Re: not quoting signatures on reply
On Fri, Aug 11, 2000 at 02:40:18AM +0200, Michael Tatge wrote: Caster muttered: It's a valid escape sequence, and I think I've heard of it before. It might not, however, be fully documented :-) In fact I don't see any mention of it in the docs (I mean in refer to the name of the edited file). It's there: 5.3.3.3. Command Expansion The various commands defined in the mailcap files are passed to the /bin/sh shell using the system() function. Before the command is passed to /bin/sh -c, it is parsed to expand various special parameters with information from Mutt. It would be nice to mention that $editor also passes through this expansion stage (and add a reference to this section in 6.3.35. which talks about $editor). Are there more places other than mailcap and $editor that accept these keywords? Marius Gedminas -- Uncle Cosmo, why do they call this a word processor? It's simple, Skyler. You've seen what food processors do to food, right?
Re: unset metoo not working
On Wed, Aug 09, 2000 at 11:26:06AM -0400, Michael Soulier wrote: Hey guys. I know why unset metoo isn't working, I'm just wondering about a workaround. My external email address is [EMAIL PROTECTED] However, internally, the mail servers change the domain to something else. Because they don't match, I believe that mutt assumes that it's not me, so even though unset metoo is in my .muttrc file, I'm still included on replies. Is there a way around this? Check out `alternates' setting (section 6.3.6 in my copy of the manual). Marius Gedminas -- Never assume the reader has read the subject line.
Re: Writing Return-Path
On Tue, Aug 08, 2000 at 06:11:05PM +1000, Rob Watkin wrote: I don't want the recipient to ever know the actual Linux account name which actually handled the mail but the recipient gets this: Return-Path: [EMAIL PROTECTED] Is there any way to alter this field in .muttrc? Should I be trying to do all this in a completly different way? IIRC this header field is added by an MTA and contains the envelope from address. If you want it to match From header, set Mutt's variable $evelope_from. This variable appeared in 1.1.x development series, and if you don't want to upgrade to 1.2, you'll have to manually set $sendmail to add `-f address' to its default value. HTH, Marius Gedminas -- If something has not yet gone wrong then it would ultimately have been beneficial for it to go wrong.
Re: Missing References (was: Some small problems...)
On Tue, Aug 08, 2000 at 10:11:06AM +0200, Martin Schröder wrote: I'm curious: Why has the mail by Jesper no References:- or In-Reply-To:-Header? Is it possible to disable this in mutt? AFAIK the only way to do it is to have $edit_headers set and then remove them in your editor. (That's what I do when I want to start a new thread). Marius Gedminas -- Those parts of the system that you can hit with a hammer (not advised) are called hardware; those program instructions that you can only curse at are called software. -- Levitating Trains and Kamikaze Genes: Technological Literacy for the 1990's.
Re: Some small problems...
On Mon, Aug 07, 2000 at 10:09:29PM +0200, Jesper Holmberg wrote: Well this is exactly what I've experienced as well. On RedHat 6.1, end didn't work. After an upgrade to RedHat 6.2, the problem disappeared. Now, on Mandrake 7.1, as my previous question indicated, neither home, end, ALT nor F1-F12 work. Could anyone explain this? These three systems are indeed rather similar. OK, you asked for it... rant Various terminal control sequences, including the ones that the terminal sends to the program on various key presses, are described in the terminfo database (distributed with ncurses). The program knows what type of terminal you have by checking the $TERM variable. This is a nice and clever system. Unfortunatelly, very often the database and the actual terminal control sequences get out of sync. This is due to the sad fact, that different versions of a terminal program and even different kinds of them often advertise themselves with the same $TERM value ("xterm" is the most popular), but send and interpret quite different control sequences. Besides, even the same version of the same program can be compiled with different #defines and have various configuration files that change the control sequences, but terminfo description stays the same. The result is a mess. I've yet to see a single Linux distribution which doesn't need terminfo hacking for all keys to be recognised correctly in all terminal emulators (things usually work fine in the Linux console... except the notorious Backspace/Delete issue). I admit that I haven't seen much of them (only different versions of Slackware, Red Hat and Mandrake). All keys work for me now. This is because I spent weeks configuring Eterm: hacking its terminfo, playing with configure options and various #defines, and sending bug reports to the author (which was quick to respond, thanks, Michael). It took so long mostly because I was doing it by trial and error, and had to learn about this terminfo thing and how it works. I think I could fix a broken[1] terminfo entry in a couple of minutes now: $ infocmp $TERM $TERM.ti $ vi $TERM.ti (with a couple of shells to do `man terminfo') $ tic $TERM.ti --- [1] It's hard to say which is broken -- the terminfo entry, or the terminal program, which doesn't correctly emulate the terminal described in terminfo. But it is easier to change the terminfo entry than the program. /rant Good luck, Marius Gedminas -- If C gives you enough rope to hang yourself, C++ gives you enough rope to bind and gag your neighborhood, rig the sails on a small ship, and still have enough rope left over to hang yourself from the yardarm.
Re: problem bouncing messages
On Wed, Aug 02, 2000 at 12:41:45AM +, [EMAIL PROTECTED] wrote: These solutions are not perfect, as they override $envelope_from (I think... what would sendmail do if it got two -f options?), I've checked -- sendmail barfs if it gets -f twice. [...] But I just tested it and it seems to have a problem: "manually" doing the job $envelope_from performs isn't easy if you wish to have different "From " headers (based on "From:")... I guess this problem was what you mean in the first place when you said that this solution wasn't perfect... perhaps you were right :( That's what I meant. Thank you for your help, Manuel PS: Probably this is trivial to implement: I guess the team of developers can easily make $envelope_from also assure that "From "=="From:" when the user bounces a message... afterall, it already works flawlessly when sending, replying and forwarding! Maybe something like this would help: this should make envelope from same as the added Resent-From: header. The patch is against 1.2.4 and completely untested. --- sendlib.c.orig Sat May 20 08:38:59 2000 +++ sendlib.c Wed Aug 2 13:03:52 2000 @@ -1890,7 +1890,7 @@ mutt_copy_bytes (fp, f, h-content-length); fclose (f); -mutt_invoke_sendmail (NULL, to, NULL, NULL, tempfile, +mutt_invoke_sendmail (resent_from, to, NULL, NULL, tempfile, h-content-encoding == ENC8BIT); } Marius Gedminas -- Laugh at your problems; everybody else does.
Re: Sort
On Wed, Aug 02, 2000 at 05:35:03AM +0200, Christopher Rosendahl wrote: How do i set so that mail comming from ex. [EMAIL PROTECTED] automaticly will be moved for ex. another mailbox called Freshmeat after i have read it? Use send-hooks (see the fine manual) and use the `s' key after reading a message. You may also want to use a macro like this to avoid confirmation prompts: macro pager S 'save-message\n\n' "save message to the default location" Marius Gedminas -- As far as we know, our computer has never had an undetected error. -- Weisert
Re: problem bouncing messages
On Wed, Aug 02, 2000 at 01:25:58PM +, [EMAIL PROTECTED] wrote: Maybe something like this would help: this should make envelope from same as the added Resent-From: header. The patch is against 1.2.4 and completely untested. [...] I'll recompile mutt this afternoon and let you know (must download 1.2.4i first). Thanks! IIRC 1.2.5 is the latest version. It would be time for me to upgrade too... Do you think we should drop some sort of message in mutt-dev about this? As Suresh put it, it is a really awkward "feature"... :) Good idea. Could you do it? I do not trust my literature skills much ;) Marius Gedminas -- "Linux was made by foreign terrorists to take money from true US companies like Microsoft." -Some AOL'er. "To this end we dedicate ourselves..." -Don (From the sig of "Don" [EMAIL PROTECTED])
Re: Sort
On Wed, Aug 02, 2000 at 06:09:30PM +0300, Mikko Hänninen wrote: Marius Gedminas [EMAIL PROTECTED] wrote on Wed, 02 Aug 2000: Use send-hooks (see the fine manual) and use the `s' key after reading a message. Oops! He means save-hooks, of course. I know that's what he was thinking anyway, his fingers just typed the wrong word. ;-) Thanks for correcting me. Marius Gedminas -- Computers are not intelligent. They only think they are.
Re: problem bouncing messages
On Tue, Aug 01, 2000 at 04:33:42PM +, [EMAIL PROTECTED] wrote: Hello, I have come across a minor problem with mutt-1.2.5: when I bounce a message, my ISP's SMTP server always refuses to accept it. I have a similair problem -- bounced messages are lost and never reach the destination. It tries to resolve my home computer's name, and since the address look-up always fails it refuses the connection attempt from my local MTA. Can you tell me why this happens? I think I can. I've set $sendmail to a wrapper script and noticed, that the "-f user@host" is not added when bouncing a message, despite $envelope_from being turned on. In my .muttrc I set the variable envelope_from, which fills the "From " field in all my sent/forwarded emails correctly (by "hiding" my local address), but perhaps this doesn't work with bounced emails and my local address ([EMAIL PROTECTED]) show up in the header and confuses my ISP's server. If this is the problem, how can I overcome it? You can add -fyour@address to $sendmail variable or you can make your MTA do envelope masquerading. These solutions are not perfect, as they override $envelope_from (I think... what would sendmail do if it got two -f options?), and sometimes you need to have your envelope from equal to the From: field. Marius Gedminas -- Q: Why do mountain climbers rope themselves together? A: To prevent the sensible ones from going home.
Re: Xterm versus Eterm
On Thu, Jul 27, 2000 at 08:32:26PM +0200, Jan Houtsma wrote: i have the following macro which works for xterm: macro index F2 "escVT~A\n;Wn^T~A\nescVc\n" "change-folder and mark all read" however when i use Eterm i have to change the key into: macro index esc[12~ "escVT~A\n;Wn^T~A\nescVc\n" "change-folder and mark all read" Both use TERM=xterm. Why is that? Because Eterm is not exactly an xterm, and $TERM=xterm doesn't describe it accuratelly. There's a terminfo description in Eterm sources (doc/Eterm.ti). Either install it (then it will overwrite your xterm definition and f2 won't be recognised in the actual xterm) or modify it before installing to change the name to Eterm (and then set TERM to Eterm). This might or might not work -- as I remember, Eterm.ti had some bugs which were fixed in the CVS version. I don't remember if the 0.9 release has those fixes (and I'm pessimistic about 0.8.x). I use the latest CVS version -- all bugs that annoyed me with 0.8.x and 0.9 are fixed there. Marius Gedminas -- Science is to computer science as hydrodynamics is to plumbing.
Re: colors for the indicator (more)
On Wed, Jul 26, 2000 at 12:02:58AM +0300, Mikko Hänninen wrote: No, you can't use that. The subscribe command takes a substring (well, regular expression) to match for in recipient addresses. It won't pay any attention to whatever aliases you have defined. Is it really a regexp? I'm also not sure about the substring. My muttrc contains `subscribe @bugs.guug.de', but mails sent to [EMAIL PROTECTED] (where 42 is any string) are not matched by ~l. Marius Gedminas -- Power corrupts, but we need electricity.
