Benjamin Piwowarski wrote: > Hi, > > While I was trying to update a patch for LyX, I had troubles compiling LYX > on OS X 10.9 introduces quite some changes: - QT5 is required (as > discussed on the mailing list) - libc++ (instead of libstdc++) is used by > default systemwide > > So here are some patches for compiling LyX on OS X 10.9 that take into > account the above comments:
Hi Benjamin, many thanks for the patches. I'll comment on what I understand, others hopefully can comment on the rest. > [patch 1] is a patch for compiling with clang & libc++ that fixes those > [issues: > - in strfwd.h and debug.h, if libc++ is detected, then do not use forward > std declarations (they break since libc++ define some classes in an > anonymous namespace to provoque linking clashes when linked with > libstdc++) Is <ciso646> guaranteed to exist in C++03? If yes, then this change looks safe. > - conversion from streams to boolean has to be explicit We do not use C-style casts. Could you please rewrite this, e.g. like bool const success = (os << contents) ? true : false; and add comments why it is needed? I could do that, but I can't test whether it still fixes your problem, and the chance is extremely high that I would break something. > [patch 3] QT5 and MacOs > - this adds some QT components for Mac (MacExtras in qt4 and Gui in > support [necessary to compile Package.cpp]) - use QT_OS_MACX instead of > QT_WS_MACX (the latter is not defined anymore in Qt5) - in > GuiWorkArea.cpp, use QProxyStyle instead of QMacStyle - [all platforms > with cmake] during installation, create the qt.conf file + copy the > platform plugin (qcocoa) on Mac I can't comment on the cmake part, but the C++ changes look good. To our Mac expperts: Is there any reason why they are not already in? > [patch 4] graceful libmagic error handling > This patch just report an error if the magic file detection does not work, > and proposes a workaround for the bug I had, related to an incorrect > locale setting Well, it also sets environment variables. Dou you have any documentation why this is needed? Did the fallback to the old code not work in case libmagic failed? > Note that when bundling with cmake, the file magic database should be > copied within the bundle (but it isn't for the moment) I guess this is something for later, since libmagic is not essential. First we need to solve the bug below: > With all this, I was able to have a "working" LyX application (that is, > everything seems to work but the menu - known OS X bug) Is it this one: http://www.lyx.org/trac/ticket/8925 Do you have any pointers to documentation about this bug? > Hope it helps, > Benjamin Yes, this is really helpful. Thank you very much. Georg
