On Thu, Jul 22, 2004 at 01:21:17AM -0400, Evan Prodromou wrote: > But in evaluating licenses, we have to assume that the Licensor is not > good, generous, or rational. If we can convince ourselves that the license > grants the licensees freedom _even_when_ the Licensor is possessed by > Captain Howdy and starts spewing green goo out of their eye sockets, then > we can be reasonably certain that works released under the license are > really Free. > > Unfortunately, taking this tack makes us look like mean and vituperative > a-holes. The word is "lawyers". They'd do exactly the same thing, for the same reasons, if this were a proper license being negotiated by two parties, rather than one party trying to stuff a vague and open-ended document down the collective throat of the world. > So, if Programmer Joe really wrote a program and made the documentation > available under the by 2.0, and I created a modified version and wrote in > the modified documentation: > > Programmer Joe's version of this algorithm ran in O(N^2) time, but > our > new version runs in O(NlogN) time. > > ...then, as the license is written now, Joe could request that I remove his > name from this sentence. > > Now, is this earthshatteringly bad? Not really. We could obviously work > around it, and program documentation that leaves out reference to the > original version and its authors would probably be more or less usable. > > But opinion here seems to lean to the side that letting Licensors have this > level of editorial control over modified versions of a document makes that > document non-free. Notably it fails the smoke test: this clause prohibits us from including the work in Debian, since we cannot realistically satisfy this requirement. That means it's got to be non-free. -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | `. `' | `- -><- |
Attachment:
signature.asc
Description: Digital signature