github linuxserver/docker-orcaslicer v2.5.2-ls53

one hour ago

CI Report:

https://ci-tests.linuxserver.io/linuxserver/orcaslicer/v2.5.2-ls53/index.html

LinuxServer Changes:

Full Changelog: v2.5.1-ls52...v2.5.2-ls53

Remote Changes:

What's Changed

This is the OrcaSlicer V2.5.2 release, a maintenance update on top of 2.5.1. Thank you to everyone who reported problems with 2.5.1!

It fixes crashes while slicing, STEP files that imported with faces missing, the Linux AppImage freezing on Wayland, and parts close to the size of the bed that wouldn't arrange. On Bambu Lab printers and others with the Type 1 prime tower, you can now set the Prime volume of each filament, for example to make the tower smaller. Building OrcaSlicer from source works again too.

If you print on a Prusa CORE One or a Snapmaker U1, see Printer & Filament Profiles below: CORE One presets slow down for the steepest overhangs, and the U1 checks rod clearance more strictly when printing by object.

Downloads

Platform File
Windows installer (x64) OrcaSlicer_Windows_Installer_V2.5.2_x64.exe
Windows installer (ARM64) OrcaSlicer_Windows_Installer_V2.5.2_arm64.exe
Windows portable (x64) OrcaSlicer_Windows_V2.5.2_x64_portable.zip
Windows portable (ARM64) OrcaSlicer_Windows_V2.5.2_arm64_portable.zip
macOS (Apple Silicon and Intel) OrcaSlicer_Mac_universal_V2.5.2.dmg
Linux AppImage (x86_64) OrcaSlicer_Linux_AppImage_Ubuntu2404_V2.5.2.AppImage
Linux AppImage (aarch64) OrcaSlicer_Linux_AppImage_Ubuntu2404_aarch64_V2.5.2.AppImage
Linux Flatpak (x86_64) OrcaSlicer-Linux-flatpak_V2.5.2_x86_64.flatpak
Linux Flatpak (aarch64) OrcaSlicer-Linux-flatpak_V2.5.2_aarch64.flatpak
Windows debug symbols (for developers) Debug_PDB_V2.5.2_for_developers_only.7z

🌊 New Features & Improvements

  • Set each filament's prime volume for the Type 1 prime tower by @Art-Porto in 17424
    On the Type 1 prime tower, which every Bambu Lab printer uses, the process Prime volume had no effect, because the tower purges each filament's own prime volume, and no settings page showed that value. The filament's Multimaterial page has a new Prime volume group for it. Filament change sets how much the tower purges for a filament change. Hotend change does the same for a hotend change and appears on printers whose extruder holds more than one nozzle, such as the H2C. The group shows only on printers that use the Type 1 tower, and there the process Prime volume is greyed out, with the reason shown when you hover over its label. Projects slice as before until you change a value, and the Type 2 tower is unchanged. Fixes #16350.

    A lower prime volume makes a smaller tower: on an H2D with two filaments, setting Filament change to 10 mm³ instead of the profile's 30 cut the tower from 3.26 m of filament to 1.95 m.

    Prime volume group on the filament's Multimaterial page for an H2C, with Filament change and Hotend change

🛠️ Bug Fixes

Crashes

  • No crash slicing some projects that 2.4 sliced by @HanifKoh in 17599 and 17600
    Since 2.5.0, some projects crashed OrcaSlicer while slicing. A long, nearly straight strip of internal solid infill could get a point placed metres away, and the shape that made crashed the slicer. Those strips slice correctly again. As a safety net, a region the slicer can't build toolpaths for is skipped instead of crashing the whole slice: it gets no walls or infill, and an error goes to the log. Fixes #17577.

  • No crash on an empty first layer with a prime tower by @RF47 in 17565 — with the prime tower on and a single filament, an object with nothing on its first layer, such as a cube standing on an edge, crashed OrcaSlicer and the command line. Slicing stops with the "One object has an empty first layer and can't be printed" error instead, as it does without the tower. Fixes #17554.

  • No crash when the 3D view can't start OpenGL by @HanifKoh in 17441
    When OpenGL couldn't be activated for the 3D view, for example on a PC whose monitor was plugged into the motherboard with its integrated graphics turned off, OrcaSlicer showed an empty grey view and then crashed on the first notification or slice. It no longer crashes. On Windows, a message says the 3D view can't be shown, names the graphics adapter behind that display when it can find it, and lists what to check, such as plugging the monitor into another graphics card. The 3D view still needs a working OpenGL driver to draw anything.

