Thus spake Greg A. Woods ("GAW> ") sometime Today...

GAW> I've probably seen many more network configs than you!  ;-)

I wouldn't doubt that one, but only by a little.

GAW> > NIS kind of needs a domainname in order to run properly (hint: ypbind
GAW> > doesn't take a "-d domain" argument, last I looked, although "ypserv"
GAW> > does...).
GAW>
GAW> Note I said "NIS domains" -- I didn't say "no domainname", nor did I
GAW> mean to imply that "domainname" returned an empty string.

NIS domains are bound, and they need a domainname as set/returned by
[sg]etdomainname().  Now either you are saying it is light, or I am saying
it is not dark.

GAW> Note "ypserv"'s -d option has nothing to do with domainname, but rather
GAW> is how you tell it to use the DNS for host information (which on systems
GAW> with nsswitch.conf can usually only cause more confusion ;-).

Oh, hells.  I'm sorry, you're right (ERTFM).  I was sure, though, that
ypserv could serve multiple domains.  Maybe that's just the /var/yp stuff;
yes, that would be it.

{
	int eat = open("crow", O_RDWR|O_CREAT);
	if (eat >= 0) {
		write(eat, "*sigh* Sorry about that.\n", 25);
	}
	return (0);
}

GAW>   NOTA BENE
GAW>      The YP/NIS (formerly ``Yellow Pages'' but renamed for legal reasons)
GAW>      domain name does not necessarily have anything to do with the Domain Name
GAW>      System domain name, although they are often set equal for administrative
GAW>      convenience.

I note this.  In quite a few instances this was done.  In others it was
not.  It's more convenient, sure, but in the manner of having them separated
out, it makes one a bit more acutely aware that there are two naming services
running, and that they are not necessarily connected.

				--*greywolf;
--
The only thing more traumatic than bootstrapping a compiler is bootstrapping
a debugger.
	-- from the notebooks of another heretic.