems and trying to restore them to operational excellence,” Bellotti said. She had been working with programmers who had come straight out of Silicon Valley into the U.S. Digital Services program, an elite technology unit housed within the Executive Office of the President. It provides consultation services to Federal agencies on information technology.
These programmers rather disdained legacy IT modernization and were much more comfortable and excited about working with new technology. However, Bellotti insisted that “the skills that we had to learn in order to deal with the old government systems were very useful and broadly applicable to all forms of software engineering, regardless of the age of the system.”
The book resulted from taking those lessons learned and assembling them into a package “that was accessible and valuable to all sorts of software engineers and all sorts of technical people regardless of how close to the actual programming they are,” she said in the MeriTV interview.
Bellotti said the title, “Kill It with Fire,” is not about the spider-killing video game, but instead refers to the typical first reaction to modernize a legacy system falling into obsolescence. She maintains that this approach – like the epigraphs warning about building over ruins – is a “knee-jerk approach that often burns through tons of money and time only to result in a less efficient solution.”
The title also conveys something else: Bellotti’s way of poking fun at the haughty attitude toward her work. “I’m very much a fan of old technology. I’m very fascinated by old technology,” Bellotti said, though that view may not match most people’s reaction to old technology in general. “So I was sort of wanting to give a little bit of a wink and a nod to these people’s attitudes about technology,” she said, which tend toward just burning older IT to the ground and starting over.
Bellotti’s strategy recalled my own experience as a state CIO, where so little attention was given to re-engineering and modernizing business procedures and processes that were in fact older than the legacy IT system itself. Our term for this approach was “painting over rust.”
Bellotti instead calls for a more rational approach, and expects to find some resistance to it. “I think it’s going to start some fights, which I’m not against doing, but I think the key thing is that when projects get to be so big and so ambitious, you don’t have the opportunity to verify whether or not your earlier assumptions are actually correct. You don’t have a lot of space to maneuver.”
This is why Bellotti maintains that legacy modernization in government is so difficult – because the underlying business and organizational forces are completely separate from the fact that the technology is old.
Bellotti covers much more during our interview, including her intriguing call for a “more forgiving modernization framework” to avoid asking programs to be perfect. “This is counterproductive and when you allow for a certain amount of failure and budget for a certain amount of failure, you get much better results,” she said.
She finished our interview with perhaps the two most relevant points. First, I asked if political and technology leaders in government are likely to adopt her new approach for modernization, or will they be unwilling to take on existing program bureaucracies. And finally, what has been the reception to her book among her many colleagues and friends in the IT government community?
If you’ve been around government IT for a while, I think you will enjoy her answers.