Re: M$ dog
On Mon, Jul 17, 2000 at 10:18:25PM -0400, David T-G wrote: % Since I must, I use win2k at work. I've found PC-Pine to be a marginally I hear TheBAT! is pretty good, BTW. (Yes, but I would trade it for Mutt any minute...) % acceptable mailer, but I'd love to use mutt. I've found nothing on the % web about this. My guess is that this is "by design" behavoir. Well, yes and no; the design is definitely a factor, but it wasn't done just because people hate windoze (I say "just because" :-) It boils down to mutt's UNIX-like behavior of expecting other pieces to be there; all mutt does is read mail, and it sucks less at that than anything else. The biggest problem people often quote is the lack of an MTA, since mutt does not talk directly to the MDA on the recipient machine via SMTP port 25, as I believe PINE does. It doesn't help that Win doesn't handle subshells gracefully and so things like PGP can be tricky as well, as I understand it. There's another problem: \r\n end-of-lines. Mutt doesn't use O_TEXT/O_BINARY, and gets confused when CygWin translates \r\n to \n thus unexpectedly reducing message size. At least I think that's why Mutt segfaults in the pager on Win32. I have posted stack traces on mutt-dev, but the developers have enough problems on Unix, and didn't have time to investigate Win32 issues. On Tue, Jul 18, 2000 at 07:39:26AM -0400, Todd Goodman wrote: * John Saylor [EMAIL PROTECTED] [000717 21:50]: Has anyone done anything about porting mutt to the windows world? Someone has reportedly ported mutt and sendmail to Cygwin (see http://sourceware.cygnus.com/cygwin). Check out http://www.fiction.net/blong/programs/mutt/ for the mutt patch. IIRC it requires CygWin 1.0 DLL which is commercial. I've also heard that it is open source, so maybe someone could make it available for downloads? I haven't tried it, but I've compiled Mutt with CygWin 0.something, and it worked (modulo the pager segfaults mentioned above). Marius Gedminas -- Of course I use Microsoft. Setting up a stable unix network is no challenge ;p
Re: codepages aliases
On Tue, Jul 18, 2000 at 02:54:54PM +0300, Leonid Mamtchenkov wrote: Hey everyone, Guess I have missed it in the manual, but how do I alias codepages in 1.2 and above? By using the unintuitively named `charset-hook' command. HTH. Marius Gedminas -- Unix is an operating system, OS/2 is half an operating system, Windows is a shell, and DOS is a boot partition virus. -- Peter H. Coffin
SMTP
Isn't it strange, that Mutt does not do SMTP in order to simplify things and avoid duplication of functionality, while fetchmail uses SMTP for mail delivery in order to simplify things and avoid duplication of functionality? Marius Gedminas -- We don't care. We don't have to. We're the Phone Company.
Re: SMTP
I mean -- both Mutt and fetchmail require an MTA[1], but fetchmail uses SMTP to access it (IIRC to avoid problems with different command line arguments for different MTAs etc.). Mutt also needs an MTA, but it goes with the command line solution. And there are problems with this solution -- witness the recent questions about RFC-822 addresses containing spaces and the requirement for special wrapper scripts to make them work with qmail. --- [1] It can be argued, that fetchmail needs an MDA, not an MTA... but the thing listening on port 25 is usually an MTA. On Thu, Jul 13, 2000 at 11:44:34AM -0400, David T-G wrote: The mutt equivalent of fetchmail talking to SMTP would be mutt talking to port 25 only on your machine, and therefore only after you had set up an MTA already. Exactly. And this leads to the next trick question: what's the difference between talking to localhost:25 or remotehost:25? :) It's not that I advocate adding SMTP support to Mutt [2], but I just wonder, why two programs following the same Unix philosophy (do one thing well) choose so different solutions [3] for so similair tasks -- handing an email to an MTA. --- [2] Well, that would be a nice thing to have if I try to compile it on WinNT again ;) [3] I know that fetchmail can feed the email to an external program instead of localhost:25, but the default way is SMTP. BTW in the standard Unix world (MDA/MUA/MTA), where does fetchmail fit in? Marius Gedminas -- All those who believe in psychokinesis raise my hand.
Re: sendhook
On Tue, Jul 11, 2000 at 05:42:13PM +0300, Mikko Hänninen wrote: Jerry Walsh [EMAIL PROTECTED] wrote on Tue, 11 Jul 2000: the problem is i need to be able to also set my From header to something else when i do this? Can i specify two sendhooks? or will they overwrite eachother? Two send-hooks with the exact same regexp pattern will overwrite, I believe you are wrong here. My ~/.muttrc contains three send-hooks with pattern `.', and two send-hooks with pattern `^~C \.lt$'. They all are executed. Marius Gedminas -- HOST SYSTEM RESPONDING, PROBABLY UP...
Re: read only spool??
On Thu, Jul 06, 2000 at 08:47:46AM +0200, Dirk Ruediger wrote: I user mutt on a nfs-mounted dir. But unfortunatly the systen crashed during a mutt session (mutt was innocent ;-) and now I can't write to this dir, mutt always says, that this folder is write-only. I can't find a lock file. Maybe, the remote box (Solaris) locks it and I can't convine it to stop locking (I also made `mutt_dotlock -u' on every folder, without success. (Now I have a maildir on a local filesys and a soft link to the usual dir ~/Mail). What happened and what is to be done to fix it? I don't know. I've never used NFS or Maildir (though I tried the latter once). Maybe you need to use -f or -p parameters for mutt_dotlock? (BTW the manual refers to `mutt_dotlock (8)', but mutt_dotlock man page is in section 1.) Marius Gedminas -- If all else fails, read the documentation.
Re: read only spool??
On Wed, Jul 05, 2000 at 02:28:22PM -0400, Michael Soulier wrote: This is strange. I changed my spoolfile from a local file to /var/mail/msoulier, which I have write permission to: [bmerh56e-msoulier-mail]$ ll total 1282 -rw-rw 1 msoulier mail 65556 Jul 5 14:21 msoulier And yet, mutt won't modify the contents, because it says that the file is read only?? Elm likes it fine. What would cause this? Locking. For Mutt's dotlocking to work, your /var/spool/mail directory should be writable by group `mail', and your `mutt_dotlock' should be sgid mail. HTH, Marius Gedminas -- This company has performed an illegal operation and will be shut down. If the problem persists, contact your vendor or appeal to a higher court.
Re: read only spool??
On Wed, Jul 05, 2000 at 03:05:53PM -0400, Michael Soulier wrote: On Wed, Jul 05, 2000 at 08:39:47PM +0200, Marius Gedminas wrote: Locking. For Mutt's dotlocking to work, your /var/spool/mail directory should be writable by group `mail', and your `mutt_dotlock' should be sgid mail. Where can I find information on this? On this list. :) I just grepped through the entire manual for "lock", and I didn't find anything on this. Well, the $dotlock_program variable is described there. And there's a reference to mutt_dotlock(8) man page (although the reader is expected to unserstand that "(8)" after "mutt_dotlock" makes it a reference). But you're right, the available documentation on this issue is sparse and incomplete. :-/ There could be a couple of sentences in the manual detailing Mutt's locking mechanisms, mentioning facts like that dotlocking does not work with msdos/vfat filesystems, while fcntl/flock is unreliable over NFS... [bmerh56e-msoulier-msoulier]$ cd /var [bmerh56e-msoulier-var]$ ll -d mail drwxrwxr-x 2 binmail 1024 Jul 5 14:32 mail/ Ok, the directory is writable by mail... So, what's a "mutt_dotlock"? It's a program (part of the Mutt package) that implements NFS-safe file locking. To ensure that different programs do not write to the same mailbox at once (thus corrupting it), mutt_dotlock creates a lock file in the same directory (that's why the directory has to belong to the group `mail', be group-writable, and mutt_dotlock has to be sgid mail). Check its permissions with $ ls -l `which mutt_dotlock` Marius Gedminas -- "Nuclear war can ruin your whole compile." -- Karl Lehenbauer
Re: Mutt Compile Problem
On Sat, Jul 01, 2000 at 10:20:49PM -0500, Donnie Pennington wrote: On Sat Jul 01 2000 at 07:03:30PM +0100, Lars Hecking wrote: Ok. According to this, you won't be able to compile anything with your compiler. You need to install the glibc-devel RPM which includes essential files like the startup codes for C programs (crt1.o), and libraries and include files. I installed glibc-devel and ncurses-devel. The mutt compile acted like it was working, but when I check mutt after the compile I'm still left with the old one. This sounds like you did a `make install' and it installed everyting into /usr/local, while your old mutt version is still in /usr. Try running /usr/local/bin/mutt. If it works, you can rpm -e your old mutt package. (Also check if /usr/local/bin in the $PATH, Red Hat does not add it by default IIRC). Here is the latest config.log. The problem appears to be at line 3593. That's not actually a problem. It just indicates that your /var/spool/mail is not world writable. Mutt will use a sgid mail mutt_dotlock binray in this situation. Marius Gedminas -- Unfortunately, most programmers like to play with new toys. I have many friends who, immediately upon buying a snakebite kit, would be tempted to throw the first person they see to the ground, tie the tourniquet on him, slash him with the knife, and apply suction to the wound. -- Jon Bentley
Re: Mutt Compile Problem
On Sun, Jul 02, 2000 at 01:06:29PM -0500, Donnie Pennington wrote: Well, you're absolutely correct. Mutt installed to /usr/local rather than /usr. Is something I could have tweaked to get the install to go to /usr instead? Yes, the --prefix=usr option to ./configure. However I prefer to install programs into /usr/local, leaving /usr for software installed from rpms. Marius Gedminas -- An NT server can be run by an idiot, and usually is.
Re: UTF-8?
On Fri, Jun 30, 2000 at 10:42:26AM +0200, Thomas Roessler wrote: On 2000-06-29 19:36:24 +0200, Marius Gedminas wrote: Can Mutt 1.2 handle UTF-8? Only on the receiving side. OK. Maybe this should be documented before 1.2.3 goes out? For heavily improved utf-8 support (including the use of UTF-8 as the local display's character set), I didn't know Mutt-1.3.x supported utf-8 display without additional patches. I guess it still needs a utf-8 enabled slang, right? you'll have to look at the unstable versions. Note that the publically distributed tar balls don't even need to compile... ;-) (But you read mutt-dev, I believe, so you know that already.) I play with the CVS version a bit. Marius Gedminas -- If it wasn't for C, we'd be using BASI, PASAL and OBOL
Re: those users (was Re: Reply to all???)
On Fri, Jun 30, 2000 at 04:01:47PM +0530, Suresh Ramasubramanian wrote: Why don't you post each FAQ to the list as each one is done (all in one thread) so that people can suggest changes, as well as track it's development? [...] The original author of Mutt was Michael Elkins. (I think ME still *is* the original author.) [...] One of these tweaks is setting your outbound address, ie the one that will show up in the From: line. For example, I have a linux box called whack.spammers.cluestick.org, but my address is [EMAIL PROTECTED] I don't want to send out mails as [EMAIL PROTECTED] - I wouldn't even get a single reply, and anyone trying to mail me at this very entertaining address would get a bounce. So, I have to tell mutt that my address is actually [EMAIL PROTECTED] The line in my muttrc then would read like this: my_hdr From: Suresh Ramasubramanian [EMAIL PROTECTED] This won't fix your (I mean, the user's) envelope address. I suggest mentioning $envelope_from at this point, or bounces and error messages won't reach you (and some SMTP servers won't accept non-existing domain names in the envelope -- which is often the case for dialup users). Marius Gedminas -- C is a language that combines all the elegance and power of assembly language with all the readability and maintainability of assembly language.
Re: imap_checkinterval doesn't seem to be working
On Fri, Jun 30, 2000 at 02:38:38PM -0500, Ben Beuchler wrote: I am using mutt 1.3.3 and reading my mail off a Courier IMAP server. I have imap_checkinterval = 10. My understanding is that that is how long it will wait to check mail when I am not actually pressing any keys. No, that's what $timeout is for. However, it seems to be waiting considerably longer than 10s. I just now had a new message become visible that, per the logs, was actually delivered to my mailbox about 4 minutes ago. Of course, if I press any of the navigation keys while on the index screen, new mail shows up instantly. Any suggestions? I suggest to add this question to the FAQ, as it definetely is one. Marius Gedminas -- To err is human, but to really foul things up requires a computer.
UTF-8?
Can Mutt 1.2 handle UTF-8? When I have $charset="iso-8859-13", and $send_charset="utf-8", bad things happen. BTW, I have glibc 2.1.3. Marius Gedminas -- Hoping the problem magically goes away by ignoring it is the "microsoft approach to programming" and should never be allowed. -- Linus Torvalds
Re: those users (was Re: Reply to all???)
On Fri, Jun 30, 2000 at 01:32:41AM +0100, Thomas Ribbrock wrote: On Tue, Jun 27, 2000 at 12:08:28PM -0400, David T-G wrote: [...] for starters? Heck, maybe the Muttrc included in RPMs could pop up a message saying "Don't ask; read first" until the user figures out how to get rid of it, [...] Hm, maybe for a start I could put a few big, red "Read the manual!" on the RPM Team webpages... ;-) Other than that, I'm not certain a "nag screen" is such a good idea - as tempting as the idea is. Simply too annoying, if you ask me. I'm pretty sure, that if such a message were added to Mutt, the mailing list would be flooded with questions "how do I turn this message off?" ;-) Marius Gedminas -- We don't care. We don't have to. We're the Phone Company.
