GWdym?
GOh, you want to use sed?
Jthe last part of the command is a rebase, so we need to have them fill that out
Jno it's a placeholder for the human
KColor it red, pray they read, become disappointed,????, depression
Jyeah if we leave it as is people will just paste it in and get bazzite ... which, I agree I would have done the same thing
Jbut like, shouldn't we make them pick?
Kyeah, I'm more saying like
JActaully, I'll say "just replace bazzite with whatever image you want, scroll down for the list"
K
IMAGE_NAME_HERE :latestJYES.
JYES
Jhow do I do that in markdown
Jlol
Ksame as discord
K:rotating_light:
Joh wait
Jok I had to add that last part.
EJust read up on a way to do it but it's painful. A precommit hook that uses git to pull the repo name and appends it to the README each time is the closest I've found
Joh I should add once you are on our image you don't need that again
Jok now it's ready!
JGitHub
The upstream documentation is so scattered, but we don't need full docs. Rolling back, rebasing, pinning, need to be documented with the oci method, we could put a little nvidia section in ther...
Jas I struggled to find the rollback, pin, etc. docs upstream when I eventually find them it's like all ostree-native, so I figured this would be an easy one for someone to bust out if they're looking for an easy thing that users need
XI personally find it more legible and understandable if each command is on their own line instead of a huge one liner with
&&%{Arch} should be all caps for consistency
PI prefer if needed
PThat way I can see each command in its own line, be able to copy-paste, and make sure the second command doesn’t happen if the first fails
B@89luca89 i cant seem to get nvidia running in distrobox either
JIs this on the fedora repos version or directly running off git
JI'm able to utilize my nvidia gfx card on the repo version
JI will confirm that later but pretty sure 
BIm running those from fedora repos
JFunny
BI have run multiple programs that show whether nvidia is running(eg. Nvidia-smi) including actual cuda enabled programs and all of them either report no nvidia driver or in case of cuda enabled programs dont seem to be able to run with cuda
JJBI get an error that nvidia-smi was unable to establish communication with the driver
Jis that only on bazzite
Jbecause im running sb-nvidia directly off the ublue builds and it runs just fine
BNo i see that on my own image which is based on bluefin(which is based on ublue's silverblue-nvidia)
JHmm no idea maybe rhe logs when stopping and opening a box will provide more info
J@Bigpod I updated to the latest image with kernel 6.4.4 and retried again, no such issues
Kstill an ubuntu image?
Kif so try making a new arch image
Kdistrobox create --name archtest --image archlinux:latest --nvidiaJtested ubuntu and fedora
Kout of 5 people I've had test this you're the first person with any level of success
Jgoing to see if i can reproduce on arch :>
Ji've been running it since the nvidia flag existed on distrobox never had issues 
Jguess what
