Re: xcdroast does no longer work with wodim: Who to blame?
On 03 Mar 20:51, Joerg Schilling wrote:
> Bill Unruh <firstname.lastname@example.org> wrote:
> > On Tue, 3 Mar 2009, Joerg Schilling wrote:
> > ...
> > >
> > > As a hint: "the work mkisofs" is the plain files that can be found in the
> > > sub-directory "mkisofs" in the cdrtools source tree. Other sub-directories in
> > > this source tree colletion contain _other_ independent works.
> > >
> > >
> > > You have to decide whether the GPL is a completely unusable license or whether
> > > there is no problem with mkisofs.....
> > I am afraid that we will not solve the problems of the shortcomings of GPL
> > here. The question is whether or not we can solve the problems of the
> > different reading of the GPL for the purposes of this one program, mkisofs.
> Well despite the claims from some people that try to prevent a solution,
> there in fact is only a very minor disagreement. This disagreement is based on
> the attempt from some people to interpret some meaning into the "system
> exception" that is not in the GPL text.
No, the exact problem is:
A ISO-9660/Rock-Ridge/Joliet/HFS/UDF filesystem formatter (GPL)
Note: uses libscg
A local SCSI-generic transport lib (CDDL)
What people are *clearly* telling you is that libscg is under the CDDL
(you should know that, you put it under it...) and that mkisofs is under
the GPL (you should know that too, you put it under it), and that GPL
software (with the exception of system libraries, i.e. libraries that
you'd expect to find on the system...) can not link with CDDL software
as they have incompatible licences - the source is fine to distribute -
any binary package becomes undistributable. As debian ships binary
packages (as does SuSE, RedHat, Ubuntu etc etc etc), we are not able to
ship mkisofs at all.
We're not trying to interpret any extra meaning in to the system
exception - we're telling you that libscg *does not* fit in the system
exception and that the GPL and CDDL are incompatible licences.
Also, can you fix your website to refer to the work "package" rather
than "packet" when refering to the distributed "packages" by the various