[Yum-devel] yum-utils package
Gijs Hollestelle
g.hollestelle at gmail.com
Wed Mar 23 13:19:16 UTC 2005
On Wed, 23 Mar 2005 02:22:57 -0500, seth vidal <skvidal at phy.duke.edu> wrote:
> 1. yum-downloader should use the same mechanism to determine most
> 'appropriate' package to download as yum uses
> 2. it should use the same package string parser
> 3. it should use urlgrabber as normal
>
> What else does it really need to do?
> - download to curdir?
> - download to the relevant cache dir?
>
> Dumb questions:
> How should 'yum-downloader *' work?
> Should yum-downloader know about srpms?
Hmm these comments make it appear to me as if this would be much
easier to write as a command line switch to yum (i.e. --download-only)
that woud implement almost all of the above (it would always download
to the cache dir) with the least amount of effort. Would that be
acceptable into mainline yum, Seth?
If we want to do it stand-alone I propose only having --disablerepo,
--exclude and --resolve with the default behaviour being only
downloading the latest version of the requested rpm to the current
directory and --resolve being also download all packages needed to
resolve dependencies to the current directory.
Downloading to the current directory has the advantages of being able
to run it without being root. SRPMS should be handled by a different
yum-util I think.
Greets,
Gijs
More information about the Yum-devel
mailing list