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

Re: Re: Some myths regarding apt pinning

Lloyd Zusman wrote:

Erik Steffl <steffl@bigfoot.com> writes:



[ ... ]

  but the point is that pinning is not very good because you either
  bring a number of important packages from unstable (libc6, perl etc)
  or you simply cannot use it. reading of the manual page and checking
  the apt-listchanges does not solve the problem. i.e. you recommend
  pinning, person reads the manpage, tries pinning and finds out that
  it was pretty much pointless excercise because it would upgrade large
  part of the system to unstable. or yet another wording: Adrian Bunk
  wasn't complaining about system actually upgrading packages but about
  system trying to upgrade packages.

	erik


I want to be sure that I understand the significance of this.  Are you
saying that pinning a certain package, say "randompackage", to
"unstable" in /etc/apt/preferences is worse than doing this the first
time that "randompackage" is installed? ...

  apt-get -t unstable install randompackage

Or do these two methods have equally undesirable effects?
Pinning is roughly equivalent to 'apt-get install randompackage/unstable'. Neither will try to pull in unstable dependencies, but will instead just fail if the package can't be installed with testing dependencies. 'apt-get -t unstable install randompackage' is probably the worst command to use since it will pull unstable dependencies for uninstalled packages even for unversioned dependencies. Pinning has essentially been useless for the past 2 months, but for the 10 months before that I thought that pinning *was* useful so it seems disingenuous to claim that it is always useless.
Reply to: