On Thu, Aug 14, 2008 at 7:57 PM, robert_horning@netzero.net robert_horning@netzero.net wrote:
One thing that I think is reasonable to ask for is that the GFDL is one of the licenses that must be included. I don't think there is any dispute here.
Nope, none whatsoever.
The problem comes from those editors who decide to contribute to a wikibook, but for some reason or other explicitly don't want the "dual licensing" to continue. A great deal of this can and IMHO should be in some sort of policy guideline for the project.
Yes, but I think this is just a small aspect of the larger problem. It's not just users who explicitly want GFDL-only, but also people who are lead to believe that Wikibooks is GFDL-only and edit a dual-licensed book without knowledge of it. The last thing that we need right now is to insist that all new editors familiarize themselves with the licensing intricacies of every single book before attempting an edit. Our barrier to entry is already high enough without demanding our editors navigate an uneven and haphazard legal landscape.
There are obviously some steps we could take to mediate this problem. If we altered the copyright notice at the bottom of the edit page, we could inform users that some books use more complicated licensing schemes then the default GFDL-only. However, even with warning there are people who aren't going to read that notice or understand it.
There are some book projects that there are some very legitimate reasons for wanting to continue the dual-license. We've gone over before some of those book projects that want to be dual-licensed GPL/GFDL in order to ultimately include some GPL'd software into the book in some sort of final distro. I've also seen some books that were to be distributed with other items that were of some other free-type license (such as CC-by-SA or some other similar license). In these cases, maintaining the dual-licensing can be incredibly useful as we get the benefit of having the free content on Wikibooks, but the "end users" can also use the content in a completely different context.
In the cases you mentioned, an aggregate could contain portions of content that are licensed separately. Think about how fair use excerpts are included in GFDL articles at Wikipedia. Or, think about how CC-BY-SA images are included in our books. Code snippets can be licensed separately, and we should find a way (special templates?) to support that more transparently.
I would hope that dual-licensing aren't completely dismissed from Wikibooks, and I especially would be abhorred by some "administrator" or other user randomly going through all of the current wikibooks and removing all mention of any dual-licensing content that currently exists. Such actions are IMHO contrary to a wiki spirit and this is something that should receive further debate... especially if such drastic actions are taking place.
This isn't something that is restricted to admins, any ordinary user could do it, and in a legally convincing manner. A dual license is an either-or affair: A reuser can choose one license or the other as their preference allows. All a user would need to do on a dual-licensed book is to make an edit that is GFDL-only, thus negating the second license.
Further, consider the potential pitfalls if a user decided to make a CC-BY-SA edit instead, thus negating the GFDL component of the dual-license. In a dual license, it's perfectly within the rights of the editor to select the second of the two licenses for their derivative. What a pickle we would be in if our books were all incompatible with each other.
All this said, the dual-licensing is something that should be decided when the book is established and is nearly impossible to change once multiple authors have contributed (in a significant manner) to its contents.
I'll reiterate my points: We risk creating an uneven landscape that is difficult to navigate. We risk raising the already formidable barrier to entry, asking new users to know and understand a myriad of licenses and license combinations.We run the risk of creating books that become legally incompatible with each other. +
As many opportunities and benefits as there are for flexible licensing schemes, I think we are far better served with a consistent and simple licensing scheme for now. The best I think we can do is to allow particular revisions to be flagged as dual-licensed, if all contributions previous to that revision are likewise licensed. This, of course, raises a certain amount of confusion in and of itself, and might be untenable for long-term ubiquitous use.
I just don't see how the additional flexibility we gain from this offsets all the other potential problems inherent in it.
--Andrew Whitworth