Re: [courier-users] Re: Message submission process (Re: Rejection/re-insertion methodology howto?)
On Fri, 2004-02-13 at 03:00, Sam Varshavchik wrote: [...] Well, it's not as easy as inserting something into something else. The stuff in question is a hierarchical tree, that contains offsets, field contents, etc Hello Sam Yes, I thought it wouldn't be like in Perl ;-) But (see quote) it seems to be possible. Does submit flush the structure from memory to the datafile before it passes it on to the next step in the chain (maildrop already?) ? Could you point me at the place where the rewriting mentioned below is done, so I could give it a try over the weekend? krgds /markus --- BEGIN QUOTE --- This structure is needed in the even the submit process feels it necessary to rewrite the message. A message may be rewritten for several reasons. The most common one would be to provide default values for MIME headers. If you have a system script that runs a jobs and mails its output, and your system uses the iso-8859-1 character set natively, then chances are that your system script will not bother to supply the MIME header that specifies that the messages uses the iso-8859-1 character set. Well, submit will do that for you, in this case. --- END QUOTE --- --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id56alloc_id438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Re: IMAP/POP-SSL and Comodo Certificates
On Thu, 12 Feb 2004 19:24:47 -0500 Sam Varshavchik [EMAIL PROTECTED] wrote: Stefan Hornburg writes: Hello, I would like to know how to use Comodo SSL certificates with Courier POP/IMAP etc. Setting up Apache to recognize this is a no-brainer ... For Courier I did the following steps: - put the Comodo root certificates into /usr/lib/courier/rootcerts alongside with the shipped one - run c_rehash on this directory - put the certificate and the private key for the mail domain into /etc/courier/imapd.pem - restart courier-imap-ssl Try converting Comodo's cert to PEM, and appending it to imapd.pem Well, that works like a charm. No conversion needed, just append to imapd.pem does the trick. Thanks, Sam Racke -- LinuXia Systems = http://www.linuxia.de/ Expert Interchange Consulting and System Administration ICDEVGROUP = http://www.icdevgroup.org/ Interchange Development Team --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Making user names case-insensitive?
Anand Buddhdev wrote: man courier, and search for locallowercase Thanks all, my fault, I overlooked this option. However, if I understand correctly, it works only for local accounts. Does this mean real accounts only or does it work for virtual (userdb) accounts too? -- Ciao, Flavio Stanchina Trento - Italy --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Making user names case-insensitive?
On Fri, Feb 13, 2004 at 10:32:39AM +0100, Flavio Stanchina wrote: Anand Buddhdev wrote: man courier, and search for locallowercase Thanks all, my fault, I overlooked this option. However, if I understand correctly, it works only for local accounts. Does this mean real accounts only or does it work for virtual (userdb) accounts too? It works for *all* accounts that are delivered locally, whether they are looked up from /etc/passwd, userdb, ldap, pgsql or mysql. Just make sure that the account entry is in lower case in the user database. -- Anand Buddhdev --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
AW: [courier-users] Making user names case-insensitive?
Works for userdb virtual users too. But you could just try that ;) if I understand correctly, it works only for local accounts. Does this mean real accounts only or does it work for virtual (userdb) accounts too? --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id56alloc_id438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
[courier-users] Re: dynamic delivery instruction and 550 User unknown
Nat Sakimura writes: Hi. I am using dynamic delivery instruction in /etc/courier/aliasdir/courier-default . The external program makes a lookup into a database and decides where is the final destination. I want to make it such that if the lookup fails, the mta returns 550 User unknown to the client program. How can I achieve it? If the lookup fails, have the external program generate a: | echo 550 User unknown pgp0.pgp Description: PGP signature
[courier-users] Re: Message submission process (Re: Rejection/re-insertion methodology howto?)
