Beta 2.4: acceleration edition

This isn't a stable version. Many bugs are expected. The 2.3.57.11 is the current stable (see below).

Merged with PrusaSlicer 2.4.1
So you have all the new stuff from it

• Dark mode for windows
• paint-for-supports
• gcode preview now show the gcode
• fuzzy perimters
• many more

known issues

• bug in the color.ini file: add a newline before "splash_screen_gcodeviewer"
• too many "do you want ot save that"? dialog when closing. Also missing buttons on linux.
• z-offset in shambles (if using color swap) because prusa fixed it on top of the sueprslicer one.
• profiles not up-to-date
• can crash (on startup only?) if supports are enabled.
• Bad gcode on relative extrusion mode

new

• Acceleration control (by @julianschill)
  • Also modified the speeds a bit to have the same functionalities
• Also now deceleration should happen at a better point in the travel path:
  • if too short, it use the target accel for the whole travel
  • if short, it will switch to target accel at the tipping point
  • is the path is long enough, it will switch a bit before the needed distance to decelerate
• "overlapping perimeters" now have thresholds
• scripted widget (aka fake setting): Allow to create "simple settings" that modifies the real ones.
  • one example with the return of the 'overhangs' checkbox, that activate /deactivate the thresholds
• Automatic Z-rotate for newly imported object (defined in the printer setting) by @csch0
• Preference to don't change the focus when moving the mouse over the platter by @RealDeuce
• Adding MPMDv2 Print host upload support by @drbachler
• Added Client certificate authentication to physical printer options by @Pieterv24

Can't build macos binaires for now because of an unreproductible error that only happens on the macos github runners (zstd missing)

Final 2.3

I made a mistake, it's the 2.3.57.11 ...

changes

• first_layer_infill_speed now take the first_layer_speed if set to 0
• Change the SLICE tooltip to a one-liner #2352
• Add support for libspnav for 3DXonnexion mouse on Linux ; by @RealDeuce

fixes

• Fix crash on some brim geometry (2.3.58.10 regression) #2331
• Fix thin_walls_merge (was creating extremely thin extrusions).
• Fix _shrink_contour_holes (xy_inner_size_compensation) when a hole collapse into hole & contours #2339
• Fix overhangs_width disabling overhangs_width_speed if set to 0 #2340
• small_perimeter now disable that modifier if set to 0#1948
• Allow to modify the time estimation compensation even without the limits enabled
• Fix custom var exists() function (2.3.58.10 regression)
• Fix 2.3.58.10 regression on legacy placeholder array syntax #2359
• Fix 'extrusion spacing' display error when using default extrusion width #2337
• Fix a looping_perimeter bug
• Fix icon scaling causing crash on startup #854 by @LucaRood
• Use bridge_infill_margin for all noperi options #2367
• New reliable way to see if a refresh gcode & slice previews is needed
• Prevent SuperSlicer from crashing due to a lack of a $USER env ; by @helfrichmichael
• fix typos

note: I haven't tested the linux & macos build.
For the macos-arm, you need to install it and then use xattr -cr /Applications/SuperSlicer.app. And the arm compile is still a bit buggy, creating some nan/inf somewhere.
For Windows automatic installer/updates: https://community.chocolatey.org/packages/superslicer
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

Better macro edition

Maybe last fixes for the 2.3 branch
The 2.4 build is going well even if it's still a bit buggy, there is a nightly if you want to test it.

known issue:

• The brim has a bug and may crash the application. You can disabel it or use a (more recent) nightly. Will add a new release to fix that shortly.

changes

• macro language:
  • remove default() function
  • Add default_bool(var_name, bool) function
  • Add default_int(var_name, int) function
  • Add default_double(var_name, double) function
  • Add default_string(var_name, string) function
  • Add ignore_legacy(bool) function to enable/disable the legacy format using the []
  • Now can escape the [ by \[, this will print [. same for the {.
• localization update:
  • French (full)
  • Chinese (full) by @togke
  • Czech (a bit) by @kmarty
• profiles update:
  • Vertex
  • RatRig by @top-gun

fixes

• fix crash (in bridge detector) #2261
• Fix an "External Perimeter First" that extrude each perimeter in double if only one of the two is selected. #2267
• Fix wipetower using filament_max_speed but without the mm/s to mm/min conversion #2271
• Fix unreliable wipe_only_crossing #2154
• fix min skirt length #2200
• Fix bridge offset when multiple bridge polygon on a layer #2249
• Fix gcodeviewer to symlink to real slicer name on linux&apple #1627
• Fix a problem when a bridge area is too thin to be printable
• Ironing pattern calibration: now use also the skirt for arrange detection #2246
• Fix first_layer_print_min and first_layer_print_max when complete_objects #2201
• Don't merge an extrusion with the next if different nor contiguous #2258
• Don't fake unretract after M600 but reset the retraction data properly. #2295
• remove phantom solid surfaces: too thin surfaces that won't print anything and so may create empty spaces
• Fix dense_infill when infill/perimeter encroachment is 0 #2243
• Allow version numbers to be greater than 9
• Allow configuration directory next to app bundle on MacOs by @sturmm
• update for compiling on distribution with difference in sysctl by @jaggzh
• various updates for compiling on UNIX by @RealDeuce
• typos