  • No crash on a project that assigns a filament to a missing nozzle by @HanifKoh in 17476 — on printers with more than one nozzle, a project that assigned a filament to a nozzle the printer doesn't have, as a hand-edited or old project can, crashed the slice, and a bad tool number for a nozzle could crash or hang the G-code export. Such a filament prints with the first nozzle, and the bad tool number is skipped. The same assignment from the command line (--filament-map) also slices rather than crashing.

Slicing & Print Quality

  • Only one wall on top surfaces keeps the top infill off inner walls by @lamweilun and @ianalexis in 17597
    With Only one wall on top surfaces and Top surface expansion both on, an inner wall that ran straight across a top surface, such as the long side of a wide ledge with parts rising at its ends, was kept, and the expanded top infill printed over it. Such walls are now found along their whole length, with both wall generators. Arachne cuts them back where they cross the top. Classic drops each such wall loop whole, so more loops go than before, and the band they leave gets internal solid infill when Ensure vertical shell thickness is All, or sparse infill otherwise. Fixes #17300. Wiki

    Before After
    Top surface infill printed over the inner walls on the layer where the base ends The same layer with no inner walls under the top surface infill
  • Fewer tiny moves in smoothed infill at fine resolutions by @ianalexis in 17630
    With Sparse infill smooth factor on and a very fine G-code Resolution, each rounded corner was split into so many tiny moves that a Klipper printer had to plan over a thousand of them per second, and a Qidi Q2 stopped several times along infill paths that the preview showed without pauses. Corners follow the resolution down to 0.005 mm and no finer. On a test project at a resolution of 0, the sparse infill went from 656,000 moves to 286,000, the G-code shrank from 28 MB to 18 MB, and the infill printed about 7% faster in a simulation of Klipper's planner. At 0.005 mm or coarser, the G-code doesn't change; among system profiles, only the four with a resolution of 0 are affected. Wiki

  • Tree Hybrid support keeps its expanded first layer on the bed by @HanifKoh and @hammams in 17700 — with Tree Hybrid support, no raft and a First layer expansion above 0, an overhang near the edge of the plate got first-layer support past the edge, and the plate was refused for G-code outside the printable area. The expanded first layer is trimmed to the bed and its excluded areas. Fixes #17695. Wiki

  • A plate's Skirt start point is saved with the project by @yw4z in 17457 — a Skirt start point set in a plate's settings was lost when the project was saved and opened again.

  • The max flowrate test starts at its starting flow by @ianalexis in 17699 — the first loop of the test's spiral ramped up from zero flow, so the base of the test printed almost no plastic and reached the starting flow only at the end of that loop. The whole first loop prints at the starting flow. Wiki

  • The G-code's first layer time comment reports the first layer by @lamweilun in 17328 — the ; estimated first layer printing time (normal mode) comment gave the time of the start G-code instead, close to zero when it only calls a PRINT_START macro. It holds the first layer's time, the same value saved in 3MF files. Fixes #17325.

Import, Arrange & Tools

  • STEP files import with all their faces by @RF47 in 17720
    Since 2.5.0, some STEP files imported with faces missing. In the Fusion file from the report, four bands of a tilted chimney came in as flat discs. Those faces mesh completely again. On ARM64 computers, such as Apple Silicon Macs, the repair that runs on an imported mesh with holes could also merge unrelated edges and stretch triangles, in STEP and SVG imports alike. It no longer does. Fixes #17636.

    Before After
    STEP model whose four cone bands imported as flat discs The same STEP model imported with all its faces
  • Parts almost as wide as the bed arrange again by @raistlin7447 and @yw4z in 17416
    Arrange couldn't place some parts that nearly fill the bed in one direction, though 2.4 could, such as 250 mm parts on the 256 mm bed of an A1. The arrange spacing no longer pushes such parts off the bed, and the room kept for brims and skirts is measured on each part as it is turned for packing, including by Align to Y axis. A first layer line width given as a percentage is no longer read as millimetres when leaving room for the skirt, which affected 32 bundled profiles. With an arrange spacing set, parts can also come right up to the bed margin instead of staying half the spacing inside it. Fixes #17376.

    The reporter's parts arranged by 2.5.0 on the left and by 2.5.2 on the right
  • The Measure and Assemble tools work on stretched objects by @kisslorand in 17549 and 17552
    On an object scaled by different amounts along each axis, the Measure tool still took a round edge stretched into an ellipse for a circle, so it showed a wrong diameter and a highlight that didn't follow the edge. Angles between faces were off too: a pyramid stretched along X measured 91.7° instead of 96.2°. Features are now found on the stretched shape, so angles come out right and an ellipse is no longer offered as a circle. In the Assemble tool, Parallel could also deform a part of such an object, turning a cylinder's round end oval. Fixes #17526.

    Before After
    Measure tool highlighting a circle that doesn't follow the stretched top edge of a cylinder The same stretched edge, no longer detected as a circle
    Angle between two faces of a stretched pyramid measured as 91.724 degrees The same angle measured as 96.227 degrees
  • Right clicks on empty space leave the Cut tool alone by @Noisyfox and @yw4z in 17200 — with the planar Cut tool open, a right click on empty space could change the cut preview, uncheck and grey out Cut to parts, and leave Reset unavailable. Such clicks leave everything as it was, and Reset stays available whenever you're assigning parts to a side of the cut.

  • The Texture Displacement tool shows only the selected object by @yw4z in 17656 — the tool kept the other objects and instances on the plate in view, along with the negative parts and modifiers of the selected object. It now shows only the model parts of the object you're working on.

  • Clicks between toolbar icons stay on the toolbar by @yw4z in 15548 — a click in a gap between the icons of the 3D view's top toolbar, or on one of its dividers, went through to the 3D view, which deselected the object or closed the open tool. Those clicks, right clicks included, no longer reach the 3D view, in the Assemble view too.

Interface Fixes

  • No trailing zeros in number fields in Czech and Russian by @HanifKoh in 17704 — with OrcaSlicer in Czech or Russian, numeric fields in the settings could show four decimals, such as 0.2000 instead of 0.2. Fixes #17672.

  • The Sync Nozzle status dialog fits its table in dark mode by @peachismomo in 17607 — on a printer with a nozzle rack, such as the H2C, the Sync Nozzle status dialog could open with a large empty band under its nozzle table, white in dark mode, and flash white before the table appeared. The dialog sizes the table once it has its full width, and opens dark in dark mode.

Plugins

  • Cloud plugins update on Windows by @peachismomo in 17462 — updating a cloud plugin that was in use could stop with an error that its folder wasn't empty, before the new version was downloaded. Those updates go through.

  • Plugin pages switch properly from the overflow tab by @HanifKoh in 17626 — with more plugin pages than Visible plugin pages allows, picking another page from the last tab's dropdown left the previous page drawn over it. Fixes #17624.

  • Texture packs with long file paths show up on Windows by @HanifKoh in 17730 — textures from a Plugin Hub texture pack whose full path reached 260 characters were missing from the texture picker, with nothing in the log, and how many went missing depended on the length of the Windows user name. They're now listed and load, and so do textures in folders whose names have non-ASCII characters. Fixes #17713.

Linux, macOS & Windows

  • No more Wayland freezes in the AppImage by @raistlin7447 in 17399
    Since 2.5.0, the Linux AppImage could freeze or garble its window on Wayland when you opened Preview, Device or Project, with the 3D view staying drawn over the other pages. Reporters hit it on Intel and AMD graphics with their system's WebKitGTK. On Wayland, builds outside the Flatpak now have WebKitGTK draw through shared memory instead of the GPU buffers that broke, and the Device tab keeps its smooth scrolling. The Flatpak, whose own WebKitGTK works, keeps the GPU buffers. To keep them in other builds, start OrcaSlicer with WEBKIT_DMABUF_RENDERER_FORCE_SHM=0. Fixes #17381.

  • The 3D view stays drawn under VirtualGL by @HarrierPigeon in 17508 — when OrcaSlicer runs through VirtualGL's EGL back end (vglrun -d egl), as in remote or headless sessions, 2.5.0 and 2.5.1 turned the 3D view black whenever it reused its last frame, such as on every plain mouse move in Preview. With FXAA on, every frame was black. The view draws normally again, with anti-aliasing kept.

  • "Open in OrcaSlicer" links work on macOS when OrcaSlicer is closed by @bwees in 16769 — when OrcaSlicer wasn't already running, a link from Printables or Thingiverse started it on the Home screen without loading the model. The model now loads either way. Fixes #14602.

