From: utzoo!decvax!cca!hplabs!hpda!ld@sri-unix
Newsgroups: net.followup
Title: Signing articles
Article-I.D.: hpda.210
Posted: Wed Aug 11 13:25:43 1982
Received: Sun Aug 15 02:00:33 1982

The following dribble is the raving of a pit-nicker.  Do not bother reading
it unless you are a "hacker" (i.e. you like to nose through others mail,
etc.)

In English texts, it is common practice to indicate that a
quoted sentence is incomplete by pre/post pending "..."
(this might even be a rule rather than practice, I am not an English
major).  Since an author of a uucp message is quoting only part of her/his
return address The "..." is inserted in front of an address to indicate that it
may take an unknown path to lead up to "wonderland".  If the novice user saw 
"...ucbvax!arnold" and tried to mail to "...ucbvax!arnold"
from "WEB40", [s]he would be rudely informed that there was no
such machine ("...ucbvax").  Would this novice user be any more informed by the
return address "ucbvax!arnold"?  What would stop this naive user
from mailing to "ucbvax!arnold" and getting the same obnoxious
message that there was no such machine ("ucbvax").  I am sure that
Ken's query was rhetorical (come on, isn't it obvious).

Wouldn't it be nice if mail or uucp knew that "...<important.machine>" meant
that it was supposed to look in a file, maintained by local system
management, which contained the path to "<important.machine>", and
it was supposed substitute a pre-canned path to said machine?

Better still, wouldn't it be wonderful if mail could alias machines as
well as users?

I do not understand the statement that Ken made:

   The "..." is implicit in uucp addresses, and is therefore redundant.

There is no implication to my version of uucp that "..." means anything, so it 
is not redundant (does this sentence make any more sense than Ken's?).

Remember, all paths lead to ucbvax, but not all of us live there.

		Larry Dwyer
		...!ucbvax!hpda!ld	(take that!)