From monad.swb.de!okir@aichan.swb.de Mon Dec 30 16:05:55 1996 Return-path: Envelope-to: clameter@waterf.org Delivery-date: Mon, 30 Dec 1996 16:05:55 -0800 Received: from aichan.swb.de [193.175.30.7] (root) by waterf.org with esmtp (Exim 1.59 #1) id 0verhr-0000ED-00 (Debian); Mon, 30 Dec 1996 16:05:41 -0800 Received: by aichan.swb.de id m0verbd-0007V5C (ChiBi Smail-3.2 1996-Jul-4 #2); Tue, 31 Dec 1996 00:58:57 +0100 (MET) Received: by monad.swb.de with smtp (smail3.2 #3) id m0vetGd-000HRmC; Tue, 31 Dec 1996 02:45:23 +0100 (MET) Message-Id: To: linux-kernel@vger.rutgers.edu Cc: Christoph Lameter Subject: Re: NFS Problem in Kernel 2.0.27: inode status not updated Date: Tue, 31 Dec 1996 02:45:23 +0100 From: Olaf Kirch Status: RO X-Status: On Sat, 28 Dec 1996 10:25:29 PST, Christoph Lameter wrote: > It first opens a lockfile with a unique name and then links that one > to the classic /var/spool/mail/username.lock lockfile. Then it checks the > number of links on that file. If there are two then the lock is > successful. Problem is Linux fstat call always returns 1 for the number of > links even though a link() was done immediately prior. There is nothing in the NFS spec that requires the client to update its cached link count[*]. So I would assume the author's claim means that this locking technique `works better over NFS on the systems I tested it on.' Nevertheless, I guess I'll have to support this in the upcoming NFS code... For the time being, I suggest exim users apply the following patch to fs/nfs/dir.c, function nfs_link: -------- fake patch ----- nfs_lookup_cache_remove(dir, oldinode, NULL); + NFS_CACHEINV(oldinode); iput(oldinode); ------------------------- IMHO, mounting your mail spool with the noac mount option is not a very good idea. Olaf [*] The NFS spec is deliberately vague about how and to which extent the client caches information. Vague being an overstatement here. For those interested, the bare-bones protocol spec is available as RFC 1094. There's also a (fairly expensive) specification available from X/Open that also discusses some implementation issues as well as the side effects on various system calls and utility programs. Some information about current implementation practice can also be found in the NFSv3 spec (available as RFC, and as postscript file all over the net--archie for nfsv3.ps). -- Olaf Kirch | --- o --- Nous sommes du soleil we love when we play okir@monad.swb.de | / | \ sol.dhoop.naytheet.ah kin.ir.samse.qurax For my PGP public key, finger okir@brewhq.swb.de. From jake@ibmpcug.co.uk Tue Feb 18 10:06:34 1997 Return-path: Envelope-to: clameter@waterf.org Delivery-date: Tue, 18 Feb 1997 10:06:34 -0800 Received: from nora.pcug.co.uk [192.68.174.71] by waterf.org with smtp (Exim 1.596 #1) id 0vwtwG-0001jR-00 (Debian); Tue, 18 Feb 1997 10:06:29 -0800 Received: from kate.pcug.co.uk by nora.pcug.co.uk id aa18513; 18 Feb 97 18:05 GMT Subject: Re: UUCP To: Christoph Lameter Date: Tue, 18 Feb 1997 18:05:18 +0000 (GMT) In-Reply-To: from "Christoph Lameter" at Feb 18, 97 09:50:12 am From: jake@pcug.co.uk X-Organisation: The PC User Group, Harrow, UK X-Address: 84-88 Pinner Road, Harrow, HA1 4LF, UK X-Phone: +44 181 863 1191 X-Mailer: ELM [version 2.4 PL2] Content-Type: text Content-Length: 1071 Sender: jake@ibmpcug.co.uk Message-ID: <9702181805.aa13929@kate.ibmPCUG.CO.UK> Status: RO X-Status: > I saw a discussion a while back regarding UUCP feeds with exim (domain > based of course). Has anyone successfully done so? I would like to set up > a small UUCP batch feed and would like to have some examples. > --- +++ --- +++ --- +++ --- +++ --- +++ --- +++ --- +++ --- > Please always CC me when replying to posts on mailing lists. "standard" rmail uucp is easy.. use something like this.. # a transport uucp_pipe: driver = pipe; #ignore_status, user = uucp, directory = "/tmp", restrict_to_path, path = "/usr/bin:/bin", command = "uux - -a$sender_address -r $host\!rmail ($local_part@$domain)", return_output # a router example entry uucpdomainlist: driver = domainlist, transport = uucp_pipe; search_type = dbm, route_file = "/usr/exim/table/uucpdom"; # the uucpdom file is like "wibble.com: wibblesys" Batched SMTP over uucp is also possible, I have not tried it yet. regards, -- Jake Dias jake@pcug.co.uk ...!ibmpcug!jake My PGP Key? - finger jake@pcug.co.uk or email with Subject: get pgp key From exim-users-request@lists.cam.ac.uk Wed Feb 19 08:01:21 1997 Return-path: Envelope-to: clameter@waterf.org Delivery-date: Wed, 19 Feb 1997 08:01:21 -0800 Received: from jethro.fuller.edu [206.1.27.7] (mail) by waterf.org with smtp (Exim 1.596 #1) id 0vxESi-0002zo-00 (Debian); Wed, 19 Feb 1997 08:01:20 -0800 Received: from lilac.csi.cam.ac.uk [131.111.8.44] (exim) by jethro.fuller.edu with smtp (Exim 1.596 #1) id 0vxENJ-0001TB-00 (Debian); Wed, 19 Feb 1997 07:56:06 -0800 Received: from most.weird.com [204.92.254.2] (root) by lilac.csi.cam.ac.uk with esmtp (Exim 1.58 #1) id 0vxE5j-0004YC-00; Wed, 19 Feb 1997 15:37:56 +0000 Received: by most.weird.com via sendmail with stdio id