procmail
[Top] [All Lists]

Re: plussed addresses

1997-07-14 17:47:00
At 11:34 PM 7/13/97 -0400, Timothy Luoma wrote:

hrm....

the mailer @peak.org seems to be stripping off my hostname when I send out  
mail.... even in the reply-to line.... or perhaps it is my sendmail config  
here....

Anyone else besides me think that sendmail ought to leave the ReplyTo line 
alone?

Oh well...

TjL

Hi Timothy,

Actually, in the past I've spent considerable time at multiple sites trying to
(usually, successfully) get them to *DO* such things.  The general problem
was that when user(_at_)subdomain(_dot_)domain(_dot_)com got sent out in any of 
the headers
(To, Cc, From, Reply-To, etc.), replies (or group replies) went there
and either got mishandled, bounced, or grossly delayed depending just
what "subdomain" understood about mail.  Frequently, "subdomain" was
a workstation sitting on someone's desk which might be turned off for a week
when they were on vacation (just for example).

Another problem was users expecting to be able to post to the same
mailing list from three different workstations, say.

Tiac (my ISP) does not strip subdomains, probably for historical reasons (i.e.,
they have lots of customers signed up to mailing lists, etc.) but instead
intercepts mail to xxxx.tiac.net (for all legal values of xxxx) and routes
it to a correct machine for delivery.  However, if the outgoing address is
entirely lacking a domain, it will be added unless delivered within tiac.

If you tried to send out just "Reply-To: luomat" I think you *would*
want your sendmail "fixing" that one.

I believe the RFC requirement is that the addresses be replyable; how that
gets implemented is an implementation detail of sorts.  Changing from one
way to the other would probably impact everyone using peak.org to some extent.

Each method (altering outgoing headers, or rerouting incoming mail) has its
advantages and disadvantages... probably getting well beyond the scope of the
procmail mailing list.

Cheers,
Stan

<Prev in Thread] Current Thread [Next in Thread>