From owner-ntemacs-users@june Mon Dec 30 12:48:49 1996 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] [nil "Mon" "30" "December" "1996" "15:03:11" "-0500" "bill@attmail.com" "bill@attmail.com" nil "72" "win95 networking, vm-pop, and smtpmail" "^From:" nil nil "12" nil nil nil nil] nil) Received: from joker.cs.washington.edu (joker.cs.washington.edu [128.95.1.42]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id MAA29536 for ; Mon, 30 Dec 1996 12:48:48 -0800 Received: from june.cs.washington.edu (june.cs.washington.edu [128.95.1.4]) by joker.cs.washington.edu (8.6.12/7.2ws+) with ESMTP id MAA17430 for ; Mon, 30 Dec 1996 12:48:46 -0800 Received: from cagw1.att.com (cagw1.att.com [192.128.52.89]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id MAA28115 for ; Mon, 30 Dec 1996 12:08:15 -0800 Received: from lynxhub.ho.att.com by caig1.att.att.com (SMI-8.6/EMS-1.2 sol2) id PAA20856; Mon, 30 Dec 1996 15:03:14 -0500 Received: from CARPENTER.cis.att.com ([135.46.61.210]) by lynxhub.ho.att.com (5.x/EMS-1.2 sol2) id AA22073; Mon, 30 Dec 1996 15:03:13 -0500 Message-Id: <9612302003.AA22073@lynxhub.ho.att.com> X-Mailer: emacs 19.34.1 (via feedmail 3 Q) WJC Original-From: bill@attmail.com (WJCarpenter) Reply-To: bill@attmail.com (WJCarpenter) From: bill@attmail.com To: ntemacs-users@cs.washington.edu Cc: bug-vm@uunet.uu.net, bug-gnu-emacs@prep.ai.mit.edu Subject: win95 networking, vm-pop, and smtpmail Date: Mon, 30 Dec 1996 15:03:11 -0500 I recently wrote that I sometimes get the Win95 "blue screen" when mixing POP3 (vm-pop) and SMTP (smtpmail) in an NTemacs session. (I also wrote that having separate NTemacs invocations for sending and receiving seemed to help, but that was a red herring and didn't actually help.) I think I have found the source of the problem. Win95 comes with a UNIX-like "netstat" utility, so you can look at open connections and so on. I noticed that other (non-NTemacs) applications opened connections which disappeared just about as soon as they closed. For vm-pop or smtpmail connections, the socket instead went into a TIME_WAIT state for a while. After a couple minutes, the socket would go away. I speculate that having too many of these sockets in this condition (or using unlucky socket numbers or whatever) triggers some bug in Win95 winsock. smtpmail connections aggravate the problem because each piece of email is sent with a separate connection and SMTP session (whereas vm-pop retrieves all mail in a single POP3 session). Ok, if that's the case, what to do about it? I speculated that the reason the sockets were going to TIME_WAIT was that they had some data left to be read from them, but the NTemacs stream reader wouldn't oblige (the process is already deleted). In both the vm-pop and smtpmail cases, the elisp sends a QUIT command, reads a (structured) response, and then nukes the process. My very inelegant fix is to put a sleep of 2 seconds before the delete-process. That allows the data to drain from the socket even though it doesn't actually go to the process filter (which doesn't care in this context). For vm-pop.el, the end of the function vm-pop-move-mail now looks like this in my copy: (if process (save-excursion (set-buffer (process-buffer process)) (vm-pop-send-command process "QUIT") (vm-pop-read-response process) ;; bill@att.com (sleep-for 2) ;; to drain stream (delete-process process)))))) For