Add the localization/ directory to the Flatpak source list so
run_gettext.sh can compile .po files into .mo translations.
Replace LC_ALL=C.UTF-8 with LC_NUMERIC=C in the entrypoint script
to preserve the user's language settings while still preventing
decimal-separator parsing issues.
# Description
This pull request fixes the bug where Orca Slicer enters an infinite
loop when a density of zero is introduced for the top layer solid infill
# Screenshots/Recordings/Graphs
<img width="2560" height="1392" alt="image"
src="https://github.com/user-attachments/assets/5bf5c220-a398-483f-9f69-63dadbc888af"
/>
Note: For bottom solid infill, this is not a problem because the minimum
density is 10%.
Makes sure all non-bbl printers go through the same filament reordering
code so custom filament sequence is properly handled, while skipping the
filament grouping part by mapping all filaments to group 0.
This is the same method used by BBL printer to calculate the optimal
filament grouping so should be robust.
Fix#12449Fix#12766
# Description
<!--
> Please provide a summary of the changes made in this PR. Include
details such as:
> * What issue does this PR address or fix?
> * What new features or enhancements does this PR introduce?
> * Are there any breaking changes or dependencies that need to be
considered?
-->
Happy Hare now writes mmu lane data to the moonraker db similar to AFC,
so we can normalize to the one implementation. For now, i've just
re-ordered it so that the moonraker db is checked first. I want this
because I've actually updated the happy hare data to include more data
than we had access to via the "mmu" object. For happy hare we now have
access to vendor name, which we can eventually use to further fine tune
the auto-matched filament preset.
# Screenshots/Recordings/Graphs
<!--
> Please attach relevant screenshots to showcase the UI changes.
> Please attach images that can help explain the changes.
-->
_n/a_
## Tests
<!--
> Please describe the tests that you have conducted to verify the
changes made in this PR.
-->
# Description
Qidi printers run Klipper and we don't want arc fitting enabled for
these.
This is wrongly enabled in QidiStudio as well, likely where these
profiles originated.
With certain irrational beam interlocking parameters, it will take a few
minutes to complete the interlocking generation; even worse:
cancellation doesn't work once it started.
This PR fix that by adding cancellation check inside some of the most
time consuming loops so it can be stopped ASAP.
Also cherry picked a sanity check from
https://github.com/bambulab/BambuStudio/pull/9925 by @mpaperno. Thanks!
The following test project (which was modified from the upper PR by
@mpaperno) took me more than 90s to complete with 0.1mm beam width:
[Interloack-Test2.orca.3mf.txt](https://github.com/user-attachments/files/25975637/Interloack-Test2.orca.3mf.txt)
# Description
I was just building the project and checked out the CMakeLists to see if
there are some optional features I can disable to speed up compilation
and tripped on this chinese comment.
This just translates the comment to English for consistency
# Description
The wipe tower config position (wipe_tower_x/y) could be outside the plate boundary (e.g. default y=250 on a 200mm printer). No constraint was applied at slice time, so the tower was generated out-of-bounds.
issue reported in #12731
# Screenshots/Recordings/Graphs
<!--
> Please attach relevant screenshots to showcase the UI changes.
> Please attach images that can help explain the changes.
-->
## Tests
<!--
> Please describe the tests that you have conducted to verify the changes made in this PR.
-->
The wipe tower config position (wipe_tower_x/y) could be outside the
plate boundary (e.g. default y=250 on a 200mm printer). No constraint
was applied at slice time, so the tower was generated out-of-bounds.
std::map::at() throws std::out_of_range when option.category or
"Basic information" is not present in the tab's category_icon_map,
crashing the app. Replace with a safe find()-based helper lambda
that falls back to an empty string (no icon) when the key is missing.
* Fix GIT_COMMIT_HASH not set in Flatpak builds
The env var check was gated inside the .git directory check, so
Flatpak builds (which exclude .git from the sandbox) always fell
back to "0000000". Lift the env var check to top level and inject
the commit hash into the Flatpak manifest via build-options.env.
* Switch Flatpak build to Clang/LLD via LLVM 21 SDK extension
- Add org.freedesktop.Sdk.Extension.llvm21
- Bump runtime to GNOME 49 (SDK 25.08) for llvm21 availability
* fix build errors and improving build speed for flatpak
* fxi more build errors
* Update error messages for GNOME Platform and SDK versions
The env var check was gated inside the .git directory check, so
Flatpak builds (which exclude .git from the sandbox) always fell
back to "0000000". Lift the env var check to top level and inject
the commit hash into the Flatpak manifest via build-options.env.
