The Plan for Code

Might have to do with national security. NASA is a US government agency, where the US government probably would not like all their code in the hands of countries using it in ways the US government will not like.

> Can anyone explain why software developed by a government agency would not be in the public domain, once open?

While US govt authored stuff isn't subject to copyright, I don't know how US govt stuff derived from copyrighted material is treated. (The copyright for derived work ordinarily stays with the original author, but the US govt is special.)

BTW - Which not-US countries have the "no copyright on things that we do" rule/policy/law?

Although under US copyright law all federal works are in the public domain, that does not prevent the government from having valid foreign copyrights in its works that are enforceable abroad.

Some more information is available at http://www.cendi.gov/publications/04-8copyright.html#317.

Most code is not written by government employees, but contractors. The contractors usually sell the government on reusing internally-developed code to lower the cost, but they then maintain rights to that code. My guess is that the license is an attempt to cover such things.

Also, government appropriations can be weird- money for 'changes' might not be able to be used for 'new' software, etc. Some of the language in there seems to be related to that as well.

It used to be that, if you wanted to distribute NASA-sponsored software using an open-source-like model, it had to be done through Open Channel foundation. Here's a press release which gives a flavor of how that looked:

http://www.openchannelsoftware.com/NASA_1.html

Sounds pretty bureaucratic to me.

Open Channel was an antique system (pre-WWW) in which software was "thrown over the fence" to a caretaker, with no provision for general users to contribute back. So, faced with a long process of approval and little possibility of tangible benefit, it was unusual for software to be released that way.

I'm not certain about the reason for a long approval process. Since about 1999, there is ITAR, which is a huge impediment, but open-sourcing was hard even before ITAR.

My guess is that NASA wished to (or was required to, by Congress) recover some costs by licensing the software, and this was the reason for the impediments to open-sourcing. In this way, the government was treating software more like hardware components than like documentation of research results.

NASA as an agency has always had an uneasy relationship with software. It's much more comfortable with hardware technology, and tends to treat software technology by analogy with hardware.