In case the configuration snapshot cannot be taken before installing
configuration updates (because the current configuration state
is invalid), ask user whether to continue or abort.
1) Slic3r::RuntimeError was replaced with ConfigurationError,
all exceptions thrown by the configuration layer are derived
from ConfigurationError.
2) When parsing configuration files, ConfigurationError is catched and
rethrown extended with the file name being parsed.
1) Starting with this commit, configuration block exported into G-code
is delimited by "; prusaslicer_config = begin" and "; prusaslicer_config = end".
These delimiters look like any other key / value configuration pairs
on purpose to be compatible with older PrusaSlicer config parsing from G-code.
2) Config parser from G-code newly searches for "; generated by ..."
comment over the complete G-code, thus it is compatible with various
post processing scripts extending the G-code at the start.
3) Config parser from G-code parses PrusaSlicer version from
the "; generated by PrusaSlicer ...." header and if the G-code was
generated by PrusaSlicer 2.4.0-alpha0 and newer, it expects that
the G-code already contains the "; prusaslicer_config = begin / end"
tags and it relies on these tags to extract configuration.
4) A new simple and robust parser was written for reading project configuration
from 3MF / AMF, while a heuristic parser to read config from G-code located
at the end of the G-code file was used before.
Note: When ProgressDialog has no-nullptr parent, then PrusaSlicer doesn't lose the focus at the end of model fixing,
but start to flickering during fixing of the big models.
changes in hints.ini
Hint notification: Just one button in Help menu.
Hint notification: New naming - Tip of the day
Notifications: no scrolling.
Hint notification: fadeout colors.
- The optional output_name file might contain the .pp suffix.
- In case the file contains just filename, prepend it with the
output dir. We don't want to save to current workdir.
1) New environment variable SLIC3R_PP_HOST contains one of
"File", "PrusaLink", "Repetier", "SL1Host", "OctoPrint", "FlashAir", "Duet", "AstroBox" ...
2) New environment variable SLIC3R_PP_OUTPUT_NAME contains the name
of the G-code file including path (for SLIC3R_PP_HOST == "File")
or a name of the file after upload to the host (PrusaLink, Octoprint ...)
3) The post-processing script may suggest a new output file name
(likely based on SLIC3R_PP_OUTPUT_NAME) by saving it as a single line
into a new "output name" temp file. The "output name" file name is
created by suffixing the input G-code file name with ".output_name".
Please note that the G-code viewer visualizes G-code before post-processing.
Fixes Broken PostProcessing when script changes out-filename #6042
Hints notification: enabled / disabled tags
Hints notification: suppress opening url due to preferences.
Hint notification: close after 5 minutes and button in Help menu
Hint notification bold (diff color) headline on multiple lines
Hint notification: Documentation link button.
Hints.ini header
hints logo - clippy
disabled a preffered modes and techs for hints notifications
Empty hints array check. Fix of #6774.
1. Do not show what's below the bed when a gizmo is open
2. Triangulate the cut so people don't see inside
3. If regular clipping plane is used, the cuts are mutually clipped by one another
4. Painting itself ignores the hidden part of the object
on_is_selectable and on_is_activable functions are now completely independent,
the former says if there shall be an icon in the left toolbar, the latter
says if the gizmo can be activated (by a shortcut or GLGizmoManager::open_gizmo)
Make is_converted_from_meters / is_converted_from_inches exclusive-or.
Maybe it would be better to make a single enum from the two booleans,
if they are exclusive-or?
This reverts commit d001195ebd. It makes no sense,
reload from disk is used when the file has changed, which means the paint-on
data are possibly meaningless or even completely wrong (referencing
triangles that no longer exist)
It is now possible to use e.g. --ensure-on-bed=0 for bools (meaning the same as --no-ensure-on-bed).
Using --no- prefix on non-boolean is an error (--no-ensure-on-bed=1)
Providing a value for --no- prefixed bool is an error (--no-loglevel 5)
'dont-ensure-on-bed' (which allows to override). This was the original
behaviour in Slic3r and Sli3rPE, probably broken long ago when CLI
was ported from Perl.
Also, --scale-to-fit should now work again (#5772)
when the respective object info line was clicked, the variable
layer height mode was opened correctly, but closing it through
the toolbar deactivated most of the icons as if it was just opened.