Markus Wernig writes: On Fri, 2004-02-13 at 03:00, Sam Varshavchik wrote: [...] Well, it's not as easy as inserting something into something else. The stuff in question is a hierarchical tree, that contains offsets, field contents, etc Hello Sam Yes, I thought it wouldn't be like in Perl ;-) But (see quote) it seems to be possible. Does submit flush the structure from memory to the datafile before it passes it on to the next step in the chain (maildrop already?) ? Could you point me at the place where the rewriting mentioned below is done, so I could give it a try over the weekend? You want to look at submit.C and submit2.C pgp0.pgp Description: PGP signature
[courier-users] Re: My Modest Proposal
From: Sam Varshavchik [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: [courier-users] Re: My Modest Proposal Date: Thu, 12 Feb 2004 19:07:04 -0500 Mitch (WebCob) writes: Looking forward to the benchmarks - and also secretly wanting to try modifying a message file to see what happens. Maybe we misunderstood Sam - it is possible a modification could be ok as long as it doesn't corrupt the mime envelope?? Just a thought... No. The only modification that's safe is the one that leaves the starting and the ending positions of each structural component of an E-mail message -- the headers and the body of each MIME section -- intact. Adding or removing a single byte anywhere may screw things up. It may not screw up every message, but eventually it will screw up something. What about adding SPF support to reformail? -- Matt Pavlovich [EMAIL PROTECTED] Allegiance Telecom, Inc. --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
[courier-users] Re: My Modest Proposal
Lloyd Zusman [EMAIL PROTECTED] writes:
[ ... ]
I use the following for my unique ID, encoded into base 64:
32-bit IP address
32-bit PID of submit (some systems use more than 16 bits for this)
64-bit time from gettimeofday() [down to microseconds]
32-bit incrementing counter
[ ... ]
Hmm ... I now realize that I don't need the incrementing counter, since
there will be only one message processed per call to submit. I will
remove it from my code and repost another patch once I test it.
I now have changed the way I construct the unique ID:
128-bit IP address [*]
32-bit PID of submit
64-bit time from gettimeofday()
[*] If IPv6 is enabled, this is an in6_addr; otherwise, it's
an in_addr padded with high-order zeros to 128 bits.
This is encoded into base64, resulting in a 38-character ID. Look at
the 3rd line of this header for an example:
Received: from hiram.io.com (hiram.io.com [:::199.170.88.27])
(IDENT: root)
id AP@@2LYK%vYWAAAy5CxAe3oMAA
by asfast.net with esmtp; Fri, 13 Feb 2004 09:50:26 -0500
Attached is the latest patch.
Please look this over and see if you notice any problems. Thanks.
*** courier/submit.C.orig Mon Dec 15 20:51:29 2003
--- courier/submit.C Fri Feb 13 09:19:23 2004
***
*** 12,17
--- 12,18
#include rfc2045/rfc2045charset.h
#include rfc1035/rfc1035.h
#include rfc1035/rfc1035mxlist.h
+ #include rfc2045/rfc2045.h
#include numlib/numlib.h
#include dbobj.h
#include afx.h
***
*** 35,41
--- 36,52
#if HAVE_UNISTD_H
#include unistd.h
#endif
+ #if TIME_WITH_SYS_TIME
+ #include sys/time.h
#include time.h
+ #else
+ #if HAVE_SYS_TIME_H
+ #include sys/time.h
+ #else
+ #include time.h
+ #endif
+ #endif
+ #include netdb.h
#define TIMEOUT 1800
***
*** 935,940
--- 946,1041
static void getrcpt(struct rw_info *rwi);
static void rcpttoerr(int, const char *, struct rw_info *);
+ struct unique_id_buffer
+ {
+ #if !RFC1035_IPV6
+ RFC1035_ADDR dummy[3];
+ #endif
+ RFC1035_ADDR ip;
+ int pid;
+ struct timeval calltime;
+ };
