FSF Releases Draft of Version 1.2 of the GNU FDL

bkuhn writes: "The FSF has released a draft version 1.2 of the the GNU Free Documentation License for comment by the Free Software community. Comments should be directed to <fdl-comments@fsf.org>."

Comments

The license is non-free?

[OiBoy]

From the FDL:

"Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed."

I see a bit of a dichotomy here.

Re:The license is non-free?

[bcrowell]

The GFDL is a GPL-style license, which means it's viral. You can't have a viral license if people are allowed to change the license after the fact. If you don't like GPL-style licenses, you can make your work public domain, or choose a different license.

The whole point of a GPL-style license is to take away people's freedom. For example, there might be a hacked Perl interpreter linked into the software that runs your PDA, and you might never know it, because Perl's license gives the PDA manufacturer the freedom to change Perl, use the changed version, and not publish the changes. A GPL-style license doesn't give you the same freedom.

Of course, "free" isn't necessarily the same as "good," as implicitly assumed by Stallman's rhetoric. I sometimes use the GFDL, because I want to take away certain freedoms from people who use use my books. For instance, I may not want them to have the freedom, which they would have under a BSD-style license, to use excerpts from my book in their own book, without making their book free-as-in-anything.

side-by-side diff

[bcrowell]

Here is a side-by-side diff. The main change seems to be that they've added some explicit discussion of no-warranty clauses. Not a momentous change, really -- does anyone really get sued because they gave away a book for free, and the book had a mistake in it???

Re:side-by-side diff

[Pembers]

Another difference appears to be that they've added restrictions on the lengths of Front-Cover and Back-Cover Texts (5 and 25 words, respectively).

These limits seem a little arbitrary. Also, they might be inappropriate for languages other than English. Most European languages would be OK, but I gather that something like Chinese uses many short words. At the other extreme are languages that say as much in one long word as English says in a sentence (Inuit, I think, is one example). Even Latin gets by with about half the number of words in a sentence that English normally does.

Could this cause problems, or am I just picking nits here?

free use? bad for authors?

[bcrowell]

I think this may violate general copyright's concept of fair use.
Well, the GFDL can't trump fair use. The only reason a reader would need to agree to the GFDL is to gain permission to copy the text. If they're copying under the fair use doctrine, they don't need permission from the copyright owner, and they don't need to agree to the GFDL.

Plagiarism is already illegal,...
No, it's not.

...so I don't see why this restriction is necessary...
From the FSF's point of view, I think the strategy is to built a worldwide corpus of free information. By forbidding the mixing of GFDL'd information with GFDL-incompatible information, they want to arm-twist people into GFDL-ing their information. It's the same strategy that was such a big success for the GPL and Linux.

From my point of view, the point is that if someone else wants to use one of my physics homework problems, verbatim and with credit, in their own GFDL'd physics textbook, they can. On they other hand, they can't just publish my whole book as a not-free-as-in-anything book, without dealing with me. Yes, I want to prevent this, since I sell printed copies myself.

...except to make FDL'd texts off-limits to publishers.
There are three perfectly good ways around such a problem: (1) the author can choose a BSD-style license; (2) the author can offer the book to the publisher under some separate contract, rather than under the GFDL; (3) the publisher can GFDL the stuff with which they're mixing the author's GFDL's work.

[The GFDL] isn't even very good for original text writers as they don't gain very much from the license.
Would you complain that the GPL isn't good for programmers because they don't gain much from the license? There are ways to make money from GPL'd code, but for most people it's a hobby; they have some other way to pay the rent.

Re:free use? bad for authors?

[Arandir]

It's the same strategy that was such a big success for the GPL and Linux.

That's still open for debate. Without a suitable control group, there's no way to know. You can't look to BSD for guidance, since they were under a bogus lawsuit at the time Linux got started. You can't look at GNU as-a-whole, since it never took off until Linux.

But you can look at a several smaller projects. Here we see myriad GPLd projects that bit the dust and myriad MIT/BSD-style projects that took off. In fact, of the top four most successful Open Source projects (Apache, Linux, Perl, XFree86), four are not under a copyleft license.

Call me a heretic, but from my vantage point, the success of a project has a lot more to do with the project management, structure and culture than with its licensing.