On Fri, Sep 21, 2012 at 6:04 PM, Pavel Sanda <sa...@lyx.org> wrote: > Lars Gullik Bj?nnes wrote: >> | diff --git a/src/Server.cpp b/src/Server.cpp >> | index 7ec096e..b4a0d11 100644 >> | --- a/src/Server.cpp >> | +++ b/src/Server.cpp >> | @@ -1010,12 +1010,16 @@ bool LyXComm::loadFilesInOtherInstance() >> | // Wait a while to allow time for the other >> | // instance to reset the connection >> | Sleep::millisec(200); >> | - pipefd = ::open(pipe.toFilesystemEncoding().c_str(), O_WRONLY); >> | + pipefd = ::open(pipe.toFilesystemEncoding().data(), O_WRONLY); >> >> This is wrong. > > I wonder what is the need to fix such thing in branch. For improving the code > we have trunk. Branch should remain as stable as possible, so unless we fix > some > real bug, its just asking for regressions without good reason. Pavel
After re-reading this response, I think you are suggesting that it is not worth it worrying about the warnings because they are not harmful and that by trying to fix them, something could go wrong. The reason I wanted to fix the warnings on branch is because when I compile a program as a user, I trust it more when it does not have warnings. To me it shows that the developers are careful. But I wonder if this is just the opinion of an inexperienced programmer who pays too much attention to unimportant things. For example, upon seeing that the find_matching_brace function is defined but not being used, I don't want users to have the same reaction as these folks when finding out that the airplane that they're on doesn't have a left phalange: http://www.youtube.com/watch?v=DrwVB4vMx-Q This isn't a strong opinion at all though. I'm just curious what others think. Thanks, Scott
