Brian Nelson: What does removing GFDL documents accomplish?
One thing I just don’t understand about the whole GFDL debacle is why it’s so urgently important to purge all GFDL documents from Debian. For those documents with invariant sections, I can kind of see the point. Invariant sections clearly violate DFSG3, and we’re obligated by the social contract to abide by the DFSG. Granted, we’ve been distributing non-modifiable documents since Debian’s inception, so that’s a rather shocking change in policy. Thanks to some “editorial changes” to the social contract, the Debian distribution has to change in fundamental ways.
What’s even more suprising is that the changes in policy are going far beyond the requirements in the DFSG. According to the draft position statement, there are two problems in addition to the invariant sections–the “DRM” restriction and transparent and opaque copies. Note that in the GR proposal, the invariant section problem is explained by quoting DFSG3. However, the other two problems are not given any reference to the DFSG, and for good reason–there is no clear DFSG violation in either one. Sure, they’re icky, worded poorly, and no one particularly likes them. But since when is that grounds for removal from Debian?
Even worse, overriding the interpretation that these problems are DFSG violations will require a 3:1 super majority thanks to the secretaries opinion on the matter. Apparently, we are being strong-armed into purging all GFDL documents from Debian, which little chance of reversing this decision.
So, what does this accomplish?
- Alienation of our users Users will have their documentation pulled out from under them, which is acerbated by the installer no longer adding the non-free section to the sources.list by default. To most users, this documentation will simply disappear. That includes documentation for all GNU software and all of KDE, which is some of the most crucial documentation for users.
- Alienation of the FSF Well, this bridge has probably already been burned, but removing all GFDL documentation from Debian will probably make it even less likely the FSF would take steps to improve the GFDL more to Debian’s liking.
- Heavy maintenance burden on Debian’s developers All packages in Debian containing GFDL documents will have to be permanently forked from upstream–in other words, they’ll have to be repacked with the documentation removed. Then a new source package will have to be created to contain just the documentation to be uploaded to non-free. And that’s if the maintainer even bothers with the non-free stuff. I suspect most will just drop the documentation on the floor, never to be seen again.
- Far greater reliance on the non-free archive





