as the opengl context may not be ready on some platforms (Linux)
at the time the window gets its focus for the first time.
Changed the G-code preview invalidation to trigger when the print
gets invalidated. At that time the 3D path preview switches to the old
preview, if there is anything valid left.
as the GCode is generated outside of the Print class.
Exported the GCodePreviewData as GCode::PreviewData to Perl.
When exporting the G-code with a command line Slic3r,
the GCodeAnalyzer is now supressed for performance reasons.
Removed obsolete Perl module Slic3r::GUI::Plater::3DToolpaths.
due to the differences in the M203 code
(RepRap firmware has it in mm/min, Marlin in mm/sec).
This difference is important to the G-code time estimator.
Changed the g-code flavor to Marlin for all Prusa3D bundled profiles.
The PlaceholderParser is currently used by the GCode.cpp
and by Printer.cpp to generate a new name for the exported G-code or SVG file.
The PlaceholderParser::process() will throw a runtime_error with
a comprehensive error message.
The G-code export will include these error messages into the G-code text
with !!!!!! separators, and the GUI will inform the user, that the G-code
export failed.
both single_extruder_multi_material and wipe_tower are enabled,
and the print prints with a single extruder only.
Newly the same situation will be handled through a conditional G-code
in the following format:
{if not has_wipe_tower}
; Do the priming
{endif}
fixed crashes when loading a config.ini with "compatible_printers"
disabled export of "compatible_printers" into gcode and config.ini
Enabled compatibility of printing multiple objects with support / no support
with a wipe tower.
G-codes in case of single extruder multiple material setup:
At the start of the print, the filament_gcode_start is executed
for the active extruder only, and the filament_gcode_start /
filament_gcode_end are then executed at each tool change.
When the Prusa MM wipe tower is active, the tool changes are handled
a bit differently: M900 K0 is emited before the wipe tower extrusions start,
and the filament_gcode_start code is executed after the wipe tower extrusions
are done. This rule effectively disables the linear advance over the wipe tower.
Implements https://github.com/prusa3d/Slic3r/issues/568
thanks @lordofhyphens, https://github.com/alexrj/Slic3r/pull/3275
Improved handling of custom G-code blocks: Slic3r will try to extract
the target extruder and bed temperatures from the custom G-code blocks.
Removed the Perl dependencies on Encode, Encode::Locale and Unicode::Normalize.
Added dependency on boost::locale.
Added encode_path, decode_path, normalize_utf8 functions to Slic3r.xs
Slic3r.xs has been made mostly utf8 safe by using the boost::nowide library,
thanks to @alexrj for the idea.
Simplified the encode_path / decode_path stuff:
wxWidgets are unicode already, so there is no need to decode_path() from it.
Perl / win32 interfacing is non-unicode, so decode_path() is executed
on ARGV just at the beginning of the perl scripts.
https://github.com/prusa3d/Slic3r/issues/443
Change of the PlaceholderParser:
All vector configuration values stored into the PlaceholderParser
are expected to be addressed by the extruder ID, therefore
if a vector configuration value is addressed without an index,
a current extruder ID is used.
Also a small fix of fan handling: The fan speed is set to zero
at the start of the G-code if the cooling for the initial extruder
is disabled.
perimeters if not necessary, don't take the bridging time into account
when slowing down the print.
Removed Extruder & GCodeWriter Perl bindings.
Improved Extruder for constness.
Refactored GCode::m_elapsed_time to struct ElapsedTime.
values to GCode generator, one per active extruder:
bed_temperature
bridge_fan_speed
cooling
disable_fan_first_layers
fan_always_on
fan_below_layer_time
first_layer_bed_temperature
max_fan_speed
min_fan_speed
min_print_speed
slowdown_below_layer_time
Now it remains to extend Slic3r to correctly apply these values.
(rewrote the Dijkstra shortest path algorithm to use a binary priority
heap instead of a dumb O(n^2) algorithm, added some bounding box tests
to avoid expensive in-polygon tests if possible).
Removed GCode.pm
Removed the Perl bindigns for AvoidCrossingPerimeters, OozePrevention, SpiralVase, Wipe
Changed the std::set of extruder IDs to vector of IDs.
Removed some MSVC compiler warnings, removed obnoxious compiler warnings when compiling the Perl bindings.
Removed the questionable Pressure Advance feature. It is better to use the Pressure Advance implemented into a firmware.
Added a C++ implementation of GCodeReader and SpiralVase, thanks to @alexrj
Added a C++ implementation of GCodeTimeEstimator, thanks to @lordofhyphens
When extruding supports, the support is interleaved with interface
if possible (when extruded with the same extruder).
Otherwise the base is extruded first.
https://github.com/prusa3d/Slic3r/issues/74
The way it works now is following:
Slic3r will add some negative penalty to all perimeter points near the last seam.
Once the perimeter point with minimum penalty is found, its penalty is compared
to a point closest to the last seam. If the penalty of the point
closest to the last seam is nearly as good as the minimum penalty,
the point closest to the last seam is picked instead. This heuristics
will hide the seams into corners if possible, but if not possible, it will
strive to align the seams precisely.
