[Bug]: Blank screen after startup

[justinsmid]

Describe the bug

When I try to open the application, the screen stays blank:

cinnamon-2026-03-04T125425+0100.webm

Note that I initially tried installing through snap, and ran into the same issues as #45.
There, I found the suggestion to install through downloading the .deb from releases.

I did that, and that seems to fix the errors, but as you can see, I now just get a blank screen.
It does seem to be rendering something, because I can click things like a github link, but I can't see anything.

To Reproduce

1. Run tabularis

OS Version

Linux Mint 21.3 (Virginia), Linux version 5.15.0-171-generic (buildd@lcy02-amd64-064)

Tabularis Version

v0.9.4

Relevant Log Output

$ tabularis --debug
[Logger] Capturing logger initialized successfully with level: Info
[LOG] [2026-03-04 12:59:46.710] [INFO] tabularis_lib - Tabularis application starting...
[LOG] [2026-03-04 12:59:46.710] [INFO] tabularis_lib - Debug mode enabled - verbose logging active
[LOG] [2026-03-04 12:59:46.851] [INFO] tabularis_lib::drivers::registry - Registering driver: MySQL (mysql)
[LOG] [2026-03-04 12:59:46.851] [INFO] tabularis_lib::drivers::registry - Registering driver: PostgreSQL (postgres)
[LOG] [2026-03-04 12:59:46.851] [INFO] tabularis_lib::drivers::registry - Registering driver: SQLite (sqlite)
[LOG] [2026-03-04 12:59:46.852] [INFO] tabularis_lib - DevTools opened (debug mode active)

[debba]

Hi!
Interesting, but hard to debug.
I cannot see significant logs.
Are you able to run via code?

For example using npm run tauri dev ?

[igoralentyev]

Same thing with appimage on arch, but AUR pack starts normally (queries not works but app launches)

[debba]

Hi @justinsmid @igoralentyev
I was able to reproduce the issue in Appimage:

Fatal Python error: Failed to import encodings module
Python runtime state: core initialized
ModuleNotFoundError: No module named 'encodings'

Current thread 0x00007fc5b22e3f40 [python3] (most recent call first):

[LOG] [2026-03-04 14:04:56.805] [ERROR] tabularis_lib::plugins::driver - Plugin process exited unexpectedly
Could not create default EGL display: EGL_BAD_PARAMETER. Aborting...

Will check now how can I fix it

[debba]

@igoralentyev @justinsmid
It looks a common issue related to Wayland and Tauri.
Tried launching as it and it works:

LD_PRELOAD=/usr/lib/libwayland-client.so ./tabularis_0.9.4_amd64.AppImage

Can you confirm?
Just for having a path to follow for fixing the issue.

[justinsmid]

@debba

Running LD_PRELOAD=/usr/lib/libwayland-client.so ./tabularis_0.9.4_amd64.AppImage exactly as-is gives me an error about not being able to find /usr/lib/libwayland-client.so

find /usr/lib -name "libwayland-client.so*" 2>/dev/null tells me it is actually at /usr/lib/x86_64-linux-gnu/libwayland-client.so.0 for me.

Running LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libwayland-client.so.0 ./tabularis_0.9.4_amd64.AppImage gives me the following error:

$ LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libwayland-client.so.0 ./tabularis_0.9.4_amd64.AppImage
fuse: memory allocation failed
squashfuse 0.5.2 (c) 2012 Dave Vasilevsky

Usage: squashfuse [options] ARCHIVE MOUNTPOINT

(null) options:
    -o offset=N            offset N bytes into ARCHIVE to mount
    -o subdir=PATH         mount subdirectory PATH of ARCHIVE
    -o notify_pipe=PATH    named pipe that will receive 's' (success)
                           or 'f' (failure) when the mountpoint is ready
    -o timeout=N           idle N seconds for automatic unmount
    -o uid=N               set file owner to uid N
    -o gid=N               set file group to gid N

FUSE options:
    -h   --help            print help
    -V   --version         print version
    -d   -o debug          enable debug output (implies -f)
    -f                     foreground operation
    -s                     disable multi-threaded operation
    -o clone_fd            use separate fuse device fd for each thread
                           (may improve performance)
    -o max_idle_threads    the maximum number of idle worker threads
                           allowed (default: -1)
    -o max_threads         the maximum number of worker threads
                           allowed (default: 10)
fuse: memory allocation failed
squashfuse 0.5.2 (c) 2012 Dave Vasilevsky

