Bill Studenmund <wrstuden@nas.nasa.gov> writes:

> Is there a standards ruling on this? We've had to add patches to the
> netatalk package to include <signal.h> rather than <sys/signal.h>. I want
> to feed them back to the maintianers, and wanted to be able to make an
> arguement that the change should be global. :-)

As far as userland API standards are concerned, <sys/signal.h> does
not exist.