Fix hotkeys blocked in Prepare view when notification is shown
Restore SetFocus() in the GLCanvas3D mouse-entering handler so the
canvas reclaims wxWidget keyboard focus whenever the mouse enters it
(provided the main window is active). Without this, clicking a sidebar
control or a UI update triggered by slicing could leave the canvas
without focus, causing hotkeys like Tab to stop working until the user
clicked the canvas or dismissed the notification.
* optimize resize feature on Linux
* fix the black screen issue on app startup on Linux
* Fixed an issue that Filament Grouping dialog always popup on Linux
* Fix an issue that on Linux the project name or model file name was not displayed in title bar
* optimize resize feature on Linux
* fix the black screen issue on app startup on Linux
* Fixed an issue that Filament Grouping dialog always popup on Linux
The WipeTower2 code path never called construct_mesh(), leaving
wipe_tower_mesh_data as std::nullopt. GCode export then dereferenced
it unconditionally, triggering an assertion failure / crash.
1. Dirty flags not showing for extruder options:
2. Crash when switching to non-first extruder tabs:
3. Modifying one extruder's parameter affects other extruders:
* Linux: prefer discrete GPU on dual-GPU systems via PRIME environment variables
Set DRI_PRIME=1 for AMD/nouveau PRIME setups and __NV_PRIME_RENDER_OFFLOAD/__GLX_VENDOR_LIBRARY_NAME for NVIDIA proprietary driver, only when the NVIDIA kernel module is detected. Uses replace=false to respect user overrides.
* Improves usability on Linux. No more double title bar (#12600)
* Fix double title bar on Linux by removing WM decorations for GTK
* wip
* fix
* finish
* address review comment
## Summary
- Change default bed temperature type from "By First filament" to "By Highest Temp"
- Move `bed_temperature_formula` option from develop mode to advanced mode for better accessibility
- Relocate UI control from "Basic Information - Advanced" to "Multimaterial Setup" section where it's more relevant for multi-filament printing
Using the highest temperature of all printed filaments is generally safer for bed adhesion than using the first filament's temperature, especially in multi-material prints where different filaments may require different bed temperatures.
<img width="390" height="239" alt="Screenshot 2026-02-26 at 18 44 56" src="https://github.com/user-attachments/assets/0fe9f8c7-062a-4a7a-8ab3-c52df7e74b26" />
print_machine_envelope() used get_extruder_id(extruder_id)*2 to index
machine limit arrays that only hold [Normal, Stealth] (2 entries).
For multi-extruder setups this went out-of-bounds, causing wrong M201/M203
values in the G-code which then override the estimator's correct limits.
Same class of bug as c6d1c11ebb but on the G-code writer side.
Changes:
- Remove unused extruder_id param from print_machine_envelope()
- Use .values.front() for M201/M203, matching M204/M205 in same function
- Change get_option_value() fallback from .back() to .front() so any
future out-of-bounds index returns Normal mode instead of Stealth
The extruder_id*2 offset in get_axis_max_feedrate/get_axis_max_acceleration
was cherry-picked from BambuStudio's per-nozzle limit system, which OrcaSlicer
never ported. Without that system the limit arrays only have 2 values
([0]=Normal, [1]=Stealth), so any extruder_id > 0 or the uninitialized
value (255) would overshoot the array and fall back to values.back(),
always returning stealth-mode limits and producing incorrect time estimates.
Revert to indexing by time mode only (matching v2.3.1 behavior) and simplify
the M201/M203 handlers to write only the two mode slots they actually use.
* Reapply "Switch to self hosted mac runner (#12024)" with proper fix
This reverts commit 783f9926e3.
* use conditional logic for self-hosted runners
* improve readbility
# Description
Fixes the Setup Wizard getting stuck on the "Loading......" screen when
running on Wayland sessions.
**Root Cause:** When OrcaSlicer runs under XWayland (Wayland sessions
with `GDK_BACKEND=x11`), WebKit2GTK's compositing mode can fail
silently, causing the Setup Wizard WebView to freeze or render blank.
The JavaScript signal from C++ (`userguide_profile_load_finish`) either
fails to execute or its rendering result never displays, leaving users
permanently stuck.
**Solution:**
1. Set `WEBKIT_DISABLE_COMPOSITING_MODE=1` environment variable to force
software rendering for WebKit2GTK - this is the same fix used by
BambuStudio
2. Enable the 20-second timeout fallback in `load.js` so the wizard
proceeds automatically even if the C++ → JavaScript communication fails
**Files changed:**
- `src/OrcaSlicer.cpp` - Added `WEBKIT_DISABLE_COMPOSITING_MODE` env var
(4 lines)
- `resources/web/guide/0/load.js` - Enabled timeout fallback (2 lines)
# Screenshots/Recordings/Graphs
N/A - This is a Linux/Wayland-specific fix that addresses a rendering
issue. The visual result is simply that the Setup Wizard proceeds past
"Loading......" as expected.
