Just to be more pedantic and probably bit more secure, what about making
networking in copr opt-in (at least for new projects)?
If build of my package used internet arbitrarily, I would be warned very
Also, in Koji we also do not have networking, so we can faster teach
(future) Fedora packagers.
Thanks for considering,
First of all I want to say I'm really happy for having Copr, being an
Arch user for many, many years I got really used to the AUR (Arch User
Repository) so I can see Copr potential.
And speaking about that, I would like to make some suggestions to
improve the 'community' side of Copr as I believe from what I read the
technical side is already very well covered:
1. Copr needs a place where users and packager(s) can discuss stuff
about the packages and repositories, that could be achieved by simply
letting logged people to comment on each Copr repo page;
2. Likewise it would be nice to have a way to: a) contact the
packager(s)/maintainer(s), b) flag outdated packages, inform about any
Having the chance for community to interact ultimately leads to
improved workflows and better quality projects; some folks will look
for guidance to create their first Copr repo, other people may want to
lend a hand and offer maintain a repo packed with the stuff they
usually use while others can just offer tips for improve things - or
ask for assistance as well!
Hope you guys take my suggestions in a positive way, I'm already happy
to have a great infrastructure like Copr already set and ready for
anyone to use, kudos!
While trying to create a COPR build fetching SRPMs from a github
release, COPR for some reason fails to fetch the file.
I am wondering if somebody can test this and let me know if I'm doing
something wrong here or this is in fact a COPR bug. You can give this a
go with this SRPM URL:
Alberto RuizRed Hat, Inc
I just enabled new architecture in Copr - PowerPC 64 LE (little endian).
There are those chroots available:
* fedora-21 ppc64le
This happened as co-operation of Red Hat, Brno University of Technology
We have 4 VM available on a physical machine, which is physically
located at Brno University of Technology.
This is not ideal solution - as we have to copy the bits over the ocean
- but it works. I hope that this is a temporary solution and one day we
will get an IBM machine into PHX. However that can last several months,
while this works now.
Those VM have little bit more resources compared to Intel: 8 VCPU and 8
GB RAM (comp. to 3VCPU and 5GB RAM). Storage is the same - 50 GB per
I would like to thanks D, Horak and J. Capik from Secondary Archtectures
for cooperation and making this happen.
Enjoy and happy hacking
I had a failed build and went to inspect the logs, when I clicked on the
link for the text log Chrome wouldn't display it, only download it. When I
tried to open the file everything was garbled.
On a whim I tried opening the .gz log file and it opened as plain text. I'm
guessing the files are getting swapped somehow.