On Tue, May 16, 2000 at 07:44:36AM -0400, Sam Varshavchik wrote:
1. the docs on the website say makealiases is in .../bin, but it's
actually in .../sbin. Same for all the makesomething programs.
I only found one other reference for makesmtpaccess.
Actually, that's the one I meant to refer to. Sorry. Should have been
more specific.
2. the courier local delivery component adds the return-path: and
delivered-to: headers in the opposite order to most other MTAs (qmail,
postfix), ie. delivered-to: appears right at the top, followed by
return-path:. Any particular reason?
No. The order does not matter.
DJB's serialmail package reads the first line for return-path: and
the second line for Delivered-To:. Would you consider reversing the
order?
3. The courieresmtpd component returns a list of ESMTP capabilities,
regardless of whether a HELO or EHLO was used. From what I know, only
a single line containing the hostname should be returned with HELO,
otherwise some clients have a problem, because they only expect a
single line, and instead get a multiline response.
What clients?
I don't remember now. qmail-smtpd <= 1.02 used to return multiline
responses even to HELO. Then, around April/May of 1998, there was some
discussion on the qmail list. One of these weird clients, Lotus, I think,
was having trouble with the multi-line response to a HELO; it was
expecting a single line. Bad program, I know, but that's what it did.
So in May 1998, DJB changed code so that qmail-smtpd would send a
single line for HELO and multiline for EHLO (that's how sendmail did
it, and it sort of set the standard). I've tried to go through the
qmail archives, but I can't locate the particular discussion thread
that prompted this change. It doesn't particularly affect me at the
moment, because I only use the esmtp component with netscape mail,
but I thought you might want to keep this in mind for the future.
A multiline reply is perfectly valid.
I don't dispute that. But reality is usually not 'perfect'.
4 messages in net.sourceforge.lists.courier-usersRe: [courier-users] some observations
