NASA advised to study up on what open source, free software, and permissive licenses actually mean

Turns out making code public with the right fine-print is harder than rocket science

Houston, we've had a problem: our rocket scientists don't entirely understand the nuances of software licensing.

NASA, of course, is more than just rocket scientists. It's home to software engineers and other technical types, as well as those inclined to maintenance, management, and administration, and other less storied roles.

But among those at the US space agency who deal with software – writing it, requisitioning it, glaring at it – there's less understanding of open-source software requirements than there should be.

Or so say John Haiducek, Thom Edwards, Wade Duvall, Sarah Cannon, Kai Germaschewski, and Jason Kooi – a medley of boffins from the US Naval Research Laboratory, Technical University of Denmark, University of New Hampshire, and others.

Haiducek et al. recently completed a short paper titled, "Recommendations to clarify NASA open source requirements," that was released via ArXiv. Therein the researchers observe that while NASA has a policy designed to encourage open source software development, its personnel continue to be confused about the specific meaning of terms like “open source software,” “free software,” and “permissive license.”

"Some NASA documents and policies have acknowledged the OSI and FSF definitions as widely accepted, but NASA does not always use and apply these definitions consistently," the paper explains.

"Moreover, many scientists mistakenly understand the term 'open source' to mean simply that source code is available to the public. As a result, some software products developed by scientists are advertised as 'open source' even though their licenses violate one or more of the ten criteria of the OSI definition."

Such misunderstandings in the past have kept some NASA software from being included in Linux distributions. And, the authors argue, they can trip up NASA solicitations. Proposal teams may interpret OSS requirements differently than NASA, thereby limiting the scope of their work, or expanding it beyond what NASA can accept.

"Establishing common ground as to the meaning of terms related to OSS, and increasing clarity of communications around software licensing, would benefit NASA and NASA-funded scientists," the authors argue.

Bruce isn't impressed

Bruce Perens, creator of the Open Source Definition, board partner at OSS Capital LLC Venture Capital, and CEO of an undisclosed startup, told The Register in a phone interview that he's familiar with NASA scientists through his work at the Open Research Institute, which aims to foster collaboration around technologies otherwise under national export controls.

He said he's really impressed by the degree to which NASA boffins have embraced open source software, but added there's a gap in the way developers are trained.

"As involvement in open source software expands, we reach a problem, which is you can take a four-year course in computer and never have a class in intellectual property," Perens explained.

"This isn't just a NASA problem. It's a problem across the entire software industry. Not only do programmers not really recognize what open source is or what the rules are, I'd say most have never read the license."

Perens said the recommendations proposed in the paper sound reasonable. "It's essentially saying get your shit together about intellectual property," he said.

The Register asked NASA for comment. A NASA Jet Propulsion Laboratory (JPL) spokesperson cheerfully said he'd try to find someone to respond. But as this story was filed on a Friday, when the East Coast press office is closed and many of those at JPL in California had a rostered day off, we're not expecting an immediate reply. ®