smtpmail.el, the end of smtpmail-via-smtp now looks like this: (if process (save-excursion (set-buffer (process-buffer process)) (smtpmail-send-command process "QUIT") ;; bill@att.com (sleep-for 2) ;; to drain stream (smtpmail-read-response process) ; (if (or (null (car (setq response-code (smtpmail-read-response process)))) ; (not (integerp (car response-code))) ; (>= (car response-code) 400)) ; (throw 'done nil) ; ) (delete-process process)))))) I just noticed now that I put the sleep in slightly different places in the two files, but it doesn't seem to matter. I've been using it this way all day, and no more sockets stick around in TIME_WAIT. And, so far, no more "blue screen". If anyone codes up a more elegant solution than (sleep-for 2), please send it along. I can pay the price of 2 seconds per connection, but I'm not sure whether 2 is the right number for all situations. And, of course, it doesn't feel too good to do those sleeps. -- bill@attmail.com LZ 1C-320 bill@att.com (Voice/Fax) +1.908.576.2932 AT&T Labs / AT&T WorldNet Service From owner-ntemacs-users@june Mon Dec 30 12:56:13 1996 X-VM-v5-Data: ([nil nil nil nil t nil nil nil nil] [nil "Mon" "30" "December" "1996" "15:16:50" "-0500" "bill@attmail.com" "bill@attmail.com" "<9612302016.AA22907@lynxhub.ho.att.com>" "45" "vm-pop and NTemacs heap" "^From:" nil nil "12" nil nil nil nil] nil) Received: from joker.cs.washington.edu (joker.cs.washington.edu [128.95.1.42]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id MAA29773 for ; Mon, 30 Dec 1996 12:56:13 -0800 Received: from june.cs.washington.edu (june.cs.washington.edu [128.95.1.4]) by joker.cs.washington.edu (8.6.12/7.2ws+) with ESMTP id MAA16153 for ; Mon, 30 Dec 1996 12:56:11 -0800 Received: from kcgw1.att.com (kcgw1.att.com [192.128.133.1]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id MAA28519 for ; Mon, 30 Dec 1996 12:18:38 -0800 Received: from lynxhub.ho.att.com by kcig1.att.att.com (SMI-8.6/EMS-1.2 sol2) id OAA11759; Mon, 30 Dec 1996 14:13:18 -0600 Received: from CARPENTER.cis.att.com ([135.46.61.210]) by lynxhub.ho.att.com (5.x/EMS-1.2 sol2) id AA22907; Mon, 30 Dec 1996 15:16:51 -0500 Message-Id: <9612302016.AA22907@lynxhub.ho.att.com> X-Mailer: emacs 19.34.1 (via feedmail 3 Q) WJC Original-From: bill@attmail.com (WJCarpenter) Reply-To: bill@attmail.com (WJCarpenter) From: bill@attmail.com To: ntemacs-users@cs.washington.edu Cc: bug-vm@uunet.uu.net, bug-gnu-emacs@prep.ai.mit.edu Subject: vm-pop and NTemacs heap Date: Mon, 30 Dec 1996 15:16:50 -0500 GNUemacs 19.34 on Win95; VM 5.96 (beta). One of the lucky things I did during bringing up vm-pop.el to do POP3 from my laptop was to try to download a somewhat long message (around 1.5 MB). NTemacs (on Win95) consistently croaked with a message something like "process space exhausted". Now, I have 32 MB of real RAM and another 30 MB of swap space, so I figured something had to be going on. Also, I'd had no problem editing bigger messages on the same machine (they'd been dragged over via something other than vm-pop). After a little investigation I found that vm-pop was doing the more or less standard process filter stuff of just iterating and stuffing the bits and pieces into a buffer as they came over the network. On my network, they were coming at 1-2k per iteration of the process filter. This results in about a 1000 inserts at the end of the process buffer for that 1.5 MB message. I wanted to slow vm-pop down so it would work on bigger chunks. Here is my inelegant solution. The beginning of this function in vm-pop.el now looks like this in my copy: (defun vm-pop-retrieve-to-crashbox (process crash) (let ((start vm-pop-read-point) end) (goto-char start) (while (not (re-search-forward "^\\.\r\n" nil t)) (accept-process-output process) ;; bill@att.com ... to save wear and tear on the heap (if (> (buffer-size) 20000) (sleep-for 1)) (if (> (buffer-size) 50000) (sleep-for 1)) (if (> (buffer-size) 100000) (sleep-for 1)) (if (> (buffer-size) 200000) (sleep-for 1)) (if (> (buffer-size) 500000) (sleep-for 1)) ;; bill@att.com (goto-char start)) This probably isn't needed on most other platforms, but it did the trick on Win95. The effect is to sleep more the bigger the received message is. The numbers are just completely guessed-at and are not based on some keen analysis of anything. -- bill@attmail.com LZ 1C-320 bill@att.com (Voice/Fax) +1.908.576.2932 AT&T Labs / AT&T WorldNet Service From owner-ntemacs-users@june Tue Dec 31 09:42:24 1996 X-VM-v5-Data: ([nil nil nil nil t nil nil nil nil] [nil "Tue" "31" "December" "1996" "11:51:33" "-0500" "bill@attmail.com" "bill@attmail.com" "<9612311651.AA20790@lynxhub.ho.att.com>" "31" "RE: win95 networking, vm-pop, and smtpmail" "^From:" nil nil "12" nil nil nil nil] nil) Received: from joker.cs.washington.edu (joker.cs.washington.edu [128.95.1.42]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id JAA02452 for ; Tue, 31 Dec 1996 09:42:23 -0800 Received: from june.cs.washington.edu (june.cs.washington.edu [128.95.1.4]) by joker.cs.washington.edu (8.6.12/7.2ws+) with ESMTP id JAA16952 for ; Tue, 31 Dec 1996 09:42:22 -0800 Received: from cagw1.att.com (cagw1.att.com [192.128.52.89]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id IAA00441 for ; Tue, 31 Dec 1996 08:56:17 -0800 Received: from lynxhub.ho.att.com by caig1.att.att.com (SMI-8.6/EMS-1.2 sol2) id LAA17826; Tue, 31 Dec 1996 11:51:25 -0500 Received: from CARPENTER.cis.att.com ([135.46.61.210]) by lynxhub.ho.att.com (5.x/EMS-1.2 sol2) id AA20790; Tue, 31 Dec 1996 11:51:35 -0500 Message-Id: <9612311651.AA20790@lynxhub.ho.att.com> X-Mailer: emacs 19.34.1 (via feedmail 3 I) WJC Original-From: bill@attmail.com (WJCarpenter) In-Reply-To: WJCarpenter's note of 15:03:11, 30 Dec 1996 References: <9612302003.AA22073@lynxhub.ho.att.com> Reply-To: bill@attmail.com (WJCarpenter) From: bill@attmail.com To: ntemacs-users@cs.washington.edu Cc: bug-vm@uunet.uu.net, bug-gnu-emacs@prep.ai.mit.edu Subject: RE: win95 networking, vm-pop, and smtpmail Date: Tue, 31 Dec 1996 11:51:33 -0500 wjc> Ok, if that's the case, what to do about it? I speculated that wjc> the reason the sockets were going to TIME_WAIT was that they had wjc> some data left to be read from them, but the NTemacs stream wjc> reader wouldn't oblige (the process is already deleted). In both wjc> the vm-pop and smtpmail cases, the elisp sends a QUIT command, wjc> reads a (structured) response, and then nukes the process. My wjc> very inelegant fix is to put a sleep of 2 seconds before the wjc> delete-process. That allows the data to drain from the socket wjc> even though it doesn't actually go to the process filter (which wjc> doesn't care in this context). A friend of mine, who has spent a lot more time in RFC-793 than I have, clued