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

Re: comps discussion at fudcon and the future

On Sat, 2009-01-17 at 01:40 +0100, Arnaud Gomes-do-Vale wrote:
> Toshio Kuratomi <a badger gmail com> writes:
> >> But that means that for classical uses of kickstart (i.e. run the installer
> >> with this kickstart on every machine), the package would get installed,
> >> then removed, on every single machine. :-(
> >> 
> > Yes.  But that's already the case.
> Is it? I'm pretty sure this is not the case on RHEL5 and its clones at
> least.

 I think it depends, if I'm following the code right (and I might well
not be). anaconda/kickstart basically does a yb.install() to get
packages and yb.tsInfo.remove() to get rid of them. So you can do:


...and it'll work (as it'll basically be adding and then removing the
pkg from the tsInfo), but if you do:


...then foo will be removed from the transaction, but then added back in
due to the dep.
 So it'd all depend on how the "new groups" are implemented, it
certainly wouldn't be hard to have tsInfo.remove() just take the group
out too ... or change how anaconda worked, or any number of others

James Antill <james fedoraproject org>

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