NetworkManager is driving me crazy

Matthew Saltzman mjs at clemson.edu
Sun Mar 30 20:21:27 UTC 2008


On Sun, 2008-03-30 at 09:26 -0700, don vogt wrote:
> [...]
>  I still had problems so
> I tried to "yum remove NetworkManager and dispatcher"
> and it, for dependency, removed pidgin, which I want
> to message with my daughter. I used pidgin in fc7
> without NetworkManager very easily. So I tried "yum
> install pidgin. Lo and behold it also installed
> NetworkManager.
> 
>  Sorry if I shouldn't have used this title. I just
> needed to vent a little (a lot?) I am not quite sure
> what to do next, remove NetworkManager (Gosh, I get
> tired of typing that) somehow without removing pidgin,
> 
> get a new IM program? ( I hate to do that because
> configuring pidgin was problematic for me.) Set
> Network Manager to not run as a service, turn off the
> modem when I leave?
>  I am probably doing something dumb, it tends to come
> with age, but I would appreciate any suggestions.
> 

Pidgin needs something in libnm_glib.so.0, which is provided by
NetworkManager-glib.  This doesn't sound like such a great thing, but
evolution also has that requirement.

If it bothers you, file a Bugzilla entry suggesting that the relevant
functions be broken out of NetworkManager-glib.

You should be able to remove other NetworkManager packages without
removing NetworkManager-glib.  You should also be able to just shut it
off (/sbin/service NetworkManager stop; /sbin/chkconfig NetworkManager
stop; same for NetworkManagerDispatcher--the first stops it running now,
the second prevents it from starting on boot).

I don't really understand yum's view of dependencies.  The logic of
removing a package and everything it depends on, and everything that
depends on those packages, etc., escapes me.[1]  Use rpm to remove
packages if you don't want yum to run amok deleting other stuff.

[1] I could understand removing a package and everything it depends on
that doesn't, in turn, have something else depending on it.  But this
way, it seems like removing almost anything can end up ripping most of a
functioning system out by the roots.

-- 
                Matthew Saltzman

Clemson University Math Sciences
mjs AT clemson DOT edu
http://www.math.clemson.edu/~mjs




More information about the users mailing list