[License-discuss] Government licenses

Christopher Sean Morrison brlcad at mac.com
Mon Jun 3 20:03:09 UTC 2019

>> "The prohibition on copyright protection for United States Government works is not intended to have any effect on protection of these works abroad. Works of the governments of most other countries are copyrighted. There are no valid policy reasons for denying such protection to United States Government works in foreign countries, or for precluding the Government from making licenses for the use of its works abroad."
>> https://en.wikisource.org/wiki/Copyright_Law_Revision_(House_Report_No._94-1476)

> (Note that elsewhere in this thread it was already suggested that
> government projects could add a permissive license to software that
> may or may not be copyrighted at all. That is an easy practical
> solution to the question of releasing code. The question whether the
> code is copyrighted in the first place otoh is a significant, high
> stakes policy question.)

As noted in the preceding link, prevailing view and treatment is that there is full copyright protection in some jurisdictions.

> 1. US Copyright Act § 105 says:  "Copyright protection under this
> title is not available for any work of the United States Government,
> but the United States Government is not precluded from receiving and
> holding copyrights transferred to it by assignment, bequest, or
> otherwise.”

The important part relevant to this discussion being “under this title”.  It’s not exclusive of rights inferred or reserved outside Copyright Act.  HR 94-1476 further clarified a legislative position that the exclusion does *not* apply internationally (which ropes in Berne Convention reciprocation).  It directly speaks to § 105.

> 2. To a layman open source developer the above seems to say that code
> created by US government can never be copyrighted. BUT, government
> lawyers feel that is unnecessarily generous, so they are pushing a
> narrower interpretation, where such software is only public domain
> inside the US.

Layman is in error because of the latter clause regardless.  Particularly for codes with multiple hands, copyright via assignment, unlimited rights conveyance, and sbir conversion happens regularly yet would still be perceived as a code “created by US government” to a layman.

> 3. An easy mistake to make here is to think that since government
> lawyers say this, it is the correct interpretation of the law. This is
> of course not so. The government lawyers have merely presented a
> theory that is beneficial to their employer, much like an Oracle or
> Microsoft lawyer would.

HR 94-1476 clarified this interpretation, not ad hoc lawyers.  The two last paragraphs are relevant with the 4th specifically addressing international copyright and 5th clarifying public domain intent.  

> And it is especially worth noting that actually testing this theory in
> a court decision is probably not what the US government wants to do.
> It is not at all obvious their theory will win in a foreign court. A
> much wiser strategy is to just keep asserting this theory, until
> people agree and act accordingly.*

I’ve heard from several Gov’t lawyers that this has been successful in court repeatedly.

> 4. On the other side the open source community continues to act as if
> government software is truly public domain. It can be used in other
> open source projects. Sometimes unreleased government code has been
> liberated with tricks such as using the Freedom of Information Act.

Many agencies do so for public relations, not because it’s required.  Many agencies hold a view that software is not a record (thus immune to FOIA).

NASA, takes position that software is generally not a record — NPR 2210.1C (Aug. 11, 2010).
Navy considers most software a record - SECNAV 5720.42F.
Army says most software is not a record — AR 25-55.

Courts have come down on both sides, basically saying it depends what the software does, how an agency uses it:

See, Gilmore v. United States DOE, 4 F. Supp. 2d 912 (N.D. Cal 1998)
See, DeLorme Pub. v. NOAA, 917 F Supp. 867 (D. Me. 1996)

Relying on the cases that have gone to court, Gov't software not used in decisions (i.e., most software) are typically not subject to FOIA.  Can’t hurt for citizens to file as there are Open Government initiatives and agency incentives, but there is little legal ground to stand on.

> As proponents of the government side often point out, the value of all
> software at stake is 60 billion per year! If the government's theory
> prevails, then this software is no longer open source!

It’s not Open Source either way, non sequitur.  Today, most Gov't software fails OSD #2 because source code and software are simply not available.  Information must be approved for public release.  This typically is NOT compelled by FOIA but, rather, 99.99% of the time by developers internal to the Gov’t.

Even if the U.S. supreme court decided one day that the U.S. Gov’t has no rights whatsoever, public domain everywhere carte blanche, it would not change the state of source code coming from the Gov’t.  It’d actually probably make public release harder without any rights to protect out of liability and tort fears.

> 5. Which side is right or wrong is rather irrelevant. What matters is
> whether we can in practice continue to use such code in open source
> software.


> Currently we can and do.

We predominantly do not.

Most software is not even available.  How to support its availability (as Open Source) should be of paramount importance to the OSI.

> Note that whether some NASA projects could be released as open source
> is a drop in the ocean compared to what's actually at stake (#4). It
> would be a horribly irresponsible trade for the OSI to make.

NASA probably has the largest repository and it doesn’t even represent 1% of Gov’t source code.  Follow the money — U.S. DoD dominates the software acquisition space.

I don’t see a trade of any sort taking place.  Developers at NASA, DoD, 18-F, and other government agencies are trying their hardest to get sources released as Open Source under constructs that fully comply with the OSD as it’s understood, under any OSD-conforming mechanism available.  If they could even release as public domain, it would have happened, but that historically only happens when a code is no longer of any interest (e.g., GRASS GIS by the Army CoE).

There are myriad complexities and Gov’t players encounter not just a lack of support, but antagonistic and ill-informed opinions pervasive.  As it stands GOSS is continuing to grow, despite a general lack of support and understanding, but I do believe we and the OSI can do better, can do more, and it will only help Open Source.


More information about the License-discuss mailing list