Filtering of unprintable regions in multi-material segmentation depends on if gap-fill is enabled or not. So sliced object is invalidated when gap-fill was enabled/disabled by option "gap_fill_enabled" or by changing "gap_fill_speed" to force recomputation of the multi-material segmentation.
For Apple's on Arm CPU computed triangle normals inside fragment shader using dFdx and dFdy has the opposite direction. Because of this, objects had darker colors inside the multi-material gizmo.
Based on https://stackoverflow.com/a/66206648, the similar behavior was also spotted on some other devices with Arm CPU.
1) Polished up wording of the error messages.
2) Made some messages in the SysInfo dialog localized.
3) Renamed LibraryCheck.cpp/hpp to BlacklistedLibraryCheck.cpp/hpp
4) CPPized the BlacklistedLibraryCheck WIN32 C code.
It was reworked graph generation for multi-material segmentation. Now only oriented arcs in one direction are added to the graph for input polygons. This direction matches the direction of the lines in the input polygons.
ImGuiWrapper::slider_float now clamps the value returned by imgui by default. Bare ImGui::SliderFloat
allows entering off-scale values when entered by keyboard, which is not nice. The protection can
be turned off by the last optional parameter.
ImGui does not want the io flags to change in between NewFrame and EndFrame.
We did that - e.g. after a key down-key up combination with no render in between,
or when key down and mouse move event were processed with no render in between.
An assert was added in imgui to detect this between 1.75 and 1.83, which
made the issue visible.
Solution: only call the new_frame function in update_key_data/update_mouse_data
when imgui actually consumes the input. This forces immediate render so
EndFrame will be called.
To the "physical_printer" added "preset_name" option as a duplicate of the "preset_names" but like one string.
Note: It will works just for printer preset names without ";" (see #5503)
../src/slic3r/GUI/Gizmos/GLGizmoSimplify.cpp:72:81: error: cannot pass non-trivial object of type 'std::string' (aka 'basic_string<char, char_traits<char>, allocator<char> >') to variadic function; expected type from format string was 'char *' [-Wnon-pod-varargs]
../src/slic3r/GUI/Gizmos/GLGizmoSimplify.cpp:72:87: warning: format specifies type 'int' but the argument has type 'size_t' (aka 'unsigned long') [-Wformat]
updated ImGuiWrapper::selectable function (It was a copy of ImGui::Selectable
with slight changes. The function was changed a lot in ImGui 1.83, I copied
the new version and reapplied the changes.)
string or an UTF-8 string. Let's convert it to wxString before
showing an error message box.
boost::system shall return UTF-8 localized error messages as long
as we set -DBOOST_SYSTEM_USE_UTF8 in the top most CMakeFiles.txt,
which we do.
+ ProjectDropDialog: wxRadiloBox was changes to the wxStaticBox with wxRadioButtons because of dark mode colors update
+ Preferences: Fixed update of the text colors
It results in a six-fold reduction of the amount of data transferred into GPU because there is no need to duplicate vertices for every triangle. Also, normals aren't needed to be transferred to GPU.
* MSW specific: First implementation of the Notebook control to support Dark/Light color modes
* MSW specific: Set mode sizer to the Notebook control.
* MSW specific: Added icons to the Notepad control
+ There is no need to restart application after the changing of the color mode
* Fixed non-MSW build
* Updated color for SavePresetDialog
+ Added wrapper to wxMessageBox for mom-MSW platforms
key in the multi-material painting gizmo.
Fixed also a similar issue in support and seam gizmos where an object was displayed with the wrong color after pressing the ESC key.
sets of optional boolean parameters, the cut function "keep upper",
"keep lower" and "flip lower" boolean parameters were converted into
a single type safe enum_bitmask. Such a coding style is certainly
wordier than the original code, but much safer and more readable
than the error prone "boolean, boolean, boolean" function call
parameter list.
config bundles, project files (3MFs, AMFs). When loading these files,
the caller may decide whether to substitute some of the configuration
values the current PrusaSlicer version does not understand with
some reasonable default value, and whether to report it. If substitution
is disabled, an exception is being thrown as before this commit.
If substitution is enabled, list of substitutions is returned by the
API to be presented to the user. This allows us to introduce for example
new firmware flavor key in PrusaSlicer 2.4 while letting PrusaSlicer
2.3.2 to fall back to some default and to report it to the user.
