On Sat, May 1, 2010 at 00:08, Tom H <tomh0...@gmail.com> wrote: > On Fri, Apr 30, 2010 at 1:54 PM, Boyd Stephen Smith Jr. > <b...@iguanasuicide.net> wrote: >> On Friday 30 April 2010 12:10:45 James Stuckey wrote: >>> On Fri, Apr 30, 2010 at 5:44 PM, Boyd Stephen Smith Jr. < >>> >>> b...@iguanasuicide.net> wrote: >>> > On Friday 30 April 2010 06:16:22 James Stuckey wrote: >>> > > The unstable/sid doesn't have to be comment out. Setting the default >>> > > release will keep the system tracked to, in this case, testing. >>> > >>> > Er, mostly. >>> > >>> > If there is a versioned dependency that can be satisfied from sid but not >>> > testing, you will get the package from sid. This shouldn't happen given >>> > the >>> > way testing is managed, unless you installed at least one package from >>> > sid. >>> >>> I installed eclipse from sid, since there isn't eclipse in testing. >> >> It may have pulling in some dependencies from Sid, then. >> >> I know the official line is to use '-t $something' as arguments to apt- >> get/aptitude for pulling in packages from Sid/experimental/backports, but I >> think it is better to use the '$package=$version' format. (After getting the >> version from something like (apt-cache policy $package).) >> >> My instinct is that '-t $something' effectively increases the priority of all >> packages from the $something repository, which may make the dependency >> resolver pull more from that repository than is absolutely necessary. > > If you are running stable > aptitude install <package>/testing > will install <package> from testing and try to satisfy dependencies > from stable whereas > aptittude install -t testing <package> > will install <package> from testing and try to satisfy dependencies > from testing. > > I assume that > aptitude install <package>=testing_version > behaves like > aptitude install <package>/testing > and that in both these methods the dependencies might not be satisfied > (I had that problem in December with Firefox 3.6). > > > -- > To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org > with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org > Archive: > http://lists.debian.org/s2g6d4219cc1004301138i6947c1e1n6c12c267d8e83...@mail.gmail.com > >
Thanks for this nice information, Tom -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/j2nd0bf7b0b1004301943o2dd0f30fwb811b71e1ef86...@mail.gmail.com
