[License-discuss] Government licenses

[Christopher Sean Morrison]

> The work-around they are trying to use, of having contracts for the distribution of creative works circumvent limitations and exceptions to copyright, should be clearly understood as more harmful to the FLOSS community than any amount of software released by any particular government could be of benefit.  [and from later]
> License agreements which seek to circumvent existing limitations/exceptions and/or expand the scope of copyright, should be rejected (including any existing approval repealed).

Except it’s not really a work-around, it’s the widespread standard practice that has persisted for longer than OSS has existed.  Contracts / Agreements are the manner in which all Gov't creative works are shared that are in active use, at least by the USG.  Think of a UAV (with software, without weapons) distributed to an ally.  There’s a contractual agreement in place.  This is nothing new and unlikely to change.  The practice predates software.

> The existence of a problem in government procurement policy that allows the government acquisition of proprietary derivatives of GOSS software should never be used as a justification to create new problems by promoting ways to circumvent limitations and exceptions to software copyright.  It should be a reason for those government employees to work within their domestic government to fix the procurement policies, and to help ensure that such circumvention of limitations/exceptions to copyright is prohibited by law.

In my circles, avoiding buying back their own Open Source is not a principle concern any more — it definitely was a consideration in the late 90’s.  Maybe it’s still an issue to some, but there has been so many successful Open Sourcings in the time since, under myriad methods, that those fears have mostly disappeared.  Today’s concerns are dominated by untested legal postulates (uncertainty in licensing), sheer process complexity (lack of formal guidance), and the speed of acquisition — Open Source is viewed as a way of developing software faster.

There are groups working hard on both from a regulatory and policy perspective (code.gov, code.mil, data.gov, codeforamerica.org, etc), but none of that is likely to change Title 17 or the fact that all except three (maybe four?) OSI-approved licenses are contingent on copyright.  The Gov’t isn’t going to suddenly release everything into the public domain in order to participate — they could have already.  They’ll simply continue with signed agreements or choose not release.

> My strongest hope is that licenses which might have the effect of circumventing limitations/exceptions to copyright, or expanding copyright (to interfaces, etc) should be rejected.   The process needs to be clear, the outcomes documented, but they should still be rejected.

I don’t know of a single agency that is trying to circumvent exceptions to copyright.  That’s why Gov’t lawyers say “it’s public domain code” … even though that’s only to US citizens and only if the agency decides to release it.

As for licensing, what the code.gov guys are testing is particularly interesting.  They’re suggesting agencies go ahead and use a copyright license, but with a disclaimer that it might not apply to some recipients (because some/none/all code may be public domain to them).  That will almost certainly create enough uncertainty with recipients that the end result will be to treat the work as OSS and, through time, will result in eventual substantial derivation that is OSS without caveat.

It’s a gamble as the prevailing legal opinion prior was that would invalidate the license, which would have a massive chilling effect on GOSS.  The fact that it comes from the White House (and now across two administrations) gives it some credibility, albeit still untested.

Cheers!
Sean