On Mon, 17 Aug 2009 14:09:26 +0200 Rubén Rodríguez Pérez <ru...@gnu.org> wrote:
> El sáb, 15-08-2009 a las 10:55 +0930, Karl Goetz escribió: > > In order of preference, I'd go with: > > > > - Modify driver to print a warning. > > - Modify driver to print a warning (disable loading files) > > - Disable driver at compile time. > > - Remove driver (source) completely. > > >From that list, only two look "free enough" to me: > > -Remove driver (source) completely. > -Modify driver to print a warning (disable loading files) > > I prefer the first, I don't see the point of distributing a useless > program that still depends on a non-free part that now cannot be > loaded. Linux-libre chooses this solution, stating that the modified Unless it doesnt depend on non-free parts, because there is free firmware. > error message can be used to teach the user about freedom. That way > the program is not useless, but I still prefer to remove it. > > > > > > > > -Should they be kept, but disabling the file load call? > > > > > > > > what if someone develops a free file replacement? > > > > > > Then the maintainer of the freed version of the kernel can stop > > > removing that module. > > > > This would be easier if it was compiled out then if it was rm'd out. > > Difficulty can only be used to choose between valid solutions. Same We agree on this part then. > goes for any other practical concern. If you choose not to compile > the driver to fix an ethical issue, then you shouldn't distribute the > code either. The only disagreement seems to be when a bit of code is no longer allowed to be shipped. kk -- Karl Goetz, (Kamping_Kaiser / VK5FOSS) Debian contributor / gNewSense Maintainer http://www.kgoetz.id.au No, I won't join your social networking group
signature.asc
Description: PGP signature