for exim-users@lists.cam.ac.uk; Wed, 19 Feb 1997 10:37:30 -0500 (EST) (Smail-3.2.0.92-pre 1997-Jan-21 #8 built 1997-Jan-21) Message-Id: Date: Wed, 19 Feb 1997 10:37:30 -0500 (EST) From: woods@most.weird.com (Greg A. Woods) To: exim-users@lists.cam.ac.uk Subject: Re: UUCP In-Reply-To: Philip Hazel's message of "Wed, February 19, 1997 09:21:33 +0000" regarding "Re: UUCP" id References: Reply-To: woods@weird.com (Greg A. Woods) X-Mailer: ViewMail (vm) Version 5.96 (beta) with GNU Emacs 19.34.1 (m68k.68881-sun-sunos4.1.1, X toolkit) of Thu Sep 12 1996 on most Organization: Planix, Inc.; Toronto, Ontario; Canada Status: RO X-Status: [ On Wed, February 19, 1997 at 09:21:33 (+0000), Philip Hazel wrote: ] > Subject: Re: UUCP > > People are certainly doing this, but I have no examples myself. I have > added some extra features to the 1.60 release that may make this easier. > It is possible in that release to cause batching of messages to local > transports without the use of batch SMTP, and the list of addresses is > available in a variable. So you can arrange, for example, for a message > with addresses in some specific domain to cause one execution of a pipe > command such as > > |uux hostname!rmail address1 address2 ... > > which I understand is the way UUCP is frequently run. Just FYI: One of the important configuration options (compile-time, unfortunately) in the ancient smail-2.x was a limit to the total length of the argument list passed to uux(1). In some older UUCP's this limit was as low as 127 characters. It might be of some value to see if there is any limit in the Taylor-UUCP (GNU) implementation of uux or uuxqt (other than the maximum command-line limits which unfortunately may be different in the remote machine, though in theory uuxqt could cause multiple invocations of rmail to cope, though this breaks the uux interface just a bit). BTW, the 1.60 mechanisms seem generally applicable to many uses! -- Greg A. Woods +1 416 443-1734 VE3TCP robohack!woods Planix, Inc. ; Secrets of the Weird From owner@bugs.debian.org Wed Mar 29 21:56:56 2000 Envelope-to: mark@aziraphale.demon.co.uk Subject: Bug#61331: exim: example for dialup mail configuration Reply-To: Thomas Schoepf , 61331@bugs.debian.org Orignal-Sender: Thomas Sch?pf X-Debian-PR-Message: report 61331 X-Debian-PR-Package: exim X-Debian-PR-Keywords: Date: Wed, 29 Mar 2000 16:13:42 +0200 From: Thomas Schoepf To: submit@bugs.debian.org User-Agent: Mutt/1.1.9i package: exim severity: wishlist I'm running a single machine that's only temporarily connected to the Internet. I'm using a free email service for my emails and I tried to configure exim to recognize just this specific email address (my@address.com) as a local email address and all others (e.g. others@address.com) as remote emails. After a whole day of reading exim's manual, I've finally found a solution. I've put these lines into the Routers Configuration section in /etc/exim.conf: local_indeed: driver = domainlist domains = ${lookup{$local_part}lsearch{/etc/email-local}{$value}} self = local route_list = "* 127.0.0.1 byname" The file /etc/email-local contains key:value pairs that tell exim which addresses should be considered local: schoepf: domain1.org:domain2.com:...:domain3.com Maybe you could include this as an example into the exim package so that others who want such a behavior don't have to reinvent the wheel. Thank you! Thomas -- 1024D/B0FA4F49: FA38 2D7E 408F 61E4 BF49 B48F 04BD F5BE B0FA 4F49 2048g/C631AF6E: B89D 7BF4 AA6B 569B D9D1 4BF6 3459 66AB C631 AF6E To: exim-announce@exim.org From: Nigel Metheringham Content-Type: text/plain; charset=us-ascii Message-Id: Subject: [Exim-Announce] Windows generic executable filter - Version 0.09 Sender: exim-announce-admin@exim.org Date: Tue, 10 Oct 2000 10:18:11 +0100 Status: RO Content-Length: 1300 Lines: 36 [I did announce this on the exim-users list 10 days ago, but am resending to this list] I have made a new version of the Exim Windows generic executable filter, taking it to version 0.09. The filter, which runs as a system filter, attempts to detect messages with attachments that may contain executables under windows. These could potentially be trojans, virus infected or other problematic content. A number of windows based email worms/viruses send out messages which will be detected and blocked by the filter. The method of detecting MIME attachments using regular expressions is very limited. However its a low cost and so far quite effective means of blocking attacks from the various windows email problems. Instructions for installation are included within the file. The filter is available at ftp://ftp.exim.org/pub/filter/ http://www.exim.org/system_filter.exim The MD5 is 6d17e5999c36e2a2297bfdc7a0bb771d system_filter.exim Nigel. -- [ - Opinions expressed are personal and may not be shared by VData - ] [ Nigel Metheringham Nigel.Metheringham@VData.co.uk ] [ Phone: +44 1423 850000 Fax +44 1423 858866 ] -- ## List details at http://www.exim.org/mailman/listinfo/exim-announce Exim details at http://www.exim.org/ ##