+ #define UNIQUE_BUFFER_SIZE (sizeof (unique_id_buffer))
+
+ static int write_unique_bytes(const char *p, size_t l, void *vp)
+ {
+ char **cp = (char **) vp;
+
+ while (l 0)
+ {
+ switch (*p)
+ {
+ case '\r':
+ case '\n':
+ case '=':
+ p++;
+ l--;
+ continue;
+ case '/':
+ *p++;
+ **cp = '@';
+ break;
+ case '+':
+ *p++;
+ **cp = '%';
+ break;
+ default:
+ **cp = *p++;
+ break;
+ }
+ ++*cp;
+ l--;
+ }
+
+ return 0;
+ }
+
+ static char* unique_id()
+ {
+ static char result[ (UNIQUE_BUFFER_SIZE + 3) / 3 * 4 + 1 ];
+
+ struct unique_id_buffer unique_id_buff;
+ char* rp = result;
+ struct rfc2045_encode_info encodeInfo;
+ char hostbuff[256];
+ RFC1035_ADDR *addrs;
+ unsigned int naddrs;
+
+ gethostname(hostbuff, sizeof (hostbuff) - 1);
+ int rc = rfc1035_a(rfc1035_default_resolver, hostbuff, addrs,
+ naddrs);
+
+ if (rc == 0 || addrs == NULL || naddrs 1)
+ {
+ #if !RFC1035_IPV6
+ unique_id_buff.ipdummy[0] = 0;
+ unique_id_buff.ipdummy[1] = 0;
+ unique_id_buff.ipdummy[2] = 0;
+ #endif
+ memcpy((char*) unique_id_buff.ip, (char*) addrs[0],
+ sizeof (unique_id_buff.ip));
+
+ }
+ else
+ {
+
+ memset((char*) unique_id_buff.ip, 0xff,
+ sizeof (unique_id_buff.ip));
+ }
+
+ unique_id_buff.pid = getpid();
+
+ gettimeofday(unique_id_buff.calltime, NULL);
+
+ rfc2045_encode_start(encodeInfo, base64, write_unique_bytes, rp);
+ rfc2045_encode(encodeInfo, (char*) unique_id_buff,
+ UNIQUE_BUFFER_SIZE);
+ rfc2045_encode_end(encodeInfo);
+ *rp = 0;
+
+ return (result);
+ }
+
static void getrcpts(struct rw_info *rwi)
{
struct mailfrominfo *mf=(struct mailfrominfo *)rwi-udata;
***
*** 1087,1092
--- 1188,1197
line += identinfo;
line += ')';
}
+
+ // Now, always insert a unique ID
+ line += \n id ;
+ line += unique_id();
line += \n by ;
line += config_me();
--
Lloyd Zusman
[EMAIL PROTECTED]
[courier-users] Courier + Majordomo
I am trying to get courier and majordomo working together. Can anyone provide me a link or information on how to get this done? Thank You --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
[courier-users] response codes
This is not necesarily a courier question, but since thats what i use What reason is there to respond to a user unknown with transient error code instead of a permanent failure ? ..like this: : 450 [EMAIL PROTECTED]: User unknown in local recipient table i've seen this a couple of times now and the mail hangs in the queue for 4 days until it's bounced ... /thomas -- Dignity and an empty sack is worth the sack. -Ferengi Rule of Acquisition No. 109 --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier + Majordomo
Forgive my ignorance -- well I know and have used majordomo lists before -- but what are the major differences between majordomo and couriermlm? I'm using couriermlm now for a few lists and it seems to do the job quite admirably... Russell Premont wrote: I am trying to get courier and majordomo working together. Can anyone provide me a link or information on how to get this done? Thank You --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
[courier-users] Re: response codes
Thomas von Hassel writes: This is not necesarily a courier question, but since thats what i use What reason is there to respond to a user unknown with transient error code instead of a permanent failure ? ..like this: : 450 [EMAIL PROTECTED]: User unknown in local recipient table i've seen this a couple of times now and the mail hangs in the queue for 4 days until it's bounced ... This message comes from the remote mail server, and has nothing to do with Courier. pgp0.pgp Description: PGP signature
[courier-users] no entries being parced in smtpaccess but localhost?
