Every interactive prompt takes a var_<app>_<name> override and treats a
closed stdin like Enter (`|| true`), so an install keeps going after the
console that started it is gone. Values the script cannot do without end
with exit 64 and name the variable to set. Validation loops stop asking
when nobody can answer instead of spinning on an empty read.
The Ollama release binaries carry no SYCL build, so the 2.8 GB download
was never used. Its handled failure was also what later systemctl
failures were reported as.
* Open WebUI: stop when the install fails three times
The retry loop fell through to "Installed Open WebUI" after the third
failure, so the service was created without a binary and the URL never
answered (#18020). The last attempt now runs under the error trap, which
shows uv's output.
* Open WebUI: keep the retry inline, no function in the script
* Open WebUI: same retry on the update paths, and drop the numba constraint
The migration and update loops in ct/openwebui.sh fell through to
"Updated" after three failures just like the install did. The numba
constraint from #11874 guarded openai-whisper's old pin; 0.12.x depends
on faster-whisper and never resolves numba.
* formatting
Thirty-six yes/no prompts for optional components in 26 install
scripts could only be answered at the terminal, so an unattended
install stopped at the first one. Each now takes var_<app>_<option>
first and still asks when it is unset. The four conditions that only
matched a single y now take yes as well, since that is what the website
sends.
* Scripts: close every msg_info block with msg_ok
Past-tense msg_info calls that should have been msg_ok, notices that opened a block before a prompt, and blocks without a closing msg_ok. These already left a stale spinner; with core's block stack they would resume it after every later msg_ok.
* Generate passwords with random_password
openssl rand -base64 | tr -dc | head -c returned fewer characters than asked for, and the unfiltered | cut variants put / and + into passwords that end up in DSNs and sed expressions. Secrets an app decodes as base64 are unchanged. Requires community-scripts/core#61.
* Keep data directories through CLEAN_INSTALL instead of copying them
create_backup copied uploads, storage and similar directories twice per update and needed their size again in free space. CLEAN_INSTALL_KEEP moves them aside instead. Only directories the upstream release does not ship, in scripts that restored right after the fetch. Requires community-scripts/core#61.
* Drop the 300s uv timeout overrides
setup_uv exports UV_HTTP_TIMEOUT=600 now; the scripts' 300 only lowered it. Requires community-scripts/core#61.
* openwebui: Ensure required dependencies
Added zstd and build-essential as dependencies for the script.
* Update dependencies in openwebui-install.sh
Added build-essential and libmariadb-dev to dependencies.
Update scripts to use Python 3.12 for uv tool setup and Open-WebUI installs/upgrades. Add a numba constraint (--constraint <(echo "numba>=0.60")) to uv tool install/upgrade commands to ensure compatibility. Changes applied to ct/openwebui.sh and install/openwebui-install.sh for both fresh installs and update paths.
* fix(jellyfin): use setup_hwaccel for robust hardware acceleration
Replaces manual hardware acceleration setup with the centralized
setup_hwaccel function from tools.func.
This fixes the installation failure in privileged containers where
/dev/dri does not exist (e.g., when no GPU is passed through).
The setup_hwaccel function includes:
- Proper error handling for missing /dev/dri
- GPU vendor detection (Intel, AMD, NVIDIA)
- Graceful fallback when no GPU is available
Fixes: Installation fails with 'chgrp: cannot access /dev/dri'
when creating privileged containers without GPU passthrough.
* refactor(hwaccel): standardize hardware acceleration across all install scripts
Migrated all install scripts to use the centralized setup_hwaccel function:
- plex-install.sh
- emby-install.sh
- ersatztv-install.sh
- frigate-install.sh
- tdarr-install.sh
- unmanic-install.sh
- channels-install.sh
- ollama-install.sh
- immich-install.sh (added error handling)
Enhanced setup_hwaccel function in tools.func:
- Added -d /dev/dri check before setting permissions
- Added error handling (2>/dev/null || true) for all /dev/dri operations
- Added adduser error handling for video/render groups
- No longer fails if no GPU is detected (graceful skip)
- Added intel-media-va-driver for newer Intel GPUs
- Improved AMD APU support with firmware packages
- Better NVIDIA handling (warning instead of failure)
This fixes installation failures in privileged containers without GPU
passthrough, where /dev/dri does not exist.
Supports: Ubuntu, Debian 12 (Bookworm), Debian 13 (Trixie)
GPU Support: Intel, AMD, NVIDIA (manual driver)
* refactor(hwaccel): complete migration for all GPU apps
Migrated remaining GPU apps to setup_hwaccel:
- fileflows-install.sh
- openwebui-install.sh (added setup_hwaccel - was missing)
- tunarr-install.sh
Also fixed tools/pve/hw-acceleration.sh:
- Added error handling for /dev/dri operations
- Added chmod 660 /dev/dri/* that was missing
- Added error suppression for adduser commands
All 13 GPU apps (var_gpu=yes) now use centralized setup_hwaccel:
jellyfin, plex, emby, ersatztv, frigate, tdarr, unmanic,
channels, ollama, immich, fileflows, openwebui, tunarr
* feat(hwaccel): complete Intel non-free driver support and GID sync
Enhanced setup_hwaccel function:
- Auto-detect Intel GPU generation (Gen 9+ for non-free drivers)
- Debian 12 (Bookworm): Add non-free repo + intel-media-va-driver-non-free
- Debian 13 (Trixie): Add non-free repo + libvpl2 + mesa-opencl-icd
- Ubuntu: Use ubuntu repos with intel-media-va-driver
- Fallback to open drivers if non-free fails
- GID sync for video/render groups (moved from install scripts)
OpenWebUI: Added Intel oneAPI support when installing Ollama
- Intel Level Zero GPU support
- Intel oneAPI Base Toolkit
- Same setup as standalone Ollama install
Cleanup:
- Removed duplicate GID sync from tdarr-install.sh
- Removed duplicate GID sync from unmanic-install.sh
* fix(ersatztv): remove duplicate HW acceleration code
Removed manual Intel HW acceleration setup that remained after
setup_hwaccel migration. The non-free driver prompt is no longer
needed as setup_hwaccel auto-detects Intel GPU generation.
* Migrate Open WebUI to uv-based installation
Added migration logic to detect and upgrade legacy Open WebUI installations to a uv-based setup, including data backup and service recreation. Updated install and update scripts to use uv for installing Open WebUI and adjusted service configuration and cleanup steps accordingly.
* english
---------
Co-authored-by: Tobias <96661824+CrazyWolf13@users.noreply.github.com>
MickLesk