Jbut not on flathub unfortunately
KInteresting
KDoes look very nice
MGitHub
Thank you for contributing to the Universal Blue project!
Please read the Contributor's Guide before submitting a pull request.
Please read the Contributor's Guide before submitting a pull request.
Jfixed the whitespace
M@j0rge @1/4 Life
Mdamn web editor
Jsent a suggestion
JI only use the web editor lol
Mfixed
Hweb editor is for emergencies for me
K@j0rge Sir Castro?
I've noticed, that the Goku, Vegeta and Bob emojis look a bit weird with the background... I edited them on ublue image, with gimp
I've noticed, that the Goku, Vegeta and Bob emojis look a bit weird with the background... I edited them on ublue image, with gimp
K:happy_green_trees::goku::vegeta:
K
K
K
Jfixed, thanks!
Hthis is still my favorite emoji of all the servers im in
Gthoughts on reimplementing ublue-update in golang?
Bas a lover of Go, i'd ask - what's the intended purpose of rewriting? Is there anything wrong with the current ublue-update?
LI personally have had issues where it doesn't seem to run the updates. The logs don't indicate anything out of the ordinary and it just exits with non-zero code
LLI've had at least 2-3 other people in the BlueBuild server have the same issue
LSo I don't know if there's some hardcoded check for a Ublue specific image name or something, but it's been like this for a while for me. I just have to run
rpm-ostree upgradeAdded my comment to https://github.com/ublue-os/ublue-update/issues/113
GitHub
Checking the status it reports a failure exit code, though all of the checks appear to succeed: Brian@fedora ~> systemctl status ublue-update.service × ublue-update.service - Universal Blue Upda...
LLooking into it, seems that it might have something to do with the lock file
Gwell yeah there's just a whole bunch of issues with the project, it still requires unit tests, logging is bad, error handling is pretty bad, it feels pretty brittle. There are a couple things we could do to add tests, vendor topgrade, and make things better but it's still kinda ehhh for me. With go everything would be super simple and portable
Gyeah I'm making a CI PR right now but feel free to investigate that and possibly add more logging
LYeah I think we could add a try catch on the function call so that we can log the error that caused the exception so that it would be clearer as to what caused the program to fail
LI'll put something out soon
Galr
Goh man try catch
Ggotta love it haha
LI don't miss it lol
LRust has spoiled me
Gme neither
Gerrors as values is great
LSuch a better paradigm
Gbut work with what we have rn, I'll see about rewriting it
MOne feature request I would then have is for brew updates to work via it.
Went the systemd timer/service route for brew due to the whole needing a special path and to run as a single designated user
Went the systemd timer/service route for brew due to the whole needing a special path and to run as a single designated user
Gyeah handling multiuser is hard
Gthinking of switching to systemd-run
Gright now there's a whole bunch of environment variable hacks + sudo
MSystemd-run would probably be better and already available.
Bif you're doing most of the heavy lifting with the update service and you think Go is the Way, I'm on board to help, support, contribute as needed.
GThanks! Yeah one thing I'm on the fence about is how we would vendor topgrade, as installing it with pip is pretty jank
BGitHub
Upgrade all the things. Contribute to topgrade-rs/topgrade development by creating an account on GitHub.
Bso my take on "vendoring" is ublue-update lists it as a dependency and the distro (for whatever definition of distro we use) is responsible for having it installed.
LIf you mean bundling it with , I think you would have to change the license to be compatible cause it's GPL-3.0