[Lunar-bugs] [Lunar Linux 0000296]: "lrm -p MODULE" & optional_depends. (lrm -p works in dager way)
Lunar bug reports list
lunar-bugs at lunar-linux.org
Fri Nov 2 14:31:38 CET 2007
A NOTE has been added to this issue.
======================================================================
http://bugs.lunar-linux.org/view.php?id=296
======================================================================
Reported By: hdw
Assigned To: sofar
======================================================================
Project: Lunar Linux
Issue ID: 296
Category: lunar
Reproducibility: always
Severity: minor
Priority: normal
Status: assigned
Moonbase Version: 2007MMDD.NN (I'm an idiot and didn't edit this
field.)
Core Tools: Theedge
Core Tools Version: 2007MMDD (I'm an idiot and didn't edit this field.)
======================================================================
Date Submitted: 10-07-2007 21:10 CEST
Last Modified: 11-02-2007 14:31 CET
======================================================================
Summary: "lrm -p MODULE" & optional_depends. (lrm -p works in
dager way)
Description:
** The situation **
a module AAA is optionally depends on BBB,
and you install/compile AAA _without_ BBB.
and you install BBB after AAA.
if you invoke "lrm -p BBB" then,
the command uninstall not only BBB but AAA as well.
An actual situation I encountered was,
1. install "xscreensaver" without an optional_depends "fortune".
2. install "fortune".
3. lrm -p "fortune".
4. xscreensaver was also removed with fortune.
I had never thought removing "fortune" also remove "xsreensaver",
so I was bit surprised...
Also this is easily avoidable,
In my opinion, this mechanism should be changed.
Sorry for ugly sentence in Summary section.
======================================================================
----------------------------------------------------------------------
hdw - 11-02-07 14:31
----------------------------------------------------------------------
Hi.
I'm sorry, now I wanna withdraw this bug report.
I realised that I had not given an enough thought about this.
I still think the way the command works is bit insecure,
but I also can see a merit of current way and
I can't come up with better idea anyway..
Issue History
Date Modified Username Field Change
======================================================================
10-07-07 21:10 hdw New Issue
10-07-07 21:10 hdw Status new => assigned
10-07-07 21:10 hdw Assigned To => sofar
10-07-07 21:10 hdw Moonbase Version => 2007MMDD.NN (I'm an
idiot and didn't edit this field.)
10-07-07 21:10 hdw Core Tools => Theedge
10-07-07 21:10 hdw Core Tools Version => 2007MMDD (I'm an
idiot and didn't edit this field.)
11-02-07 14:31 hdw Note Added: 0000683
======================================================================
More information about the Lunar-bugs
mailing list