The GFDL

Authors Jason Self

Plaintext

jxself.org

The GFDL Home

Sun, 16 Oct 2016 Linux-libre
I've previously written about things that don't allow modification (aka "derivative
works".) This can be implemented in various ways, among any license. One of them GitWeb
that has one particular implementation of this is the GNU Free Documentation License
("GFDL"), which calls them Invariant Sections. These sections can't be modified (or How To
even removed.)
Articles
My understanding of the original motiviation is that Stallman began including copies of
things like the GNU Manifesto along with the Emacs documentation as a way to tell
RSS Feed
people about why that software was made, what free software was and why it was
important, and that he wanted to find a way to be sure that people who didn't believe
in these values couldn't remove (or change) that part. (And for that reason, the GFDL About Me
also conditions invariant sections on being exclusively about "the relationship of the
publishers or authors of the document to the document's overall subject (or to related Contact Me
matters)." Invariant sections were born.
GPL enforced
There are those that, rightly, object to invariant sections. As I've said before all
creative works should be free. There are still others that go even further though and
claim that, since the GFDL allows for invariant sections, that it should be considered a If you appreciate any of the things I
non-free license. I'm not linking to that page in order to avoid promoting it further but I am doing you can make a donation.
will argue why such a position is ridiculous.

The first thing to keep in mind is that invariant sections are an optional feature of the
GFDL and there's no requirement that they be used. If they are not then there are no
freedom problems.

The first argument I call the Alternate Universe Theory.

The problem I see with calling it non-free is that it is based on considering what could
have happened, not what actually happened. The logic is that, since invariant sections
could have been added, it's non-free.

I argue that we should base decisions on if something is free or not based what they
actually did (i.e. did they actually use invariant sections, etc.), not what could have
happened in some alternate universe.

The reason for this is because, under current laws, any given copyright holder "could"
do anything. They "could" have made it entirely non-free for example. This is actually
the default for copyright: All Rights Reserved. So, anyone "could" have not applied a
free license. Using this same proposed standard where we must consider what the
author "could" have done, we can't consider anything to be free all because of the
possibility that they "could" have done something else.

I call my other argument The Copyleft Theory.

Using the same logic that the GFDL is non-free "because it allows for so-called
invariant sections to be added to your documentation which cannot be modified or
deleted", we'd have to also consider any other license that allows for non-free
modifications to be made to similiarly be a non-free liense. That's going to be pretty
much any non-copyleft license. Since they'd all allow non-free things to be added into
an otherwise free program, are we to consider those similiarly non-free? It would
mean eliminating large swaths of free software and free culture works merely over
something that could have happened, but didn't.

Do we really want to adopt the standard of thinking that, as long as someone could
have made something non-free, that we should consider it non-free in actuality? I
don't think so.