mailx <-> smail incompatibilities
Prentiss Riddle
riddle at woton.UUCP
Sun Apr 10 02:14:50 AEST 1988
I am in the process of installing smail 2.5 on our AT&T 3B15 running
SVR2.1.1. Our front end for mail is "mailx", the SysV clone of
Berkeley Mail. We don't have sendmail. I'm having a couple of
problems which someone out there may have insight into.
(1) I would like to eliminate a long list of mail aliases from mailx's
"mailx.rc" file and put them instead in smail's "/usr/lib/aliases"
file, since smail is more flexible in its aliasing scheme and can
expand aliases in incoming network mail which mailx never sees.
Unfortunately there's a hitch: if a user types "mailx foobar" where
"foobar" is an smail alias, mailx notices that there is no user
"foobar" on the system and exits with the error message "Can't send to
foobar". However, if the user types "mailx woton!foobar" then mailx is
content to defer the interpretation of what it thinks is a network
address to the mail delivery program.
Is there any way, short of keeping redundant alias lists in both the
"mailx.rc" and "aliases" files, to convince mailx to pass simple
unrecognized addresses on to smail? (I'm encouraged by the fact that
I've also installed smail on our 3B1 UNIX PCs and haven't noticed this
problem with the 3B1 version of mailx.)
(2) Has anyone noticed a case of mailx developing indigestion when
replying to mail sent by smail? Our mailx goes into an apparently
infinite loop if a user tries to reply with the "r" command to all
recipients of a previous message. This generates a "To:" field like
"user1 at host1.UUCP (John Doe ) user2 at host2.UUCP (Mary Smith )". My
guess is that the parenthesized names are the source of the problem,
but oddly enough this only occurs with multiple recipients -- a single
"user at host.UUCP (John Doe )" works great.
-- Prentiss Riddle ("Aprendiz de todo, maestro de nada.")
-- Opinions expressed are not necessarily those of my employer.
-- riddle%woton.uucp at cs.utexas.edu {ihnp4,uunet}!ut-sally!im4u!woton!riddle
More information about the Comp.unix.questions
mailing list