me in to what was probably actually happening. Since the POP3 client and the SMTP client are both sending "QUIT" and then reading the response from the server, there is a race condition about shutting down the socket. Both the client and the server shut down the socket at nearly the same time. If the client happens to see the FIN packet from the server, it can just nuke the socket. If it doesn't see it, it must send a FIN and enter the TIME_WAIT state. Since the server side is already gone, the TIME_WAIT state has to wait to time out (a couple minutes). My hack to sleep 2 seconds gives enough time that the server is always seen as shutting down the socket, and the TIME_WAIT is avoided. This probably doesn't matter most places, but I imagine there is a problem with Win95 coping with (too many? unlucky? bad vibes?) sockets in that state. -- bill@attmail.com LZ 1C-320 bill@att.com (Voice/Fax) +1.908.576.2932 AT&T Labs / AT&T WorldNet Service From bill@attmail.com Mon Jan 6 07:38:17 1997 X-VM-v5-Data: ([nil nil nil nil t nil nil nil nil] [nil "Mon" " 6" "January" "1997" "10:36:33" "-0500" "bill@attmail.com" "bill@attmail.com" "<9701061536.AA28768@lynxhub.ho.att.com>" "23" "Re: vm-pop and NTemacs heap" "^From:" nil nil "1" nil nil nil nil] nil) Received: from cagw2.att.com ([192.128.52.90]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id HAA14437 for ; Mon, 6 Jan 1997 07:38:16 -0800 Received: from lynxhub.ho.att.com by caig2.att.att.com (SMI-8.6/EMS-1.2 sol2) id KAA28932; Mon, 6 Jan 1997 10:42:13 -0500 Received: from CARPENTER.cis.att.com ([135.46.61.210]) by lynxhub.ho.att.com (5.x/EMS-1.2 sol2) id AA28768; Mon, 6 Jan 1997 10:36:35 -0500 Message-Id: <9701061536.AA28768@lynxhub.ho.att.com> X-Mailer: emacs 19.34.1 (via feedmail 7-beta[work-in-progress] I) WJC Original-From: bill@attmail.com (WJCarpenter) In-Reply-To: Geoff Voelker's note of 15:59:14, 4 Jan 1997 References: <9612302016.AA22907@lynxhub.ho.att.com> <199701042359.PAA25647@joker.cs.washington.edu> Reply-To: bill@attmail.com (WJCarpenter) From: bill@attmail.com To: voelker@cs.washington.edu (Geoff Voelker) Subject: Re: vm-pop and NTemacs heap Date: Mon, 6 Jan 1997 10:36:33 -0500 gv> I'm not sure I understand why the default behavior works for other gv> platforms but doesn't work for Win95; I suppose I also don't gv> understand why 1000 inserts would use significantly more memory gv> than one large insert. Is there some state that is being kept gv> around on each insert? And is this state not garbage-collectable? I don't know that it doesn't fail for other platforms since I have only UNIX and Win95 to test on (and, you know how that UNIX takes a lickin' and keeps on tickin'). I'm just guessing that it might be failing under Win95 due to some bug in Win95 memory management or whatever. I figured if it were failing under NT, someone else would've complained by now. Maybe not; you have to download a pretty big message before it fails. As far as I can see, there is no per-iteration state being kept. VM is using a process filter that is just inserting the chunks at the end of a buffer as they come in. The only other thing it's doing is looking for the end of the data. VM's behavior looks proper and ordinary to me. -- bill@attmail.com LZ 1C-320 bill@att.com (Voice/Fax) +1.908.576.2932 AT&T Labs / AT&T WorldNet Service From bill@attmail.com Mon Jan 6 11:07:51 1997 