note: I haven't tested the linux & macos build.
For the macos-arm, you need to install it and then use xattr -cr /Applications/SuperSlicer.app. And the arm compile is still a bit buggy, creating some nan/inf somewhere.
For Windows automatic installer/updates: https://community.chocolatey.org/packages/superslicer
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

Ok vase mode edition

Another bugfix release while waiting for the 2.4 merge

changes

• 'retract only when crossing perimeters' now also consider top surfaces as external area, so it won't cross it without a retract.
• Spiral vase: allow to choose the seam start position. #2114

fixes

• Fix gcode writer when writing a float number rounded to an int #2172
• remove trailing '.' after int values in gcode #2151
• Fix vase mode: no seam gap. #2114
• connect solid rectilinear lines (same a prusaslicer, the max link length has been disabled for now) #2141
• fix 'print external perimeter first' also print thinwall first #2168
• Fix crash when using concentric (filled) as support interface #2124
• fix broken travel (update of last pos on wipe) #2112
• fix quirky gapfill (u-turn at the tip of it) that creates a nan values #2102
• fix filament override #2110
• some tooltip changes, etc.

note: I haven't tested the linux & macos build.
For the macos-arm, you need to install it and then use xattr -cr /Applications/SuperSlicer.app. And the arm compile is still a bit buggy, creating some nan/inf somewhere.
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

Avoid more perimeters edition

PrusaSlicer 2.4 is out. The change will be merged into the next version.
This release contains more fixes to further stabilize the current branch.

changes

• Custom macro variables: two new keywords: 'exists' and 'default', to be able to fallback when using custom variables defined in filament & print settings.
  • 'default(var_name,value)' check if the variable exist and affect it the value if not.
    The variable can be boolean 'true' 'false', int '0' '12', double '1.0' '42.24' or string '"a string"'.
  • 'exists(var_name)' return true if present. It also creates a dummy boolean variable if not already set.
    Note that if you're using the variable in a formula, it may crash even if guarded by a {if} as everything is interpreted (no lazy check).
    So be sure to add a 'default' so it gets the right type and the formula can be interpreted (unless you just print it, in this case an 'exists' is enough).
    Using 'default' doesn't change the return value of 'exists'.
• Add a new setting 'wipe_only_crossing' to diable/enable wipe when you're using 'avoid perimeter crossing'.

fixes

• Fix 'avoid crossing perimeters' when wipe enabled. #1069
• When 'wipe on retract' on an external perimeter, continue to wipe inside the object instead of coming back to the middle of the perimeter. #2061
• profile update: Fix extrusion_spacing for Anycubic #2084
• Fix medial axis creating points on top of each other. #2099 #2102
• Some tooltip updates (thank you @blex-max for the help) #2097
• Updated ConfigManipulation (disabled/enabled) for width & spacing.
• Fix 'merge thin walls with perimeter' (unneeded travels) #2092
• 0 extrusion_spacing will always create 0-auto-extrusion_width. #2084
• fix fan kickstart reverting when it shouldn't #2076
• Fix Concentric (Filled) infill: honour gap_fill_min_area #2052
• perimeter_round_corners: disabled for the first perimeter #2046
• fix div/0 in libnest2D #2053

note: I haven't tested the linux & macos build.
For the macos-arm, you need to install it and then use xattr -cr /Applications/SuperSlicer.app. And the arm compile is still a bit buggy, creating some nan/inf somewhere.
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

Remembering Sanjay Edition

If someone wonders why this title: https://e3d-online.com/blogs/news/sanjaymortimer
Most of my nozzle comes from E3D. He was such an inspirational man, and so enthusiastic in his videos.

The PrusaSlicer 2.4 may go out of beta in a not-distant future. The next version may well be the 2.4.58.
This 2.3.57 branch will continue to get fixes in the coming months. It's now the "long term support" one.

changes

• macos-arm build, thanks to @claytono. Tell me if it doesn't works.
  • to allow it to run : xattr -cr /Applications/SuperSlicer.app
• Dense infill: fix the full fill density bug & rework detection settings:
  • AutoNotFull: now print the dense infill only if the area to cover doesn't fill the sparse area.
  • AutoOrEnlarged: now print the auto dense infill, unless the area to cover is wider than (nozzle diameter*2/infill ratio) and the enlarged version is smaller than the auto
  • New: AutoOrDisable: now print the auto dense infill, unless the area to cover is wider than (nozzle diameter*2/infill ratio) or the area to cover fill the sparse area.
• Bridge improvements:
  • Don't connect if too far away. Prevent printing a bridge over another one, following the perimeter.
  • Better connection direction, as both are now tested.
  • Fix some other rectilinear issues, crash, bugs #1889
  • Bridge detector now compute much quicker than in the previous version.
  • Fix wrong bridge offset (can lead to no bridge or some missing)
• Fix too detailed gapfill
  • Add resolution_internal to control it (and gyroid)
  • Disallow gapfill lines to be split in the middle when printed (now start at an end and not in the middle)

fixes

• Fix brim-raft bug. The behaviour isn't correct yet, but at least it's now useable. #1927
• Fix 'small perimeter' speed change #1948
• Fix missing lift/retraction #1967
• fix missing prusa sla #1963
• fix disjointed fill surface #2038
• Gapfill "star" should be less frequent
• Fix: allow extra wipe for 1-perimeter configs #2029
• Fix some cases where "no perimeter on bridge" won't activate #1917
• Avoid unneeded unretract followed by a retract in mid-air for sparse wipe tower. #1467
• Plater auto-center now ignore unprintable objects #2008
• Fix seam object for multiple instances with different rotations #2015
• infill_first also modifiable if sparse infill #1971
• Fix "grey" machine limits field for export. #1960
• Fix Calibration hyperlinks: open the os browser #2005
• Fix 'Internal error: ConfigBase::get_abs_value()' (bad initialization of the wizard) #1938
  *Mmulti-line for milling custom gcode fields #2027
• Don't consider ' ' or '\n' for custom toolchange gcode detection #1368
• Don't allow 0-speed retraction #2033
• Update/fix "sliced info" to have the info per extruder & color
• Fix enforce_full_fill_volume for smooth pattern, when it splits from the half-offset.
• Tooltip, typos, ... #1981
• Fr localization update #2004 #1840

note: I haven't tested the linux & macos build.
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

Difficult bridging edition

known bug

• dense infill is at 100% instead of 40%, fixed in the nightly and next release

new

• bridge_overlap fix & redesign (now bridge density)
  • now has a min setting.
  • The density is now chosen between the two values to try to cover as much as possible the area to bridge, with the highest density possible.
  • Remove the hardcoded invisible 87.5% density modifier. Replaced by setting by default the max at 90% and min at 80%
  • Try to place the bridge at the right position (only works if the shape is simple).
  • Known issue: the algo used for bridging can add a return line on top of an already extruded line. Have to do something about that.
  • If you want to get back the old behaviour, set the two values to 87.5
• Add total layer count in the gcode footer #1886
• Add current_position variable for gcode macro: current_position[0] is x and current_position[1] is y #1898
• updated Vertex profiles (from vertex)

fixes

• fix external_perimeter_cut_corners (was extruding too much at the second segment and not enough at the 3rd) #1903
• typo & doc update (thanks oretzl)
• Print & printer custom variables are now scalar and not arrays.
• fix retraction calibration when temp%5 != 0 #1881
• fix crash on monotonic #1889 #1893
• other bugfixes #1877 #1891 #66

note: I haven't tested the linux & macos build.
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

Custom variables edition

Hopefully, most of the bugs from the new feature should have been ironed out.
Don't forget to go watch all the new things from the previous 2.3.57 builds below.
You may still want to use the 2.3.56.9.

known issue:

• crash when you are to far away from the origin. Please update to the 2.3.57.6

new

• custom variables:
  • on each tab, you can add variables (under the note page) that will be available for custom gcode macros.
  • be careful with the syntax. Don't hesitate to launch the slicer with logs (--loglevel=5) to be able to see what's parsed.
  • known issue: print & printer variables are 1-item arrays (will be fixed)

fixes

• Fix fan interactions & hints. #1820
• Fix import config with bridge flow < 1 #1818
• Add a fallback method (in windows) for getting the current exe path to open the gcodeviewer (or the opposite) #1778
• Fix for using unicode paths for loading html in calibrations #1815
• Fix for M73 for silent mode #1819
• Fix the new bridge detector (and improve the compute speed) #1824
• Fix acceleration in % of machine_max_acceleration_x #1829
• fix min/max layer height loading check. #1832
• Try fix wxwidget problem with '<' and '>' #1837

note: I haven't tested the linux & macos build.
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

Travelling edition

No more feature on this branch, will become the new stable, hopefully.

new

• Enhanced avoid_crossing_perimeter: now try to follow the second perimeter where possible #1350 #1210 #1794 #1793 #1790 #1349 #1061 #810 #173 #20
  • Trigger at retract_before_travel or when the travel is longer than 3 nozzle
  • Can trigger without only_retract_when_crossing_perimeters
  • travel can now slow down to avoid going out of the max_gcode_per_second
  • travel acceleration is now bound by machine_max_acceleration_travel and not machine_max_acceleration_extruding
  • Don't split travel for deceleration if there is no need for

• Add lift_min: minimum height to travel between objects #1783 #1775 #1575 #599 #429 #395 #241
  • At the first move don't split the travel #1560
  • At the first move, if start_gcode_manual and no layer_gcode, don't lift the nozzle so the printer won't "z-dance", whatever where the nozzle is.
  • retract_lift_first_layer is gone back to the old simple behavior (revert b16ecbf)
  • removed auto extra lift for first layer, now that lift_min exists.
  • with complete_object, don't unlift at object/first layer change, to avoid Z-dance
  • lift_min will be used between brims, skirts, objects but not between object-skirt, object-brim and their object
• Allow to choose between M73 and M117 for remaining_times #1693
• [linux] Add gtk3 builds, should have less GUI problem than GTK2 #1677
• Menu entry for importing & exporting prusa config & project file (some conversions occurs)
• Ironing: first pass now use filled monotonic instead of filled rectilinear, and it is allowed half overlap. #1708
  • Ironing is also considered as a top for "lift only on top surfaces" #1773
• Add an option to show alternative "custom tooltip", to bypass the time limit on some os. It's a half-assed solution, but at least you should be able to read them (tested on windows & linux) #1720 #1291
• update build scripts: now there is an option for gtk3 & debug build in BuildLinux.sh
  
• Updates to dense infill: #1657 #1671
  • reduced unneeded usage of "full fill"
  • Don't allow holes in infill_dense
  • Print them in the correct order
  • little overlap with everything and not only perimeter, to be able to be anchored into something.
  • Ensure that infill_every_layers and infill_only_where_needed are disabled, because it can't tolerate voids (they create different regions)
  • now multithreaded

• Better bridge angle computation, and ready for tweaking weights (currently 70% coverage, 15% median length, 15% max length, 5% bonus for following a perimeter) #565 #234 #149
  • When removing too thin bridge, use it as solid instead of sparse.
• Search: display tooltip in tooltip, search full sentence in tooltip (low score).
• Preference for hiding settings lock&dot icons #1729
• Setting to put thumbnails at the end of the gcode file
• Update to some default values
• Update to profiles
• max_layer_height & min_layer_height now can be percent.

bugfixes

• fan_speedup now works. #1742
  • don't move fan from custom_gcode
  • don't enter the start_gcode if start_gcode_manual
  • Fix for 'only for overhangs'
• fix brim/skirt issue with support raft: #1745
  • fix skirt distance, to take into account the support
  • fix brim ears: not printed between raft extrusions, but around the area.
• Fix 'one perimeter on first layer' #1687
• Add Colour Change G-code, Pause Print G-code, Template Custom G-code into the GUI (merged from PS2.3) #1681
• Fan speed override now speed up if low layer time #1707
• Fix layer_z for feature_gcode
• Fix fan speed for first layer when full_fan_speed_layer > 0 and disable_fan_first_layers == 0 #1727
• Fix gap - infill voids
• Avoid bridge/infill mess when the bridge area is a bit thin.
• Changes limits for % or float fields. Can now be changed in the UI file.
• Fix crash when using complete_object
• Don't engage "no solid area over X perimeters" before the Y first solid layers.
• Add ironing to "lift only on top surfaces" #1773
• fix brim/skirt issue with support raft: #1745
  • fix skirt distance, to take into account the support
  • fix brim ears: not printed between raft extrusions, but around the area.
• When removing too thin bridge, fill with solid infill instead of sparse. #1766 #1781
• Fix sawtooth bug #1791
• Don't engage "no solid area over X perimeters" before the Y first solid layers
• Calibration: disable ironing post-process, as it's detrimental or just useless.
• Fix fill_exactly when multiple surfaces and no infill_overlap (Was overextruding by 30%). #820
• Fix too thin gapfill when the external perimeter is much thinner than the internal one. #1640
• Fix other little bugs #1711 #1634 #1728 #1056 #1754 #1641 #1781 #1641 #1781 #1631 #1758

note: I haven't tested the linux & macos build.
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button

No more unwanted infill, but fan kickstart

The 2.3.57 branch will be the next stable after some more bugfixes.

bugfixes

• Fix for fan speed-up time, should now work correctly
• Fix for fan kickstart, should now work
• Fix for auto dense infill:
  • better coverage
  • printed in the right order
  • has a 20% overlap with everything for a good grip (and not only perimeters as before)
• crash fix when layer_height is set at 0.
• better reliability for the clipper bug mitigation #1641

note: I haven't tested the linux & macos build.
You can have daily builds in the action menu of github.
You can support me with the "sponsor" button