When slicing from command line, substutions are performed by default
and reported into the console, however substitutions may be either
disabled or made silent with the new "config-compatibility" command
line option.
Substitute enums and bools only. Allow booleans to be parsed as
true: "1", "enabled", "on" case insensitive
false: "0", "disabled", "off" case insensitive
This will allow us in the future for example to switch the draft_shield
boolean to an enum with the following values: "disabled" / "enabled" / "limited".
Added "enum_bitmask.hpp" - support for type safe sets of options.
See for example PresetBundle::load_configbundle(...
LoadConfigBundleAttributes flags) for an example of intended usage.
WIP: GUI for reporting the list of config substitutions needs to be
implemented by @YuSanka.
notification with some locales, we don't want PrusaSlicer 2.3.0/2.3.1
to show this notification. On the other hand, we would like PrusaSlicer
2.3.2 to show an update notification of the upcoming PrusaSlicer 2.4.0.
Thus we will let PrusaSlicer 2.3.2 and couple of follow-up versions
to download the version number from an alternate file until
the PrusaSlicer 2.3.0/2.3.1 are phased out, then we will revert to
the original name.
that configuration could be recovered in the case PrusaSlicer.ini
is corrupted during saving. The config is first written into a temp file
marked with a MD5 checksum. Once the file is saved, it is
copied to a backup file first, then moved to PrusaSlicer.ini.
When loading PrusaSlicer.ini fails, the backup file will be loaded
instead, however only if its MD5 checksum is valid.
The following "Fixes" comments are for github triggers. We implemented
a workaround, not a fix, we don't actually know how the data corruption
happens and why. Most likely the "Move file" Windows API is not atomic
and if PrusaSlicer crashes on another thread while moving the file,
PrusaSlicer.ini will only be partially saved, with the rest of the file
filled with nulls. We did not "fix" the issue, we just hope that our
workaround will help in majority of cases.
Fixes prusaslicer wont open 2.3 windows 10 #5812
Fixes Won't Open - Windows 10 #4915
Fixes PrusaSlicer Crashes upon opening with "'=' character not found in
line error" #2438
Fixes Fails to open on blank slic3r.ini %user%\AppData\Roaming\Slic3rPE
Quite some time ago, many of the TBB components were deprecated in favor
of their near-equivalents in the STL or, in the case of task_scheduler_init,
were broken up and reconstituted under a less ad-hoc logic. Every time a header
file marked deprecated gets included, a rather loud warning is emitted, which
leads to a complete TBB's domination over the stderr stream during build time,
making it harder to notice _legitimate_ warnings.
Instead of merely muting the output with TBB_SUPPRESS_DEPRECATED_MESSAGES,
perform a genuine migration away from the deprecated components with the added
benefit of achieving a source compatibility with oneTBB, the successor to TBB
which has dropped the deprecated API for good.
What got replaced for what?
| Deprecated | Replacement |
| ------------------------------------- | --------------------------------------------- |
| `tbb::atomic` | `std::atomic` |
| `tbb::mutex` | `std::mutex` |
| `tbb::mutex::scoped_lock` | `std::scoped_lock<std::mutex>` |
| `tbb::mutex::scoped_lock` (empty) | `std::unique_lock<std::mutex>` (deferred) |
| `tbb::task_scheduler_init` | `tbb::global_control` |
| `tbb::this_thread` | `std::this_thread` |
Signed-off-by: Roman Beranek <roman.beranek@prusa3d.com>
../src/libslic3r/QuadricEdgeCollapse.cpp:110:39: warning: suggest parentheses around '&&' within '||' [-Wparentheses]
../src/libslic3r/QuadricEdgeCollapse.cpp:394:38: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
../src/libslic3r/QuadricEdgeCollapse.cpp:395:38: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
../src/libslic3r/QuadricEdgeCollapse.cpp:514:26: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
../src/libslic3r/QuadricEdgeCollapse.cpp:514:42: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
structured exceptions (hard crashes, segmentation faults...),
converts them to Slic3r::HardCrash exceptions and displays the exception
using the usual PrusaSlicer way.
The SEH handler is installed on the main background slicing thread
as of now, therefore hard crashes in TBB worker threads are not handled.
Renamed its_create_neighbors_index() / its_create_neighbors_index_par() to its_face_neighbors() / its_face_neighbors_par().
