I have written this CLI application which lets you install NVIDIA drivers and much more easily from the RPM Fusion repository. I wish to package this in order to be able to help people with this who find trouble going through the installation process.
I recently moved into Arch Linux briefly in my secondary laptop and want make the build happen from there. I doubt if it would be possible to make COPR builds on Arch so is it possible to make builds entirely from CLI of a Fedora virtual machine installed on it?
Exactly, that is the neat part. If functions as a repository, just as any other Fedora repo on your system.
You mean control the copr built processes from CLI of your local (virtual) machine? If yes, then https://developer.fedoraproject.org/deployment/copr/copr-cli.html
Docs maybe of interest as well.
Some more questions though but I’d rather PM them to you than to derail this thread.
Cool. That was exactly what I was looking for. Thanks!
I think you can continue using the thread. In general, for copr, try #fedora-buildsys on IRC. Always best to get answers from the primary source
COPR provides lots of ways of building in a “continuous integration” type pipeline. Best to go through the docs once:
The question originated from the NVIDIA Auto Installer tool thread so @florian asked me to create this thread instead of PMing him with more questions as the discussion would eventually help other readers too.
Cool. I will look into it.
I was thinking of keeping off from automated builds - so is “continuous integration” really something that must be delved into to get through this?
Yes, remember that we “default to open”. So private conversations are discouraged—all discussion, as far as possible, should happen in the open, public channels.
Nope, check the docs. You can upload an SRPM each time if you wish. Totally up to you.
The SPEC file I believe. Cool, I’d get to CI when it is absolutely needed to. There’s already much on the plate anyway.