On Wed, Jun 29, 2011 at 06:03:01PM +0200, Michael Biebl wrote: > Am 29.06.2011 12:36, schrieb Steve Langasek: > > On Wed, Jun 29, 2011 at 04:31:55AM +0200, Michael Biebl wrote:
> >> Could you also explain to me (or point me to the relevant documentation) > >> why python-avahi changed from arch all to arch any. > > Sorry for the lack of documentation. Unfortunately I didn't even document > > it well enough for myself locally, so I'm struggling to remember now why I > > did this. I *think* it had to do with the fact that libavahi-common-data is > > Architecture: any / Multi-Arch: same, because the gdbm database files are > > not architecture-independent, and you need a python-gdbm that corresponds to > > the architecture of libavahi-common-data that you have installed. But since > > you can't actually install more than one copy of python on your system at a > > time, and python-avahi depends on python (obviously), I can't now work out > > after the fact why I thought this was needed. It may just be a bug in my > > thinking at the time - if everything installs ok using an Architecture: all > > python-avahi, then feel free to drop this change. > Maybe this sounds stupid, but how can I test if everything installs ok? Build the source package for two architectures; grab the dpkg from the git branch listed at <http://wiki.debian.org/Multiarch/Implementation#Using_multiarch> and configure it to know about your foreign arch; test-install the Multi-Arch: same packages for both architectures using dpkg -i. If everything installs, I think that's enough to show that the Arch: any wasn't needed. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. Ubuntu Developer http://www.debian.org/ slanga...@ubuntu.com vor...@debian.org
signature.asc
Description: Digital signature
