The more weird output there is, the harder for bug reporters to know what to report... nomicons/wink

I dont know if this is of any help, but this is output in the terminal from where i start Qet and load my schematic:

QObject::connect: Cannot queue arguments of type 'QVector<int>'
(Make sure 'QVector<int>' is registered using qRegisterMetaType().)
Frozen Label "Tillfällig, gummislang"
Frozen Label "Tillfällig, gummislang"
Frozen Label ""
Frozen Label ""
Frozen Label ""
Frozen Label ""
Frozen Label ""
Frozen Label ""
QObject::connect: Cannot queue arguments of type 'QVector<int>'
(Make sure 'QVector<int>' is registered using qRegisterMetaType().)
QObject::connect: Cannot queue arguments of type 'QVector<int>'
(Make sure 'QVector<int>' is registered using qRegisterMetaType().)

Ah, thanks i did not think of trying it outside QtCreator.  There it works nomicons/smile
-  now I have to understand how to use it nomicons/wink

Yes i may help a little bit with the wiki
Hmm i find no english section yet?

scorpio810 wrote:
Morganol wrote:

Until then, clang do think there are at least potential problems with Qet code.  Have you checked?

Yes of course, and clang is very verbose.
It 's found ~ 45 warnings, only ~ 4 with GCC

Ah, good nomicons/smile

..and another crash when saving, with debug source screenshot

I am trying to get this working.

I suggest the descriptions for getting this to work should be summoned up in a wiki article.
And the Qet program show a link to it when the menu chouce to launch it is made and it is not installed.
(now it shows a few instructions specific to ubuntu/debian)

The instructions should be distro agnostic...

Here on Mageia6 i tried the following:
pip is here called pip3 if it is the python 3 version we want.
After finding and installing that i did:

# pip3 install qet_tb_generator
Collecting qet-tb-generator                                                                  
  Downloading qet_tb_generator-0.5.3-py3-none-any.whl                                        
Collecting pyqt5 (from qet-tb-generator)                                                     
  Could not find a version that satisfies the requirement pyqt5 (from qet-tb-generator) (from versions: )                                                                                 
No matching distribution found for pyqt5 (from qet-tb-generator)                             
You are using pip version 8.0.2, however version 8.1.2 is available.                         
You should consider upgrading via the 'pip install --upgrade pip' command.                   
# pip install --upgrade pip
bash: pip: kommandot finns inte                                                              
# pip3 install --upgrade pip                                
Collecting pip                                                                               
  Downloading pip-8.1.2-py2.py3-none-any.whl (1.2MB)                                         
    100% |████████████████████████████████| 1.2MB 235kB/s 
Installing collected packages: pip                                                           
  Found existing installation: pip 8.0.2                                                     
    Uninstalling pip-8.0.2:                                                                  
      Successfully uninstalled pip-8.0.2
Successfully installed pip-8.1.2
[root@svarten qelectrotech-trunk]# pip3 install qet_tb_generator
Collecting qet_tb_generator
  Using cached qet_tb_generator-0.5.3-py3-none-any.whl
Collecting pyqt5 (from qet_tb_generator)
  Downloading PyQt5-5.7-cp35-cp35m-manylinux1_x86_64.whl (89.8MB)
    100% |████████████████████████████████| 89.8MB 15kB/s 
Collecting sip (from pyqt5->qet_tb_generator)
  Downloading sip-4.18.1-cp35-cp35m-manylinux1_x86_64.whl (60kB)
    100% |████████████████████████████████| 61kB 790kB/s 
Installing collected packages: sip, pyqt5, qet-tb-generator
Successfully installed pyqt5-5.7 qet-tb-generator-0.5.3 sip-4.18.1

So it was so easy??
Well, i guess it was supposed to, but:

When i in Qet r4736 select Menu "Project" > "Launch the Terminal Block creation plugin" i get no visual response, but an error output (seen in QtCreator from where i run Qet in debugging mode)

This application failed to start because it could not find or load the Qt platform plugin "xcb"
 in "".
 Available platform plugins are: eglfs, linuxfb, minimal, minimalegl, offscreen, xcb.
 Reinstalling the application may fix this problem.

I have pretty many *cxb* packages installed... could it be somewhat more specific... and what does it mean by it look for it in "" ?

OK I guessed so nomicons/smile
XML i nice
I may need netlist in future so i made a feature request:

Crash and other issues when editing an arc * REPEATABLE *

A crash when saving

(I asked in another thread but the question drowned)

Now I am writing descriptions, bunches of text, on Qet drawings.
It there a way to make the text line wrap automatically?

I want the textblock to have a certain width, and wrap the lines automatically.

Great, Joshua nomicons/smile
I am running 4736 in the same way now, will report if anything happens.
Please update Bug 98

Hooray, i caught a crash: https://qelectrotech.org/bugtracker/view.php?id=98
(a type i have not experienced before, but anyway...)

Another crash at saving document after i saw that a conductor did not get properly deleted.
As i see it we have at least two problems

1) something goes wrong in the in-memory drawing

2) it crash when/after saving a drawing with the the in-memory drawing
It is interesting and lucky that the saved drawing is OK

The debugger seem unable to spot 1)

And when crashing i only see dissassembly - not part of Qet source code?

Yes i have heard others complaining about Qt quality...

I will try to catch bugs with Qet source code.

Until then, clang do think there are at least potential problems with Qet code.  Have you checked?

( I am no programmer of this kind, so do not know much about it... just playing with the tools...)


(14 replies, posted in News)

Great looking nomicons/smile

As for support myself is rather broke but pleased to test and feedback, and some elements later

BTW, noticed a couple details in content that need be updated @ https://www.qelectrotech.org/contact.html
Under "Technical choices" it still say Qt4.4, and in "Current state" it say 0.4 is latest.

While fooling around in the menus i decided to try "Valgrind analyzer with gdb".
In the QtDesigner "Application Output" i then see things like

==9512== Conditional jump or move depends on uninitialised value(s)
==9512==    at 0x10A298CA: ??? (in /usr/lib64/libfreetype.so.6.12.6)
==9512==  Uninitialised value was created by a stack allocation
==9512==    at 0x10A29652: ??? (in /usr/lib64/libfreetype.so.6.12.6)
==9512== (action on error) vgdb me ...

Then i clicked the |> button a couple times more, got i.e
==9512== Syscall param writev(vector[...]) points to uninitialised byte(s)
==9512==    at 0x6E7D7CD: ??? (in /usr/lib64/libc-2.22.so)
==9512==    by 0xA865BC8: ??? (in /usr/lib64/libxcb.so.1.1.0)
==9512==  Uninitialised value was created by a stack allocation
==9512==    at 0xEA2E3F0: non-virtual thunk to QXcbWindow::setVisible(bool) (in /usr/lib64/libQt5XcbQpa.so.5.6.1)
==9512== Conditional jump or move depends on uninitialised value(s)
==9512== at 0x416DE2C: QtConcurrent::BlockSizeManagerV2::timeBeforeUser() (in /usr/lib64/libQt5Concurrent.so.5.6.1)
==9512== Uninitialised value was created by a stack allocation
==9512== at 0x67428D: QtConcurrent::IterateKernel<QList<ElementCollectionItem*>::iterator, void>::forThreadFunction() (qtconcurrentiteratekernel.h:217)

... and more.  About seven stops before QET is up and ready to use.
Are those issues harmless or problems waiting to happen?

Now before reading the above post i was using it and the bug triggered: i deleted a conductor and made another conductor, suddenly the deleted conductor was visible again.
I could "select" it so it gets red with blue handles but can not do anything with it.
If i select and move an element the conductor is still shown like selected.
Moving an element on top of conductor deletes that covered part visually.
Then i decided to take a screen-shot, but on screenshot the whole deleted conductor is repainted, looking like normally selected.
I then decided to save the drawing (under new name): crash.

Opening the schematic it is correctly saved; the deleted conductor is not there, and the new conductor drawn after the deletion it there OK.

This may narrow down the problem searching a bit:
Interestingly, the Menu > "File" > "Recently Opened" list do NOT contain the new file.
So it crashed after saving everything to new file, but before updating wherever it stores that list.

Now i will try to make QtCreator debug show source code... (beginning by reading the above nomicons/wink )
So far I have gdb (and also valgrind) installed and configured in QtCreator (at least correct executable paths)

EDIT: Note to self: also installed lib64qt5platformsupport-devel as i saw when started from terminal that QtDesigner was looking for libQt5PlatformSupport.prl  /EDITp, li { white-space: pre-wrap;

I have never used Qt Creator...
Is there some way i can compile qet to include debugging information?
(and then just easily run it in gdb)

Anyhow... i installed qt-creator, and opened qelectrotech.pro in the ...trunk folder, started debugging, and as excerise just closed QET  it to get the problem of previous post.  See screenshot below.

EDIT: OH, and there was also this in the Application Output window:
QObject::connect: Cannot queue arguments of type 'QVector<int>'
(Make sure 'QVector<int>' is registered using qRegisterMetaType().)

Do that tell you something?
What would you like me to do now?

Until answered i just try tu use QET as normal - run in debug this way, but i do not know how to get interesting parts when it crashes.

OK, iĺl try more later.
For now just starting it in gdb and closing by the (x) window button, it catches this:

Thread 1 "qelectrotech" received signal SIGSEGV, Segmentation fault.
0x00007ffff7340313 in QIcon::~QIcon() () from /lib64/libQt5Gui.so.5

OK i start launching it with gdb.

I compile locally, as per https://wiki.mageia.org/en/QElectroTech
Is there something i can do to increase valuable output?

Right now i got a crash in *element* editor while by mouse changing the ovality of an arc, in a element i was modifying (so i have been deleting another arc and spossibly a line in it too, before crash)

I also an hour ago got the crash when saving after having deleted a line that was still visible, like in https://qelectrotech.org/forum/viewtopi … 5081#p5081 that you already know.

It rarely crashes, but it do.   version 0.51 trunk svn 4734

I can not define a repeatable crash instruction yet, but it may have something with group copy, move copied group, undo, undo, and then move a part of the original group

In journalctl i see this:

okt 10 06:56:53 svarten audit[2731]: ANOM_ABEND auid=10702 uid=10702 gid=10790 ses=3 pid=2731 comm="qelectrotech" exe="/usr/local/bin/qelectrotech" sig=6
okt 10 06:56:53 svarten kernel: audit: type=1701 audit(1476075413.331:758): auid=10702 uid=10702 gid=10790 ses=3 pid=2731 comm="qelectrotech" exe="/usr/local/bin/qelectrotech" sig=6

Is there some more metrics i can provide?

Can i run it in some debug mode to help catching the cause better?

label in front:  https://qelectrotech.org/bugtracker/view.php?id=87  <---- UPDATE: this is fixed now, thank you! nomicons/smile

ofset: https://qelectrotech.org/bugtracker/view.php?id=91

...actually i see that when changing the order of folios, their folio id IS automatically updated nomicons/smile

Thank you Yuri for taking your time to explain it so well nomicons/smile