Hi @CodeFreaks

I would like to address your design. It is extremely confusing. It is great you are checking port existence.

Since you close them before you re-open them as-needed, I would 1) Not worry about doing new QSerialPort into a temp pointer and 2) connecting it at this point.

Just create a stack-instance (or use 'new'), do your check, close the port and if you used 'new' delete it right away. This will ensure you keep your resources clear and free. All you need to do is make your like of valid ports.

When you finally open the port for use, do your connect. This will ensure your signal and its sender are valid. Doing your connect on the temp pointer (which you delete) frees that connection. Once you are sure you have your ports opened correctly you can use sender () in your slot as follows:

QSerialPort* port = dynamic_cast <QSerialPort*> (sender ()); // You can also use qobject_cast
if (port) {
...
}

I use multiple ports with auto-baud detection constantly along with CAN and UDP so you are close and you have no fear! This is all good.

@dheerendra ... Thank you! Sounds like a good idea. I finally ended up using "PRAGMA schema_version" as described
here.

@ManiRon
As I said, you will have to look, I think, at: instead of making the QTextDocument send the PDF it can produce straight to the printer without change, rather you must "intercept" it and do whatever you need to the PDF header first.

@abanksdev
like its done in the Drawer docs.
Don't apply a transition to x, but simply calculate the width analogous to the code from the example.

@pyuxing
That looks OK. Suggest you rename that variable to first_row, because that's what is.

All, the mystery is solved after I asked Qt support about the issue. Turns out, it is associated with the lupdate routine not finding the header files for the source files scanned. In order to get the correct and fully namespace qualified context, you'll have to run this (assuming your sources are in a folder src):

lupdate ./src -ts test.ts -I ./src

With the -I flag, the include path for the header file is handed to lupdate and with this information it produces a correct ts file.

Thanks all for trying and your help!

If you always use auto connection you can use Q_ASSERT(this->thread() == qApp->thread());

You can also use Q_ASSERT(QThread::currentThread() == qApp->thread());

@jsulm Thank you.It's working.

I have the same problem and it seems like the problem is caused by this firefox bug: https://bugzilla.mozilla.org/show_bug.cgi?format=default&id=1323302
(and the resulting change: https://hg.mozilla.org/mozilla-central/rev/90a06cdcb48f)

It seems like xdg-open uses firefox in background to display the PDF files.

Unfortunately it seems like the firefox team has not provided a possibility to bypass this behavior.
(Chrome Browser for example provides "--no-sandbox" option to start it as root in a non-root user session)

@jsulm , @SGaist , @JKSH , Thank you guys

Thanks for reply! I got it.

=) I wasn't trying to be mean any, just humorous/wow/ how'd he even find that one factor haha.

You're right @MrShawn, I missed the plugins trick, thanks so much.
I will refer more on this topic, and when it's clear I will mark this topic as solved "إن شاء الله".

Update (Some progress).

I added (setSizeAdjustPolicy)(QAbstractScrollArea::AdjustToContents) to my initialization.
Now:

When MyDialog is first displayed all columns of MyTable are showing. (i.e. MyDaalog has increased it width). No horizontal scroll bas is displayed.

But:

I repopulate MyTable. On of the columns fields (Item) is now wider than before. MyDialog does not widen to accommodate the new width (the horizontal scroll bas appears) I see in my sizeHint() that MyTable is actually reporting itself as wider.

I am still at a loss!

thanks

Ok in fact the return of QTest::qExec kind of give the number of Failed Unit Tests.
So we can do manually something like this:

int main(int argc, char *argv[]) { QTEST_SET_MAIN_SOURCE_PATH QList<QString> failedTests; ushort nbTests = 0, nbTestsOK = 0, nbTestsKO = 0; uint nbFailedUnitTests = 0; #ifdef __Launch_Test1__ { Test1 test; int res = QTest::qExec(&test, argc, argv); qDebug() << "[Test1] QTest::qExec return: " << res; if (res == 0) ++nbTestsOK; else { ++nbTestsKO; failedTests << "Test1"; } ++nbTests; nbFailedUnitTests += res; } #endif #ifdef __Launch_Test2__ { Test2 test; int res = QTest::qExec(&test, argc, argv); qDebug() << "[Test2] QTest::qExec return: " << res; if (res == 0) ++nbTestsOK; else { ++nbTestsKO; failedTests << "Test2"; } ++nbTests; nbFailedUnitTests += res; } #endif qDebug() << "\n\n\n\n"; qDebug() << "###############################################"; qDebug() << "# Overall Tests Results:"; qDebug() << "# - number of macro Tests: " << nbTests; qDebug() << "# - number of macro Tests OK: " << nbTestsOK; qDebug() << "# - number of macro Tests FAILED: " << nbTestsKO; if (!failedTests.isEmpty()) { qDebug() << "# - number of Unit Tests FAILED: " << nbFailedUnitTests; qDebug() << "# List of FAILED macro Tests: " << failedTests.join(", "); } qDebug() << "###############################################"; return 0; }

we just can't have the global number of unit tests that have been run...
bit a shame but I'll live with it ;)

Iterate of all top level items with the help of QTreeWidget::topLevelItem(int) and go down recursively to the third level

@Asperamanca thanks.it sloved.