(solved) Bug in fillet or chamfer all function (0.21 master)

[user1234]

I found a minor bug in the fillet or chamfer all function.

- load file
- you see a chamfer (checkbox enabled, list grayed out if you go in the task dialog) at the end of the Body which fillet all
- change the the value of occurrences of linear pattern higher (example from 2 to 3)
- looks OK
- doubleclick on the chamfer to go in the taskdialog of the chamfer
- uncheck fillet all
- Body in now not complete chamfered, because the list is still the same of linear pattern before the change (it does not refresh the list)

- same behavior on fillet

fillet_chamfer_all_bug.FCStd

(80.74 KiB) Downloaded 35 times

Code: Select all

OS: Debian GNU/Linux 11 (bullseye) (X-Cinnamon/lightdm-xsession)
Word size of FreeCAD: 64-bit
Version: 0.21.0.32938 (Git)
Build type: Release
Branch: master
Hash: 1f6b257bfa90117299d3aead8c7f263545886ca9
Python 3.9.2, Qt 5.15.2, Coin 4.0.0, Vtk 9.0.1, OCC 7.7.1
Locale: English/United States (en_US)
Installed mods: 
  * QuickMeasure 2022.10.28
  * Assembly4 0.12.6
  * CurvedShapes 1.0.4
  * Curves 0.6.8

Greetings
user1234

[Shalmeneser]

No problem for me : after linear modification, 'all the edges' ON or OFF will keep all edges .

Code: Select all

OS: Windows 10 Version 1809
Word size of FreeCAD: 64-bit
Version: 0.21.0.32457 (Git)
Build type: Release
Branch: master
Hash: 85216bd12730bbc4c3cbf8f0bc50416ab1556cbb
Python 3.10.9, Qt 5.15.8, Coin 4.0.0, Vtk 9.1.0, OCC 7.6.3
Locale: French/France (fr_FR)
Installed mods: 
  * fasteners 0.4.55
  * fcgear 1.0.0
  * Manipulator 1.5.0
  * sheetmetal 0.2.63

[chrisb]

I found a minor bug in the fillet or chamfer all function.

I wouldn't rate this as a bug either:
- UseAllEdges is checked -> all edges are chamfered - OK
- UseAllEdges is unchecked -> the edges in the list are chamfered - OK
- creating a new chamfer with selecting a whole feature instead of an edge -> the list is populated with all edges - OK.

[adrianinsaval]

yeah I think this is expected behavior

[user1234]

OK, no bug, but i think it is not expected and very inconsistent. Either the list is always up to date, or the list should be always empty when this use all edges is checked.


Greetings
user1234

[GeneFC]

OK, no bug, but i think it is not expected and very inconsistent. Either the list is always up to date, or the list should be always empty when this use all edges is checked.

I think that would create a potential coding headache. If the list was cleared when "use all" was selected, then the list would need to be stored somewhere to be put back in place if the "use all" was then unselected.

Possible, of course, but why bother?

Gene

[user1234]

If the list was cleared when "use all" was selected, then the list would need to be stored somewhere to be put back in place if the "use all" was then unselected.

That is the point, the list do not update, so the list is useless. When you uncheck it, the output in 3D is random. When you change it from to fillet all to all except one edge, you have to: uncheck it, wait till it fails, check it again, wait, uncheck it again, and then unselect the unwanted edge. On larger model it is very time consuming, besides it is inconsistent. Many new here blame it inconsistent and some developers wrote, we should address it. Now i address one and everyone said, nah .....


Greetings
user1234

[chrisb]

That is the point, the list do not update, so the list is useless.

I don't see it as useless, it allows to switch between a selection and all edges. That may not be what you currently need, but that's a different thing. What you want is somethng similar to what is now implemented as filter function in Sketcher, but is it really worth the effort?

[user1234]

I don't see it as useless, it allows to switch between a selection and all edges.

Have i really explained it that complex, that no one get the issue? When you have to update one exception of the list (or rearrange), while a normal recompute take 30s on a larger model, you have redo recompute the steps minimum 3 times unnecessary to get all edges in the list again, which makes the list per se useless, except it updates on every geometry change (which it does already in the background, but not the list). May not an issue for small models, for me it is a real life issue and, as i said, inconsistent.

What you want is somethng similar to what is now implemented as filter function in Sketcher, but is it really worth the effort?

This have nothing to do with a filter.


Greetings
user1234

edit: stroke and blue

[chrisb]

Have i really explained it that complex, that no one get the issue?

I think we understood the issue very well, but we come to different conclusions.

What you want is somethng similar to what is now implemented as filter function in Sketcher, but is it really worth the effort?

This have nothing to do with a filter.

Of course its not a filter, and edges are not constraints, I know this very well. It shows one possibility to have an existing selection and the possibility to select all of these.
I see your use case, but unlike you I don't agree that the current behaviour is inconsistent or a bug. What you want is a feature request, where all edges are added to the list.