[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Apt-Get or Aptitude



On Thu, Nov 01, 2007 at 11:09:07AM -0800, Ken Irving <fnkci@uaf.edu> was heard to say:
> > > On Tue, Oct 30, 2007 at 08:19:58PM -0700, Daniel Burrows wrote:
> >   No, I just come down hard on this meme because it seems to have taken
> > on a life of its own and I'd like to squash it before it grows up into a
> > full-blown urban legend.
> 
> That sounds good, but is it different now than it used to be?  I haven't
> tried it lately, but it used to "seem" to want to remove lots of things.
> I'm aware of the workarounds (keep-all or whatever), have followed most
> of the threads (even instigated some...), but am still a command-line
> apt-get user waiting for a reason to change.  Two problems I have with
> aptitude are the lack of "source" functionality and my inability to spell
> it as easily as apt-get. ;-)

  There were bugs in some past versions.  As far as I know, the worst
ones (e.g., #411123) were fixed in etch.  There were some new bugs
introduced in unstable with the switchover to using apt to track unused
packages (where aptitude would even want to remove packages it had just
installed), but those should be fixed in 0.4.7.

  There are a few corner cases in which aptitude will do the wrong
thing.

  * Marking a package for removal in aptitude, exiting, removing it with
    apt-get, installing it again with apt-get, then running aptitude.
    aptitude will still remember that you want to remove the package.

  * If you interrupt aptitude before it writes its state database, it
    will sometimes get confused about the system state, especially if
    you proceed to run apt-get before aptitude. (I can't remember the
    precise sequence of events that have to happen to trigger this off
    the top of my head)

  Those are the only ways I can think of offhand to get aptitude to
remove packages you didn't ask it to.  Unfortunately, there's no
reliable way to tell if someone else has fiddled with a package
(#429438), so as long as aptitude tries to save and restore the current
state, there will be a few edge cases like this.

  Anything I didn't list above is a bug that I don't know about.

  Daniel



Reply to: