General and Desktop

@l3u_ Could you mark this thread as solved, please?

That solved it. I'm not sure what happened, but something got messed up my build dir.

@AndyBrice I'm using the latest version of Qt and macOS/iOS 26. Nothing is compatible with Qt 6.9 and everything renders in the old version style.

@JonB, that's what I was looking for! Thanks.

Hi and welcome to devnet, Does it happen with only a single file or are they all affected ? It there somewhere a sample file that could be used to test this issue ? Which OS are you running your application on ?

@sierdzio said in Best practice for QTs parent/child memory model and clang-tidy: auto widget = QPointer(new QGraphicsWidget); This just creates a QPointer which gets destroyed on leaving the scope for the sake of satisfying a tool. This is anti-programming in my pov.

That's how I understood the code.

@rcx11 it uses a similar approach to gRPC/QtRemoteObjects in that the files that are compiled are generated using calls to the generators via cmake. I checked Qt6.10.0 on windows 11 and indeed, until you configure the file cannot be found: The file that generated (for reference of others reading this thread): [image: 8bfbacc8-18a1-419d-9f48-7757554d7203.png] I'd recommend marking as resolved with @Paul-Colby 's response.

Hello, is there any update for this issue? I'm also facing the problem that the QEnums properties are not loaded by QtDesigner

It works fine for me with Qt6.9.1 with the windows11 style: int main(int argc, char *argv[]) { QApplication app(argc, argv); QTableWidget tw; tw.setRowCount(5); tw.setColumnCount(1); auto item = new QTableWidgetItem("Item"); tw.setItem(0, 0, item); QTimer::singleShot(2000, &tw, [&]() { tw.item(0, 0)->setCheckState(Qt::Checked); }); tw.show(); return app.exec(); } Please provide a minimal, compilable example of your problem.

@SuhasKrishanamurthy hi, Since you marked your thread as solved, can you share what you did / change to fix the situation ? It might help other people.

@SGaist Thanks for your help. I already have access to QtWebKit. Its there in my Qt installation. Crash is intermittent so inferring what content triggers crash might not be easy. I am trying to figure out again if I can instead use QWebEngine. Lets see, hopefully I wont have to use QtWebKit.

hi @aha_1980 Thanks for the reply. Before I got it, I'd just started doing a thing where I was counting the bytes read back in and only starting to take any notice of the data when I had read in the number of bytes sent out, so I was along the right lines. It's a bit of a pain, but with what you're saying about bytesReceived being called not being a guarantee it's gone out on the line, I guess it's not going to help anyway. For a while I have been thinking I need to get to grips with QSerialPort but as I know qextserialport so well I've been lazy. I do know that in v6 qext... is no longer available. Thanks for the post.

@AndyBrice said in Cache locality and QString: Ok, thanks for the feedback. It looks like there is no straightforward ways to improve performance, while keeping the flexibility I need. Correct. The optimization will come at a cost of working only when on a predictable subset of real world data. Because you've stated that you need a generalizaed solution, the optmization tricks wont work reliably. If you can assign hard limitations to your dataset...THEN you can consider what kinds of optimiztions make sense.

@yonizaf You should not create a new ticket. You commented in the existing, let's see whether this triggers any actions :-)

Code: #include <QCoreApplication> #include <QLocale> #include <QDate> #include <QtDebug> int main(int argc, char* argv[]) { QCoreApplication a(argc, argv); QLocale locale1(QLocale::German, QLocale::Germany); for (int m = 1; m <= 3; m++) { QDate date1(2025, m, 1); qDebug() << "locale1=" << locale1 << " date1=" << date1 << "locale1.toString( date1, \"MMMM\" )=" << locale1.toString(date1, "MMMM"); } const auto locales = QLocale::matchingLocales(QLocale::AnyLanguage, QLocale::AnyScript, QLocale::AnyCountry); for (const QLocale& locale2 : locales) { if (locale2.country() == QLocale::Germany && locale2.language() == QLocale::German) { qDebug() << "locale1 == locale2 =" << ( locale1 == locale2 ); for (int m = 1; m <= 3; m++) { QDate date2(2025, m, 1); qDebug() << "locale2=" << locale2 << " date2=" << date2 << "locale2.toString( date2, \"MMMM\" )=" << locale2.toString(date2, "MMMM"); } } } return 0; } Results: Qt 5.15.18/MSVC2019/Windows: Qt 5.15.18/XCode/macOS: locale1= QLocale(German, Latin, Germany) date1= QDate("2025-01-01") locale1.toString( date1, "MMMM" )= "Januar" locale1= QLocale(German, Latin, Germany) date1= QDate("2025-02-01") locale1.toString( date1, "MMMM" )= "Februar" locale1= QLocale(German, Latin, Germany) date1= QDate("2025-03-01") locale1.toString( date1, "MMMM" )= "März" locale1 == locale2 = true locale2= QLocale(German, Latin, Germany) date2= QDate("2025-01-01") locale2.toString( date2, "MMMM" )= "January" locale2= QLocale(German, Latin, Germany) date2= QDate("2025-02-01") locale2.toString( date2, "MMMM" )= "February" locale2= QLocale(German, Latin, Germany) date2= QDate("2025-03-01") locale2.toString( date2, "MMMM" )= "March" ^^^Month is incorrect for locale2^^^ Qt 6.7.3/MSVC2019/Windows: Qt 6.7.3/XCode/macOS: locale1= QLocale(German, Latin, Germany) date1= QDate("2025-01-01") locale1.toString( date1, "MMMM" )= "Januar" locale1= QLocale(German, Latin, Germany) date1= QDate("2025-02-01") locale1.toString( date1, "MMMM" )= "Februar" locale1= QLocale(German, Latin, Germany) date1= QDate("2025-03-01") locale1.toString( date1, "MMMM" )= "März" locale1 == locale2 = true locale2= QLocale(German, Latin, Germany) date2= QDate("2025-01-01") locale2.toString( date2, "MMMM" )= "Januar" locale2= QLocale(German, Latin, Germany) date2= QDate("2025-02-01") locale2.toString( date2, "MMMM" )= "Februar" locale2= QLocale(German, Latin, Germany) date2= QDate("2025-03-01") locale2.toString( date2, "MMMM" )= "März" ^^^ Month is as expected^^^ So the localised month is wrong (but only inside the loop) for Qt 5.15.18. Weirdly, day of week ("ddd" or "dddd") is correct in both cases. I can work around it, but it is a very odd bug.