Hello all. I am trying to allow relaying for ip address 151.199.16.40. In smtpaccess/default I have the following line: 151.199.16.40 allow,RELAYCLIENT When the person attempts to send a message with ESMTP, I get a message in my syslog that says that relaying is denied for the message he was trying to send. However, relaying is working fine from localhost. I have this exact same line in smtpaccess/default for localhost: 127.0.0.1 allow,RELAYCLIENT Does anyone know what could be happening here that is denying relaying for 151.199.16.40? Could it be Verizon trying to block the use of outside SMTP servers? However, if its Verizon trying to block outside relaying, then theoretically I would not have seen his connection attempt. P.S: Comcast has stopped filtering traffic on port 25, so I can now use my server with no problems. mail2web - Check your email from the web at http://mail2web.com/ . --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id56alloc_id438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
[courier-users] Re: no entries being parced in smtpaccess but localhost?
[EMAIL PROTECTED] writes: Hello all. I am trying to allow relaying for ip address 151.199.16.40. In smtpaccess/default I have the following line: 151.199.16.40 allow,RELAYCLIENT When the person attempts to send a message with ESMTP, I get a message in my syslog that says that relaying is denied for the message he was trying to send. However, relaying is working fine from localhost. I have this exact same line in smtpaccess/default for localhost: 127.0.0.1 allow,RELAYCLIENT Does anyone know what could be happening here that is denying relaying for 151.199.16.40? Could it be Verizon trying to block the use of outside SMTP servers? However, if its Verizon trying to block outside relaying, then theoretically I would not have seen his connection attempt. P.S: Comcast has stopped filtering traffic on port 25, so I can now use my server with no problems. Did you run 'makesmtpaccess' after editing the file? pgp0.pgp Description: PGP signature
[courier-users] make: invalid option when using DESTDIR with install-configure
Hi. I have compiled and am attempting to install courier into a packaging directory. # make DESTDIR=/some/dir install-configure gives the following result. make: invalid option -- D make: invalid option -- E make: invalid option -- T make: invalid option -- D Usage: ... ... make: *** [install-configure] Error 2 It does not do it with the install target, ie. make DESTDIR=/some/dir install Any idea why it is treating DESTDIR as options when I use a target that has a dash in it? I am using GNU make-3.79 on FreeBSD-5.2 I found one other reference from google searches where somebody else had the same problem but I saw no solution. Regards, Vincent --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] make: invalid option when using DESTDIR with install-configure
That's what make does. It's working fine. To change the installation directory you need to set it when you configure the install. ./configure --prefix=/some/dir On 14/02/2004, at 3:25 PM, Vincent wrote: Hi. I have compiled and am attempting to install courier into a packaging directory. # make DESTDIR=/some/dir install-configure gives the following result. make: invalid option -- D make: invalid option -- E make: invalid option -- T make: invalid option -- D Usage: ... ... make: *** [install-configure] Error 2 It does not do it with the install target, ie. make DESTDIR=/some/dir install -- Phillip Hutchings [EMAIL PROTECTED] http://www.sitharus.com/ smime.p7s Description: S/MIME cryptographic signature
Re: [courier-users] make: invalid option when using DESTDIR with install-configure
Thanks for the reply. On Sat, Feb 14, 2004 at 03:59:00PM +1300, Phillip Hutchings wrote: That's what make does. It's working fine. I don't understand. Why doesn't it do on on other targets like install-strip? To change the installation directory you need to set it when you configure the install. ./configure --prefix=/some/dir I am using prefix, but that isn't related to installing into a build directory for building a binary package. That requires DESTDIR. Is courier not set up to be able to create a binary package that works out of the box on more than one machine? If it is, I should not have to run make install-configure from a source directory after installing. I need to be able to use DESTDIR so that I can have the package all pre-configured before creating a binary package to take to another machine. On 14/02/2004, at 3:25 PM, Vincent wrote: Hi. I have compiled and am attempting to install courier into a packaging directory. # make DESTDIR=/some/dir install-configure gives the following result. make: invalid option -- D make: invalid option -- E make: invalid option -- T make: invalid option -- D Usage: ... ... make: *** [install-configure] Error 2 It does not do it with the install target, ie. make DESTDIR=/some/dir install -- Phillip Hutchings [EMAIL PROTECTED] http://www.sitharus.com/ --- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356alloc_id=3438op=click ___ courier-users mailing list [EMAIL PROTECTED] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