Re: feature request: delayed delete
On Wed, Jun 28, 2000 at 02:14:28AM -0700, Eugene Lee wrote: Actually, I'd like to have add some kind of search pattern to the message-tagging functions. For example, if I knew that I have a bunch of messages from Januaary and February 2000, I'd like to be to be able to tag them, then do with them as I will. AFAIK, this isn't a feature in Mutt. I know I could tag by searching all the "Date:" fields, but those dates aren't quite standard (the same datetime could come in different formats or in different timezones). Besides tagging messages by absolute datetimes, this could be extended to your specific problem by allowing relative datetime patterns. So you could do things like tag messages that are 14 days old or older. So what's wrong with ~d and ~r? Marius Gedminas -- When does summertime come to Minnesota, you ask? Well, last year, I think it was a Tuesday.
Re: suggestion
On Tue, Jun 27, 2000 at 09:25:10PM -0700, Shawn D. McPeek wrote: And mention that a 32 line sig on a 3 line message is a bit excessive :) Quoting style should get some attention too... ;) Marius Gedminas -- Science is to computer science as hydrodynamics is to plumbing.
Re: save-hook not working
On Tue, Jun 27, 2000 at 08:35:01PM -0700, jon rust wrote: I've got save-hook =customer.in =archived/customer in my dot-muttrc file. However, when I press 's' while reading a message, or while in the folder index (in customer.in) it guesses at the save file name (based on the From: address). What am I missing? Unfortunately, save-hooks match folder names. You could use a folder-hook to set the default save location (that is, redefine save-hooks within folder-hooks). Or you could use different way of determining the save location (e.g. matching by To: field). See also: section 4.2 of the fine manual. Marius Gedminas -- The reason computer chips are so small is that computers don't eat much.
[OT] Paranoia
Something just went wrong. My mails do not reach [EMAIL PROTECTED] any more. This is a test. Marius Gedminas -- The UNIX philosophy basically involves giving you enough rope to hang yourself. And then a couple of feet more, just to be sure.
Re: Qmail and Mutt
On Wed, Jun 28, 2000 at 12:47:32AM +0200, Mark Weinem wrote: On Tue, Jun 27, 2000 at 11:46:43AM -0700, Jason Helfman wrote: [Ignoring qmail+procmail questions] 4 In mutt:set check_new=yes set mailbox_type=Maildir Does "mailbox_type" really work? I use "mbox_type". Several configuration variables have aliases. You can find them with grep DT_SYN /wherever/your/mutt/source/is/init.h But there is no mailbox_type. I guess that was a typo. Marius Gedminas -- Alan Turing thought about criteria to settle the question of whether machines can think, a question of which we now know that it is about as relevant as the question of whether submarines can swim. -- Dijkstra
Re: Suggestion: aka command
On Wed, Jun 21, 2000 at 01:00:47AM -0400, David T-G wrote: I, for one, had trouble following your aka proposal. Do I understand that a configuration like aka [EMAIL PROTECTED] [EMAIL PROTECTED] would automatically readdress any email for [EMAIL PROTECTED] to [EMAIL PROTECTED]? My understanding was that *replying* to a mail from [EMAIL PROTECTED] would address the reply to [EMAIL PROTECTED] On Wed, Jun 21, 2000 at 08:52:15AM -0400, Stan Ryckman wrote: Unfortunately I deleted the original post, but I'm pretty sure that it was something like: Given, in some situations (such as replying), what would be: [EMAIL PROTECTED] Convert (via this aka thing) to produce the header: To: Joe Cool [EMAIL PROTECTED] in otherwords, fill in the "unneeded" part of the address. Not just the "unneeded" part, the whole address. In other words, it defines a "canonical" address for replying. I can find uses for this too. Some mailing lists around here can be reached via multiple aliases -- [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] When list-replying to several messages, sometimes the To: field contains several of these aliases. Something like $alternates for arbitrary addresses would be nice, e.g. aka canonical-address list-of-addresses or aka canonical-address regexp While gathering addresses for replying, Mutt would check all these aka's and if the address matches the regexp, it would replace it with canonical-address (and elminate resulting duplicates). This cannot be solved with reply-hook (or send-hook with the `%' modifier) -- IIRC `my_hdr To:' within send-hooks is ignored. Besides, that wouldn't help when replying to multiple recipients. Marius Gedminas -- Immanuel Kant but Kubla Khan.
Re: 1.2.2i problems
On Wed, Jun 21, 2000 at 11:34:45AM -0600, Vincent Danen wrote: -rwxr-xr-x1 root root 1262377 Jun 21 11:29 /usr/bin/mutt* -rwxr-sr-x1 root mail36607 Jun 21 11:29 /usr/bin/mutt_dotlock* -rwxr-xr-x1 root root 6668 Jun 21 11:29 /usr/bin/muttbug* -rwxr-xr-x1 root root 418000 Jun 21 11:24 /usr/bin/mutt* -rwxr-sr-x1 root root 7588 Jun 21 11:24 /usr/bin/mutt_dotlock* -rwxr-xr-x1 root root 6668 Jun 21 11:23 /usr/bin/muttbug* Anyone know why there's such a discrepancy? I think the RPM might be stripping the binaries but I don't know if that would make such a big difference... I think it could. Try to strip your binaries and look at their sizes then. (That's only 3x difference, C++ programs may decrease in size ten times after stripping.) System: Linux 2.2.15-4mdk [using slang 10400] I also still have a problem with colors, regardless of which binary I use. My color config looks like: [...] When I comment it all out, then I get mono but I can see what I'm doing. Otherwise the colors are severely messed up. Any ideas about this one? The one thing I like about mutt is the color support, but it kinda needs to work too... =( This is Unix. Extreme flexibility and many possible points of failure. Your terminal program could be buggy. It could have different RGB values assigned to colour numbers. Its terminfo description may be wrong (e.g. earlier versions of Eterm specified setf=/setb= instead of setaf=/setab=). Your $TERM may not match your terminal program (all those xterm clones usually are different in this regard, and xterm itself has many versions.) This might be a bug in slang/ncurses (even the latest stable ncurses version (5.0) is buggy). This might be a bug in Mutt (unlikely). Several of the above may be true. I've had *lots* and *lots* of problems with Mutt and colours. All investigations and lots of time spent debugging them showed that the problems weren't Mutt's fault. I finally fixed my terminfo and settled on ncurses 5.0, since it distored my colours less than slang or other ncurses versions. YMMV. (I've been told that latest ncurses developement versions have my problem fixed. I still haven't tried them...) Marius Gedminas -- If it wasn't for C, we'd be using BASI, PASAL and OBOL
Re: Fwd: failure notice
Hello, On Sun, Jun 18, 2000 at 11:53:23PM -0600, Vincent Danen wrote: Hi there. I can't seem to figure this one out. It's like sendmail is re-writing the Return-Path header and it's making my posts to some mailing lists bounce for some reason. [...] Sounds like you need to fix your envelope From_ address. Mutt 1.2 has a nice `set envelope_from' option for this, but you can emulate it with set sendmail='/usr/sbin/sendmail -oem -oi [EMAIL PROTECTED]' (IIRC). Of course, the best solution would be to fix your sendmail settings to do proper masquerading (but I'm too lazy to do that ;) Marius Gedminas -- If something has not yet gone wrong then it would ultimately have been beneficial for it to go wrong.
Re: [bug?] pager_index_lines
On Fri, Jun 16, 2000 at 02:08:51AM +0200, Marius Gedminas wrote: Suggested documentation patch: --- init.h 2000/06/15 21:37:05 2.96 +++ init.h 2000/06/16 00:07:36 @@ -1057,9 +1057,9 @@ ** remain to be read in the current thread. One of the lines is reserved ** for the status bar from the index, so a \fIpager_index_lines\fP of 6 ** will only show 5 lines of the actual index. A value of 0 results in - ** no index being shown. If the number of messages in the current folder - ** is less than \fIpager_index_lines\fP, then the index will only use as - ** many lines as it needs. + ** no index being shown. If the number of visible messages in the current + ** folder is less than \fIpager_index_lines\fP, then the index will only + ** use as any lines as it needs. ^^^ Should be `many'. Thanks, David. Marius Gedminas -- Bumper sticker: Alcohol and calculus don't mix. Never drink and derive.
[bug?] pager_index_lines
Today a strange thing occured to me: setting pager_index_lines to 4 or 10 (I have macros for this) gave me just one index line on top. I've checked (with :set pager_index_lines=tab) that the value was correct. After restarting mutt things were back to normal. Has anyone else experienced this? I'm using Mutt 1.2. Marius Gedminas -- (mental note: stop installing red hat. everytime i do so, it takes ages to fix my system again.) -- from the sig of Martin Högman
Re: muttrc.el
On Thu, Jun 15, 2000 at 10:33:33AM -0600, Charles Curley wrote: - P.S. Please don't use non-standard $indent_string. It prevents - editors with mail modes (like jed) from rewrapping long lines in - quotes while preserving the quote characters. What are the standard quote characters? An RFC reference will do fine. However, I doubt I'm the only person who uses this, so perhaps you could add it to the appropriate regex for your editor? The default $indent_string is " " and the manual says: You are strongly encouraged not to change this value, as it tends to agitate the more fanatical netizens. The Jargon File also mentions " " as the standard inclusion leader. I've grepped RFC-1855 (Netiquette Guidelines), but couldn't find anything about quoting styles. Marius Gedminas -- The last good thing written in C was Franz Schubert's Symphony #9. -- Erwin Dietrich
Re: [bug?] pager_index_lines
On Thu, Jun 15, 2000 at 11:53:46PM +0530, Suresh Ramasubramanian wrote: Marius Gedminas proclaimed on mutt-users that: Today a strange thing occured to me: setting pager_index_lines to 4 or 10 (I have macros for this) gave me just one index line on top. I've checked (with :set pager_index_lines=tab) that the value was correct. After restarting mutt things were back to normal. Has anyone else experienced this? I'm using Mutt 1.2. I have not been able to repro this on mutt 1.3 (standard ./configure and make install) and I hate the two window view, so am removing that line now from my muttrc ;) The only reason why I use it is keeping track of threads. There are no `%' macros for status_format that would show the number of messages in current thread and the number of the current message relative to the start of thread. Marius Gedminas -- "I may not understand what I'm installing, but that's not my job. I just need to click Next, Next, Finish here so I can walk to the next system and repeat the process" -- Anonymous NT Admin
Re: quoted-printable problems
On Thu, Jun 15, 2000 at 10:38:32AM -0700, Cesar Crusius wrote: Hi. I am having problems with quoted-printable messages. Whenever someone sends me with quoted-printable encodings (=E40 and the likes), mutt invariably decodes all the encoded characters to '?,' and not to the iso-8859-1 equivalents. The term handles 8-bit characters correctly, and in fact when I edit the message with vim I can see the accented characters correctly. Any suggestions? You have to set your locale (the LANG, or at least LC_CTYPE environment variable) correctly. In the default POSIX (aka C) locale all characters above 127 are non-printable. If your OS does not have a correct locale database, you can recompile mutt with --enable-locales-fix. Marius Gedminas -- Read what I mean, not what I write.
Re: [bug?] pager_index_lines
On Fri, Jun 16, 2000 at 12:16:28AM +0200, Michael Tatge wrote: Are you sure you had more than 1 message in the folder? On Fri, Jun 16, 2000 at 02:33:56AM +0300, Mikko Hänninen wrote: I've seen this situation when the folder has only one message, the mini index at the top doesn't use up the full assigned space in that case. If the folder had more than one email, then I don't know... I think I now understand what had happened. The folder had more than ten messages, but a limit pattern matching only a single message was active. Suggested documentation patch: --- init.h 2000/06/15 21:37:05 2.96 +++ init.h 2000/06/16 00:07:36 @@ -1057,9 +1057,9 @@ ** remain to be read in the current thread. One of the lines is reserved ** for the status bar from the index, so a \fIpager_index_lines\fP of 6 ** will only show 5 lines of the actual index. A value of 0 results in - ** no index being shown. If the number of messages in the current folder - ** is less than \fIpager_index_lines\fP, then the index will only use as - ** many lines as it needs. + ** no index being shown. If the number of visible messages in the current + ** folder is less than \fIpager_index_lines\fP, then the index will only + ** use as any lines as it needs. */ { "pager_stop", DT_BOOL, R_NONE, OPTPAGERSTOP, 0 }, /* Marius Gedminas -- "Only wimps use tape backup: _real_ men just upload their important stuff on ftp, and let the rest of the world mirror it ;)" (Linus Torvalds, about his failing hard drive on linux.cs.helsinki.fi)
Re: Managing multiple accts with mutt + fetchmail
On Sat, Jun 10, 2000 at 01:37:59AM +0300, Mikko Hänninen wrote: Marius Gedminas [EMAIL PROTECTED] wrote on Fri, 09 Jun 2000: Use `set from=...' instead of `my_hdr From: ...'. Right, this is the way to go. $realname takes precedence over $from, but not over my_hdr. Oops! I wanted to say $reverse_name, not $realname. g Marius Gedminas -- When in trouble or in doubt, run in circles, scream and shout.
