[It is possible this belongs more on the development list, but I thought I'd play safe and post it here first]
Qt uses a QT_NO_EXCEPTIONS preprocessor macro to decide whether to enable various exception-related code, both in headers and in its source code. qmake defines QT_NO_EXCEPTIONS if and only if you tell it to disable exceptions *for your application* (regardless of whether Qt itself was compiled with exceptions). When compiling with GCC, qglobal.h will also define QT_NO_EXCEPTIONS if -fno-exceptions was passed. This means that QT_NO_EXCEPTIONS matches whether the user of Qt is being compiled with exceptions, rather than Qt itself, and so the headers may not always match the library code. An obvious example is that you may end up with QException declared but not defined, and various header-defined methods of, say, QFutures look like they could call the potentially-undefined QFutureInterfaceBase::reportException. So my question: is this sane/intended behaviour? Or am I missing something? I mainly ask because we currently mess with the QT_NO_EXCEPTIONS define in the standard CMake setup for KDE software, and I have a suspicion that both we and qmake are Doing It Wrong. Alex _______________________________________________ Interest mailing list Interest@qt-project.org http://lists.qt-project.org/mailman/listinfo/interest