EDIT (removed wrong solution). EDIT (to add this other option):
Another way to use it would be subclass QThread since it has protected *sleep methods.
QThread::usleep(unsigned long microseconds);
QThread::msleep(unsigned long milliseconds);
QThread::sleep(unsigned long second);
Here's the code to create your own *sleep method.
#include <QThread>
class Sleeper : public QThread
{
public:
static void usleep(unsigned long usecs){QThread::usleep(usecs);}
static void msleep(unsigned long msecs){QThread::msleep(msecs);}
static void sleep(unsigned long secs){QThread::sleep(secs);}
};
and you call it by doing this:
Sleeper::usleep(10);
Sleeper::msleep(10);
Sleeper::sleep(10);
This would give you a delay of 10 microseconds, 10 milliseconds or 10 seconds, accordingly. If the underlying operating system timers support the resolution.
As an update of @Live's answer, for Qt = 5.2 there is no more need to subclass QThread, as now the sleep functions are public:
Static Public Members
QThread * currentThread()Qt::HANDLE currentThreadId()int idealThreadCount()void msleep(unsigned long msecs)void sleep(unsigned long secs)void usleep(unsigned long usecs)void yieldCurrentThread()
cf http://qt-project.org/doc/qt-5/qthread.html#static-public-members
C++11 has some portable timer stuff. Check out sleep_for.
So this question is nearly 10 years old, but it popped up on one of my searches, and I think that there are better solutions when programming in Qt: Signals & slots, timers, and finite state machines. The delays that are required can be implemented without sleeping the application in a way that interrupts other functions, and without concurrent programming and without spinning the processor - the Qt application will sleep when there are no events to process.
A hack for this is to have a sequence of timers with their timeout() signal connected to the slot for the event, which then kicks off the second timer. This is nice because it is simple. It's not so nice because it quickly becomes difficult to troubleshoot and maintain if there are logical branches, which there generally will be outside of any toy example.
A better, more flexible option is the State Machine infrastructure within Qt. There you can configure an framework for an arbitrary sequence of events with multiple states and branches. An FSM is much easier to define, expand and maintain over time.
Source: Stackoverflow.com