  • The "Check the reason" panel lays out correctly on macOS by @raistlin7447 in 17366 — in the print, send, calibration and printer binding dialogs, the error panel that Check the reason opens drew all of its rows on top of each other. Fixes #17065.

Printer Connectivity

  • The Device tab loads the printer's page only when you open it by @Art-Porto in 17580 — since 2.5.0, for printers whose Device tab shows the printer's web interface, OrcaSlicer loaded that page about 20 seconds after starting, even if you never opened the tab, and kept it running until you quit. A camera on that page streamed the whole time: on a Creality K2, each open OrcaSlicer cost the printer about 14% of a CPU core. The page loads when you first open the tab.

  • Filament sync reads the CFS Lite on a Creality SPARKX i7 again by @HanifKoh in 17702 — since 2.5.1, filament sync found no slots on a SPARKX i7 with a CFS Lite, because the printer first reports that no CFS is connected and sends the slots right after. OrcaSlicer now waits briefly for them. Fixes #17675.

🧰 Platform & Build

  • Building from source works again by @raistlin7447, @RF47, @Noisyfox and @HanifKoh in 17742, 17743, 17725, 17596 and 17601
    Since the wxWidgets fork that OrcaSlicer builds against moved on November 20, every build from source stopped at the wxWidgets download, including the Flatpak and distribution packages built from the 2.5.1 source tarball. Builds fetch the pinned wxWidgets version again. The dependencies also build with CMake 4.5, which the GNOME 50 SDK ships, with MSVC on Windows systems set to a non-UTF-8 code page such as Simplified Chinese (936), and with MSVC 14.52 on Windows on Arm. Fixes #17740.

🖨️ Printer & Filament Profiles

Profile Updates & Fixes

  • Prusa CORE One presets slow down for the right overhangs by @HanifKoh in 17605
    Many CORE One and CORE One L process presets had PrusaSlicer's overhang speeds in the wrong slots, so a steeper overhang could print faster than a gentler one. "0.20mm STRUCTURAL @core One 0.4", for example, printed 25% overhangs at 30 mm/s and 50% overhangs at 45 mm/s. On ten presets, mostly for 0.6 mm nozzles, the steepest overhangs printed at a speed meant for mild ones, about 91 mm/s on "0.20mm SPEED @core One 0.6". The speeds move to the right slots with their values unchanged, and the CORE One MMU3 presets take the corrected speeds from the presets they're based on. Fixes #17604.

  • Snapmaker U1 rod clearance from Snapmaker's own measurements by @KuzuriAo and @ianalexis in 17276 and 17099
    The U1's Height to rod drops from 27.5 to 26 mm, Snapmaker's CAD value of 27.05 mm with a 1 mm safety margin, and Distance to rod is 41.2 mm instead of the default 40. When printing by object, a plate is now refused if an object between 26 and 27.5 mm tall shares a Y band with an object printed after it, since the rod can touch it or come within a millimetre. The Snapmaker U1 (0.4+0.6 nozzle) profile, which set no clearances of its own, gets the U1's values instead of a 36 mm rod height and a 65 mm radius. A typo in the comments of the U1 start G-code is fixed too. Fixes #17091.

  • AnkerMake M5 printers show the right print time by @MinDBreaK in 17548 — the print time written for the printer held only the seconds part of the total, so the printer showed the wrong time. The fix was tested on the M5, and the M5C, which shares its start G-code, gets it too.

🗪 Localization

  • All 22 languages translate every new string in 2.5.2, by @HanifKoh in 7324a6f

  • Korean uses the same word for "extruder" in the left and right extruder labels as everywhere else, by @rlarjsdn-kgw in 17612


❤️ Support OrcaSlicer

OrcaSlicer is — and will remain — free and open source.

Every release represents countless hours of behind-the-scenes work: building features, fixing bugs, validating profiles, and testing across real machines.

If OrcaSlicer helps you bring ideas to life, please consider supporting the project. Donations are used to support the open-source contributors who help build and maintain OrcaSlicer, and to cover project costs such as servers, development tools, test machines, filament, printer parts, and other infrastructure.

Your support helps us keep improving OrcaSlicer for everyone — while keeping the project free, open, and community-driven.

Thank you for being part of the OrcaSlicer journey. 🙏

Support on Ko-fi

Donate via PayPal

Don't miss a new docker-orcaslicer release

NewReleases is sending notifications on new releases.