AppImage not running on Fedora 41
What happened?
I downloaded the AppImage from the Website, gave it permission to be executed and ran it. It showed only a white screen, while when after closing it, it doesn't open again. I made sure all dependencies are installed as described on the installation guide.
What did you expect to happen instead?
The UI should show up.
How can we reproduce the bug? (as minimally and precisely as possible)
Run Fedora 41, try to open the AppImage.
My devcontainer.json:
Doesn't exist yet.
Local Environment:
DevPod Provider:
None yet.
Here is the AppImage console output from Fedora 41
Gtk-Message: 19:37:30.173: Failed to load module "appmenu-gtk-module": 'gtk_module_display_init': /usr/lib64/gtk-3.0/modules/libwindow-decorations-gtk-module.so: undefined symbol: gtk_module_display_init
/usr/bin/xdg-mime: line 885: qtpaths: command not found
Gtk-Message: 19:37:31.110: Failed to load module "appmenu-gtk-module": 'gtk_module_display_init': /usr/lib64/gtk-3.0/modules/libwindow-decorations-gtk-module.so: undefined symbol: gtk_module_display_init
**Could not create default EGL display: EGL_BAD_PARAMETER. Aborting...**
I do have this error as well with fedora 41. With some research i found out that a lot of other apps have the same problems. like :
GitButler : https://github.com/gitbutlerapp/gitbutler/issues/5282
It looks like the issue come from webkit itself : https://www2.webkit.org/show_bug.cgi?id=280239
downgrade dependency look like the temporary solution.
I'm using Arch and had to downgrade to:
mesa-1:24.1.0-1llvm-libs:17.0.6-2
I'm using Arch and had to downgrade to:
* `mesa-1:24.1.0-1` * `llvm-libs:17.0.6-2`
Thank you for your reply.
We were facing the same issue on Fedora 40 and downgrading mesa (sudo dnf downgrade mesa-libEGL to revert back to version 24.0.5-1.fc40.x86_64) allowed us to start devpod once again until a better solution can be found.
Thanks for your help, unfortunately mesa-libEGL-0:24.2.4-1 is the oldest version available in the fedora 41 repos. I'm still hoping for a fix soon, but at the moment thankfully don't really need it.
This also applies to Fedora 40 as well:
Error loading the injected bundle (././/lib/x86_64-linux-gnu/webkit2gtk-4.1/injected-bundle/libwebkit2gtkinjectedbundle.so): ././/lib/x86_64-linux-gnu/webkit2gtk-4.1/injected-bundle/libwebkit2gtkinjectedbundle.so: cannot open shared object file: No such file or directory
Could not create default EGL display: EGL_BAD_PARAMETER. Aborting...
$ rpm -qa|grep EGL
mesa-libEGL-24.1.7-1.fc40.x86_64
mesa-libEGL-devel-24.1.7-1.fc40.x86_64
mesa-libEGL-24.1.7-1.fc40.i686
I'm getting this on Fedora 41 KDE:
/usr/bin/xdg-mime: line 885: qtpaths: command not found
** (WebKitWebProcess:93347): WARNING **: 11:48:57.901: Error loading the injected bundle (././/lib/x86_64-linux-gnu/webkit2gtk-4.1/injected-bundle/libwebkit2gtkinjectedbundle.so): ././/lib/x86_64-linux-gnu/webkit2gtk-4.1/injected-bundle/libwebkit2gtkinjectedbundle.so: cannot open shared object file: No such file or directory
[/home/runner/.cargo/registry/src/index.crates.io-6f17d22bba15001f/tauri-plugin-deep-link-0.1.2/src/linux.rs:104:29] buffer = "version"
thread '<unnamed>' panicked at src/workspaces.rs:175:72:
called `Result::unwrap()` on an `Err` value: SendError { .. }
stack backtrace:
0: 0x558900aac0da - <std::sys::backtrace::BacktraceLock::print::DisplayBacktrace as core::fmt::Display>::fmt::ha4a311b32f6b4ad8
1: 0x558900adb8b3 - core::fmt::write::h1866771663f62b81
2: 0x558900aa7063 - std::io::Write::write_fmt::hb549e7444823135e
3: 0x558900aabf22 - std::sys::backtrace::BacktraceLock::print::hddd3a9918ce29aa7
4: 0x558900aad1dc - std::panicking::default_hook::{{closure}}::h791f75256b902d7d
5: 0x558900aad022 - std::panicking::default_hook::h82cc572fcb0d8cd7
6: 0x558900aad7b7 - std::panicking::rust_panic_with_hook::he21644cc2707f2c4
7: 0x558900aad64a - std::panicking::begin_panic_handler::{{closure}}::h42f7c414fed3cad9
8: 0x558900aac5b9 - std::sys::backtrace::__rust_end_short_backtrace::ha26cf5766b4e8c65
9: 0x558900aad2dc - rust_begin_unwind
10: 0x5588ffb29f30 - core::panicking::panic_fmt::h74866b78e934b1c0
11: 0x5588ffb2a376 - core::result::unwrap_failed::h899ed7ab2ccb8159
12: 0x5588ffeecf5f - std::sys::backtrace::__rust_begin_short_backtrace::h65aab9bfb1572dfe
13: 0x5588ffc7ca74 - core::ops::function::FnOnce::call_once{{vtable.shim}}::h516ceddbb71339f7
14: 0x558900ab3dbb - std::sys::pal::unix::thread::Thread::new::thread_start::h14f1eb868ff90fc9
15: 0x7f2390f98168 - start_thread
16: 0x7f239101c14c - __clone3
17: 0x0 - <unknown>
Just FYI, as a workaround until this gets fixed, I've been able to start DevPod on Fedora by using the Targz instead of the AppImage, after a sudo dnf install libappindicator-gtk3 to fix this:
thread 'main' panicked at /home/runner/.cargo/registry/src/index.crates.io-6f17d22bba15001f/libappindicator-sys-0.9.0/src/lib.rs:41:5:
Failed to load ayatana-appindicator3 or appindicator3 dynamic library
libayatana-appindicator3.so.1: cannot open shared object file: No such file or directory
libappindicator3.so.1: cannot open shared object file: No such file or directory
libayatana-appindicator3.so: cannot open shared object file: No such file or directory
libappindicator3.so: cannot open shared object file: No such file or directory
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Same error happens on Ubuntu (pop os). The workaround proposed by @vorburger worked for me too.
This issue is stale because it has been open for 60 days with no activity.
not stale
This issue is stale because it has been open for 60 days with no activity.
still not stale
also fails on fedora 42
Also, why does the app not crash when it fails to create the main window? Oh, I know why: because ignoring errors is the default behaviour in Go 🙄
Still having same issue on Fedora 43
version v0.7.0-alpha.34 on Arch fails with the same issue
Gtk-Message: 13:59:02.715: Failed to load module "appmenu-gtk-module"
Could not create default EGL display: EGL_BAD_PARAMETER. Aborting...
Works when using "tar.gz" way with prebuild executables