General and Desktop

@MulderR That goes without saying. Obviously we're searching for a way that does not require human translators. QLocale internally has the needed data but somehow fails to address this use case. @Wilk I don't see how that class is relevant to this thread.

sorry, for the late reply, I put the file in "bin \ debug \ platforms " and your build worked with no errors, so I guess it's something wrong with my SDK installation not the QT build (as the error still happens with the X86 version I downloaded from here), I tried to uninstall visual studio SP1 to reinstall the SDK (as microsoft said) but couldn't, so I'm reinstalling windows right now.. after I finish with all the installation and compilation I will inform you with the result, again thanks for your help.

"BUG report":https://bugreports.qt-project.org/browse/QTBUG-28878

Such information will be powered by a database. The best one is to reimplement QTableView (for the view) and QStyledItemDelegate (for displaying information) "QStyledItemDelegate":http://qt-project.org/doc/qt-4.8/qstyleditemdelegate.html "Example - Star":http://qt-project.org/doc/qt-4.8/itemviews-stardelegate.html

Listen for the finished() signal, or process.waitForFinished(-1) Read this: http://qt-project.org/doc/qt-5.0/qtcore/qprocess.html

Why not save the time worrying about it and just write a pseudo random number generator yourself in 10 seconds? Then you know what you've got and for benchmarking you probably don't need high quality random numbers. @r = (r*1103515245+12345) & max;@

more info about my problem: in my setHeaderData() function, I have "emit headerDataChanged(Qt::Horizontal, section, section);", this signal triggered slot QSortFilterProxyModelPrivate::_q_sourceHeaderDataChanged(). But based on QAbstractItemModel Class Reference, "When reimplementing function setHeaderData(), the headerDataChanged() signal must be emitted explicitly." so seems there is nothing wrong with my code.

[quote author="utcenter" date="1356798564"]But is it more unprofessional than providing nothing at all? Especially since it is actually doable, unlike porting the actual source code. This is the one and only line that fixed the project to be Qt5 compatible: @greaterThan(QT_MAJOR_VERSION, 4): QT += widgets@[/quote]That works for a "Qt Gui Application" only, but not the other project types supported by Qt Creator. IMHO, it doesn't make sense to bundle a tool that only works on a small subset of possibilities -- you'd then have to explain to users when they can/can't use the tool. And if users has to read something, why not let them read a short guide on how to port it themselves (which Qt does already provide, so it's not "nothing at all")? I do agree with you that automated porting would be very nice, but it's not the role of Qt Creator. We're going off-topic now, but someone made a start on a standalone tool for source code porting: http://www.kdab.com/automated-porting-from-qt-4-to-qt-5/

The problem does not appear on qt 4.8.3

removing the deletes worked. thank you :)

P.s. It work correctly for Qt 4.7.4 and I think for lesser versions too. In 4.8.4 dialog close after clicking on 1 file, so it need some other hints.

I think the compiler does a fairly good job at inlining on its own, since doing it explicitly results in nothing, either that, or the method is too complex to inline. bq. In high performance code, appending to a string in the inner loop is insane. I reserve plenty of space to ensure I avoid potential reallocation. What I found to be odd is that passing a QChar reference is actually faster than passing a copy, even thou I am using a 64bit build and QChar is only 16bit, worst case scenario performance should be the same but it takes a small hit when passing by value. I compared this algorithm to the performance of word count in MS Word. Keeping in mind word only counts characters and words and does not even go into the details of unique words and word usage count, analyzing the same text in MS Word takes about 4-5 seconds, for my algorithm it takes only 230 msec, and I still have the option to make it multithreaded. But hey, I am pretty happy with having "ONLY" 25x times better performance than a top tier professional text editor that doesn't even do as much work :)

maybe I just understood your saying little wrong. But the imagination to rewrite everything in QPainters little scared me, seems like lots of work but on ther other hand bq. built in collision detection. that sounds actually pretty awesome if it is what I think what it is. My first aim was actually creating a Tower Defense game where you can build the towers anywhere you like and the monster had to find a way through your "maze". but I gave up inventing such kind of algorithm, seemed to be a way too difficult, but if I can use a built in collision detection I may still implement later. But right now I have to postpone my plans to February after exams, QT is just one of six other subjects. So long story, thanks for the information. I'll take a close look at them 4 weeks later.

Dear Gerolf and Andre... Thank you for the links... They are to the point and very informative. Solved many of my confusions. I understood why the things in above code is not working. [and sorry for repeating the same question again n again... :) ] marking this as solved

i have the same problems.