On 2006-01-26 00:03:27 +0900, Austin Ziegler wrote: > On 25/01/06, nobu / ruby-lang.org <nobu / ruby-lang.org> wrote: > > At Tue, 24 Jan 2006 23:45:35 +0900, > > noreply / rubyforge.org wrote in [ruby-core:07195]: > > > if a program now includes unistd and ruby.h you get the following error: > > > /usr/include/unistd.h:266: error: declaration of 'int eaccess(const char*, int) throw ()' throws different exceptions > > > > Why `throw' in C? > > It may be compiled with a C++ compiler and "throw" is somehow defined > to do nothing ;) > > IMO, Ruby shouldn't be defining POSIX functions except on platforms > that don't support them, and probably not even then because of > possible conflicts. Look what happened with someone who tried to build > something with both Ruby and the APR recently and a definition of > pid_t instead of rb_pid_t. yes. as mentioned on #ruby-lang we triggered it with compiling a koffice svn snapshot with ruby support. darix