On 10 April 2014 19:20, Dirk Hohndel <[email protected]> wrote: > > Thiago, > > I believe this is largely borrowed from your code - can you work with > Joshua on good / correct attribution of the code origin. Also, would you > do a quick review of the code? > > I asked Joshua to send it here so I could test it and everyone could > provide input. Robert (or any of the other main Mac based developers), > can you make sure this makes sense? Lubomir, what about your native > builds on Windows, does this work? > > Thanks everyone for your help
joshua, here are some errors that i get on windows. errors: ---------- 1) QSysInfo::WV_NT should be used instead of WV_NT and similar. ---------- 2) there is a typo near: - case WV_WINDOWS8: - ros_details.versionString = "8"; ^ 'ros instead of 'os' ---------- 3) 'WV_WINDOWS8_1' is not part of Qt 4.8. i think it should be wrapped in: #if QT_VERSION >= QT_VERSION_CHECK(5, 0, 0) ... #endif ---------- 4) all the switch() cases 'under #ifdef Q_OS_WIN' need 'break;' or it will fall to: os_details.versionString = "CE6"; ---------- non-errors / ideas: 1) i don't think these are needed, as i don't see actual debug code or error handling: #if !defined(Q_OS_WINCE) #include <errno.h> #if defined(Q_CC_MSVC) #include <crtdbg.h> #endif also we don't really support MSVC. ---------- 2) i'm pretty sure that #ifdef Q_OS_UNIX will branch OSX as well, so perhaps Q_OS_MAC / Q_OS_LINUX have to be used. also a quick google search for similar code shows people not using "/etc/os-release" for OSX detection in favour of a shell script: "sw_vers -productVersion" ---------- i've tried building on linux but i get some linker errors with: libgit2-0 0.18.0 will have to find why is that. lubomir -- _______________________________________________ subsurface mailing list [email protected] http://lists.hohndel.org/cgi-bin/mailman/listinfo/subsurface
