the move from 20.04 to 24.04 seems to be breaking a lot of things
#308
Comments
This was referenced Feb 15, 2025
|
update: 827d1ab did not fix the issue for the person who reported the for anyone with a similar problem that finds this issue report: pinning to https://github.com/linuxdeploy/linuxdeploy/releases/tag/1-alpha-20240109-1 seems to be a valid workaround |
The release binaries did not get built for some reason and I don't want to break the Git history to work around that.
I don't see how the change of the base OS (remember: the binaries are built in a Docker container) could have any influence on the produced binaries. |
|
Please provide some build logs, links to issues, and stop guessing yourself. |
|
changing the base image back to the steam deck crash issue is here HarbourMasters/Shipwright#5068 i had the user test out a build from HarbourMasters/Shipwright@adfc68a (this workflow run https://github.com/HarbourMasters/Shipwright/actions/runs/13342732636) that was using a release created on my fork after i pushed 827d1ab the part of the logs from that build that involve and the latest build logs (from https://github.com/HarbourMasters/Shipwright/actions/runs/13342345010/job/37269343662 as of writing) using upstream continuous are here: |
|
as for
if the kernel has an impact on the produced binaries then the base OS could have an impact (because docker containers use the host kernel) |
Very likely they have appimagelauncher installed, which will show that error. The stable releases of appimagelauncher don't support the static appimage runtime, which is what was changed here. The user would need to either remove it or update to the alpha builds
This is odd, does that person get the crash if they run the extracted appimage instead? |
that makes sense, but telling people to uninstall appimagelauncher or update it to alpha just to play a game isn't a viable solution for HarbourMasters projects. pinning to https://github.com/linuxdeploy/linuxdeploy/releases/tag/1-alpha-20240109-1 is
i'm not sure, but i'm hesitant to ask them to test that as i've already asked quite a lot of them. i'll see if i can do some investigation on my own and report back |
Note that by doing that you will stuck with the dynamic runtime, which has an EOL |
i tested this and was able to reproduce the crash very quickly using an appimage generated using latest |
In that case it is likely a bug in the type2-runtime |
|
Yep, sounds like it. Could you share that exact AppImage and information on the system it's crashing on so that we could try to reproduce the issue? If the runtime is to be blamed, an issue over there is needed. |
|
the AppImage i tested was from this actions run https://github.com/HarbourMasters/Shipwright/actions/runs/13348583888 i tested it on a steam deck running the latest (as of writing) stable version of SteamOS
with the AppImage added as a "non-steam game" - running in "Gaming Mode" (as opposed to running from the KDE desktop) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
https://github.com/HarbourMasters uses
linuxdeployto package appimages, and we've always used thecontinuousrelease. as of c28054b multiple problems have popped up:execv error: No such file or directoryi forked the repo, pushed 827d1ab to my fork, and kicked off a ci build for the person getting the steam deck crashes to test. they were unable to reproduce the crashes with that change.
i know that github is retiring the
20.04runners, so reverting to those doesn't seem like a reasonable solution.so i'd like to propose what feels like a reasonable workaround to me: would it be possible to tag/create a release for the most recent version before c28054b?
edit: the name on https://github.com/linuxdeploy/linuxdeploy/releases/tag/1-alpha-20250213-1 threw me off, i see that is likely exactly what i'm looking for
edit 2: i now see that there's a tag for https://github.com/linuxdeploy/linuxdeploy/releases/tag/1-alpha-20250213-1 but not a release. if a release could be created that would be huge!
The text was updated successfully, but these errors were encountered: