US Government Now Has An Official Open Source Software Policy
from the about-time dept
Earlier this year, we noted that the federal government was looking to further embrace open source software in its process of contracting out for (or creating in house) code. It released a draft policy which was good, though we hoped the final product would be much stronger (for example, it pushed for a portion of any code to be released under an open source license, but didn’t consider that to be the default. I was also concerned about it allowing software developed by federal government employees to be locked up by a license — something that I’m pretty sure is not allowed, since works created by federal government employees are automatically in the public domain.
On Monday, the White House’s Chief Information Officer, Tony Scott, revealed the finalized official “Federal Source Code” policy, and you can read the whole thing. Because the original was posted to GitHub, you can also easily see what’s changed. On top of that, as part of this, the government also launched a new site at code.gov, which will act as a repository for open source code from the government.
Much of the focus of the policy, understandably, is on enabling reuse of code within the government, so that different agencies and departments aren’t reinventing the wheel (and paying hundreds of millions of dollars) for projects that others are already working on. Lots of people and agencies weighed in on the draft proposal, including some interesting/surprising ones. Homeland Security, of all organizations, worried that simply pushing government agencies to release 20% of their software as open source, without understanding how that might be most useful to the wider community, would be a waste. It preferred pushing government agencies to refactor code into reusable modules, with a focus on what would be the most reusable. Others, like the Consumer Financial Protection Bureau favored (as I suggested) a default open source policy, rather than the 20% solution.
Unfortunately, the plan sticks with this “pilot program” of only having to open source 20% of code, and how well that works will be evaluated over time. It appears to have “fixed” the problem of lumping in-house developed code into the policy (since that code is public domain) by now focusing the policy solely on custom developed code by third parties (at least that’s my read on the new policy). While it’s still disappointing that the policy didn’t move to a “default to open source absent a compelling interest” standard, at least it didn’t go in the other direction either. And that’s in the face of complaints from the likes of the Software Alliance (a major Microsoft lobbying group) that whined about the need for such a policy in the first place.
In the end, this looks like a good step forward. It could have gone much farther, but it’s still a step in the right direction. Hopefully the pilot program will lead to even bigger steps towards embracing more open source (and public domain!) software.