My Tapeworm-thread of observed FreeCAD bugs

BIM Bo · Post by **BIM Bo** » Mon Dec 07, 2020 11:41 am

As mentioned in https://forum.freecadweb.org/viewtopic. ... 20#p455303
It's basically just a stack - really. Until further notice, observed in this version/environment.

Code: Select all

OS: Arch Linux (/usr/bin/startxfce4)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.19.23141 (Git) AppImage
Build type: Release
Branch: master
Hash: 0eba78d40096720f7a637a9a73c382a08b820bca
Python version: 3.8.6
Qt version: 5.12.5
Coin version: 4.0.0
OCC version: 7.4.0
Locale: English/United States (en_US)

Zoom working - up to a certain point only

When zooming into my model with the mouse wheel, everything works up to a certain point. Then, visual zooming stops. FreeCAD still seems to take the mousewheel zooming into account internally (but not visually), because the longer I try, the longer it also takes to scroll the mousewheel in the opposite direction (to zoom out) to actually have some visual feedback again. UI is not blocked or anything.

Fun fact: sometimes - not reproducible (so far) - when I toggle active part so it's alone on the screen, further zooming does happen. But not always, r should I say "with 0.19.23141 in the minority of the cases now".

I did search the forums
https://forum.freecadweb.org/viewtopic.php?t=17207
https://forum.freecadweb.org/viewtopic.php?t=50231
but none of these seem applicable/a solution.

Tools > Edit Parameters - a desert

There are lots (read: the majority) of entries that actually do not contain any parameters. So the question is why show them at all?
e.g. Preferences: "AddOns" has some params, but then
AppImage, Bitmaps/Theme, Browser, DAGView, Dialog => nada
they are simply just in the way.
Expression, NaviCube, Macro/Macros, Plugins/parts_library etc. => nada

Maybe don't show it if it has no information value.

While we are at it, the "Parameter Editor" window has an inconsistent windows handle. See, if I call Edit > Preferences, I can actually close that window with the 'x'. Not so in the Parameter Editor, where I fullscreen the window. Every. Single. Time. You got me there.

Right mouse contextual menu - the Jester

How about a little magic trick? I'm gonna make this pencil disappear...

Let's say you want to change the color of an object. What do you do? You right-click mouse, see "Appearance", click that, change the color.
Now you want to do the same with the next object. Naturally you right-click mouse, see ... nothing. Tadaa!
Compare its_there.jpg and its_gone.jpg attached.

That is the Houdini feature of FreeCAD. But there is a not-so-elegant workaround: You "simply" Toggle active part of some other object and voila! It's there again.

drmacro · Post by **drmacro** » Mon Dec 07, 2020 1:38 pm

BIM Bo wrote: ↑Mon Dec 07, 2020 11:41 am Right mouse contextual menu - the Jester
...
Compare its_there.jpg and its_gone.jpg attached.

Is the Appearance dialog open? When I try to duplicate this, Appearance is not gone from the list, but is greyed out until I actually close the Appearance dialog.

Post by **Kunda1** » Mon Dec 07, 2020 1:45 pm

This type of post specifically report all bugs in one post, are notoriously well known for becoming so convoluted and difficult to track. That's why we recommend to create a one bug for one post. Each each post is commented on and vetted by the community which then decides if to add it to the bugtracker.

BIM Bo · Post by **BIM Bo** » Mon Dec 07, 2020 2:36 pm

@Kunda1 - right now it's not "reporting" bugs, but as I said a stack. The Germans also call it a "Stoffsammlung" and it's usually notoriously known as the first step - before structuring. So stay in a safe distance for now and observe/or not. Indulge me a little monologue. We can (at some later point) extract specific stuff from here to a separate thread, or do a GitHub ticket or forget it.

