On Mon 14 Nov 2016 at 14:05:46 (+0100), steve wrote:
> Hi David,
>
> Coming back to this problem now that I have a bit more time.
>
> As I have been making a lot of testing, this message is a bit long,
> sorry for that.
>
>
> >If you follow my recipe, any packages counted twice (as eg in both
>
Dear Axel,
I just wrote an extensive answer to David *before* reading your reply
and it happens that it answers pretty much to what I discovered, which
is that if a backport package exists but the stable one is installed,
the search with ~A$a~i are counted twice. Seems strange at first thought
bu
Hi David,
Coming back to this problem now that I have a bit more time.
As I have been making a lot of testing, this message is a bit long,
sorry for that.
If you follow my recipe, any packages counted twice (as eg in both
the stable and jessie searches) will show up in the diff with a "-".
steve wrote on 11/06/16 20:47:
> Le 06-11-2016, à 10:43:58 +0100, Jörg-Volker Peetz a écrit :
>
>> What is the output of
>>
>> aptitude -F "%p" '~o'
>
> gives an error (unknown command « ~o »)
>
Sorry, this should be
aptitude -F "%p" search '~o'
>> ? Seems to me, there are packages installed
On Sun 06 Nov 2016 at 20:53:32 (+0100), steve wrote:
> Le 06-11-2016, à 07:05:15 -0600, David Wright a écrit :
>
> >>so 3507 ≠ 3349. Both figures should be equal as I understand. It seems
> >>that some packages are counted two or more times or my calculation is
> >>plain wrong.
> >>
> >>Thoughts?
Hi Steve,
Steve wrote:
> Thank you for your explanation, I understand a bit better the logic. But
> there is still something that doesn't quite match. Please consider the
> following.
[...]
> a= aptitude search ~A[$a]~i | wc -l
> dpkg -l | grep ^i
Le 06-11-2016, à 07:05:15 -0600, David Wright a écrit :
so 3507 ≠ 3349. Both figures should be equal as I understand. It seems
that some packages are counted two or more times or my calculation is
plain wrong.
Thoughts?
Piping to wc -l throws most of the information away.
I don't see why *h
Le 06-11-2016, à 10:43:58 +0100, Jörg-Volker Peetz a écrit :
What is the output of
aptitude -F "%p" '~o'
gives an error (unknown command « ~o »)
? Seems to me, there are packages installed on your system that don't belong to
any architecture in sources.list ("obsolete").
However, running
On Sun 06 Nov 2016 at 08:18:52 (+0100), Steve wrote:
> Hi Sven and Axel,
>
> Thank you for your explanation, I understand a bit better the logic. But
> there is still something that doesn't quite match. Please consider the
> following.
>
> apt-cache policy | grep 'a='
> release a=now
> re
What is the output of
aptitude -F "%p" '~o'
? Seems to me, there are packages installed on your system that don't belong to
any architecture in sources.list ("obsolete").
Regards,
jvp.
Hi Sven and Axel,
Thank you for your explanation, I understand a bit better the logic. But
there is still something that doesn't quite match. Please consider the
following.
apt-cache policy | grep 'a='
release a=now
release v=14.04,o=LP-PPA-opencpn-opencpn,a=trusty,n=trusty,l=OpenCPN,c=m
Hi again,
Axel Beckert wrote:
> The main reason is that ~A does (as most of aptitude's patterns)
> substring matching,
[…]
> * aptitude's ~A pattern matches substrings of the Archive value in the
> Release file.
Sven is of course right, and it's regular expression matching and not
(only) substr
Hi Steve,
steve wrote:
> I'm trying to understand why
>
> aptitude search ~Ajessie~i | wc -l
> 240
>
> is different from
>
> aptitude search ~Astable~i | wc -l
> 3243
>
> but is the same as
>
> aptitude search ~Ajessie-backports~i | wc -l
> 240
The main reason is that ~A does (as most of apt
13 matches
Mail list logo
