rpm groups and fedora: a modest proposal

David Kewley kewley at cns.caltech.edu
Wed May 26 18:15:37 UTC 2004


Kenneth Porter wrote on Wednesday 26 May 2004 11:01:
> --On Wednesday, May 26, 2004 10:49 AM -0400 Brad Smith <brads at redhat.com> 
> wrote:
> 
> > So there is a need for grouping info even for "standalone" packages
> > because once they are installed they become part of a group of packages,
> > effective management of which is very important.
> 
> Instead of grouping, might it not make sense to use a keywords field? When 
> packaging I find either that something fits in no existing group or
> several existing groups, so I often want multiple entries and often need a
> new one.

I think the idea in this subthread is that Group: is good for indexing.  
Either at install time, or at cleanup time, or at "what sorts of packages 
do I have on my system" time.

So maybe the question isn't "What should we do with Group:?", but rather 
"How can we best provide indexes of available or installed packages?"  I 
think keywords and indexing tools are a very good idea here.

David





More information about the fedora-devel-list mailing list