Re: Managing multiple accts with mutt + fetchmail
(Extremely long lines wrapped to 72 columns. Hint, hint.) On Fri, Jun 09, 2000 at 03:50:06PM -0400, jeremy redburn wrote: I would like to compose messages by default from [EMAIL PROTECTED] not [EMAIL PROTECTED] because I don't want mail actually sent directly to my local machine -- I want to receive mail only via fetchmail. I can accomplish this with: my_hdr From: Jeremy Redburn [EMAIL PROTECTED] my_hdr Reply-to: Jeremy Redburn [EMAIL PROTECTED] The only problem is that the my_hdr commands seem to override the set reverse_name directive which I need since I would like to reply to mails depending on which account they were delivered to. Use `set from=...' instead of `my_hdr From: ...'. $realname takes precedence over $from, but not over my_hdr. BTW I've heard that setting $from does not work well within send-hooks. Marius Gedminas -- An algorithm must be seen to be believed. -- D.E. Knuth
Re: Colors, a wish ...
On Thu, Jun 08, 2000 at 10:01:42AM +0200, Gerhard den Hollander wrote: now what would be nce if I could do something like (pseudosyntax) color index from_the_boss yellow default color index priority=urgent bright default which means that normal messages are default message that are tagged urgent are set in the bright default color messages from the boss in yellow and urgnet messages fom the boss in bright yellow You can already do that: color index yellow default '~f [EMAIL PROTECTED]' color brightdefault default '~h "^x-priority: urgent"'# or whatever color brightyellow default '~f boss ~h "^x-priority: urgent"' # both Check sections 3.7 and 4.2 for complete documentation. IIRC if multiple patterns match the last one is used. Unfortunately the manual is silent about this. Marius Gedminas -- "Cogito ergo I'm right and you're wrong." -- Blair Houghton
Re: Mutt/Procmail/Fetchmail question NEW colour question
On Wed, Jun 07, 2000 at 09:48:50AM -0700, Timothy Grant wrote: I'm not quite sure I know how to get the colours working. I have run it in both a Eterm and an rxvt session and there is no colour. Do I have to do something to turn colour on, or do I need to do something to my terminal settings to make it work (ls --color works in both sessions, so I know my terminals both are configured for colour support) Set $TERM to `xterm-color' and you should get colour. Marius Gedminas -- Your eyes are weary from staring at the CRT. You feel sleepy. Notice how restful it is to watch the cursor blink. Close your eyes. The opinions stated above are yours. You cannot imagine why you ever felt otherwise.
Re: In-line color in body but not header
On Tue, Jun 06, 2000 at 06:16:32PM -0700, Jeff Krueger wrote: color header cyan default "[-a-z_0-9.+]+@[-a-z_0-9.]+" color body cyan default "[-a-z_0-9.+]+@[-a-z_0-9.]+" The intent is to color just e-mail addresses as defined by this regex differently from the rest of the line. It works fine in the body, even in quoted text - the address is in cyan and the rest of the line is whatever color it is supposed to be. In the header, however, this colorizes the entire line cyan. Am I defining this improperly somehow, or is in-line coloring not yet available? Not available. However once I posted a patch (against Mutt-1.1.1) to add a new color item (`hdrpart') that gives precisely this functionality, in addition to current whole-line coloring of `header'. If anyone is interested, I could blow the dust off it and test it with current 1.2/1.3 versions. Marius Gedminas -- If you are good, you will be assigned all the work. If you are real good, you will get out of it.
Re: new folders
On Tue, Jun 06, 2000 at 06:45:36PM +0300, Mikko Hänninen wrote: [EMAIL PROTECTED] [EMAIL PROTECTED] wrote on Tue, 06 Jun 2000: forgive silly question: is mutt able to create new folders?, just point me to the relevant part of manual please, might be I have overlooked it ;-) If you save or copy a message to a folder that doesn't exist, it will be created automatically. If you're in the file selector and want to pick a new file/folder name, you can press n for selecting a new file that doesn't yet exist. Actually, that should be `N'. `n' stands for search-next in default key bindings (althoug I've heard in a recent thread that using IMAP support binds that key differently). Marius Gedminas -- Do not meddle in the affairs of sysadmins, they are quick to anger and have no need for subtlety.
Re: internationalization
On Thu, Jun 01, 2000 at 09:23:32PM -0400, Sergei Gerasenko wrote: When I do "set charset=CP1251" from inside mutt, it doesn't work, but as soon as I defined a charset hook "charset-hook windows-1251 CP1251", it worked immediately. Do any of you know why? The message you're looking at probably has Content-Type with charse=windows-1251. However glibc implementation of iconv does not recognize `windows-*' charset names. Adding a charset-hook tells it that windows-1251 is the same as cp1251, and glibc recognizes the latter one. Marius Gedminas -- I doubt, therefore I might be.
Re: (OT) editor
On Thu, Jun 01, 2000 at 09:23:38AM -0600, Charles Curley wrote: There is another way to do this, I believe. I seem to recall a tool that lets you launch a program into a given virtual console. Then you can switch between mutt and emacs by toggling virtual consoles, which I believe was your original question. Anyone recall its name? I think Keyboard-and-Console-HOWTO mentions that. It's called open/openvt (IIRC there are a couple of versions, and one of them comes with console-tools). Marius Gedminas -- The Second Law of Thermodynamics: If you think things are in a mess now, just wait! -- Jim Warner
Re: Exclusive matching
On Sun, May 28, 2000 at 10:33:52PM -0400, David T-G wrote: % Ah, I see. It's ^~L "(alice|bob|carol)". Precedence would be a good % topic for the manual to go over quickly. I think most Unix users (after struggling with bash, grep, etc.) develop a feeling on where and how and what to quote. Whoops. Pardon me for jumping in here, but shouldn't any parser be able to detect a ( and put everything up until the matching ) in a separate precedence? It's not so simple as it sounds. There are many quoting issues, like \( or [abc)def], that are not exactly matching parentheses. I still don't see why the parser would break the expression into three ORs with an ( in one and a ) in another unless the code is just plain broken -- but I can't read C enough to have any idea :-) Note that that's not a flame of the code or developers, especially since I'm in no position to provide a patch to "fix" the behavior... The code is indeed broken. Try, for example, this pattern: (~b "default[)]") Mutt just scans for the first matching `)' ignoring any quotes, escapes, etc. OK, I've got to go. Expect a patch in the evening. Marius Gedminas -- "Only wimps use tape backup: _real_ men just upload their important stuff on ftp, and let the rest of the world mirror it ;)" (Linus Torvalds, about his failing hard drive on linux.cs.helsinki.fi)
Re: receive-hook?
On Mon, May 29, 2000 at 07:11:59PM +0200, Daniel Bauke wrote: I can't figure out how to set some variables (i.e. my_hdr From and Organization or .signature) upon the address I have received the e-mail from. send-hook '~t bonkey@work' 'set signature=".sig_work"' send-hook '~t bonkey@home' 'set signature=".sig_home"' obviously doesn't work If you use reverse_name, try replacing ~t with ~f. Unfortunately, there's no way to execute send-hooks dependent on the message you are replying to. Marius Gedminas -- C is quirky, flawed, and an enormous success. -- Dennis M. Ritchie
Re: .fetchmailrc
On Sun, May 21, 2000 at 08:37:21PM -0300, Richard Spencer wrote: I then decided to switch to Mutt, and with some effort, configured my .muttrc to my satisfaction. I used this config file to name my smtp server, Huh? SMTP is for sending mail, and Mutt does not support it. and settled for checking only mail in my primary account. But today I tried to use Mutt to check the other account as well. Here's what I did: created a .fetchmailrc like this... poll pop3.uol.com.br proto pop3 user [EMAIL PROTECTED] pass secret1 fetchall keep poll pop.a001.sprintmail.com proto pop3 user [EMAIL PROTECTED] pass secret2 fetchall keep smtphost smtp.uol.com.br ^^^ AFAICU this line tells fetchmail to forward all messages received from pop.a001.sprintmail.com account to smtp.uol.com.br instead of delivering them locally. I'm not quite sure how this option works -- I've never used it. But I was wrong; the mail for the second account didn't end up in /var/spool/mail/$USER Can anyone tell me: 1) where the mail from sprintmail might be? 2) which files to check for clues? man fetchmail 3) a good way to tell Mutt to read my inbox in /var/spool/mail first? (I already tried making my /home/$USER/Mail/inbox a symlink to /var/spool/mail/$USER, and that seemed not to work as well; the inbox was labeled in a peculiar fashion--it became 'inbox@'. That `@' is only shown to inform you that `inbox' is a symlink. It is not actually a part of the name. I can tell you I'm running Red Hat 6.0 on a Pentium Laptop, and haven't taken any steps to configure sendmail, although it always runs as a process. 430 ? S 0:00 sendmail: accepting connections on port 25 Perhaps you'll want to specify a smart host if your computer do not have a registered domain name. Otherwise many servers will reject SMTP connections from you. But maybe using a simpler MTA is a better choice. Marius Gedminas -- An NT server can be run by an idiot, and usually is.
Re: set print
On Wed, May 24, 2000 at 03:54:23PM -0400, Michael Soulier wrote: Ok, according to my manual, the $print variable, if set... print Type: quadoption Default: ask-no Controls whether or not Mutt asks for confirmation before printing. This is useful for people (like me) who accidentally hit ``p'' often. I have "set print" in my .muttrc file, but when I hit "p", it doesn't ask for confirmation. Does it need to be set to something like 1, or "y", or something? Yes. Use set print=ask-yes # or ask-no Marius Gedminas -- When in danger, or in doubt, run in circles, scream and shout. -- Robert A. Heinlein
Re: Idea: saving vs. deleting
On Wed, May 24, 2000 at 03:36:09PM -0400, Rob Reid wrote: I like the idea, but d is already used to indicate messages with deleted attachments. How about s? That's too already used for PGP signed but unverified messages. Maybe `w' (written)? I only see (capital) S whether or not I've tried to verify the signature. Is this a new feature ( 0.95)? (I'm giving the alpha/beta version extra testing...otherwise known as waiting for 1.2.1.) Perhaps. I'm using 1.2. I don't remember when that was introduced. Marius Gedminas -- As easy as 3.14159265358979323846264338327950288419716
Clearing macros
While reading another thread here I playd a bit with this macro: macro editor ,t test Now I don't know how to get rid of it (short of restarting mutt). There's no `unmacro' command. Section 3.6 of the manual says nothing about undefining them. Marius Gedminas -- We have enough youth, how about a fountain of SMART?
Re: mutt process doesnt terminate
On Wed, May 17, 2000 at 10:12:26PM +0200, Jan Houtsma wrote: Now for the sequence of events that make mutt go wild: -- 1) In my own X-session my wife starts xterm (which is of course owned by my id) and in that xterm she runs the script 'caroline' and logs into her own account. 2) Then she starts mutt. 3) Then she kills the xterm with the cross button!! Now xterm window terminates but mutt keeps running like crazy (88% cpu)!!! Note that it's something related to mutt. When she starts pine or elm this does NOT happen at all and everything is nicely deleted from the process list. Another wierd thing It ONLY happens if i run the 'su - caroline' via that script! If i start an xterm and manually run the 'su - caroline' and then mutt, all IS nicely deleted after killingthe xterm. I've got a similair problem. Eterm + `su -' + Midnight Commander = 100% CPU usage after closing the window. But this only happens sometimes. Sometimes it dies in a couple of seconds, and sometimes it just runs until I kill it. (BTW, I run su - directly (`Eterm -e su -' is bound to a hot key in my IceWM keys file)). The only common thing between your problem and mine seems to be `su'. BTW, do you have mutt compiled with slang? That would be another common thing since that's what mc uses. Marius Gedminas -- Premature optimization is the root of all evil. -- D.E. Knuth
Re: How add to address for mailing list semi-automatically?
