On 9-8-2004 18:58, "Matthew Garrett" <[EMAIL PROTECTED]> wrote:
> Freek Dijkstra <[EMAIL PROTECTED]> wrote: > >> So if indeed netatalk contains executable code from openssl, then according >> to #2, the redistributed binary must have the openssl-licence, and that is >> not allowed. However, just thinking about how dynamic libraries work, I >> belief there is no executable code of the library being called involved, >> even though the header files are used (since they only contain definitions, >> but don't lead to executable code). >> >> Could someone confirm or deny this? > > The FSF disagree - their claim is that even with dynamic linking the > libraries must be GPL compatible. Nobody has so far been willing to have > a lawsuit over this, so it's not possible to confirm or deny this. > Believing the FSF is safer than not doing so, so we take the low-risk > approach. I understand the low risk thing. However, in this case, I'd say it's worth the lawsuit. Though I have supported FSF, in the particular case I'd hope they lose :-) (I'm sure Richard Stallman doesn't agree with me). For those interested, there was an discussion on this topic on gnu-misc-discuss last month. Read this mail for interesting pointers: http://lists.gnu.org/archive/html/gnu-misc-discuss/2004-07/msg00070.html For example, the first pointer, page 16 of it on how Richard Stallman and Eben Moglen (general counsel of FSF) disagree on what is "derivate works". However, that being said, I claim it does not apply to this particular scenario! In this case, I suggested to distributed a binary of netatalk, including the UAMS linked with OpenSSL under GPL. To see if this is allowed you have to look at the *OpenSSL-licence*, NOT at the *GPL*. (You could for that matter have looked at the LGPL as well, which explicitly would have allowed dynamic linking). Well *I* do not see anything in the OpenSSL licence which specifically forbid dynamic linking against it. So I think it is allowed. (If you like, I am more then willing to contact the OpenSSL on this matter). Would you agree that if the OpenSSL licence doesn't forbid this, then it is OK for netatalk to link against OpenSSL? Or do I still miss something? Regards, Freek Dijkstra