----------------------

I thought: Why not "firing up a compiler" as suggested elsewhere, so I did. Observations follow with this version:

Code: Select all

OS: Arch Linux (/usr/bin/startxfce4)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.19.23254 (Git)
Build type: Unknown
Branch: master
Hash: 959c214513b8016eeee48a9ae1c6ec571cef35e8
Python version: 3.9.0
Qt version: 5.15.2
Coin version: 4.0.1
OCC version: 7.4.0
Locale: English/United States (en_US)

ad "Zoom working - up to a certain point only":

I managed to circle this to specific models.

https://www.victronenergy.de/support-an ... re-octo-gx
and (to greater extent)
https://www.victronenergy.de/support-an ... smartshunt
do this (both imported as STP models) when in the view. If they are outside the view, zooming virtually does not stop.

Contrary to that
https://www.victronenergy.de/support-an ... re-quattro
here the Quattro 48V 10000VA 140-100/100 230V (stp)
does not exhibit the problem.

So that might finally provide some repeatability workbench.

---------------------

It's not all bad

Usually one does not mention "the good things - that worked" encountered during testing. The backside of this is that feedback quickly might look like "rah rah it's all bad". It isn't.

I actually managed to kill 0.19-23058 quite easily while playing with links (linked copies?) and trying to force cyclic links upon FreeCAD:

Code: Select all

Origin can only be in a single OriginGroup                                                                                                                                          
Can't find Origin for "Unnamed#Body001"                                                                                                                                             
Can't find Origin for "Unnamed#Body001"                                                                                                                                             
Can't find Origin for "Unnamed#Body001"                                                                                                                                             
Can't find Origin for "Unnamed#Body001"                                                                                                                                             
Can't find Origin for "Unnamed#Body001"                                                                                                                                             
Can't find Origin for "Unnamed#Body001"                                                                                                                                             
Can't find Origin for "Unnamed#Body002"                                                                                                                                             
Can't find Origin for "Unnamed#Body002"                                                                                                                                             
Can't find Origin for "Unnamed#Body002"                                                                                                                                             
Can't find Origin for "Unnamed#Body002"                                                                                                                                             
Can't find Origin for "Unnamed#Body002"                                                                                                                                             
Can't find Origin for "Unnamed#Body003"                                                                                                                                             
Can't find Origin for "Unnamed#Body003"                                                                                                                                             
Can't find Origin for "Unnamed#Body003"                                                                                                                                             
Can't find Origin for "Unnamed#Body003"                                                                                                                                             
Can't find Origin for "Unnamed#Body003"                                                                                                                                             
8.47e-07 SoFCUnifiedSelection.cpp(1521): Cyclic scene graph: Unnamed#Link                                                                                                           
Program received signal SIGSEGV, Segmentation fault.                                                                                                                                 
#0  /usr/lib/libc.so.6(+0x3d6a0) [0x7fe86660a6a0]                                                                                                                                    
#1  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Gui.so.5(+0xf7b6b) [0x7fe866ec7b6b]                                                                                                
#2  0x7fe866ec865a in QGuiApplication::font() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Gui.so.5+0x24                                                                       
#3  0x7fe8673511cd in QApplication::font() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0x9                                                                       
#4  0x7fe86735216c in QApplication::font(QWidget const*) from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0xc2                                                        
#5  0x7fe86736eb87 in QWidgetPrivate::naturalWidgetFont(unsigned int) const from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0x33                                     
#6  0x7fe867374199 in QWidgetPrivate::resolveFont() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0x2b                                                             
#7  0x7fe86737eb4e in QWidget::setParent(QWidget*, QFlags<Qt::WindowType>) from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0x31e                                     
#8  0x7fe8673808b3 in QWidgetAction::releaseWidget(QWidget*) from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0x4f                                                    
#9  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5(+0x29cd51) [0x7fe86747ad51]                                                                                           
#10  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5(+0x29cda7) [0x7fe86747ada7]                                                                                          
#11  0x7fe86737ab04 in QWidget::~QWidget() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0x188                                                                     
#12  /tmp/.mount_FreeCAeYdKk4/usr/lib/python3.8/site-packages/PySide2/QtWidgets.cpython-38-x86_64-linux-gnu.so(+0x5d6b7f) [0x7fe83f829b7f]                                           
#13  0x7fe866b41062 in QObjectPrivate::deleteChildren() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Core.so.5+0x5a                                                            
#14  0x7fe86737ad2f in QWidget::~QWidget() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libQt5Widgets.so.5+0x3b3                                                                     
#15  /tmp/.mount_FreeCAeYdKk4/usr/lib/python3.8/site-packages/PySide2/QtWidgets.cpython-38-x86_64-linux-gnu.so(+0x4166ef) [0x7fe83f6696ef]                                           
#16  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/./libshiboken2.cpython-38-x86_64-linux-gnu.so.5.13(+0x14d7c) [0x7fe865ebfd7c]                                                           
#17  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(+0x1c0ffa) [0x7fe868872ffa]                                                                                         
#18  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(+0x1364a4) [0x7fe8687e84a4]                                                                                         
#19  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(+0x1c0e24) [0x7fe868872e24]                                                                                         
#20  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(+0x1364ef) [0x7fe8687e84ef]                                                                                         
#21  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/./libshiboken2.cpython-38-x86_64-linux-gnu.so.5.13(+0x158d6) [0x7fe865ec08d6]                                                           
#22  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(+0x1472ac) [0x7fe8687f92ac]                                                                                         
#23  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(+0x1fbc0e) [0x7fe8688adc0e]                                                                                         
#24  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(PyGC_Collect+0x58) [0x7fe8688f2008]                                                                                 
#25  /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libpython3.8.so.1.0(Py_FinalizeEx+0x70) [0x7fe86890bf60]                                                                                
#26  0x7fe868103744 in Base::InterpreterSingleton::finalize() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libFreeCADBase.so+0x14                                                    
#27  0x7fe86856eac7 in App::Application::destruct() from /tmp/.mount_FreeCAeYdKk4/usr/bin/../lib/libFreeCADApp.so+0x1e7                                                              
#28  /tmp/.mount_FreeCAeYdKk4/usr/bin/freecad(+0x3868) [0x555c83dac868]                                                                                                              
#29  /usr/lib/libc.so.6(__libc_start_main+0xf2) [0x7fe8665f5152]                                                                                                                     
#30  /tmp/.mount_FreeCAeYdKk4/usr/bin/freecad(+0x4521) [0x555c83dad521]