On Fri, May 12, 2000 at 08:03:37AM +0200, Frank Derichsweiler wrote: Now I would like to be abe to create a header with To: mailinglist-adress by pressing e.g. M. Which mailing list? Should it depend on the folder/current message/phase of the moon? You could, for example, define a macro in a folder-hook. I have already set the subscribe option, I can send reply to the list. But then I get the In-Reply-To: header field which is stupid for a new thread in the list ... If you set edit_headers, you can simply delete In-Reply-To: in your editor (that's what I sometimes use for new threads). Marius Gedminas -- A Law of Computer Programming: Make it possible for programmers to write in English and you will find that programmers cannot write in English.
Re: printf like sequences for folder_format
On Wed, May 10, 2000 at 11:33:43PM +0300, Mikko Hänninen wrote: [... Maildir is not for everyone ...] less convenient (no progress indication while opening and no line counting). Huh? I get progress indication when opening a Maildir folder, every 10 messages, which means the counter is basically just a blur... I mean there's no percent display. With mbox you see: `Reading foo... 310 (10%)', with Maildir you only see the number of messages. Psychologically this makes the waiting seem longer ;) No line counting? That's just a procmail rule away. Although if you don't want to use procmail (or maildrop), I guess it's a problem then. (Actually, I think all my saved mails have Lines: headers, so it's not an issue for me). Maildir sure has many of advantages (like faster saving when you delete a single message from the beginning of a folder, etc.), but they are not applicable to my situation. Mbox seems better for mail archives (faster, smaller, and they are mostly read-only anyway). Marius Gedminas -- "Linux: the operating system with a CLUE... Command Line User Environment". (seen in a posting in comp.software.testing)
Re: printf like sequences for folder_format
On Wed, May 10, 2000 at 06:50:56PM +0200, Christian Ordig wrote: You have a point here, and that could be a killer feature for mbox format ;-) right... but who really uses mbox when it comes to have more than 1000 or 2000 messages in a folder? I do. I've tried Maildirs and found them to be slower to open (about twice for the first time and only very slightly slower when everything is in the cache), bigger (a single folder with 3100 messages took 13 Mb in mbox, and 21 Mb in Maildir format -- on ext2 fs with 4K blocks), and less convenient (no progress indication while opening and no line counting). Perhaps I'll switch to Maildir when I'll start using Reiserfs (or NFS). Marius Gedminas -- Favourite MS-DOS error message: "Drive C: not ready, close door."
Re: viewing html emails in mutt
On Mon, May 08, 2000 at 12:03:15PM +0100, John P . Looney wrote: Has anyone noticed that sometimes, when viewing a mail with Lynx, through mutt, that they fight over the terminal ? Not with mutt, but lynx used to fight with Midnight Commander this way on my system. The reason was an extra `' in /usr/lib/mc/mc.ext that sent lynx into background unnecessarily (well, maybe that did work with gmc, but definitely not with mc in an xterm). I suggest you to check your mailcap for spurious ampersands. Marius Gedminas -- Hoping the problem magically goes away by ignoring it is the "microsoft approach to programming" and should never be allowed. -- Linus Torvalds
Re: Looking for a nice color scheme
On Sun, May 07, 2000 at 11:40:55PM -0500, Corey G. wrote: Mutt Group, I am looking for what I would call a nice color scheme for Mutt. I currently use the scheme listed below but I would like to see what other people have come up with. I tinkered around but I have very little artistic ability. Looks like it's time to create mutt.themes.org :-) Seriously, it would be nice to have a web page with screenshots of various color schemes and corresponding muttrc files. However my artistic abilities are limited by bugs in slang/ncurses that prevent my colors from appearing correctly. Marius Gedminas -- Remember, drive defensively! And of course, the best defense is a good offense!
Re: viewing html emails in mutt
On Sat, May 06, 2000 at 11:08:50AM -0400, Nollaig MacKenzie wrote: What would I do to get mutt to use w3m? I tried (mailcap) text/html; w3m %s text/html; w3m -T text/html -F %s text/html; w3m -T text/html -F -dump %s; copiousoutput Marius Gedminas -- The *REAL* Y2K is the year 2048.
[bug/misfeature?] Display problem? (was Re: Request for a volunteer)
Mutt's pager chokes on this message (Message-ID: [EMAIL PROTECTED]). All I see is [-- Error: unable to create PGP subprocess! --] -END PGP SIGNATURE- However everything is visible with view-attachment, and the text is available when replying. The relevant headers look like this: Content-Type: application/pgp; x-action=sign; format=text Content-Disposition: inline; filename="msg.pgp" I have no problems with other signed messages (which have more formally looking Content-Type: multipart/signed; micalg=pgp-md5; protocol="application/pgp-signature"; boundary="aX6oBa4COn3eIhlv" Content-Disposition: inline headers). BTW I do not have PGP/GPG installed at the moment, therefore I'm not surprised to see "unable to create PGP subprocess". I think Mutt should at least display the text. Marius Gedminas -- For Sale: Parachute. Only used once, never opened, small stain.
Re: OT? Proposal: Third sorting key
On Thu, Apr 06, 2000 at 03:50:42PM -0400, Michael Tatge wrote: On Wed, Apr 05, 2000 at 09:18:47PM +0200, Marius Gedminas wrote: Yes. I'd like to sort by To: field, then by thread, and then by date ^^ Do you mean Subject: field here? If not how do you want to sort mails by To: field and sorted in threads at the same time??? You can sort threads against other threads by the To: field of cause. Well, that's what I wanted to say. I was abusing "by thread" as a way of saying "perform the threading at this point". (and I also want to group messages with the same Message-Id together). ??? Have you ever received duplicate mails? When they appear in the middle of the thread, usually a whole subthread separates them. Anyway, it is unlikely that this problem could be solved with sorting, so it's kinda offtopic... (Of course, since I began using procmail, I just set it up to find duplicates and add a X-Note: header to them. color index red ... takes care of everything else ;) One of my secret wishes for 1.3 (it's too big for 1.2) is to see current sorting interface changed to a more consistent set sort=key1, key2, ... What should be the semantic of this $sort variable then? E.g. sorting without threading: set sort="to, reverse-date" Sorting and threading: set sort="to, date, thread, reverse-date" (everything to the left of "thread" shows sorting order for threads, and everything to the right -- for messages within threads). I think it's better to have three different sorting vars. Like $sort (how to sort mails), $sort_aux (secound sorting key) and $thread_sort(if $sort=threads, how to sort threads internally), for instance. $sort and $sort_aux could be merged into a new $sort= key1, key2 maybe. Maybe. $sort, $thread_sort and $show_threads (yes/no)? Marius Gedminas -- Microsoft -- because God hates us.
Attach-message + attachments
Hi, Suppose, Alice sent me a message with a large attachment. Suppose, that Bob sends me an email asking for Alice's message. I reply to it and attach-message Alice's message. However I do not want to send Bob full message with that huge attachment, I want just plain text. How do I do that? Playing with mime_forward_* was fruitless. I guess they really are only about forwarding with forward-message. Am I abusing attach-message? Should I just reply and then forward Alice's mail in a separate message? Or maybe make a copy of it in a temporary mailbox, go there, delete that big attachment and then forward/attach the resulting message? (Not forgetting to put Bob's email in the clipboard, and to delete the temporary mailbox afterwards.) Marius Gedminas -- MCSE == Must Consult Someone Experienced
[bug?] Limiting patterns and Boolean algebra
Once I thought about asking for a new command -- reverse-limit. Then I realized that it was not necessary -- one could just write a macro like this: limithome!(end)enter So far so good. However there's a strange bug/feature: foo limits to messages matching pattern foo !(foo)limits to messages not matching pattern foo !(!(foo)) limits to messages not matching pattern foo too! But (!(!(foo))) limits to messages matching pattern foo Now if I omit the parentheses, the behaviour is what I'd call `normal': foo limits to messages matching pattern foo !foo limits to messages not matching pattern foo !!foo limits to messages matching pattern foo But this won't work with foo == "~C x | ~C y", for example. (BTW one cannot reverse "really simple patterns" with this method -- $default_hook is then simply used to match for "~f !foo !~P | (~P ~C !foo)"...) Marius Gedminas -- If A equals success, then the formula is A = X + Y + Z. X is work. Y is play. Z is keep your mouth shut. -- Albert Einstein
Re: OT? Proposal: Third sorting key
(Hello, everyone, again. Somehow I got unsubscribed from this list -- maybe my subscription had expired? ;-) Or maybe just my ISP's mail server went down at the wrong moment.) On Wed, Apr 05, 2000 at 04:52:44PM -0400, Michael Tatge wrote: I just played around with the $sort_aux variable and I came to the opinion that a third sorting key is needed. It's a bit annoying that $sort_aux controls how threads are sorted against other threads _and_ the thread internal sorting. [...] What do you think about this? Yes. I'd like to sort by To: field, then by thread, and then by date (and I also want to group messages with the same Message-Id together). One of my secret wishes for 1.3 (it's too big for 1.2) is to see current sorting interface changed to a more consistent set sort=key1, key2, ... Marius Gedminas -- There is a 70% probability of tomorrow. (Actual weatherman quote, 1988)
Re: why is mutt better?
On Fri, Mar 10, 2000 at 10:48:01AM +0100, Ralf Hildebrandt wrote: On Fri, Mar 10, 2000 at 11:28:59AM +0200, Marius Gedminas wrote: Bat! does have a couple of features that Mutt lacks, but those are the features that I do not need. Maybe others do. What features does it have? It can be a "mail server" for the local network (I don't quite remembet what kind of server is that -- maybe it's just The Bat! server + a couple of The Bat! clients), it has some automatic form processing (whatever that means). And, of course, it has lots of things builtin, like mail filtering, HTML rendering, multiple POP3/IMAP mailboxes, direct sending via SMTP, randomized one-line signatures, builtin editor (that is actually useful), address book, etc. Mutt relies on external programs for most of these (and I don't have problems with that... except maybe that it is hard to get all Win32 versions of all required programs ;) Maybe there's a full list somewhere at www.ritlabs.com. Unfortunatelly, I cannot use Mutt under Windows NT: it dumps core in the pager. I've posted stack backtrace on mutt-dev, but nobody seems to can't you use an external pager? I can, but I don't want to. Marius Gedminas -- This is an object-oriented system. If we change anything, the users object.
Re: why is mutt better?
On Thu, Mar 09, 2000 at 06:39:45PM +, J McKitrick wrote: I just got in a debate over email clients, and my windows friend argues anything i can do in mutt, he can do in TheBat! just as easily. I checked the feature list, and it is extensive. Most of what mutt offers, thebat offers. Why is the advantage of mutt, or any text-based email client? I use The Bat! at work. It is good, but Mutt is much better (for me, at least). I miss hooks, macros, patterns, mailing list support, replying to multiple e-mails with one message, the limit command, etc. The Bat! does have a couple of features that Mutt lacks, but those are the features that I do not need. Maybe others do. Unfortunatelly, I cannot use Mutt under Windows NT: it dumps core in the pager. I've posted stack backtrace on mutt-dev, but nobody seems to have noticed it (and I don't blame the developers -- they are busy finishing Mutt 1.2 under Unix). Marius Gedminas -- Microsoft does have a Year 2000 problem. We're it.
