On Wed, 26 Jun 2013 19:45:52 +0200 David Kalnischkies wrote: > On Wed, Jun 26, 2013 at 6:47 PM, Francesco Poli > <invernomuto@paranoici.org> wrote: > > On Wed, 26 Jun 2013 18:06:53 +0200 David Kalnischkies wrote: > >> as it > >> looks like it pins the wrong package if its using v2 currently and I am also > >> not really sure if the pin-logic shouldn't be changed drastically as a bug > >> usually effects all architectures and not just one, but my ruby knowledge > >> is non-existent, so I could be completely wrong on this. > > > > Let's leave implementation details alone (specific to Ruby or to any > > other language). > > As you probably know, apt-listbugs pins packages by adding stanzas > > to /etc/apt/preferences: as a consequence, what apt-listbugs can put > > there depends on the syntax apt and aptitude will be able to understand. > > Is there any way to qualify the architecture in Pin stanzas? > > I glanced over apt_preferences(5), but couldn't find anything about > > this topic. > > You can pin a specific architecture just like you can request a specific > architecture on the commandline with <packagename>:<architecture> > If no architecture is provided it will refer to the native architecture. [...] This is useful information, thanks! > > That's why I believe apt-listbugs pins the "wrong" package as if I e.g. have > foobar:i386 installed (on a amd64 machine) the hook will give you foobar > as package name and the pin will be issued for foobar which means the pin > has no effect, as it isn't effecting "foobar:i386". I presume pinning > "foobar:*" would be better as most bugs will effect all architectures. At a first thought, this scenario seems to be indeed possible. I'll have to perform some tests in order to check whether this actually works as you imagined, but, in theory, it really looks like what you said is going to happen. > > But I am not an apt-listbugs user, No?!? You should totally install it, it's the coolest package in the whole Debian archive, don't cha know? ;-p > so this is just guessing based on a quick > look into the code, which as I said, could be completely wrong. > (and hence I just cc'ed you rather than report a bug or anything formal) OK. > > > > Please note that, as I said in the bug log for #688506, I am > > unfortunately quite ignorant about multiarch. > > Where can I find a crash course on multiarch theory and practice? > > You can find various documentation on https://wiki.debian.org/Multiarch Good, thanks for the hint. [...] > If you know what the package > ia32libs was and why it was needed: MultiArch is replacing it with a cleaner, > more general and overall better solution, otherwise: MultiArch is enabling > the installation of packages from different architectures on the same machine > so you can run e.g. applications compiled for i386 on amd64. Good, these few lines summarize my whole knowledge about the topic. So now you should realize that when I said I am quite ignorant about multiarch, I really meant it! ;-) > > The only really important thing is that the package name isn't a unique key > anymore: You can easily have a bunch of packages with the same name > installed (those packages will be marked: MultiArch: same). > Anything else are more or less boring details (actually, even this is a > boring detail, but many things assumed that the package name is a unique > key, so it has to be mentioned as important). Now you are beginning to add something that I didn't know (the meaning of MultiArch: same). Thanks a lot for this! Bye and thanks for the useful discussion! It's really appreciated. -- http://www.inventati.org/frx/frx-gpg-key-transition-2010.txt New GnuPG key, see the transition document! ..................................................... Francesco Poli . GnuPG key fpr == CA01 1147 9CD2 EFDF FB82 3925 3E1C 27E1 1F69 BFFE
Attachment:
pgpuu5GjlbKZP.pgp
Description: PGP signature