Gentoo Archives: gentoo-user

From: Dale <rdalek1967@×××××.com>
To: gentoo-user@l.g.o
Subject: Re: [gentoo-user] problems with depclean
Date: Thu, 17 Oct 2019 11:58:07
Message-Id: 6e3cee37-4224-b1bb-9b55-52e6502d144e@gmail.com
In Reply to: Re: [gentoo-user] problems with depclean by John Covici
1 John Covici wrote:
2 > On Thu, 17 Oct 2019 03:23:07 -0400,
3 > Neil Bothwick wrote:
4 >> [1 <text/plain; US-ASCII (quoted-printable)>]
5 >> On Wed, 16 Oct 2019 21:28:18 -0400, John Covici wrote:
6 >>
7 >>>> Can you post an example output from emerge, the eix entry from that
8 >>>> package and the entry in @world?
9 >>> In the case I see before me, depclean wants to unmerge
10 >>> dev-util/glib-utils -- its not in my world file, but if I do
11 >>> qdepends dev-util/glib-utils I get
12 >>> dev-util/glib-utils-2.58.3: dev-libs/libxslt app-arch/xz-utils
13 >>>> =dev-lang/python-exec-2:2/2=[python_targets_python3_6(+),-python_single_target_jython2_7(+),-python_single_target_pypy(+),-python_single_target_pypy3(+),-python_single_target_python2_7(+),python_single_target_python3_6(+)]
14 >>>> !<dev-libs/glib-2.56.2:2 dev-lang/python:3.6
15 >>>> app-text/docbook-xsl-stylesheets
16 >> That is odd, I use emerge -cpv rather than qdepends and it shows many
17 >> reverse dependencies.
18 >>
19 >>> but if I do eix of that it says:
20 >>> [U] dev-util/glib-utils
21 >>> Available versions: 2.58.3{tbz2} 2.60.6 (~)2.60.7
22 >>> {PYTHON_SINGLE_TARGET="python3_5 python3_6 python3_7"
23 >>> PYTHON_TARGETS="python3_5 python3_6 python3_7"}
24 >>> Installed versions: 2.58.3{tbz2}(08:28:38 AM
25 >>> 05/10/2019)(PYTHON_SINGLE_TARGET="python3_6 -python3_5
26 >>> -python3_7" PYTHON_TARGETS="python3_6 -python3_5
27 >>> -python3_7")
28 >> It may not be relevant, but you don't have the latest version installed
29 >> and it is recommended to run depclean after a full world update.
30 >>
31 >>> But if I do
32 >>> script -c "emerge --update --deep
33 >>> --with-bdeps=y --changed-use --backtrack=500 --keep-going
34 >>> world" /usr/src/world_update.txt
35 >>>
36 >>> It says 0 packages and stops.
37 >> So it's odd to see that portage considers your system up to date.
38 >> Although this is consistent with portage considering glib-utils to be
39 >> unnecessary. Does
40 >>
41 >> grep -r glib-utils /etc/portage
42 >>
43 >> show anything. Is the package listed in the output from emerge -ep @world?
44 > No output from the grep or emerge -ep @world for that package.
45 >
46
47
48 I'm going to mention something I ran into recently.  It may not have any
49 effect but it may.  Recently I ran eix-test-obsolete and was cleaning
50 out old entries in /etc/portage.  On occasion, I would run emerge -auDN
51 world to make sure it was reporting things that shouldn't change
52 anything.  For the record, I have some default options in make.conf. 
53 The odd thing is, it started updating packages.  Most of them was KDE
54 because I run a mixed system, mostly stable but some
55 unstable/keyworded.  System tools are stable, such as things needed for
56 booting etc.  KDE and some other packages I run unstable.  Anyway, after
57 I did one set of upgrades, I continued running eix-test-obsolete and
58 cleaning out old cruft.  At times, it would update some more packages. 
59 I did a rinse and repeat for a while.  Eventually I got the files
60 cleaned out and got done with updates. 
61
62 I couldn't figure out why cleaning out old cruft that shouldn't change
63 anything would trigger updates but it did.  My mask file is short so
64 almost all changes were to the package.keyword file.  I could understand
65 if I removed something and it wanted to downgrade packages but I
66 couldn't figure out why it was stopping it from upgrading. 
67
68 Could there be entries in package.* files that are affecting this but
69 doesn't register in our way of thinking?  Maybe you have a line there
70 that is stopping something but it is done in a way that isn't obvious. 
71 You may want to run eix-test-obsolete and start cleaning out some
72 entries and see if things change.  Maybe emerge is calculating things
73 differently and we don't know, heck, the dev keeping emerge updated may
74 not even realize the change.  It could even be a bug that no one noticed
75 yet. 
76
77 As I said, it may not be the problem but it seems y'all are running out
78 of ideas.  It's the only thing I can think of that just may be worth
79 looking into.  As I've said before, when grasping at straws, just grab
80 all you can until one works. 
81
82 Hope that helps.
83
84 Dale
85
86 :-)  :-) 

Replies

Subject Author
Re: [gentoo-user] problems with depclean Wols Lists <antlists@××××××××××××.uk>