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] [nil "Mon" " 6" "January" "1997" "10:48:40" "-0500" "bill@attmail.com" "bill@attmail.com" nil "26" "RE: win95 networking, vm-pop, and smtpmail" "^From:" nil nil "1" nil nil nil nil] nil) Received: from cagw1.att.com ([192.128.52.89]) by june.cs.washington.edu (8.8.3+CSE/7.2ju) with SMTP id LAA01339 for ; Mon, 6 Jan 1997 11:07:06 -0800 Received: from lynxhub.ho.att.com by caig1.att.att.com (SMI-8.6/EMS-1.2 sol2) id KAA20156; Mon, 6 Jan 1997 10:45:30 -0500 Received: from CARPENTER.cis.att.com ([135.46.61.210]) by lynxhub.ho.att.com (5.x/EMS-1.2 sol2) id AA29345; Mon, 6 Jan 1997 10:48:42 -0500 Message-Id: <9701061548.AA29345@lynxhub.ho.att.com> X-Mailer: emacs 19.34.1 (via feedmail 7-beta[work-in-progress] I) WJC Original-From: bill@attmail.com (WJCarpenter) In-Reply-To: Geoff Voelker's note of 16:07:47, 4 Jan 1997 References: <9612302003.AA22073@lynxhub.ho.att.com> <9612311651.AA20790@lynxhub.ho.att.com> <199701050007.QAA25418@joker.cs.washington.edu> Reply-To: bill@attmail.com (WJCarpenter) From: bill@attmail.com To: voelker@cs.washington.edu (Geoff Voelker) Subject: RE: win95 networking, vm-pop, and smtpmail Date: Mon, 6 Jan 1997 10:48:40 -0500 gv> I apologize for answering your mail with more questions. I'm gv> still not sure I understand the scenario yet. You have a POP3 gv> client and an SMTP client (to read and send mail at the same gv> time?). Are they both using the same socket to the server? I gv> would have thought that they would each have their own connection. I have both an SMTP client (distribution smtpmail.el) and a POP3 client (vm-pop.el from standard VM). I don't use them at the same time or to the same socket or whatever. I use them in the standard ways (SMTP port 25, POP3 port 110). The fact that I happen to be using both within the same GNUemacs invocation probably doesn't cause the problem; it just aggravates it. So, I still speculate that the problem is caused by some bug in Win95 networking. It happens to be stimulated by the race condition inherent in the SMTP and POP3 protocols, where it's not clear who's going to close the connection (both ends close it, but it's not synchronized). My hack is just a workaround that eliminates the race condition and makes it look like the server side always shuts down the connection first. -- bill@attmail.com LZ 1C-320 bill@att.com (Voice/Fax) +1.908.576.2932 AT&T Labs / AT&T WorldNet Service From owner-ntemacs-users@june Sat Jan 25 22:39:53 1997 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] [nil "Sun" "26" "January" "1997" "01:03:39" "-0500" "Kyle Jones" "kyle_jones@wonderworks.com" nil "39" "VM 6.05 freezes in the middle of downloading mail" "^From:" nil nil "1" nil nil nil nil] nil) Received: from joker.cs.washington.edu (joker.cs.washington.edu [128.95.1.42]) by june.cs.washington.edu (8.8.5+CS/7.2ju) with SMTP id WAA16495 for ; Sat, 25 Jan 1997 22:39:53 -0800 Received: from june.cs.washington.edu (june.cs.washington.edu [128.95.1.4]) by joker.cs.washington.edu (8.6.12/7.2ws+) with ESMTP id WAA27334 for ; Sat, 25 Jan 1997 22:39:51 -0800 Received: from crystal.WonderWorks.COM (crystal.WonderWorks.com [192.203.206.1]) by june.cs.washington.edu (8.8.5+CS/7.2ju) with ESMTP id WAA15664 for ; Sat, 25 Jan 1997 22:04:33 -0800 Received: by crystal.WonderWorks.COM id QQcacu05236; Sun, 26 Jan 1997 01:03:39 -0500 (EST) Message-Id: In-Reply-To: <199701260311.WAA18056@ds1.gl.umbc.edu> References: <199701260311.WAA18056@ds1.gl.umbc.edu> From: Kyle Jones To: jeff.covey@pobox.com (jeffrey covey) Cc: bug-vm@uunet.uu.net, ntemacs mailing list Subject: VM 6.05 freezes in the middle of downloading mail Date: Sun, 26 Jan 1997 01:03:39 -0500 (EST) jeffrey covey writes: > i've been having a very frustrating problem with vm for about > a week. sometimes it works perfectly, but at least once a day > i try to download the mail from my pop server and vm just > freezes. Late last month Bill Carpenter reported that VM's pop client tends to make ntemacs exhaust memory with retrieving large messages, probably due to the heap being excessively fragmented. Emacs might seem frozen while it thrashes memory. > i telnet in, and when i open pine i just see one > message. this is a snapshot of it from when i presented the > problem to my systems people (they didn't know what to do): > > N 1 Jan 22 jcovey1@gl.umbc.e (220k) /umbc/general/v17/jcovey1/.mailspool > > (the display is trimming off the last bit, "/jcovey1"). > > when i open it, i have a mountain of messages listed one after > the other (as though i'd just piped a mail folder through > more). once i move the jcovey1 file out of .mailspool, all is > well again. also, a file of 0 length named .jcovey1.pop has > appeared in .mailspool; i've never seen it before. This seems unrelated to Emacs or VM. Best guess would be that whatever it is that's delivering mail into your mailbox, or perhaps whatever it is that's assembling a mailbox for the POP server's use isn't always putting the correct message separators between the messages. The POP server sees one gigantic message (as does pine) and sends it to VM as such. VM tries to handle it, fragments the heap and brings Emacs to its knees. I have no solution. Bill Carpenter's solution was to insert delays into VM's POP read loop, which caused it to retrieve larger chunks of data from the server per iteration, which reduced heap fragmentation. I'm not happy about making a pretty slow POP client even slower, so I'm hoping a better idea will come to me. It hasn't so far. From owner-ntemacs-users@june Sat Jan 25 19:44:28 1997 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] [nil "Sat" "25" "January" "1997" "22:11:55" "-0500" "jeffrey covey" "jeff.covey@pobox.com" nil "193" "VM 6.05 freezes in the middle of downloading mail" "^From:" nil nil "1" nil nil nil nil] nil) Received: from joker.cs.washington.edu (joker.cs.washington.edu [128.95.1.42]) by june.cs.washington.edu (8.8.5+CS/7.2ju) with SMTP id TAA11089 for ; Sat, 25 Jan 1997 19:44:28 -0800 Received: from june.cs.washington.edu (june.cs.washington.edu [128.95.1.4]) by joker.cs.washington.edu (8.6.12/7.2ws+) with ESMTP id TAA24152 for ; Sat, 25 Jan 1997 19:44:27 -0800 Received: from ds1.gl.umbc.edu (ds1.gl.umbc.edu [130.85.3.11]) by june.cs.washington.edu (8.8.5+CS/7.2ju) with ESMTP id TAA10097 for ; Sat, 25 Jan 1997 19:12:08 -0800 Received: from JEFFREY-COVEY (jcovey1@umbc10.umbc.edu [130.85.3.14]) by ds1.gl.umbc.edu (8.8.5/8.6.9) with SMTP id WAA18056; Sat, 25 Jan 1997 22:11:55 -0500 (EST) Message-Id: <199701260311.WAA18056@ds1.gl.umbc.edu> X-Mailer: emacs 19.34.1 (via feedmail 3 I) From: jeff.covey@pobox.com (jeffrey covey) To: bug-vm@uunet.uu.net CC: ntemacs mailing list Subject: VM 6.05 freezes in the middle of downloading