On Thu, Feb 20, 2014 at 9:42 AM, Martin Jansa <[email protected]> wrote: > On Thu, Feb 20, 2014 at 09:22:03AM -0300, Otavio Salvador wrote: >> Martin, >> >> On Thu, Feb 20, 2014 at 9:18 AM, Martin Jansa <[email protected]> wrote: >> > On Wed, Feb 19, 2014 at 12:22:36AM -0300, Otavio Salvador wrote: >> >> When looking for Qt5 components and files we ought to use the Qt5 >> >> specific path. In case the distribution wants to have a flat tree it >> >> can set QT_DIR_NAME and it will still works as expected. >> > >> > NAK >> > >> > I don't think this is good idea, end-user applications shouldn't install >> > their data, archdata, examples etc in qt5 subdirectories. >> > >> > https://github.com/meta-qt5/meta-qt5/wiki/Building-with-OE#wiki-bad-include-paths-and-build-failing-to-find-qt5-headers >> > >> > If you don't care about qt4/qt5 separation just set empty QT_DIR_NAME >> > and all qtmodules will fold into default directories and you won't need >> > >> > OE_QMAKE_PATH_HEADERS = "${OE_QMAKE_PATH_QT_HEADERS}" >> > >> > listed in above wiki. >> >> and what about we change just the headers? I want to avoid every >> recipe using Qt5 and qmake to require setting the headers var. > > That is still quite bad work around, because ideally end-user apps > shouldn't install headers in qt5 dirs as well. > > So having this line explicitly in the recipe at least shows that app > headers will be installed in weird qt5 location (so your other recipes > should search for it there).
If a system is splitting Qt5 and Qt4 I don't think this is wrong. I understand your point but I think this causes no harm for user and easy the general usage. -- Otavio Salvador O.S. Systems http://www.ossystems.com.br http://code.ossystems.com.br Mobile: +55 (53) 9981-7854 Mobile: +1 (347) 903-9750 _______________________________________________ Openembedded-devel mailing list [email protected] http://lists.openembedded.org/mailman/listinfo/openembedded-devel
