Impact of having loghost defined in DNS...?

[Jonathan de Boyne Pollard]

GH> There is no "domain" for "localhost" that is a standard.

That's a mis-statement, based upon the erroneous foundation that domain 
names comprise a "host part" and a "domain part".  They don't.  A 
correct statement is that "localhost." is a top-level domain name, 
reserved (to be effectively a "host local" portion of the DNS namespace) 
by RFC 2606, and that RFC 1912 section 4.1 superficially describes some 
of the issues surrounding it.

GH> (Linux uses "localdomain" but this is ALSO only valid for the local
GH> machine since there is no TLD called "localdomain".)

That's also a mis-statement.  Some mis-configured applications on Linux 
use, and (worse) look up, "localhost.localdomain.".  But this is 
specific to the applications and not generic to Linux.

GH> "localhost" is tied by the OS (as in HARD-WIRED or HARD-CODED) to the
GH> loopback interface.

False.  Domain names are not directly associated with network 
interfaces.  It is IP addresses that some operating systems hard-wire to 
loopback network interfaces.  And not all operating systems actually do 
this.