But I haven't been able to kill 0.19-23141 or 0.19.23254 with the same malevolence (yet). So this definitely is a plus.

If some "real bugreports" pop out of this thread in a distilled form, I'd like to make sure they are observable across several revisions and not some transient regression.

Post by **Kunda1** » Mon Dec 07, 2020 3:02 pm

So stay in a safe distance for now and observe/or not. Indulge me a little monologue.

There seems to be an explicit level of 'snark' here. Am i right? If so, lets lose the atitude, shall we? The mods are not on a power-trip here, we're keeping the forum moderated and organized.

Zolko · Post by **Zolko** » Mon Dec 07, 2020 3:45 pm

Kunda1 wrote: ↑Mon Dec 07, 2020 3:02 pm There seems to be an explicit level of 'snark' here. Am i right?

Yes you're right, but still, one should listen when such posts appear quite often, 3 in the past 2 days:

My Tapeworm-thread of observed FreeCAD bugs
basic draw...
FreeCAD is no where near ready for serious modeling

dismissing them all as trolling may be missing the possibility that there IS some feedback to consider ?

BIM Bo · Post by **BIM Bo** » Mon Dec 07, 2020 4:02 pm

If you allow, I'd like to ignore the "how" and return to the "about".

------

ad "Right mouse contextual menu - the Jester"

IDK ... it get's worse. (Version: 0.19.23254) now only shows a very very limited menu - see menu_pruned.jpg. Not sure if that could be a configuration mishap between the 0.19-23141 and the newer version as they share the same config files in my home. And of course I do not run both processes in parallel. Although that could be an interesting test... (see below *boom*)

and return to 0.19-23141 ... and yeah. Now that very limited contextual menu remains in 23141 also.

0.19.23254 broke - if not the internet - my contextual menus.

