* CAM: fix G0 regression in drilling
At some point the initial rapid to Safe Height got dropped from drilling cycles.
This leads to excessive air drilling.
It also alters "OLD_Z" which sets the Retraction plane under G98.
The defect is causing much confusion and time wasting in the roadmap discussions
This PR fixes the G0 regression.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Fixes#22858
Ensure that the default CAM asset location exists and is created in the user data location
* Update src/Mod/CAM/Path/Preferences.py
Co-authored-by: Chris Hennes <chennes@pioneerlibrarysystem.org>
---------
Co-authored-by: Chris Hennes <chennes@pioneerlibrarysystem.org>
* CAM: correct grbl_post retract heights
Some small corrections to retraction heights in expansion of G8x drilling cycles for GRBL machines.
Add comment blocks from NIST definitions to aid maintenance and understanding.
Rename some variables for clarity and ease of future maintenance.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* correct typo which prevented final retract move
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* [CAM] fix biggest performance problems in ramp dressup
key items:
- finding the index of the current edge with edeges.index(edge) was very
expensive; fixed by tracking the index while looping
- checking which edges were rapids with edge equality was also
expensive; fixed by keeping a list of indexes of rapid input edges,
and tagging output edges with whether or not they are rapids
* [CAM] comment out Path.Log.debug in hot segments of ramping code
Even when low level logs are supposed to be suppressed, Path.Log.debug
takes take invoking traceback.extract_stack. This time adds up if logs
are invoked in frequently run loops.
* Fix CAM test
* [CAM] reimplment ramp method 1 with faster code
* [CAM] reimplement ramp methods 2, 3, and helix
* [CAM] patch to make output match original
* [CAM] ramping full performance + functionality fix
* [CAM] extract tool controller ui elements into their own file
* [CAM] make the changes in the tool controller UI only apply when ok is clicked
* [CAM] Add tool controller edit panel to the Profile operation
* [CAM] Add copy button to in-operation tool controller editor
* [CAM] clean up changes
* [CAM] Add tool controller edit UI to all operations
Notes on changes that were not a simple copy/paste job from the changes
I made for Profile:
- Deburr: changed TC/coolant rows from 1 and 2 to 0 and 1
- Probe: didn't work at all initially due to bug in main where ShapeName
changed to ShapeType. I added a utility for reading either a ShapeType
or a ShapeName (check for both properties, convert ShapeType to lower
case) and applied it to probe and camotics
- Drilling: moved Keep Tool Down checkbox up from row 8 to row 2 (all
intermediate rows were missing) and added the edit checkbox in row 3
below it
- VBit, Probe (or anything else that requires a specific tool type): in
Base.py setupToolController(), I added a check to see if the currently
selected tool is an invalid type, and if so and there is a valid tool,
then change to that one. This fixes two UI bugs. Plausibly pre-existing,
if there is one valid tool and an invalid tool is selected, it's
impossible to switch to the valid one because you can't generate a combo
box change event for the new tool. Definitely new: if an invalid tool is
selected and there are no valid tools, the combo box will be empty but
the new tool controller edit utility will let you edit the current TC
anyway.
- Thread Milling: replaced the Tool Controller GroupBox with the
standard QFrame layout, and added the checkbox. Note that this
operation doesn't have a UI element for coolant -- likely a bug, but I
didn't look into it
- Surface: Changed from form layout to grid layout. Deleted an old
SurfaceEdit.ui file -- it was replaced with PageOpSurfaceEdit.ui in
2017 but not deleted (commit 77af19e7489e1fc637a68cdad220e5dd430dc2b9)
- Waterline: Changed from form layout to grid layout
* [CAM] Bug fixes
setupUi() wasn't called on the tool controller editor, preventing
changes in its UI from being written back to the object immediately.
This caused weird behavior where if you edited a field twice it would
reset the second time it was focused
Added a hook to automatically update the TC combo box when the TC name
(or anything else about it, since that was easier) changes
* Fix usage of QSignalBlockers
* [CAM] Block scroll events on tool number and spindle direction when not
focused
Specifically, if you mouse over either of these UI elements and use the
scroll wheel, it used to focus the element and change its value. This
commit makes it do neither of those things, for these specific elements,
as a measure against users accidentally changing these values.
* disable tcNumber edit field in operations panel
* Add "New tool controller" option to TC combo box
When selected, it opens (toggles, technically) the tool bit dock and
returns to the previous selection. Adding a new tool controller using
the dock (already, before this commit) automatically switches the
operation's tool controller to the new one
* Add "Copy" option to tool controller combo box
* Copy TC function only in combo box, no button
* [CAM] update in-operation "new tool controller" function to use a dialog
* [CAM] make the tool selector always a dialog and never a dock
* remove spacer from ToolControllerEdit.ui to fix exces white space
* [CAM] change tool dialog default sizing/spacing
* [CAM] fix bug where copying tool controller doesn't copy all values
* CAM: revert grbl_post regression
I recently noticed a bug in grbl_proc post processor where it was not correctly setting OLD_Z if current machine_z is below G81 retract plane in R parmeter.
On digging into my original commit is looks like the two lines which did this have been commented out.
I can never find the Blame feature when I need so if someone else can find the culprit feel free.
2bd617a2b3
This PR simply reinstates the missing lines and adds a snippet of NIST as a comment block to prevent any over-zealous editing in the future.
This regression means that drilling cycles start with z=clearance height instead of z=safe height leading to much costly air cutting.
This similarly messes up the the retract plane for G98 leading to excessive machine time wasting and incorrect application of G98 returns.
Indeed this regression was noticed because some were claiming the Safe height was redundant in drilling ops since it was never used.
There is much discussion about all this at the moment which is being disrupted by people assuming the current broken paths are what FreeCAD is intended to do.
Let's fix this ASAP, so we can stop this confusion and get discussions back to reality.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit fixes how origin (coordinate system) indicator looks. Before
it showed as disc, and now it shows as point which is more approperiate
given the context.