Usage: squashfuse [options] ARCHIVE MOUNTPOINT

(null) options:
    -o offset=N            offset N bytes into ARCHIVE to mount
    -o subdir=PATH         mount subdirectory PATH of ARCHIVE
    -o notify_pipe=PATH    named pipe that will receive 's' (success)
                           or 'f' (failure) when the mountpoint is ready
    -o timeout=N           idle N seconds for automatic unmount
    -o uid=N               set file owner to uid N
    -o gid=N               set file group to gid N

FUSE options:
    -h   --help            print help
    -V   --version         print version
    -d   -o debug          enable debug output (implies -f)
    -f                     foreground operation
    -s                     disable multi-threaded operation
    -o clone_fd            use separate fuse device fd for each thread
                           (may improve performance)
    -o max_idle_threads    the maximum number of idle worker threads
                           allowed (default: -1)
    -o max_threads         the maximum number of worker threads
                           allowed (default: 10)
squashfuse 0.5.2 (c) 2012 Dave Vasilevsky

Usage: squashfuse [options] ARCHIVE MOUNTPOINT

(null) options:
    -o offset=N            offset N bytes into ARCHIVE to mount
    -o subdir=PATH         mount subdirectory PATH of ARCHIVE
    -o notify_pipe=PATH    named pipe that will receive 's' (success)
                           or 'f' (failure) when the mountpoint is ready
    -o timeout=N           idle N seconds for automatic unmount
    -o uid=N               set file owner to uid N
    -o gid=N               set file group to gid N

FUSE options:
    -h   --help            print help
    -V   --version         print version
    -d   -o debug          enable debug output (implies -f)
    -f                     foreground operation
    -s                     disable multi-threaded operation
    -o clone_fd            use separate fuse device fd for each thread
                           (may improve performance)
    -o max_idle_threads    the maximum number of idle worker threads
                           allowed (default: -1)
    -o max_threads         the maximum number of worker threads
                           allowed (default: 10)
Can't open squashfs image: Bad address

Cannot mount AppImage, please check your FUSE setup.
You might still be able to extract the contents of this AppImage 
if you run it with the --appimage-extract option. 
See https://github.com/AppImage/AppImageKit/wiki/FUSE 
for more information
execv error: No such file or directory

Some brief googling tells me extracting the appImage first might help, but no luck there either:

$ ./tabularis_0.9.4_amd64.AppImage --appimage-extract
This doesn't look like a squashfs image.
Failed to open squashfs image

Any idea what else I might need to do in order to get the appImage running?
Mind you, I have next to no idea what I'm doing here with appImages. I might just be doing something wrong 😅

Are you able to run via code?
For example using npm run tauri dev ?

I will give this a go later today.

[debba]

@justinsmid
appimage-extract works to me
I am investingating ...

[ScottMorris]

Checkout this PR within Tauri tauri-apps/tauri#12491, it impacts app images and blank screen starts.

[debba]

Thanks @ScottMorris will check it

[justinsmid]

Just tried running from source using npm run tauri dev and I get the same blank screen doing it that way.

[debba]

Tried the method suggested by @ScottMorris , but no luck also in v0.9.5 .
I modified here:
https://github.com/debba/tabularis/blob/5dda28920add5e4e61c9e979b1915286cedcb0b9/.github/workflows/release.yml#L54C1-L56C117

[ScottMorris]

Did you add the env var? I don't see it in your workflow file.

TAURI_BUNDLER_NEW_APPIMAGE_FORMAT="true

[debba]

@ScottMorris

Here:

tabularis/.github/workflows/release.yml

Line 66 in 5dda289

TAURI_BUNDLER_NEW_APPIMAGE_FORMAT: ${{ runner.os == 'Linux' && 'true' || '' }}

i have a doubt here:

https://github.com/debba/tabularis/blob/5dda28920add5e4e61c9e979b1915286cedcb0b9/.github/workflows/release.yml#L63C2-L76C35

Because I tried what suggested here in my local machine: tauri-apps/tauri#12491 (comment)

So using

cargo build

instead of

npm run tauri build

appimage building looks different, but at the end I had:

