----- Original Message -----
Hi,
Discussing with Seth today, he was telling me that the current idea
would be to kind of loop through a list of packages to find the one
requested by the user.
While I was actually more thinking of something like
--enablerepo=copr-<repo_name>
Even if we do a
--enablerepo=copr-all
which would eventually still do the loop up through all packages in
all
repo.
I think the copr-<repo_name> has several advantages though:
- no extra infrastructure needed
- faster (only one call)
- allow multiple copr to have the same rpm (with eventually different
option in the build or tweaks or <insert here your idea if you follow
mine>)
Just my 2cts on the subject :)
Pierre
I had a bit different idea:
All the coprs will have the same URL, but the owner's name and copr name will be
variable. That gives us an option to create a single repofile with one URL like
baseurl=http://copr-be.cloud.fedoraproject.org/$username/$coprname
We would then have a plugin that would provide functionality like '--enablecopr
bkabrda/foo', that would finalize the above URL and would be able to install packages
from that repo.
(Please note that to identify the copr, you need both owner's name and copr name; more
users can have equally named coprs.)
--
Regards,
Bohuslav "Slavek" Kabrda.