Re: Mutt colours in xterm
On Wed, Mar 01, 2000 at 09:01:44AM +, Glyn Millington wrote: I spent a happy half-hour building mutt-1.7.1i last night,(with ncurses) fired it up in console mode - colours as usual - brilliant! Trying it this morning in an x-term I'm reduced to mono. Not a big problem but I like the colours and they do make reading easier. Is there anything I need to tweak either in the system Muttrc or in the building process that I am overlooking. You need to check if your $TERM variable is set to something that supports color (e.g. `xterm-color') or maybe set the $COLORTERM variable. I'm running Mutt on Mandrake Linux 6.1 (beautiful) - and my Mutt-1.0.1 rpm supplied with that system does produce colours in an xterm, which suggests, on reflection, that the trick is in the compiling, but where? AFAIR Mutt from Mandrake RPMs is compiled with slang, not ncurses. Unfortunatelly, all versions of S-Lang that I've seen have problems with Alt+key combinations (as witnessed in the Mutt's INSTALL file), so I'd suggest trying `export TERM=xterm-color'. Marius Gedminas -- MSDOS didn't get as bad as it is overnight -- it took over ten years of careful development. -- [EMAIL PROTECTED]
Re: Russian charsets and automatic recoding
On Wed, Mar 01, 2000 at 04:52:19PM +, Edmund GRIMLEY EVANS wrote: It's in "windows-1251". Presumably your mutt doesn't know about that charset. I think you can tell whether mutt knows about a charset by going to the compose menu and doing ":exec change-charset" on any text part. If the charset you specify is not known to mutt, mutt gives an explicit error. I'm not sure if there's an easier way of asking mutt whether it knows about a charset. My mutt does know about windows-1251. I'm using it for the body of this message, I think. Mine doesn't. Actually, it's not Mutt who doesn't know this charset, it's my /usr/share/i18n/charmaps/ colection. However :charset-hook windows-1251 cp1251 helps: (Saluton! Kiel vi fartas?) (I don't have a cyrillic font and ICONV (or whatever Mutt 1.1.x uses) is smart enough to translate everything into ASCII). Once again -- could `charset-hook' be renamed to `charset-alias' before 1.2 please? It's not a hook. Definetely. Not a hook at all. Not even close to that. Marius Gedminas -- First rule of public speaking. First, tell 'em what you're goin' to tell 'em; then tell 'em; then tell 'em what you've tole 'em.
Re: [fwd] Can not run mutt.exe (from: siddhivinayak.nirvaneshwar@tatainfotech.com)
On Thu, Feb 24, 2000 at 05:49:13PM +, Edmund GRIMLEY EVANS wrote: If there's demand for it, perhaps one of the people who has succeeded in building mutt on Windows could provide an unofficial executable. Or does the Cygwin version require a library that can't be redistributed? Sounds like a good idea. CygWin is GPL-ed, according to their FAQ. (Although I don't quite understand how they can sell CygWin 1.0 for $99 w/o any download options...) Today I somehow managed to build Mutt 1.1.4 w/ CygWin B20.1 and ncurses 5.0, however it isn't fully functional (it dumped core on me a couple of times). It would also be nice to have binaries for fetchmail (does it compile on CygWin?) and perhaps ssmtp. I don't have much time for playing with all this at work, though (and I don't use Windows at home). Marius Gedminas -- $ fortune $3,000,000
Re: home/end/pageup/pagedown don't work
On Tue, Feb 22, 2000 at 07:48:20PM +0100, Jens Wilhelm Wulf wrote: in Mutt, you can press : for "enter command", and then press esc, and then the key for which you want to see the char sequence. The reason why you need to press esc is that the first character from your home/pageup/whatever key is likely to be esc, and Mutt won't display that -- but if it receives two esc's, it will display the second on the command line. shitdidn't workwhatever I type before those keys, it's always the same: home/end won't display anything but stop the input, while pageup/down just make mutt beep without closing the prompt. Tried to hack around in the code a bit (mutt_getch in curs_libs.c) but now I know as much as before...however, I'll give up for now, there are much more important things to do. Tried slang version 1.3.10 instead of 1.4, too. You should try hacking terminfo instead. Use $ infocmp $TERM and look at khome, kend, kpp and knp. They should match the sequences your term is generating for Home, End, PgUp and PgDn respectively. For example, Eterm 0.9 on my system is configured to send this: $ keystrings ^D exits \e[1~ \e[4~ \e[5~ \e[6~ \4 (keystrings is a small program whose sole purpose in life is to display nicely formatted character sequences. I hacked it from examples in libc.inf). If they don't match, you can redirect the output of infocmp to a file, modify it and reinstall with `tic'. Good luck. Marius Gedminas -- "Nobody will ever need more than 640k RAM!" -- Bill Gates, 1981 "Windows 95 needs at least 8 MB RAM." -- Bill Gates, 1996 "Nobody will ever need Windows 95." -- logical conclusion
Attaching an attachment
Hello, In the spirit of recent queries in this list: can Mutt attach an attachment from another message? I guess this question needs some elaboration: Suppose, Alice sends me a message with an attachment X. I save it in my mailbox. A week later Bob sends me another message. I reply to it and I want to attach X in the reply. I have tried attach-message, but it only allows me to include whole Alice's message. view-attachments does not work inside attach-message, so I cannot save X to a file and latter attach it. I must start another copy of mutt, find Alice's message, save its attachment to a file and return to the first mutt to attach it (or play with postponing the first message, navigating, saving to a file, etc.). Is there a simpler way? Would it be hard to enable view-attachments in attach-message mode? And maybe even allowing tagging some of the attachments there? Marius Gedminas -- Any sufficiently advanced magic is indistinguishable from technology.
Re: Adding fields to (MH) messages
On Sat, Feb 12, 2000 at 08:57:49PM +0100, Jos Vos wrote: When I use Mutt for MH folders, it adds Status and Content-Length fields to each individual message and thus changes all files, even whena you have only opened the folder and not touched any messages. When you have MH folders with thousands of messsages (4000+), this is not really a desirable behaviour :-(. Is there an option with which I can prevent Mutt adding those fields to messages, _at least_ when the message is not read or whatever? If not, do you also consider this a useful feature? You can always open the folder in read-only mode or just disable writing by hitting `%' before quitting. Marius Gedminas -- Laugh at your problems; everybody else does.
Re: mail does not reach certain hosts
On Thu, Feb 10, 2000 at 03:11:36AM -0500, Michael Tatge wrote: Hi, I have a major problem with Mutt 1.0i. Mail to some hosts does not reach its destination without me getting an error message. I use sendmail as MTA and it seems that sendmail is correctly configured. When I use netscape or kmail this does not happen though I also use sendmail as MTA then. The thing I can't understand why it's only some hosts that do not receive my mail, while with other it's on problem! It could be caused by a wrong envelope `From '. If it is set incorrectly, your error messages from remote SMTP servers will go to the wrong address. You can try adding this to your .muttrc: set sendmail="/usr/sbin/sendmail -oem -oi [EMAIL PROTECTED]" However fixing this in sendmail configuration is the Right (and hard) Way to do it. I've heard that Return-Path: header has something to do with all this, but I don't know for sure. Marius Gedminas -- $ fortune $3,000,000
Charset-hook?
I suggest renaming `charset-hook' to `charset-alias'. It's not really a `hook' after all. Marius Gedminas -- An algorithm must be seen to be believed. -- D.E. Knuth
Re: save-hook regexp two TLDs
On Thu, Feb 10, 2000 at 09:48:08AM -0600, freix wrote: issue: One of the mailing lists I'm on has two domains merit.edu nanog.org i have save-hook "~C nanog@(nanog|merit)\.(edu|org)" +nanog as the epxression, and when i start mutt it says: Error in /home/freix/.muttrc, line 245: Unmatched ( or \( source: errors in /home/freix/.muttrc This is very similar to how i have all my $alternates setup, but this one doesn't seem to like it. Again, same as last issue, if I do it on two seperate lines it works no prob, but I'm sure there's a more elegant way of doing it. Anyone feel free to smack me with the hammer of enlightenment. I see two problems with this hook: 1) it also accepts merit.org and nanog.edu 2) it gets parsed this way: "(~C ...) | (...) | (...)" I'u use save-hook "~C 'nanog@(nanog\.org|merit\.edu)'" +nanog # ^ ^ (BTW it's easy to get lost in various ways of quoting. I'm not sure if `\.' shouldn't be replaced to `\\.' in this case. A couple of my save-hooks look like this: save-hook '~C "\\foo@bar"' +foo I've found the need for double backslashes by trial and error. Perhaps `[.]' would be a better and more universal solution?) Marius Gedminas -- I doubt, therefore I might be.
1.1.3 notes
While browsing through diff -u manual.sgml-1.1.2 manual.sgml-1.1.3 I noticed this suspicious line deletion. Looks like it's accidental. @@ -3416,7 +3429,6 @@ p Specifies the password for your IMAP account. If unset, Mutt will -prompt you for your password when you invoke the fetch-mail function. bfWarning/bf: you should only use this option when you are on a fairly secure machine, because the superuser can read your muttrc even if you are the only one who can read the file. There's a typo: s/dfault_hook/default/hook/ in section `Matching Messages with Hooks'. Some tildes are missing in manual.txt (e.g. 6.3.3. alias_file Type: path Default: " /.muttrc" ). I suppose `~' characters in init.h should be translated to `tilde;'. I think changing sgml_fputc in makedoc.c would do the trick (I haven't tried it -- Mandrake 7.0 does not have sgml-tools for some reason): --- makedoc.c.orig Tue Feb 1 20:35:33 2000 +++ makedoc.c Thu Feb 10 21:03:53 2000 @@ -589,6 +589,7 @@ case '"': return fputs ("dquot;", out); case '[': return fputs ("lsqb;", out); case ']': return fputs ("rsqb;", out); +case '~': return fputs ("tilde;", out); default: return fputc (c, out); } } Speaking about makedoc.c -- I get these warnings when compiling mutt: make[3]: Entering directory `/opt/mutt-1.1.3' gcc -DSHAREDIR=\"/usr/local/share/mutt\" -DSYSCONFDIR=\"/usr/local/etc\" -DBINDI R=\"/usr/local/bin\" -DHAVE_CONFIG_H=1 -I. -I/usr/local/include -I/usr/include /ncurses -I./intl -Wall -pedantic -g -O2 -c makedoc.c makedoc.c: In function `main': makedoc.c:112: warning: assignment discards qualifiers from pointer target type makedoc.c:114: warning: passing arg 2 of `getopt' from incompatible pointer type makedoc.c: In function `print_confline': makedoc.c:675: warning: enumeration value `F_NONE' not handled in switch makedoc.c: In function `print_it': makedoc.c:960: warning: enumeration value `F_NONE' not handled in switch I would be nice to get them removed before releasing 1.2, wouldn't it? There are too many blank lines in manual.txt before some of the tables (e.g. 6.4.2, 6.4.3). It's probably a bug/feature of SGML to text converter. Can anyone shed some light on this issue? Marius Gedminas -- Never assume the reader has read the subject line.
Re: Different mail accounts
On Sun, Feb 06, 2000 at 05:09:50PM +0200, Mikko Hänninen wrote: [...] Mutt 1.0 doesn't support as an explicit feature setting the envelope sender; however you can specify the $sendmail variable to whatever you want, so you can set it like that ("set sendmail=sendmail -fsender@address"). Again, in the developement versions there's a variable, I forget which, that sets the envelope sender to your current From: address (IIRC). Once there was a patch for this in mutt-(.*)@mutt\.org, but I see no traces of it in 1.1.2. The variable was called `envelope_from'. BTW grepping for /envelope in the manual reveals one incorrect use of the word. AFAIK envelope and headers are quite different parts of an e-mail, right?. Here's a patch for this: --- manual.sgml.head.orig Fri Jan 28 21:42:29 2000 +++ manual.sgml.headSun Feb 6 22:10:35 2000 @@ -1717,7 +1717,7 @@ language for matching messages in hook commands. This works in exactly the same way as it would when em/limiting/ or em/searching/ the mailbox, except that you are restricted to those operators which match information -from the envelope of the message (i.e. from, to, cc, date, subject, etc.). +from the headers of the message (i.e. from, to, cc, date, subject, etc.). For example, if you wanted to set your return address based upon sending mail to a specific address, you could do something like: Marius Gedminas -- If you are good, you will be assigned all the work. If you are real good, you will get out of it.
Re: Setting From when replying
On Fri, Feb 04, 2000 at 01:07:06AM +0100, Terje Elde wrote: New challenge: My box is set with my email addr as [EMAIL PROTECTED], because that's what I want when I just get lazy and pipe things into mail, and when I reply to news postings using slrn and such. However, for mutt I want the default email to be [EMAIL PROTECTED], which I've solved by setting my_hdr From: Terje Elde [EMAIL PROTECTED] That's not normally a problem, but what happens when I want that for my default email, but I want [EMAIL PROTECTED] when I reply to emails sent to that addr? Normally my_hdr overrides any good done by reverse_name :( Since you're using 1.1.x, you can use `set from' instead of `my_hdr From:'. AFAIK $from was introduced in order to solve this problem (my_hdr overriding reverse_name). Marius Gedminas -- If you are angry with someone, you should walk a mile in their shoes... then you'll be a mile away from them, and you'll have their shoes.
