Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!usc!snorkelwacker.mit.edu!bloom-beacon!eru!hagbard!sunic!mcsun!ukc!cam-eng!glasgow!daemon
From: idc@cs.hw.ac.uk (Ian Crorie)
Newsgroups: mail.uk-sendmail-workers,comp.mail.sendmail
Subject: Ultrix sendmail problems
Message-ID: <18403.9104031739@odin.cs.hw.ac.uk>
Date: 3 Apr 91 17:39:05 GMT
Sender: uk-sendmail-workers-request@cs.hw.ac.uk
Organization: Glasgow University Computing Science Dept.
Lines: 43
Organisation: Computer Science, Heriot-Watt Univ., Scotland
X-Mailer: Sendmail/Ream v4.12b (The Choice of the Old Generation too), mail-news 2.0.5


Martin Tomes (mt00@eurotherm.co.uk) asked me about a problem he is having:

[Martin]
| We are on some mailing lists and by the time messages get to us they
| have passed though many machines and the hop count is high, we then
| bounce the message and Joe Stoy at Oxford gets them and wishes he
| didn't.  Is there a way to persuade sendmail to allow more hops?  The
| memximum appears to be 17.  It is sendmail on an Ultrix Decstation
| which is bouncing the messages.

I discovered this problem with Sun sendmail a couple of years ago
and cured it using the (undocumented?) Oh configuration option i.e.

Oh30

added to the config file set MAXHOP to 30.   Unfortunately it doesn't
seem to be supported by Ultrix sendmail.

I suppose it might be possible to use the -h flag on the command line
to reset the hop count when the message arrives.   This presumes that
you have the source for an interface program between whatever transport
system you used to take mail into the site and sendmail (i.e. rmail.c as
supplied with UKsendmail).

I've never tried this but assuming it *does* reset the hop count, overriding
the number of Received: lines in the message header, then it should work.
It would mean problems if a message was bouncing between you and an
outside site however - it would bounce forever.   Perhaps it would be
possible to check that the message hadn't already been through your
site by looking through the Received: lines and only resetting the hop count
if it hadn't.

Does anyone have any comments on this or any better solutions (like where
he could poke the sendmail binary with adb)?



--- Support Desk:
	Two, four, six, eight
	Who do we appreciate?
	DESK!
---