[Aptitude-devel] Bug#485008: aptitude: 'why' doesn't seem to work

[Manuel A. Fernandez Montecelo]

Control: tags -1 + moreinfo


Hi Sam,

2013-01-09 03:01 Axel Beckert:
>Sam Morris wrote:
>> On Sat, 2008-06-07 at 11:43 -0700, Daniel Burrows wrote:
>> > Could you run "aptitude-create-state-bundle aptitude-state-bundle.tar.bz2"
>> > and post the output somewhere that I can get to it?
>>
>> Sure, I have put it at
>> <http://robots.org.uk/stuff/aptitude-state-bundle.tar.bz2>, however it
>> may be too late as I removed those packages by hand. If this happens
>> again I will be more careful. :)
>
>Unfortunately that file is broken (and with 224K far too small):
>
>bzip2: Compressed file ends unexpectedly;
>        perhaps it is corrupted?  *Possible* reason follows.
>bzip2: Inappropriate ioctl for device
>        Input file = (stdin), output file = (stdout)

Have you seen this problem more recently with other packages?

I suspected that maybe it was a problem with being virtual, but this one
works fine:

  $ aptitude why aspell-bin
  i   reportbug  Suggests   claws-mail (>= 3.8.0)
  p   claws-mail Recommends aspell-en | aspell-dictionary
  p   aspell-br  Provides   aspell-dictionary
  p   aspell-br  Suggests   aspell-bin


I then suspected that maybe the package is of priority
required/standard/etc, maybe that's why the packages depended on that
one explicitly and "why" could not find a reason.  But at least the
version that I pulled from snapshots, it's not:

 $ wget -nv http://snapshot.debian.org/archive/debian/20080118T000000Z/pool/main/g/gcj-4.1/gcj-4.1-base_4.1.2-19_amd64.deb
 ...
  
 $ dpkg --info gcj-4.1-base_4.1.2-19_amd64.deb | grep Priority
  Priority: optional

However, when the package is installed automatically and nothing depends
on it, it fails to say that it's because of the priority:

  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=729349


So I am not sure what was going on there.

I made some tests with several other packages, esp. of these cases that
I mentioned (required like gcc-5-base, and virtual) and everything seems
to work fine now.

Maybe the underlying problem was fixed without this bug being closed.


Cheers.
-- 
Manuel A. Fernandez Montecelo <manuel.montezelo at gmail.com>