Hi Bruno, > Your email is hardly readable, because
I'm sorry for that. I hope it is fixed now. > > _spawnvp(), or _wspawnvp() are not returning a pid. It is a process handle. > > No one claimed that _spawnvp() is returning a pid. What I wan't to point out is that in gnulib on WIN32 API pid_t is not a process id like it is on linux. Functions in gnulib that are using pid_t eg waitpid() accepting a process handle instead. I'm using parity[1] in an gentoo prefix environment to compile in linux like style win32 as win64 applications. parity is a wrapper around the visual stdio compiler and it defines pid_t as int. Because getpid() return also int. And GetCurrentProcessId() as well. So if I compile a project that is using gnulib I might get a compiler error because of the different definition of pid_t. (In best case. Depending on the situation, the programm will simple crash) For my use case a better solution would be a compile time test in the configure script that tests the existance of pid_t. Another solution would be not using pid_t in gnulib at all. Just redefining it in a way like #if ( defined(_WIN64) || _defined(_WIN32) ) && !defined(__CYGWIN__) intprt_t GNULIB_PID_HANDLE #else pid_t GNULIB_PID_HANDLE #endif What do you think? Martin [1] https://github.com/mduft/parity
