home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.unix.sys5.r4:510 comp.mail.misc:3699
- Path: sparky!uunet!ogicse!uwm.edu!zaphod.mps.ohio-state.edu!news.acns.nwu.edu!casbah.acns.nwu.edu!skrenta
- From: skrenta@casbah.acns.nwu.edu (Rich Skrenta)
- Newsgroups: comp.unix.sys5.r4,comp.mail.misc
- Subject: Re: Email header problems on SVR4.
- Message-ID: <1992Nov15.233842.2811@news.acns.nwu.edu>
- Date: 15 Nov 92 23:38:42 GMT
- Article-I.D.: news.1992Nov15.233842.2811
- References: <1992Nov14.074815.15655@clark.edu>
- Sender: usenet@news.acns.nwu.edu (Usenet on news.acns)
- Organization: Northwestern University, Evanston, Illinois
- Lines: 33
-
- ade@clark.edu (Adrian Miranda) writes:
- > I am having some problems with mail on ESIX SVR4.0.4.
-
- The Release 4 mailer is a variant of Bell Lab's upas mailer. It
- has a really neat configuration file that uses regular expressions
- to rewrite delivery addresses.
-
- Unfortunatley, the folks who designed this mailer don't care much
- for, or at least don't understand, RFC822-style headers. For instance,
- the bizarre quoting of From_ that you mention, or the lack of proper
- From:, Message-Id, etc.
-
- Also, there is quite a bit of cruft associately with the R4 mailer,
- since the people who added it to R4 didn't just adopt upas as their
- new mailer, they glued it onto the side of the existing 3.2 /bin/mail
- program. The result is something of a mess, although the nifty
- /etc/mail/mailsurr configuration file almost makes up for it.
-
- Another example: SVR4 mail will only attempt to generate RFC822
- headers if it invokes the smtp transport. If you send mail locally
- or via uucp, forget it -- no From:, those aren't "RFC822" transports.
- Actually, the smtp transport program tries to kludge together RFC822
- compliant headers. The /bin/mail program doesn't understand them at all.
-
- The only way to fix this is with source, unfortunately.
-
- I recently wrote a mailer which uses a similar configuration file
- format as the R4 mailer, but does RFC822 headers correctly, and
- leaves out the cruft. If it gets robust enough I'll post it to
- alt.sources.
-
- --
- Rich Skrenta
-