New variant of its_face_edge_ids() to create edge IDs from face neighbors.
Fixed some incorrect use of _NDEBUG, it should be NDEBUG.
PrintObject::slice_support_volumes() returns newly Polygons, which are cheaper than ExPolygons.
Updated SeamPlacer and SupportMaterial to use regions defined as Polygons, not ExPolygons.
TriangleSelector::get_facets_strict() returning a patch with T-joints retriangulated.
New slice_mesh_slabs() - slicing projections of a triangle patch into top / bottom layers of slices, for MMU top / bottom segmentation.
TriangleMeshSlicer - use 64 mutexes instead of one when scattering sliced triangles into layers. This makes a big difference on modern many core desktop computers.
When applying MM segmented regions to input regions, the split regions are now re-merged with 10x higher positive offset epsilon to avoid creating gaps.
When testing for existence of paint-on supports or seam, use a more efficient has_facets() test, which does not deserialize into the expensive TriangleSelector tree structure.
GLIndexedVertexArray newly uses Eigen::AlignedBox<float, 3> for efficiency instead of our double based BoundingBoxf3.
Improved MMU painting refresh speed by optimizing generation of the vertex buffers.
Refactored MMU segmentation - projection of painted surfaces from top / bottom.
1) Parallelized.
2) Using the new slice_mesh_slabs() instead of projecting one triangle by the other and merging them with Clipper.
* MSW specific: Dark Mode: First implementation
* Use menu instead of NoteBook
* Implemented MessageDialog
+ Fixed DarkMode for all dialogs and ColorPicker
* MSW DarkMode: Added missed updates for the switching between modes
* MSW DarkMode: Updated all existed context menus after switching of the mode
+ Added markers for the menu item witch is related to the selected tab
* Used wxFrame instead of wxDialog for SettingsDialog
(this change allow us to use menu bar in SettingsDialog)
+ fix for #6548 - Prusa Slicer 2.3.1 not activating non-modal settings window if settings window is minimized
* Implemented "Always use Dark mode colors" preference option
* Fixes for non_MSW build
* Next fixes for non-MSW builds
* Preferences: Fixed selection of the Settings Layout for non-MSW platforms
+ Updated DarkMode for colorpickers
* Windows DarkMode next fixes
* MSWDarkMode: Suppress to use system color to the PrusaSlicer
Select "Preferences -> Use Dark color mode (experimental)" to allow dark mode for the application
* Fixed MSW build
* MSWDarkMode: Upadteed color mode for ExtruderSequenceDialog and for dialogs related to the DoubleSlider
* Implemented Auto recreation of the PrusaSlicer when color mode is changed.
* Preferences: Added option "Set settings tabs as menu items (experimental)"
1) Fixing yesterday's regression in deserialization of older painted 3MFs
(order of triangle children is now reversed, thus the serialization
/ deserialization has to take it into account).
2) WIP extraction into facets to triangulate T-joints.
multi_material_segmentation_by_painting is now returning only the painted region. Regions with default colors that aren't painted by multi-material gizmo aren't returned.
This commit also fixed the following issues:
1) After loading a 3MF with painted triangles using the MMU painting gizmo, the painted triangles might not be displayed correctly in the MMU painting gizmo.
2) The MMU segmentation was unnecessarily executed for all layers and not just for the painted layers.
3) Object's base color wasn't changed when the assigned extruder for that object was changed while the MMU paint gizmo was opened.
4) Changing the base color of an object was only possible by removing all painted triangles.
The cracks were caused by using glPolygonOffset to resolve Z-fighting. All triangles of the object were firstly rendered entirely with a base color. And then paint triangles were drawn over these already rendered triangles.
1) When splitting a triangle, vertices are now properly shared with
the neighbor triangles, if these are already split.
Please note that the splitting may not be regular if the splitting
thershold is changed between splitting calls. Still the new code
shares the vertices accross shared edges properly.
2) Triangles resp. vertices are newly reused after deleted using
linked lists of released triangles resp. vertices.
This mechanism replaces the old mechanism of reusing already split
triangles.
Based on information in https://stackoverflow.com/a/46115028, structured bindings in C++17 don't name variables, and because of that, it is impossible to capture these variables in lambda functions.
TriangleSelector::valid refactored to m_valid and valid()
and bool members moved next to each other to shrink data structure
size due to memory alignment.
