In the mlt instructions, we are passing
"--qt-includedir=$QT5DIR/include". This is wrong. Before adding
something like this in the future, please test (or ask somebody else to
test) for the system install. That said, from what I can see, it is
probably not necessary anyway.
My read of src/modules/qt/configure says pkg-config should find it (and
it does on a system install). Can somebody see if those last two lines
are still necessary if QT is installed into /opt and PKG_CONFIG_PATH is
updated (Qt5Gui.pc)?
As an aside, and mostly for Bruce (but others are absolutely welcome to
comment), what would you say to unifying the install of QT5, making it
match a system install? This breakage in the book brings to the
forefront why it might be nice. This would basically mean supplying the
following to configure (for both camps):
./configure -prefix $QT5PREFIX-sysconfdir /etc/xdg \
-confirm-license -opensource -dbus-linked \
-openssl-linked -system-harfbuzz -system-sqlite \
-nomake examples -no-rpath -skip qtwebengine \
-archdatadir $QT5PREFIX/lib/qt5 \
-bindir $QT5PREFIX/bin \
-plugindir $QT5PREFIX/lib/qt5/plugins \
-importdir $QT5PREFIX/lib/qt5/imports \
-headerdir $QT5PREFIX/include/qt5 \
-datadir $QT5PREFIX/share/qt5 \
-docdir /usr/share/doc/qt5-&qt5-version; \
-translationdir $QT5PREFIX/share/qt5/translations \
-examplesdir /usr/share/doc/qt5-&qt5-version;/examples &&
Obviously, the above needs to be tested yet, and I wouldn't want to do
it until at least the next QT update, but I understand the reluctance to
do so WRT how pretty the install looks in /opt.There may be some room to
go the other way as well - really, we could drop all of the qt5
suffixes, only headerdir is really invasive in the /usr hierarchy (funny
that is the variance that caused the breakage in the book). That would
result in, I believe, only headerdir, docdir, and examplesdir being
passed in.
--DJ
--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
