From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.io/gmane.emacs.gnus.general/31765 Path: main.gmane.org!not-for-mail From: Harry Putnam Newsgroups: gmane.emacs.gnus.general Subject: Re: Gnus Access mail spool by ssh Date: 12 Jul 2000 12:22:39 -0700 Sender: owner-ding@hpc.uh.edu Message-ID: References: <87lmz7d93d.fsf@sonera.com> <87hf9vcsc6.fsf@sonera.com> NNTP-Posting-Host: coloc-standby.netfonds.no Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: main.gmane.org 1035168139 15953 80.91.224.250 (21 Oct 2002 02:42:19 GMT) X-Complaints-To: usenet@main.gmane.org NNTP-Posting-Date: Mon, 21 Oct 2002 02:42:19 +0000 (UTC) Return-Path: Original-Received: from fisher.math.uh.edu (fisher.math.uh.edu [129.7.128.35]) by mailhost.sclp.com (Postfix) with ESMTP id 47729D051E for ; Wed, 12 Jul 2000 15:38:29 -0400 (EDT) Original-Received: from sina.hpc.uh.edu (lists@Sina.HPC.UH.EDU [129.7.3.5]) by fisher.math.uh.edu (8.9.1/8.9.1) with ESMTP id OAC05654; Wed, 12 Jul 2000 14:34:02 -0500 (CDT) Original-Received: by sina.hpc.uh.edu (TLB v0.09a (1.20 tibbs 1996/10/09 22:03:07)); Wed, 12 Jul 2000 14:32:59 -0500 (CDT) Original-Received: from mailhost.sclp.com (postfix@sclp3.sclp.com [204.252.123.139]) by sina.hpc.uh.edu (8.9.3/8.9.3) with ESMTP id OAA06342 for ; Wed, 12 Jul 2000 14:32:47 -0500 (CDT) Original-Received: from mail.networkone.net (mail.networkone.net [209.144.112.75]) by mailhost.sclp.com (Postfix) with SMTP id D57E0D051E for ; Wed, 12 Jul 2000 15:33:20 -0400 (EDT) Original-Received: (qmail 20959 invoked from network); 12 Jul 2000 19:33:19 -0000 Original-Received: from adsl-117-113.ln.networkone.net (HELO reader.ptw.com) (209.144.117.113) by mail.networkone.net with SMTP; 12 Jul 2000 19:33:19 -0000 Original-Received: (from reader@localhost) by reader.ptw.com (8.9.3/8.9.3) id MAA30701; Wed, 12 Jul 2000 12:33:18 -0700 X-Authentication-Warning: reader.ptw.com: reader set sender to reader@newsguy.com using -f Original-To: ding@gnus.org In-Reply-To: Nuutti Kotivuori's message of "12 Jul 2000 17:54:01 +0300" User-Agent: Gnus/5.0807 (Gnus v5.8.7) Emacs/20.5 Original-Lines: 74 Precedence: list X-Majordomo: 1.94.jlt7 Xref: main.gmane.org gmane.emacs.gnus.general:31765 X-Report-Spam: http://spam.gmane.org/gmane.emacs.gnus.general:31765 Sorry, I'm still confused here.... Your description stops right at the punch line. Nuutti Kotivuori writes: > Local machine runs fetchmail, which runs ssh to connect to the remote > machine and run imapd, which the fetchmail will access. . . . What does fetchmail do with it. Deliver to /var/mail/$USER or put it on standard out of ssh command or what? > Local .fetchmailrc: > > poll mail.host.foo with proto IMAP and options no dns > preauth ssh plugin /usr/home/xxx/bin/fetchmail-imap-wrapper OK .. I follow the .fetchmailrc recipe. Is this wrapper necessary if ssh-agent is engaged? > Local bin/fetchmail-imap-wrapper: > > #!/bin/sh > exec ssh -i $HOME/.ssh/identity-imap -l xxx $1 /opt/net/etc/imapd You aim ssh at a specific id in ~/.ssh -- login as xxx.... what does $1 do? First argument to what... ssh? --- imapd gets invoked and passes new mail to fetchmail? Which then does what? I've tried this with the following scripts: .fetchmailrc: poll reader.local.lan with proto IMAP and options no dns preauth ssh plugin /home/reader/scripts/ssh-fetch.sh ssh-fetch.sh #!/bin/sh exec ssh reader@reader /usr/sbin/imapd I can't tell if its going to work because some other mess with my FreeBSD install crops up to wreck it. $ fetchmail /usr/libexec/ld-elf.so.1: Shared object "libssl.so.1" not found Fetchmail installed from FreeBSD install with no complaints. This is starting to look like too much pain in the ass for such a simple chore. Running just the shell script (ssh-fetch.sh) gives: bsd > scripts/ssh-fetch.sh * PREAUTH reader.ptw.com IMAP4rev1 v12.264 server ready whoopee .... now what? `man imapd' you say... Well the one I have is just one tiny cut above useless, lacking descriptive information. I think I like the straight forward approach laid out by Kai and Lars. What are the chances on a local network that the `cat' command in the mail-sources prescript will die an ugly death? I'm beginning to think what ever risk, it is much less problem than sorting out a bunch of complicated soft ware for this one task. Plus, it seems one could insert some kind of `trap' that would make it safer. Especially since I have an overall backup of all incoming mail that keeps the latest 1000 messages.