------

Starting two instances of FreeCAD at the same time

Yeah baby, if you're looking for trouble call me! So what happens when one starts two instances of FreeCAD? I mean it happens to the best of us - I have 10 virtual desktops and hell do I know what's happening on the other 9.

So let's start an 0.19-23141 AppImage, and then let's start the compiled 0.19.23254. The latter does not complain about another instance running, so surely there can be no problem? Let's change the Language in 0.19-23141 from English to German and save.

No change in 0.19.23254 visible... ok, so let's click uh ... oh:

Code: Select all

$ QT_SCALE_FACTOR=0.5 FreeCAD
FreeCAD 0.19, Libs: 0.19R23254 (Git)
© Juergen Riegel, Werner Mayer, Yorik van Havre and others 2001-2020
FreeCAD is free and open-source software licensed under the terms of LGPL2+ license.
FreeCAD wouldn't be possible without FreeCAD community.
  #####                 ####  ###   ####  
  #                    #      # #   #   # 
  #     ##  #### ####  #     #   #  #   # 
  ####  # # #  # #  #  #     #####  #   # 
  #     #   #### ####  #    #     # #   # 
  #     #   #    #     #    #     # #   #  ##  ##  ##
  #     #   #### ####   ### #     # ####   ##  ##  ##

Cannot find icon: :/icons/Sketcher_EditSketch.svg
Cannot find icon: :/icons/Sketcher_LeaveSketch.svg
connect failed: No such file or directory
Start Editing Enclosure_for_OctoGX_rev00_3D
Program received signal SIGSEGV, Segmentation fault.
#0  /usr/lib/libc.so.6(+0x3d6a0) [0x7fe9de4816a0]
#1  /usr/lib/libpython3.9.so.1.0(+0x14ab00) [0x7fe9e0beeb00]
#2  /usr/local/lib/libFreeCADBase.so(+0x17bd20) [0x7fe9e13a2d20]
#3  /usr/local/lib/libFreeCADBase.so(+0x17bd64) [0x7fe9e13a2d64]
#4  0x7fe9e13a3345 in Py::_XDECREF(_object*) from /usr/local/lib/libFreeCADBase.so+0x18
#5  0x7fe9e33b122c in Py::Object::release() from /usr/local/lib/libFreeCADGui.so+0x1c
#6  0x7fe9e33b135c in Py::Object::~Object() from /usr/local/lib/libFreeCADGui.so+0x2a
#7  0x7fe9e39a2d5a in Gui::ViewProviderPythonFeatureImp::~ViewProviderPythonFeatureImp() from /usr/local/lib/libFreeCADGui.so+0x11a
#8  0x7fe9ac96ae0d in Gui::ViewProviderPythonFeatureT<PartGui::ViewProviderPart>::~ViewProviderPythonFeatureT() from /usr/local/lib/PartGui.so+0x37
#9  0x7fe9ac96ae76 in Gui::ViewProviderPythonFeatureT<PartGui::ViewProviderPart>::~ViewProviderPythonFeatureT() from /usr/local/lib/PartGui.so+0x18
#10  0x7fe9e3435554 in Gui::Document::~Document() from /usr/local/lib/libFreeCADGui.so+0x368
#11  0x7fe9e34357c0 in Gui::Document::~Document() from /usr/local/lib/libFreeCADGui.so+0x18
#12  0x7fe9e33be690 in std::default_delete<Gui::Document>::operator()(Gui::Document*) const from /usr/local/lib/libFreeCADGui.so+0x28
#13  0x7fe9e33b7b70 in std::unique_ptr<Gui::Document, std::default_delete<Gui::Document> >::~unique_ptr() from /usr/local/lib/libFreeCADGui.so+0x52
#14  0x7fe9e33a34ca in Gui::Application::slotDeleteDocument(App::Document const&) from /usr/local/lib/libFreeCADGui.so+0x1b2
#15  0x7fe9e33f9e03 in boost::_mfi::mf1<void, Gui::Application, App::Document const&>::operator()(Gui::Application*, App::Document const&) const from /usr/local/lib/libFreeCADGui.so+0x71
....