Re: a few questions
On Sun, Jan 30, 2000 at 11:37:43AM -0800, Joshua Haberman wrote: I guess the gist of my first question is this: I upgraded from 0.85 or so (I wish I could remember the exact version) to 1.0.1i and the colors no longer work. I made no other changes to the system configuration in the mean time (same terminal program, same .muttrc, etc...) It linked with ncurses during the make, so I can't figure out why the colors don't work. What version of ncurses do you use? I had problems with pre-4.0 versions of ncurses; 4.2 and currently 5.0 seem to work fine (except for one little problem which I'm now trying to solve in ncurses mailing list). I've just installed Mandrake 7.0 and noticed that mutt is linked with slang there. Perhaps you should try recompiling mutt with slang? Note, that if you use stock version of slang, Alt+key will not work. See patch.slang-1.2.2.keypad.1 in contrib/ directory. (Mandrake comes with slang 1.3.8 and this problem is not yet fixed! Do slang maintainers know about this?) [... snip ...] Just out of curiosity, do you generally have mail from different mailing lists filtered into different folders? It seems like it'd be much easier to manage that way... It seems most of people prefer it that way. Personally I use a couple of mutt macros that give me "virtual folders" by using the (l)imit command. So I can see all my incoming mail in one place (when there are just a couple of messages) or just hit f3...f11 and look at messages from different mailing lists (f2 is "no limits", f12 is for personal messages). offtopic Sometimes I whish mutt would accept limit patterns which have no matching messages... After all, if you delete all visible messages and hit $, your limit patter does not magically go away. Also, sometimes I wish there were a command for inverting currently active pattern... /offtopic Marius Gedminas -- Iki, If you are good, you will be assigned all the work. If you are real good, you will get out of it.
Re: mutt-1.1X
On Sun, Jan 30, 2000 at 03:02:27PM -0500, Brian wrote: I have recently made the switch from mutt-1.0 to mutt-1.1x and have come across a few "issues" with colors. I am unable to make the default text colour be a dark grey. I would much prefer it to be white. Any suggestions? So do you want it to be dark gray (color normal brightblack ...) or white? (color normal brightwhite ...)? Also, there might be color problems depending on your version of ncurses/slang. Feel free to point me towards a colour sampling for 1.1x #ifdef SHAMELESS_PLUG You can look at http://www-public.osf.lt/~mgedmin/mutt/. Everything is in Lithuanian there, but you could look at a couple of screenshots and see my color definitions in a copy of my .muttrc. BTW colors in .muttrc do not match those in the screenshots ;) I've replaced all `blue' backgrounds to `default' to see Eterm's background bitmaps. Also, slang 1.3.8 was unable to display my colors properly, so I recompiled mutt with ncurses 5.0, but haven't (yet) updated the screenshots. Perhaps I'll make an English version of those pages. So far I'm using them to lure local Linux newbies to mutt :-) #endif Good luck, Marius Gedminas -- When I'm not in my right mind, my left mind gets a little crowded.
Re: line wrap
On Sat, Jan 29, 2000 at 02:31:20PM -0500, David T-G wrote: ...and then Jim Breton said... % Do people generally find it annoying if I don't use line wrap in my editor Actually, yes. It's only because everyone here is so polite that you haven't been flamed for it already, but it's a pretty big peeve. Oh, wait; this is the mutt-users list. I guess folks have just been too busy ;-) There's one more reason: mutt and vim (and I think this is the most common setup for people on this list; substitute vim with any other decent editor) can handle long lines nicely. Both wrap them when displaying and vim allows to easily and correctly wrap quoted text. If people used less advanced mailers, Jim would already be flamed to death. I still remember the times when I used a DOS-based mailer (Demos Mail for UUCP) and how I felt about scrolling left and right single column at a time (and that means hundreds of keypresses at 30 cps autorepeat rate) when someone composed a message using Microsoft LookOut. Marius Gedminas -- Everything you read in newspapers is absolutely true, except for that rare story of which you happen to have first-hand knowledge. -- Erwin Knoll
Re: Solved: aterm mutt problem with home/end keys
On Fri, Jan 28, 2000 at 09:40:06AM -0600, [EMAIL PROTECTED] wrote: Awhile ago I wrote about having a problem running the email program Mutt in an aterm and not getting it to obey the Home/End keys, even though everything else seemed to work (PageUp/PageDown, Delete) in Mutt, and also these keys worked everywhere else (notably, in tcsh). I finally traced down the problem to this: snip Cool, man. That gave an idea - I'm going to add Troubleshooting page ( http://aterm.sourceforge.net/troubleshooting.php ) to aterm's site and I'll be posting things like that on this page (if you don't mind). Please, be sure to send me other workaround that you will find for aterm. I just thought I should mention I had the same problem (among numerous others) with Eterm. And I could have used the same solution, but instead I modified terminfo/termcap entries for xterm. I just switched strings for khome (kh in termcap) and kfnd (@0), and for kend (@7) and kslt (*6). Perhaps Mutt should recognize KEY_FIND and KEY_SELECT as synonyms for KEY_HOME and KEY_END? Marius Gedminas -- If it weren't for the last minute, nothing would get done.
Regexp dangers (was Re: regexp help?)
On Wed, Jan 26, 2000 at 02:13:23PM -0600, David DeSimone wrote: By the way, something to watch out for in .mutrrc's: set reply_regexp = "^(re|aw):[ \t]*" This doesn't do what you think it does; Mutt sees the "\t" as simply "t", because backslashes are parsed within double-quotes. So the regexp comes out as "^(re|aw):[ t]*", and so a subject like "Re: Tuesday" comes out with a real subject of "uesday". :) I always wondered, how many backslashes I need in such places ;). However my ~/.muttrc contains this: set reply_regexp="^(re|aw)(\\[[0-9]+\\])?:[ \t]*" # TheBat! uses Re[%d]: and when I query Mutt (:set ?reply_regexp), I see reply_regexp="^(re|aw)(\[[0-9]+\])?:[ .]*" I suppose that `.' in brackets is the actual tab character, so your warning is not necessary in this case. OTOH things like send-hook '~C "\\foo[^ ]*@"' ... or color index red default '~h "X-Note: Message-ID seen before recently\."' are a different issue. I'm not quite sure how many `\'s should there be before that `.' and I'm a little too lazy to experiment to find it out. :-) (I'm sure a single \ there is not enough, but even an unescaped dot would work for me there, and I don't expect many false matches.) Marius Gedminas -- F U cn rd dis U mst uz Unix.
Re: Colour loss when running via an exec command...
On Thu, Jan 20, 2000 at 02:11:40PM -0600, Jamie Novak wrote: [...] Eterm --trans --shade "40%" -T Mutt -n "Electronic Mail" -e mutt [...] What happens, though, is that mutt defaults to a monochrome colour scheme when I open it via an exec like that. (It does the same for rxvt terms, etc., as well.) Does anyone know how to get around this? It's rather annoying. Any suggestions would be greatly appreciated. Edit your Eterm theme and in the section misc add or modify term_name to say `xterm-color'. (BTW I use (slightly modified) mutt theme that came with Eterm instead of specifying all arguments every time in every menu) Marius Gedminas -- "If it ain't broke, don't fix it." - Bert Lantz
sendmail -U
Browsing through sendmail(8) man page I stumbled upon this option: -U Initial (user) submission. This should always be set when called from a user agent such as Mail or exmh and never be set when called by a network delivery agent such as rmail. What does it do and why isn't it used in the default $sendmail value? Just curious. Marius Gedminas -- If two wrongs don't make a right, try three wrongs.
Re: Charset in mutt 1.0i
On Fri, Jan 21, 2000 at 01:02:42PM +0100, Dirk Pirschel wrote: On Fri, 21 Jan 2000, Ryszard Lach wrote: Hi! I have a problem with "charset" variable in mutt. Mutt seems to ignore this variable (I've set it to iso-8859-2), if $LANG environment variable isn't set to The charset variable only affects the Content-Type header of _sent_ mails. (AFAIK) No, that's what send_charset does that in unstable versions. "pl_PL", and I'd like to keep it set to "C". Do you have any suggestions ? I use LANG=en_us.ISO-8859-1 as workaround to display the german "o "u "a. Maybe you can use en_US-ISO-8859-2. I don't use LANG at all (so it defaults to "C", IIUC) but I set my LC_CTYPE to lt_LT.iso-8859-13 and everything works perfectly. BTW I also have charset set to "iso-8859-13" in .muttrc (and I have a couple of macros to change it an load the appropriate font if I encounter an occasional cyrillic mail). AFAIU Mutt does not use locale information for determining charset at all, it just uses it to determine which characters are printable (otherwise it displays them as quotation marks). Marius Gedminas -- Given enough eyeballs all bugs are shallow. -- Eric S. Raymonds, "The Cathedral and the Bazaar"
Re: problem with bounce
On Fri, Jan 21, 2000 at 03:19:01PM -0600, David DeSimone wrote: I imagine the problem is in the sendmail config, something like this: Mlocal, P=/usr/bin/rmail, F=lsDFMAw5:/|@m, S=10/30, R=20/40, T=DNS/RFC822/X-Unix, A=rmail -d $u This is my particular setting for the "local" MDA, in my sendmail.cf. If I remember my sendmail configuration rightly, it is the "F" option in the "F=" section which determines that a "From " header will be added to the message before it is piped into the MDA. Now, for your case, you probably have P=/usr/bin/procmail, and since procmail creates its own message separator, having sendmail add it is superfluous; in that case, removing the "F" might be the way to handle this. My /etc/sendmail.cf reads: Mlocal, P=/usr/bin/procmail, F=lsDFMAw5:/|@qShP, S=10/30,R=20/40, T=DNS/RFC822/X-Unix, A=procmail -a $h -d $u but everything works correctly, despite that `F' in `F='. I've never seen any bogus From lines. Marius Gedminas -- "Only wimps use tape backup: _real_ men just upload their important stuff on ftp, and let the rest of the world mirror it ;)" (Linus Torvalds, about his failing hard drive on linux.cs.helsinki.fi)
Mutt + colors
I have `color status black cyan' in my ~/.muttrc. Stepping through mutt ensures that ColorDefs[MT_COLOR_STATUS] is 0x0A00 where color-pair 10 was initialized with init_pair(10, 0, 6). This should produce normal black on cyan, however ncurses (version 4.2) for some strange reason outputs \e[0;10;1m\e[30m;\e[46m (reset everything, set bold, set foreground to black and background to cyan), making it look like `brightblack cyan'. Is this a bug in mutt (e.g. not specificallty resetting A_BOLD which was set by a previous attrset?), or a bug in ncurses (or maybe in terminfo?)? Note that the preceding sequences are \e[0;10m\e[?25l\e[H (reset everything, make cursor invisible, move cursor to 1,1), so bold is turned on purposedly just for displaying help line. Note that this affects both help and status line in the index, and only help in the pager. Status line in the pager displays normal black on cyan. This occurs both on the linux console and Eterm. Other colors seem to work as expected (with ncurses; there's a wholy different story with slang...). Marius Gedminas -- Favourite MAC error message: "Not enough memory to eject disk!"
Mutt with PDCurses?
