(1) bsd.prog.mk already includes bsd.subdir.mk, and
(2) including bsd.subdir.mk first screws up the clean/cleandir targets
in such a way that the program's objects don't get blown away
when you make clean.
This is a compromise between the Solaris 2.5 naming of this file
(/var/yp/binding/domain/ypservers and /var/yp/binding/domain/vers)
and traditional NetBSD naming (/var/yp/binding/domain.vers).
Changing to the Solaris naming makes the C library incompatible
with ypbind. While it's easy to change the C library, I don't feel
right about introducing a gratuitous incompatibility.
The current name (/var/yp/domain/ypservers) conflicts with the name
of the YP map listing slave servers for a domain.
Per discussion w/ Chris Demetriou and Luke Mewburn.
of YP servers a client should bind to, mostly verbatim, but slightly
modified for better semantics when nagging servers if a ypset has been
issued. Default to broadcast mode if no .ypservers file is present.
Documentation changes to match, slightly tweaked by Scott Reynolds and
myself.
Closes PR #1759.
Implement a better scheme where we `find' the login ttys by looking in
/etc/ttys. Of course this scheme breaks when /etc/ttys changes, but such
is life.
- some cleanups...
lpd run in a mode where the it listens only to the local unix domain
socket and not to the network. Changes are similar but not identical
to the supplied patches.
it to "."
- include sockio.h if needed to define SIOCGIFCONF (for svr4)
- use POSIX signals and wait macros
- add -S silent flag, so that the client does not print messages unless there
is something wrong
- use flock or lockf as appropriate
- use fstatfs or fstatvfs to find out if a filesystem is mounted over nfs,
don't depend on the major() = 255 hack; it only works on legacy systems.
- use gzip -cf to make sure that gzip compresses the file even when the file
would expand.
- punt on defining vsnprintf if _IOSTRG is not defined; use sprintf...
To compile sup on systems other than NetBSD, you'll need a copy of daemon.c,
vis.c, vis.h and sys/cdefs.h. Maybe we should keep those in the distribution?
cgd