I was looking at why gnome-core is not yet installable on arm64 (currently checking with both debian-ports and debian official in sources.list). One of the problems I ran into was the depchain.

gnome-core->eog->libpeas->seed

Seed currently has four RC bugs, three of them are duplicates of each other and 
relate to failures on big endian architectures which does not appear to be 
getting much attention upstream and one general FTBFS bug that has a fix 
upstream (and has had for some time).

I have been informed that libpeas is already built without seed support on hurd-i386 mips powerpc s390x sparc and there are outstanding requests to add hppa and alpha support to that list ( https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=757241 ).
Looking at the seed packages with apt-cache rdepends I get

libseed-gtk3-0
Reverse Depends:
 libpeas-1.0-0
 seed
 libseed-gtk3-dev
 libpeas-1.0-0

libseed-gtk3-dev
Reverse Depends:
 gnome-core-devel

seed
Reverse Depends:

seed-doc
Reverse Depends:
 libseed-gtk3-dev

Build-rdeps shows a similar story with libpeas being the only source package in 
debian listed.

The dependency of gnome-core-devel on seed is also architecture qualified (as 
not mips) so presumablly seed is optional there too.

I see three possible ways to get things back to a sane (non rc-buggy) state

1: Fix all the FTBFS bugs in seed.
2: Apply the fix the general FTBFS in seed. Check what architectures the 
resulting package builds on and either fix or declare seed unssupportable on 
any remaining architectures. On the architectures where seed has been declared 
unsupportable adjust libpeas and gnome-core-devel to eliminate dependencies on 
it and then request removal of the old seed binaries from those architectures.
3: get rid of seed completely.

Option 1 would obviously be prefferable in an ideal world but it doesn't seem like there is anyone prepared to do the work Option 2 seems like the most sensible course of action to me.
Option 3 is probablly over-aggressive.


--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to