Can Mutt be compiled with pdcurses in place of ncurses? Yours, Marius Gedminas -- For those who like this sort of thing, this is the sort of thing they like. -- Abraham Lincoln
Re: Various unimportant nitpicks
On Sat, Jan 15, 2000 at 09:04:42PM +0100, Byrial Jensen wrote: On Sat, Jan 15, 2000 at 18:54:59 +0200, Marius Gedminas wrote: There is no `help' function in editor context (BTW `editor' is a very misleading name. Until today I thought it meant that mutt had a very basic (maybe like ed(1)) internal editor for composing messages :-) It has. Use "set editor=builtin" to use it. This doesn't work in 1.1.2i. After :set editor=builtin trying to edit a message results in an error from the shell builtin: /tmp/mutt-mg-6883-75: not a shell builtin BTW I also tried :set editor="" and mutt segfaulted: Core was generated by `mutt'. Program terminated with signal 11, Segmentation fault. Reading symbols from /usr/lib/libslang.so.1...done. Reading symbols from /lib/libm.so.6...done. Reading symbols from /lib/libc.so.6...done. Reading symbols from /lib/ld-linux.so.2...done. Reading symbols from /lib/libnss_files.so.2...done. #0 0x400d6a93 in strchr () from /lib/libc.so.6 (gdb) bt #0 0x400d6a93 in strchr () from /lib/libc.so.6 #1 0x400 in ?? () #2 0x808e73a in mutt_expand_file_fmt (dest=0xbfffe69c "\220\226\r\bçÿ¿ÿ", destlen=1024, fmt=0x0, src=0xbfffec94 "/tmp/mutt-mg-6883-114") at muttlib.c:584 #3 0x8057545 in mutt_edit_file (editor=0x0, data=0xbfffec94 "/tmp/mutt-mg-6883-114") at curs_lib.c:133 #4 0x808f3da in edit_one_message (ctx=0x80ba440, cur=0x80efcd8) at editmsg.c:91 F2#5 0x808f872 in mutt_edit_message (ctx=0x80ba440, hdr=0x80efcd8) at editmsg.c:193 #6 0x805bb39 in mutt_index_menu () at curs_main.c:1568 #7 0x806ac46 in main (argc=1, argv=0xbb94) at main.c:724 Marius Gedminas -- Hacking graphics in X is like finding sqrt(pi) with roman numerals. -- man xdaliclock
Re: installing mutt
On Sun, Jan 09, 2000 at 10:05:26AM -0800, --A. Santoyo-- wrote: hello I'm having trouble getting mutt to run. when I run the mutt command at the prompt I get an error that says /var/spool/mail/root is not a mail folder. what can I do about this. Any other tips to get this app going would be very usefull. This probably means that the file is not present (i.e. not yet created). You should probably send some mail to the relevant user (`root' in this; however I support the advice not to do anything as root when possible). Marius Gedminas -- To err is human, but to really foul things up requires a computer.
Re: Compile error
[ long lines in quotations wrapped to 72 columns (hint hint) ] On Fri, Dec 24, 1999 at 10:40:11PM -0600, Bram Shirani wrote: Howdy all... I've checked the FAQ pages and a few friends of mine to see if they've gotten the following compile error, but to no avail. Just wondering if anyone here has had the same trouble, or could point me in the right direction to fixing it. Thanks. I get this error while trying to compile mutt-1.0i on a Slackware 7.0 machine with glibc2.1, and ncurses installed: DHAVE_CONFIG_H=1 -I. -I/usr/local/include -I/usr/include/ncurses -I./intl -DH VE_CONFIG_H=1 -Wall -pedantic -g -O2 -c help.c gcc: Internal compiler error: program cpp got fatal signal 11 Have you tried compiling it a couple more times? Do you always get this error in the same location? Are there similar symptoms when compiling other, especially large, projects? If the answers to these questions are yes, no, and yes, you might want to check the GCC-SIG11-FAQ at http://www.bitwizard.nl/sig11/ (or at `locate GCC-SIG11-FAQ`, which under Red Hat gives me /usr/doc/FAQ/txt/GCC-SIG11-FAQ). Merry Christmas! Marius Gedminas -- What goes up, must come down. Ask any system administrator.
Re: Mutt and Lynx squabbling in a window.
On Tue, Dec 14, 1999 at 01:42:00PM +, Telsa Gwynne wrote: On Tue, Dec 14, 1999 at 12:56:31PM + or thereabouts, John P . Looney wrote: On Tue, Dec 14, 1999 at 11:37:07AM +, Telsa Gwynne mentioned: or xterms. Sorry. However, I can imagine the effects as I once managed to get both lynx and mutt apparently attempting to run in the same gnome-terminal. I think I may still have a screenshot of the resulting I've done this loads of times. It's really wierd. I think I tryed to quit out of lynx, but hit the wrong key, and then, depending on how many times I pressed Ctrl-L, it looked like lynx or mutt. Must try track down what's happening the next time I get a HTML mail.. [...] I think I either: messed up with urlview shelled out of lynx into mutt shelled out of mutt into lynx When I try to run lynx from inside Midnight Commander (the tty version, not gmc) from inside Gnome Terminal, I often get this weird behavior. If I just type lynx foobar from the command line it works nicely, however if I hit Enter on a .html file (and thus run... hey, why does it say `(lynx %f )' in the mc.ext file?), both programs try to update the screen simultaneously and both struggle for keystrokes. The ampersand sure looks suspicious. (Note, that both lynx and mc are compiled with slang. But vim uses ncurses, so perhaps that's not relevant.) I also tried the same trick in plain console, and lynx was then properly `stopped (tty input)'. Mutt seems to be innocent. Its more likely a problem in xterm (not suspending background tasks). Best Regards, Marius Gedminas -- MCSE == Must Consult Someone Experienced
Re: Multiple POP accounts and personalities...
On Mon, Dec 06, 1999 at 01:46:27PM -0600, David DeSimone wrote: Ronny Haryanto [EMAIL PROTECTED] wrote: Some people doesn't have permanent internet connection, and it even costs them by the minute. Therefore selective downloading could make sense. Certainly. I have to dial up to the Internet myself. Fetchmail is configured to only poll POP servers when I am actively connected, and while I don't have it configured that way, it can be told to skip messages that are beyond a certain size. Thus, the desired behavior can be easily automated with fetchmail. Not quite. Fetchmail cannot download only headers of oversized messages, nor can it delete them. I do not know a way to achieve this in Linux short of telnetting to port 110. (Or, for that matter, going to work and using a Windows based mailer... Offtopic: hasn't anyone succeeded compiling a recent version of Mutt under CygWin?) That's really the main reason that Mutt's POP3 support is so lame: Because fetchmail does it better, so there's no point in doing all the work to improve Mutt's support. Fetchmail is not interactive. Mutt could use this advantage over fetchmail. If it doesn't, I do not see reasons for POP3 client in Mutt at all. Perhaps this job would be perfect for a different standalone tool, however Mutt already does something like this with IMAP folders. (I do understand that IMAP would be better, but not every mail provider supports it. Most of them don't even support APOP...) Best Regards, Marius Gedminas -- First rule of public speaking. First, tell 'em what you're goin' to tell 'em; then tell 'em; then tell 'em what you've tole 'em.
Re: index colors
On Thu, Dec 02, 1999 at 04:05:49PM -0600, Jeremy Blosser wrote: Subject: this is the subject ^^^ this is highlighted This is yet another concept from the ones discussed above. Just so we keep track of what's being talked about, there have been 3 color issues mentioned so far: 1) Q: Can Mutt color different parts of the same line different colors in the message index? A: No. 2) Q: Can Mutt color different headers different colors in the pager? A: Yes. and now we have: 3) Q: Can Mutt color different parts of the same header line different colors in the pager? A: No. Let's add another one: 4) Q: Can Mutt color different parts of the header like it can do with the body (color body foo bar regexp)? A: No. (3) is very similar to (4) and could be accomplished like this: color header color2 ... ^Subject:.*$ color header-part color1 ... ^Subject: color header-part color3 ... my@email\.address # to show it's more # powerful ;) Would result in this: Subject: foo bar [EMAIL PROTECTED] baz color1 color3 Everything else is in color2. It would be pretty easy to implement. I think. A patch is attached. Well, maybe not that easy. The problem is that MT_COLOR_HEADER overloads syntax[0].color for its default color. And continuation lines use syntax[0].first/last for their own purposes. I added a new field to the syntax_t structure (int def_color) to overcome the first problem. And I think the second is not really a problem because continuation lines are already successfully used with syntax chunks in the body. However I'm not sure if these two lines are needed for anything else besides providing default color for MT_COLOR_HEADER: pager.c line 306 (after applying my patch), function append_line: (lineInfo[n+1].syntax)[0].color = (lineInfo[n].syntax)[0].color; pager.c line 713, function resolve_types: (lineInfo[n].syntax)[0].color = (lineInfo[n-1].syntax)[0].color; A patch is included (against vanilla mutt-1.1.1i). This is my first try to hack mutt, so I'd appreciate comments. I should have probably subscribed to mutt-dev before sending patches, but this one was not planned. After writing "it would be pretty easy to implement", I decided to check if this was really so ;) Marius Gedminas -- I doubt, therefore I might be. diff -urN mutt-1.1.1.orig/color.c mutt-1.1.1/color.c --- mutt-1.1.1.orig/color.c Wed Jul 7 00:40:25 1999 +++ mutt-1.1.1/color.c Fri Dec 3 12:47:47 1999 @@ -29,6 +29,7 @@ int ColorQuoteUsed; int ColorDefs[MT_COLOR_MAX]; COLOR_LINE *ColorHdrList = NULL; +COLOR_LINE *ColorHdrPartList = NULL; COLOR_LINE *ColorBodyList = NULL; COLOR_LINE *ColorIndexList = NULL; @@ -82,6 +83,7 @@ { "tilde", MT_COLOR_TILDE }, { "markers", MT_COLOR_MARKERS }, { "header", MT_COLOR_HEADER }, + { "hdrpart", MT_COLOR_HDRPART }, { "body",MT_COLOR_BODY }, { "message", MT_COLOR_MESSAGE }, { "attachment", MT_COLOR_ATTACHMENT }, @@ -679,7 +681,8 @@ /* extract a regular expression if needed */ - if (object == MT_COLOR_HEADER || object == MT_COLOR_BODY || object == MT_COLOR_INDEX) + if (object == MT_COLOR_HEADER || object == MT_COLOR_HDRPART || + object == MT_COLOR_BODY || object == MT_COLOR_INDEX) { if (!MoreArgs (s)) { @@ -714,6 +717,8 @@ if (object == MT_COLOR_HEADER) r = add_pattern (ColorHdrList, buf-data, 0, fg, bg, attr, err,0); + else if (object == MT_COLOR_HDRPART) +r = add_pattern (ColorHdrPartList, buf-data, 1, fg, bg, attr, err, 0); else if (object == MT_COLOR_BODY) r = add_pattern (ColorBodyList, buf-data, 1, fg, bg, attr, err, 0); else if (object == MT_COLOR_INDEX) diff -urN mutt-1.1.1.orig/doc/manual.sgml.head mutt-1.1.1/doc/manual.sgml.head --- mutt-1.1.1.orig/doc/manual.sgml.headThu Sep 23 23:01:43 1999 +++ mutt-1.1.1/doc/manual.sgml.head Fri Dec 3 12:47:37 1999 @@ -949,8 +949,9 @@ itembody (match em/regexp/ in the body of messages) itembold (hiliting bold patterns in the body of messages) itemerror (error messages printed by Mutt) -itemheader (match em/regexp/ in the message header) +itemheader (match em/regexp/ in the message header; highlight whole line) itemhdrdefault (default color of the message header in the pager) +itemhdrpart (match em/regexp/ in the message header) itemindex (match em/pattern/ in the message index) itemindicator (arrow or bar used to indicate the current item in a menu) itemmarkers (the ``+'' markers at the beginning of wrapped lines in the pager) diff -urN mutt-1.1.1.orig/doc/muttrc.man.head mutt-1.1.1/doc/muttrc.man.head --- mutt-1.1.1.orig/doc/muttrc.man.head Wed Aug 18 08:54:13 1999 +++ mutt-1.1.
send_charset does not work for Subject line
Scenario (Mutt-1.1.1i): set charset=iso-8859-13 set send_charset=iso-8859-4 I send a letter to myself and include some iso-8859-13 characters both in the body and in the Subject line. I get these headers: Subject: =?iso-8859-13?Q?skipped?= Content-Type: text/plain; charset=iso-8859-4 Now the bug is this: the characters in skipped *are* translated to iso-8859-4, but Mutt says they're in iso-8859-13. Best Regards, Marius Gedminas -- In order to get a loan you must first prove you don't need it. P.S. I just switched to Mutt-1.1.1 and, of course, forgot that `lists' command changed semantics... So my other mail about an error in the manual (which incorrectly states to send the word `subscribe' in the subject instead of the body to mutt-*[EMAIL PROTECTED]) went straight to the moderator for approval.
An error in the manual
The manual says: To subscribe to one of the following mailing lists, send a message with the word subscribe in the subject to [EMAIL PROTECTED] ^^^ Majordomo says: Commands must be in message BODY, not in HEADER. Commands in the "Subject:" line are NOT processed. Best Regards, Marius Gedminas -- Those who don't understand Linux are doomed to reinvent it, poorly.