(tabularis:80578): Gtk-WARNING **: 21:25:53.740: Could not load a pixbuf from icon theme.
This may indicate that pixbuf loaders or the mime database could not be found.
**
Gtk:ERROR:../gtk/gtk/gtkiconhelper.c:495:ensure_surface_for_gicon: assertion failed (error == NULL): Failed to load /usr/share/icons/Adwaita/scalable/status/image-missing.svg: Loader process exited early with status '1'Command:
 "bwrap" "--unshare-all" "--die-with-parent" "--chdir" "/" "--ro-bind" "/usr" "/usr" "--dev" "/dev" "--ro-bind-try" "/etc/ld.so.cache" "/etc/ld.so.cache" "--ro-bind-try" "/nix/store" "/nix/store" "--tmpfs" "/tmp-home" "--tmpfs" "/tmp-run" "--clearenv" "--setenv" "HOME" "/tmp-home" "--setenv" "XDG_RUNTIME_DIR" "/tmp-run" "--setenv" "XDG_RUNTIME_DIR" "/run/user/1000" "--symlink" "/usr/lib" "/lib" "--symlink" "/usr/lib" "/lib64" "--ro-bind-try" "/etc/fonts/conf.d" "/etc/fonts/conf.d" "--ro-bind-try" "/etc/fonts/fonts.conf" "/etc/fonts/fonts.conf" "--ro-bind-try" "/home/debba/.cache/fontconfig" "/home/debba/.cache/fontconfig" "--ro-bind-try" "/home/debba/.fonts" "/home/debba/.fonts" "--ro-bind-try" "/home/debba/.local/share/fonts" "/home/debba/.local/share/fonts" "--ro-bind-try" "/var/cache/fontconfig" "/var/cache/fontconfig" "--bind-try" "/home/debba/.cache/glycin/" "/home/debba/.cache/glycin/" "--setenv" "XDG_CACHE_HOME" "/home/debba/.cache/glycin/" "--seccomp" "23" "glycin-svg" "--dbus-fd" "22" (gdk-pixbuf-error-quark, 0)
Bail out! Gtk:ERROR:../gtk/gtk/gtkiconhelper.c:495:ensure_surface_for_gicon: assertion failed (error == NULL): Failed to load /usr/share/icons/Adwaita/scalable/status/image-missing.svg: Loader process exited early with status '1'Command:  "bwrap" "--unshare-all" "--die-with-parent" "--chdir" "/" "--ro-bind" "/usr" "/usr" "--dev" "/dev" "--ro-bind-try" "/etc/ld.so.cache" "/etc/ld.so.cache" "--ro-bind-try" "/nix/store" "/nix/store" "--tmpfs" "/tmp-home" "--tmpfs" "/tmp-run" "--clearenv" "--setenv" "HOME" "/tmp-home" "--setenv" "XDG_RUNTIME_DIR" "/tmp-run" "--setenv" "XDG_RUNTIME_DIR" "/run/user/1000" "--symlink" "/usr/lib" "/lib" "--symlink" "/usr/lib" "/lib64" "--ro-bind-try" "/etc/fonts/conf.d" "/etc/fonts/conf.d" "--ro-bind-try" "/etc/fonts/fonts.conf" "/etc/fonts/fonts.conf" "--ro-bind-try" "/home/debba/.cache/fontconfig" "/home/debba/.cache/fontconfig" "--ro-bind-try" "/home/debba/.fonts" "/home/debba/.fonts" "--ro-bind-try" "/home/debba/.local/share/fonts" "/home/debba/.local/share/fonts" "--ro-bind-try" "/var/cache/fontconfig" "/var/cache/fontconfig" "--bind-try" "/home/debba/.cache/glycin/" "/home/debba/.cache/glycin/" "--setenv" "XDG_CACHE_HOME" "/home/debba/.cache/glycin/" "--seccomp" "23" "glycin-svg" "--dbus-fd" "22" (gdk-pixbuf-error-quark, 0)

[ScottMorris]

Sorry for the delay. I was looking through your workflow and comparing it to mine and one thing that I noticed is you're using platform: "ubuntu-22.04". It think the AppImage support requires Ubuntu 24.04, which I noticed I'm using for my runner: ubuntu-24.04. I'm not sure if this would introduce any issues for you but your message looks like it is missing a package or library - which might not be installed or is different between versions?

I looked a bit into this and it seems:

• In Ubuntu 22.04: libgdk-pixbuf2.0-0
• In Ubuntu 24.04: libgdk-pixbuf-2.0-0

These seem to come out of the libgtk-3-dev dependency. In Ubuntu 24.04 it looks like they've transitioned the package dependencies to using more modular packages and use different tools, but this process should still work on Ubuntu 22.04 since libgtk-3-dev is available in both versions and has the dependencies setup at install time.

It sounds like there is a library cache with pixbuf, perhaps try researching that, the command I ran into (although it was surfaced by AI, so I'd research it first before using it) was:

sudo /usr/lib/x86_64-linux-gnu/gdk-pixbuf-2.0/gdk-pixbuf-query-loaders --update-cache

So maybe there is an issue on your system (if you were running the build locally and got the message you saw above)

Sorry for the wall of text. There appear to be a bunch of potentially interdependent nuances in some of the be behind the scenes libs depending on OS version, build environments, and build targets.

Here's my Desktop build job that makes AppImage files for reference if that helps.

https://github.com/liminal-hq/threshold/blob/3ed6d5ab33790ddc802d087562e01734e846c8a0/.github/workflows/release-build.yml#L149