I'm running Arch Linux. This is a rolling release distribution and so I'm running the latest GCC (v 4.7).
Unfortunately Cafu fails to compile now. The problem seems to be related to wxwidgets. If I replace wxwidgets with the latest version of it, it compiles, but afterwards Cafu fails again to build, because of compatibility issues.
I found this link and this seems to explain a solution to this problem: https://groups.google.com/forum/#!msg/w ... Fib2BZWP8J
If it would be awesome, if you could fix this asap.
Joe (teammember of Haimis staff )
PS: We created a ticket: http://trac.cafu.de/ticket/114
I'll look into it later today.
Unfortunately another issue has occured in the MainCanvas.cpp now:
Code: Select all
g++ -o build/linux2/g++/x86_64/release/Common/WorldMan.o -c -O3 -funsigned-char -Wall -Werror -Wno-char-subscripts -fno-strict-aliasing -pthread -DNDEBUG -DSCONS_BUILD_DIR=build/linux2/g++/x86_64/release -D_FILE_OFFSET_BITS=64 -D__WXGTK__ -ILibs -IExtLibs -IExtLibs/wxWidgets/build-gtk/lib/wx/include/gtk2-unicode-static-2.9 -IExtLibs/wxWidgets/include -Ibuild/linux2/g++/x86_64/release/ExtLibs/lua/src -IExtLibs/lua/src Common/WorldMan.cpp Ca3DE/Server/Server.cpp: In destructor 'ServerT::~ServerT()': Ca3DE/Server/Server.cpp:180:29: error: 'close' was not declared in this scope scons: *** [build/linux2/g++/x86_64/release/Ca3DE/Server/Server.o] Error 1 Ca3DE/MainCanvas.cpp: In destructor 'virtual MainCanvasT::~MainCanvasT()': Ca3DE/MainCanvas.cpp:149:16: error: deleting object of polymorphic class type 'SvGuiCallbT' which has non-virtual destructor might cause undefined behaviour [-Werror=delete-non-virtual-dtor] cc1plus: all warnings being treated as errors scons: *** [build/linux2/g++/x86_64/release/Ca3DE/MainCanvas.o] Error 1 scons: building terminated because of errors.
Still running Arch Linux 64bit .
I've encountered the same issue as the thread starter, and we are of the same distro, compiler and architecture (Arch Linux x86_64, GCC 4.7). The solution in http://trac.cafu.de/ticket/115 works. As with the thread starter is concerned, I've also encountered those "not declared" errors. I've managed to silence the errors by including unistd.h, though I'm not sure if that's the proper fix.
The compilation went smoothly after that. Btw, I'm compiling the vanilla source.
You are right, the error with close() is our fault, we'll look into that. The funny thing is, that Haimi is able to compile the code without facing the above error message. But I guess this might be again related to my system (GCC v4.7)...
I have installed GCC 4.4 parallelly now and I'm going to test, whether this might help.
Thanks again for your help!
Regarding #115 / the dtor, with each new major release, GCC tends to provide more thorough warnings, which is usually a good thing.
close(), it may be possible that the error is a result of their cleaning up the header files as well; similar changes have happened in the past. If so, we'd need to include the proper header, but as I have no access to GCC 4.7 at this time, I really cannot tell anything specific about it.
It's perfectly possible that including
unistd.his in fact the proper solution. If so, could you please post or submit a patch?
Thanks, but why didn't you put the changes intotuldok89 wrote:ok then. here's my patch
And why include
io.h, it's the nearest win32 equivalent of
unlink(), etc.). If it's uneeded for windows targets, then I should just include
unistd.hfor linux systems.
Users browsing this forum: No registered users and 3 guests