In FSF Takes New Aim at Enterprise Software, I ranted for fun about recent adolescent actions of the Free Software Foundation (FSF) in Boston.
It was kind of a slow-summertime/no-news kind of post, but a persistent commenter on the post felt I was not illuminating you sufficiently on the difference between free and open source software (FOSS, also sometimes called Free/Libre Open Source Software or F/LOSS by those really into distinction without a difference) and the just-plain-old open source software (OSS) concept.
I didn't do that because it was irrelevant to the FSF ragging on Microsoft, and it is something I have blogged about multiple times via interviews and podcasts with eminent open source lawyer Larry Rosen, Linux Foundation executive director Jim Zemlin, Apache Software Foundation founder Jim Jagielski and many others in the community, including folks at FSF.
It really wasn't worth another blog post in my mind. But serendipitiously, if that's a word, a new document from Larry Rosen, Open Source and Free Software, 2009, addresses the subject. Personally I have felt the distinction should not matter to IT staffs and managers, assuming that they do not plan to re-distribute any software they develop based on open source tools and libraries. Who needs the headache of having to keep track of all these different licenses? And from an IT Investment Research point of view, I have very much concluded that the FOSS vs OSS distrinction has absolutely no effect on enterprise software market dynamics because all open source is such a sliver of the market.
But thanks to Larry, you can download his recent paper and make up your own mind. He told me via e-mail that he thinks IT staffers and managers
"need to understand the law... for valuable software, and valuable legal advice, you get what you pay for."
To make the connection between the August 27 post about FSF and Larry's paper, remember that the FSF is the creator and proponent of the GNU General Public License (GPL). The GPL comes in a couple of different instantiations and with a few companion licenses to make up a set of open source terms and conditions (Ts&Cs) commonly referred to as restrictive. That is in juxtaposition to dozens of other sets of open source Ts&Cs, most notably the BSD and MIT licenses, often referred to as permissive. The various flavors of the Apache license are somewhere in between on a spectrum of restrictive to permissive.
Larry goes into great detail, in support of the Practising Law Institute (PLI) continuing education programs on "Open Source and Free Software, 2009" which will be held in November and December 2009, to point out that it is when permissively licensed and restrictively licensed, software gets mixed that some boundaries might get crossed. He says
"For example, adding new modules into Linux to create a new version of Linux is for most computer programmers somehow different from adding an Apache HTTP Server to Linux and creating a turnkey web server. Yet both involve technical acts of linking that, from a functional computer science perspective, are indistinguishable."
That makes them pretty much the same from a legal perspective, as well. The problem, he says, is that the GPL interpretation of copyright law "hasn't been tested in court." He also explains some intricacies of copyright vs. patent law, but his key message is copyrights are real, folks (my words, not his, in the last phrase).
Larry also says:
"... an attorney will often conclude that independently written open source modules can be combined and linked together in intricate ways without incurring unwanted obligations to disclose proprietary source code."
Wow, that makes the stuff I wrote about the FSF burning Windows 7 in effigy pretty mild in the eyes of the fussy FOSS folks.