Re: essential packages and Pre-Depends
Hi,
>>"Kai" == Kai Henningsen <kaih@khms.westfalen.de> writes:
Kai> sanvila@unex.es (Santiago Vila) wrote on 16.02.98 in
Kai> <[🔎] Pine.LNX.3.96.980216222233.1846A-100000@cantor.unex.es>:
>> According to this, "essential packages are the ones without which
>> the packaging system breaks".
Kai> That's how it should be, yes.
Kai> However, we have several different classes here.
Kai> First, there's stuff that, if it breaks, makes impossible to use
Kai> the packaging system to repair it. Let's call that
Kai> essential-for-dpkg.
I understand that.
Kai> Second, there's stuff that's needed for basic operations of the
Kai> packaging system, but if it breeaks, it can still be repaired
Kai> using the packaging system. Let's call that
Kai> essential-for-normal-installation.
Hmm, ok, I'll buy that.
Kai> And third, there's stuff that just happens to be in packages also
Kai> containing some of the other stuff above, or even only support
Kai> stuff for such stuff. Essential-by-infection.
Umm, we are talking about packages here. This statement does
not make sense. Either a package is in section 1 or 2 above, or it is
not. When the granularity is a package, then we can't have packages
by infection.
What am I missing?
Kai> Now, we obviously must have pre-depends for the first case. It is
Kai> much less critical for the second case, but one could argue that
Kai> it would still be useful to prevent the install tools breaking -
Kai> even if they are then repairable.
Kai> I'd say it's not needed for the third case.
If packages can belong in the third case, I think they should
be reduced from Essential to Required.
manoj
--
Let me state that programming is not the science of coding but the
art of finding solutions of non-formalized problems and expressing
these solutions in explicit and clear way. Vadim Antonov
(avg@hq.demos.su)
Manoj Srivastava <srivasta@acm.org> <http://www.datasync.com/%7Esrivasta/>
Key C7261095 fingerprint = CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org .
Trouble? e-mail to templin@bucknell.edu .
Reply to: