Cutter and Machine Presets worked without issue.
Operation presets would load, but could not be saved - `AttributeError`
This was caused by 3 errors in `AddPresetCamOperation` in `init`:
- `o.movement.movement.spindle_rotation` changed to `o.movement.spindle_rotation`
- `o.optimize` and `o.optimize_threshold` changed to `o.optimisation.optimize` and `o.optimisation.optimize_threshold`
Discovered this while trying to add a small usability patch that would add an operation if there wasn't already on in the scene, so that loading an operation preset wouldn't cause an error message.
Patch was done by adding a `setup_operation_preset` function in `utils` that would `try: except` to try to adjust an existing operation, but add one if one isn't present.
I also patched the existing operation presets, though they aren't copied to disk by default, so I added a `BoolProperty` to the `AddonPreferences` and copy logic to the `check_operations_on_load` handler so that they will only be copied once. The copy operation does not affect any additional presets the user may have, only the default set that come with the addon.
Also fixed the 'Material Preview' viewport mode.
Previously the object would disappear when switching to that mode.
Render Engines have the option to 'borrow' Material Preview mode from EEVEE via a boolean `bl_use_eevee_viewport = True`
I set that, and added the panels to the Material tab when in Cam.
Commented out `numba` import from the `__init__` for now, until the issues with crashing can be resolved.
Moved `_IS_LOADING_DEFAULTS` out of the `updateMachine` function, declared the default value = False, and grabbed it via `utils.IS_LOADING_DEFAULTS`.
(Note: It was already declared a `global` variable within the function, and I thought that would have been enough to avoid this error, but it works now)
Lastly, altered the cleanup in the `_calc_path` function in `ops.py` - previously it was checking for and deleting objects without issues.
In Blender 4.1 deleting an object would still leave behind mesh data that retained the object name, causing the cleanup function to not recognize and update old operations.
Simply deleting the mesh data solves the issue, as it will also delete the object with it, even though the reverse is no longer true.
Updated Files Organization diagram to reflect current state
Title case for titles and headings (important words capitalized)
Contributors section added.
While rearranging and cleaning up the utils imports in the initial 'Import fix' commit I deleted `utils.` from `bpy.utils.script_paths()` in the ops.py file.
This has been fixed, and I took the opportunity to reverse the autoformat issue (changed line length from 100 to 89 between commits), and restore all the code to its previous state.
I made one exception:
- comments that came at the end of the line that were moved to the line above
All functional code has been restored to its previous format.
While rearranging and cleaning up the utils imports in the initial 'Import fix' commit I deleted `utils.` from `bpy.utils.script_paths()` in the ops.py file.
This has been fixed, and I took the opportunity to reverse the autoformat issue (changed line length from 100 to 89 between commits), and restore all the code to its previous state.
I made one exception:
- comments that came at the end of the line that were moved to the line above
All functional code has been restored to its previous format.
While rearranging and cleaning up the utils imports in the initial 'Import fix' commit I deleted `utils.` from `bpy.utils.script_paths()` in the ops.py file.
This has been fixed, and I took the opportunity to reverse the autoformat issue (changed line length from 100 to 89 between commits), and restore all the code to its previous state.
I made one exception:
- comments that came at the end of the line that were moved to the line above
All functional code has been restored to its previous format.
While rearranging and cleaning up the utils imports in the initial 'Import fix' commit I deleted `utils.` from `bpy.utils.script_paths()` in the ops.py file.
This has been fixed, and I took the opportunity to reverse the autoformat issue (changed line length from 100 to 89 between commits), and restore all the code to its previous state.
I made one exception:
- comments that came at the end of the line that were moved to the line above
All functional code has been restored to its previous format.
Alain Pelletier