mail Date: Sat, 25 Jan 1997 22:11:55 -0500 (EST) i've been having a very frustrating problem with vm for about a week. sometimes it works perfectly, but at least once a day i try to download the mail from my pop server and vm just freezes. i telnet in, and when i open pine i just see one message. this is a snapshot of it from when i presented the problem to my systems people (they didn't know what to do): N 1 Jan 22 jcovey1@gl.umbc.e (220k) /umbc/general/v17/jcovey1/.mailspool (the display is trimming off the last bit, "/jcovey1"). when i open it, i have a mountain of messages listed one after the other (as though i'd just piped a mail folder through more). once i move the jcovey1 file out of .mailspool, all is well again. also, a file of 0 length named .jcovey1.pop has appeared in .mailspool; i've never seen it before. so at least once a day i have to catch this happening and jump in there and move the file and mail it to myself so i can take it apart here and see what's in it. any ideas? it's a real drag... thanks! Emacs : GNU Emacs 19.34.1 (i386-*-nt4.0) of Thu Aug 29 1996 on BANANA-FISH Package: VM 6.05 current state: ============== (setq vm-arrived-message-hook nil vm-arrived-messages-hook nil vm-auto-center-summary nil vm-auto-decode-mime-messages nil vm-auto-displayed-mime-content-types '("text") vm-auto-folder-case-fold-search nil vm-auto-get-new-mail nil vm-auto-next-message t vm-berkeley-mail-compatibility nil vm-check-folder-types t vm-circular-folders 0 vm-confirm-new-folders nil vm-confirm-quit 0 vm-convert-folder-types t vm-crash-box "~/INBOX.CRASH" vm-default-folder-type 'From_ vm-delete-after-archiving t vm-delete-after-bursting nil vm-delete-after-saving t vm-delete-empty-folders t vm-digest-burst-type "guess" vm-digest-identifier-header-format "X-Digest: %s\n" vm-digest-center-preamble t vm-digest-preamble-format "\"%s\" (%F)" vm-digest-send-type "rfc934" vm-display-buffer-hook nil vm-display-using-mime t vm-edit-message-hook nil vm-edit-message-mode 'text-mode vm-flush-interval t vm-folder-directory "~/mail/" vm-folder-read-only nil vm-follow-summary-cursor t vm-forward-message-hook nil vm-forwarded-headers nil vm-forwarding-digest-type "rfc934" vm-forwarding-subject-format "forwarded message from %F" vm-frame-parameter-alist nil vm-frame-per-completion t vm-frame-per-composition t vm-frame-per-edit t vm-frame-per-folder t vm-frame-per-summary nil vm-highlight-url-face 'bold-italic vm-highlighted-header-regexp nil vm-honor-page-delimiters nil vm-in-reply-to-format "%i" vm-included-text-attribution-format "%F writes:\n" vm-included-text-discard-header-regexp nil vm-included-text-headers nil vm-included-text-prefix " > " vm-inhibit-startup-message nil vm-init-file "~/.vm" vm-invisible-header-regexp nil vm-jump-to-new-messages t vm-jump-to-unread-messages t vm-keep-crash-boxes nil vm-keep-sent-messages 1 vm-mail-header-from nil vm-mail-hook nil vm-mail-mode-hook '(set-mail-archive-file-name sq-quote-setup) vm-mime-alternative-select-method 'best-internal vm-mime-button-face 'bold-italic vm-mime-base64-decoder-program nil vm-mime-base64-decoder-switches nil vm-mime-base64-encoder-program nil vm-mime-base64-encoder-switches nil vm-mime-external-content-types-alist nil vm-mime-internal-content-types t vm-mode-hook '((lambda nil (local-set-key "r" (quote vm-followup)) (vm-fontify)) ) vm-mode-hooks nil vm-mosaic-program "Mosaic" vm-move-after-deleting t vm-move-after-undeleting nil vm-move-messages-physically nil vm-movemail-program "movemail" vm-mutable-frames nil vm-mutable-windows t vm-netscape-program "netscape" vm-options-file "~/.vm.options" vm-pop-md5-program "md5" vm-popup-menu-on-mouse-3 t vm-preview-lines 0 vm-preview-read-messages nil vm-primary-inbox "~/INBOX" vm-quit-hook nil vm-recognize-pop-maildrops "^[^:]+:[^:]+:[^:]+:[^:]+:[^:]+" vm-reply-hook nil vm-reply-ignored-addresses nil vm-reply-ignored-reply-tos nil vm-reply-subject-prefix nil vm-resend-bounced-discard-header-regexp nil vm-resend-bounced-headers '("Resent-" "From:" "Sender:" "Reply-To:" "To:" "Cc:" "Subject:" "Newsgroups:" "In-Reply-To:" "References:" "Keywords:" "X-") vm-resend-bounced-message-hook nil vm-resend-discard-header-regexp "\\(\\(X400-\\)?Received:\\|Resent-\\)" vm-resend-headers nil vm-resend-message-hook nil vm-retrieved-spooled-mail-hook nil vm-rfc1153-digest-discard-header-regexp "\\(X400-\\)?Received:" vm-rfc1153-digest-headers '("Resent-" "Date:" "From:" "Sender:" "To:" "Cc:" "Subject:" "Message-ID:" "Keywords:") vm-rfc934-digest-discard-header-regexp nil vm-rfc934-digest-headers '("Resent-" "From:" "Sender:" "To:" "Cc:" "Subject:" "Date:" "Message-ID:" "Keywords:") vm-search-using-regexps nil vm-select-message-hook nil vm-select-new-message-hook nil vm-select-unread-message-hook nil vm-send-digest-hook nil vm-send-using-mime t vm-skip-deleted-messages t vm-skip-read-messages nil vm-startup-with-summary t vm-strip-reply-headers nil vm-summary-format "%n %*%a %-15.15F [%-3.3w %-3.3m %2d] (%4l) %I %s\n" vm-summary-highlight-face 'bold vm-summary-mode-hook nil vm-summary-mode-hooks nil vm-summary-redo-hook nil vm-summary-show-threads nil vm-summary-subject-no-newlines t vm-summary-thread-indent-level 2 vm-summary-uninteresting-senders nil vm-summary-uninteresting-senders-arrow "To: " vm-tale-is-an-idiot nil vm-temp-file-directory "/tmp" vm-trust-From_-with-Content-Length nil vm-undisplay-buffer-hook nil vm-unforwarded-header-regexp "only-drop-this-header" vm-url-browser 'vm-mouse-send-url-to-netscape vm-url-search-limit 12000 vm-use-menus nil vm-use-toolbar '(next previous delete/undelete autofile file reply compose print visit quit nil help) vm-virtual-folder-alist nil vm-virtual-mirror t vm-visible-headers '("Resent-" "From:" "Sender:" "To:" "Apparently-To:" "Cc:" "Subject:" "Date:") vm-visit-folder-hook nil vm-visit-when-saving 0 vm-warp-mouse-to-new-frame nil vm-window-configuration-file "~/.vm.windows" features '(reporter electric bbdb-com bbdb-vm bbdb t appt diary-lib cal-china lunar solar cal-dst cal-islam cal-julian cal-hebrew holidays cal-move calendar cal-menu fdb rsz-mini avoid delsel pc-select fold-isearch folding advice advice-preload generic-extras generic-mode mailabbrev tapestry vm-window vm-virtual vm-version vm vm-vars vm-undo vm-toolbar vm-thread vm-startup vm-summary vm-sort vm-search vm-save vm-reply vm-pop vm-page vm-motion vm-mouse vm-misc vm-minibuf vm-mime vm-message vm-menu vm-mark vm-license vm-folder vm-edit vm-easymenu vm-digest vm-delete sig-quote mini-cl cl cl-19 rnews mail-utils smtpmail sendmail env paren timer font-lock hilit19 ediff-hook vc-hooks lisp-float-type disp-table ls-lisp lisp-mode page menu-bar select scroll-bar mouse facemenu faces frame) ) unsolicited commercial e-mail will be billed at $250/message.