On Mon, 30 Nov 1998 01:43:38 +0100 (MET) 
 Christoph Badura <bad@ora.de> wrote:

 > I don't like that solution.  I'd prefer if the "random crap" stopped
 > including sys/signal.h.  From the perl failure I guess we are the only
 > OS, uhm, "doing it" that way.

I agree; stuff simply should NOT be including <sys/signal.h>.  It's been
that way for a long time.

Maybe we could be a little more hard-nosed about it:

#if !defined(_KERNEL) && !defined(_SIGNAL_H_)
#error You must include <signal.h> for signal function prototypes.
#endif

 > Another option would be to remove libc/arch/$MACHINE/sys/sigaction.S and
 > libc/gen/sigsetops.c from the source and bump the libc major number.
 > (The log messages on these files aren't helpful to me.)
 > That would prevent programs which fail to include signal.h from linking.
 > 
 > Of course, we could implement both changes and fix things for good. :-)

No, when we bump the major number, all of the renames go away...

Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                            Home: +1 408 866 1912
NAS: M/S 258-5                                       Work: +1 650 604 0935
Moffett Field, CA 94035                             Pager: +1 650 940 5942