## Tests
1. **On Wayland session (e.g., GNOME on Wayland):**
- Remove existing config (`~/.config/OrcaSlicer/`) to simulate first
launch
- Launch OrcaSlicer
- Verify Setup Wizard proceeds past "Loading......" to the actual wizard
page
2. **On X11 session:**
- Same test as above to verify no regression
3. **Verify env var is set:**
- Check that `WEBKIT_DISABLE_COMPOSITING_MODE=1` is in the process
environment on Linux builds
# Description
This PR lays the groundwork for future OrcaSlicer ↔ printer connectivity
enhancements by modularizing the network agent architecture. It also
adds experimental filament info sync from printers that have a material
station or multi-tool system.
## Architecture Changes
- **Introduce `IPrinterAgent` interface** — an abstraction layer that
allows per-machine printer agent switching at runtime, decoupling
printer communication from the monolithic `NetworkAgent`
- **Introduce `ICloudServiceAgent` interface** — separates cloud service
logic from printer-level communication
- **Extract `OrcaCloudServiceAgent`** — moves Orca cloud service logic
into its own implementation behind `ICloudServiceAgent`
- **Extract `OrcaPrinterAgent`** — wraps the existing BBL printer
communication behind `IPrinterAgent`
- **Add `NetworkAgentFactory`** — factory for creating the appropriate
printer agent per machine
- **Refactor `NetworkAgent`** — slimmed down from monolithic class to a
thinner coordination layer
## New Printer Agents
- **`MoonrakerPrinterAgent`** — Klipper/Moonraker-based printers
- **`QidiPrinterAgent`** — Qidi printers (with Qidi filament box
support)
- **`SnapmakerPrinterAgent`** — Snapmaker printers with filament sync
## Filament Sync (Experimental)
Syncs filament information from printers equipped with AMS-style
material systems or multi-tool changers:
- Qidi printers with Qidi box
- Armored Turtle (AFC) box via Moonraker
- Snapmaker material station
For Qidi printers with Qidi box:
<img width="1200" height="762" alt="Screenshot 2026-01-27 at 20 30 55"
src="https://github.com/user-attachments/assets/155a164f-cd08-40b0-b62b-c3ab7378224e"
/>
Armored Turtle box:
<img width="1135" height="805" alt="Screenshot 2026-01-27 at 20 32 58"
src="https://github.com/user-attachments/assets/50f6618e-eb54-46db-8e01-1197a005fbf0"
/>
# Screenshots/Recordings/Graphs
[filasync.webm](https://github.com/user-attachments/assets/e6bb7f04-8312-4014-b237-6bd3ef792215)
## Tests
<!-- Please describe the tests that you have conducted to verify the
changes made in this PR. -->
When running under XWayland (Wayland sessions with GDK_BACKEND=x11),
WebKit2GTK's compositing mode can fail silently, causing the Setup
Wizard WebView to freeze or render blank. This leaves users stuck on
the "Loading......" screen indefinitely.
Changes:
- Set WEBKIT_DISABLE_COMPOSITING_MODE=1 to force software rendering,
matching the fix used by BambuStudio
- Enable the 20-second timeout fallback in load.js so the wizard
proceeds even if the C++ -> JavaScript signal fails
* Fix active filament preset not matching wizard selection on first run
After completing the setup wizard with only a non-PLA filament selected
(e.g. Generic ABS), the active filament preset defaulted to Generic PLA
instead of the user's selection. This happened because load_presets()
falls back to Generic PLA when the initial printer differs from the
preferred printer, and the first_added_filament override was disabled.
Add post-load correction in apply_config() that switches the active
filament to the first compatible wizard-selected filament when the
current active filament is not in the wizard selection. Also fix the
guide JS (22.js, 23.js) to use the filalist attribute for building
the filament array, and add platform-specific build test commands to
CLAUDE.md.
* Replace @System filaments with vendor-specific overrides in wizard
When the wizard selects a generic filament like "Generic ABS @System",
check if a vendor-specific override exists (e.g. BBL "Generic ABS")
and use that instead. This ensures printer-tuned profiles are preferred,
preventing load_installed_filaments from adding unwanted BBL defaults.
When printers from the default bundle are also selected, both variants
are kept since those printers need the @System version.
Also adds diagnostic logging for filament loading in LoadProfileFamily.
* Guard against persisting presets for Default Printer and fix filament override logic
Prevent export_selections from saving stale preset settings for the
built-in "Default Printer" placeholder, which is only the initial state
before a real printer is loaded. Also require a non-default vendor printer
to be selected before replacing @System filaments with vendor-specific
overrides in the wizard, avoiding incorrect filament substitution when
only the default bundle is present.
SoftFever