Maybe - just maybe it'd be a good idea to touch a PID file in /var/run/FreeCAD.pid or somesuch if you can't handle parallel instances (right). At least that's how we did it back in the 90ies. Or it's the borked Python 3.9 - I mean cmake did vomit deprecation warnings for Python constructs all over the place. More investigation required.

-------

Sure - it's monkeywrenching, but how else do you want to find bugs? Not with velvet gloves for sure.
And yes, I'm here only because I'm pissed-mad at the Revit licensing policy.

Zolko · Post by **Zolko** » Mon Dec 07, 2020 4:11 pm

BIM Bo wrote: ↑Mon Dec 07, 2020 4:02 pm And yes, I'm here only because I'm pissed-mad at the Revit licensing policy.

you know what : why don't you return using Revit and leave us stupid FreeCAD users in the ignorance of your wisdom ? Why don't you display your vomit in the Revit forums ?

drmacro · Post by **drmacro** » Mon Dec 07, 2020 4:15 pm

Zolko wrote: ↑Mon Dec 07, 2020 3:45 pm ...
dismissing them all as trolling may be missing the possibility that there IS some feedback to consider ?

snark, sarcasm, outright insults, and rants. Or, drop in complain about drawing for 8 hours.
They tend to come in clumps. Typically after some mention on a wider read forum.

Of course there are plenty of things about FreeCAD that can be legitimately complained about. But, if your main complaint is it doesn't look pretty on my FruitProduct and comments about hoping to shame the devs into doing better work...now that's hard to not just ridicule.

Why is pushing back considered dismissal? (you want dismissal, I told a guy in another platform that he was posting about the wrong product and his response? It was just 2 words: "F*@k you".

)

And what feedback should be considered? "Hey you guys should fix this...because it doesn't work the way I think it should" or "hey, I can't figure out how this works, what am I missing?"

The later I'm happy to help, the former gets...something less.

BIM Bo · Post by **BIM Bo** » Mon Dec 07, 2020 4:43 pm

Zolko wrote: ↑Mon Dec 07, 2020 4:11 pm
BIM Bo wrote: ↑Mon Dec 07, 2020 4:02 pm And yes, I'm here only because I'm pissed-mad at the Revit licensing policy.
you know what : why don't you return using Revit and leave us stupid FreeCAD users in the ignorance of your wisdom ? Why don't you display your vomit in the Revit forums ?

I could actually ask you the same about staying away from this thread then. So noting here problems I observe with FreeCAD seems unacceptable to you? I fail to see where I have addressed FreeCAD users as "stupid". Actually ... I even fail to reach for that concept in my memory. (translation: I didn't even think it)
I also fail to see where I mentioned something like "Revit is better than FreeCAD". They are vastly different and I can see where FreeCAD aims at a much bigger world than Revit could ever do. I don't think I would use Revit for doing some models for a 3D printer. Ever.
But I can think of a reality where I could do a BIM project with FreeCAD 0.99.8 instead of Revit 2040.

All I mentioned was, that I am here because I am pissed at the Revit licensing policy (and elsewhere the vendor lock-in). Nothing more. And while I am here I could as well make myself useful and report some things you haven't done so far - obviously.

So my advice good man:
a) please stop the "us versus you" mindset and rhetorics (*)
b) try to do something against your halucinations or seek professional help.
c) actually try to make yourself a little bit more useful by not going into full ad hominem mode

(*) except that is the usual mode of operation here and this community really needs the "read our book - adopt the religion" approach. If it is, I'm sorry I missed the note. Of course it would help if I was able to lock and unlock this thread.

----

It is still my hope, that my feedback could be recognized as valid, because I'm certainly able to not complain about "uh ah FreeCAD doesn't look like some fluffy Apple UI guideline" F@@K that. I mean yeah: I just compiled the master branch today to have a look, I must be a troll.
Here's an idea: Let's define everyone who claims FreeCAD died with a segmentation violation on his computer is a troll.

My Tapeworm-thread of observed FreeCAD bugs

My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs

Re: My Tapeworm-thread of observed FreeCAD bugs