Hmmm. You shouldn't have to #include "config.h" in cdio-eject, because read.h includes a public portion of that called <cdio/cdio_config.h>
And as for: #ifdef HAVE_SYS_TYPES_H #include <sys/types.h> #endif I guess that should be in read.h. If this works, we should make change. Steven M. Schultz writes: > > On Sun, 15 Apr 2007, R. Bernstein wrote: > > > I think gcc is trying to say that the types off_t and ssize_t are not > > defined. These you get when unistd.h or stdlib.h are included. I use > > On some systems, such as OSX for one example, you get off_t from > <sys/types.h> > > > ubuntu and don't get these problems. Possibly you don't have the headers > > I run OSX and do get the problem - I've just been manually > patching cdio-eject.c with > > --- cdio-eject.c.dist 2007-03-10 15:17:47.000000000 -0800 > +++ cdio-eject.c 2007-03-10 15:18:17.000000000 -0800 > @@ -1,3 +1,8 @@ > +#include "config.h" > +#ifdef HAVE_SYS_TYPES_H > +#include <sys/types.h> > +#endif > + > #include <stdio.h> > #include <string.h> > #include <cdio/cdio.h> > > That works for any system that has off_t in <sys/types.h> > > Cheers, > Steven Schultz > > > > _______________________________________________ > Libcdio-devel mailing list > [EMAIL PROTECTED] > http://lists.gnu.org/mailman/listinfo/libcdio-devel > _______________________________________________ Libcdio-devel mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/libcdio-devel
