dovecot-antispam - fork of dovecot-anitspam with own trainspam extension

	Commit message (Collapse)	Author	Age	Files	Lines
*	rename mailtrain to pipe	Johannes Berg	2011-02-27	1	-364/+0
\|
*	add an API version to the backend struct	Johannes Berg	2011-02-27	1	-6/+7
\| \| \| \| \| \|	Just to make sure nobody loads a backend plugin that isn't suitable for the version of antispam itself.
*	make backend selection at runtime	Daniel Hokka Zakrisson	2011-02-27	1	-9/+18
\|
*	mailtrain: redirect stdout/stderr to /dev/null for execv	Johannes Berg	2011-02-27	1	-3/+5
\| \| \| \| \| \|	This fixes a bug where sa-learn and similar tools fail because they cannot write to stdout or stderr at all, and propagate the error (EBADF or such).
*	fix mbox "From " removal code for not-present case	Johannes Berg	2009-06-15	1	-10/+1
\| \| \| \| \| \| \|	Timo explained it to me here: http://mid.gmane.org/1245092921.21624.739.camel@timo-desktop It's been broken forever.
*	mailtrain: send correct 5 bytes if message doesn't start with "From "	Johannes Berg	2009-04-11	1	-1/+1
\|
*	clean up dovecot compat code a bit	Johannes Berg	2008-05-31	1	-11/+2
\|
*	backend preparations for 1.1	Johannes Berg	2008-04-26	1	-0/+19
\|
*	mailtrain: fix bug with stripping "From " line	Johannes Berg	2008-04-24	1	-4/+16
\|
*	add debug output to mailtrain	Johannes Berg	2008-03-20	1	-0/+2
\|
*	Properly log return value of sendmail invocation if it fails.	Steffen Kaiser	2008-03-06	1	-1/+1
\|
*	Fix file descriptor leak in mailtrain backend	Steffen Kaiser	2008-03-05	1	-2/+3
\|
*	Leaked a t_pop() call in I/O handler 0x805b350	Steffen Kaiser	2008-03-05	1	-2/+5
\| \| \| \| \| \| \| \| \| \| \| \| \| \| \|	Raw backtrace: imap [0x80ad741] -> imap [0x80ad65c] -> imap(io_loop_handler _run+0x1ce) [0x80b35ce] -> imap(io_loop_run+0x28) [0x80b29a8] -> imap(main+0x5a0) [0x8063590] -> /lib/tls/i686/cmov/libc.so.6(__lib c_start_main+0xc8) [0xb7e37ea8] -> imap [0x8055e31] child 18478 (imap) killed with signal 6 Cause is, if sendmail fails with an exit code != 0, the function returns with -1 immediately skipping the t_pop() before the end of the function. This fixes it.
*	allow arbitrary command line arguments for mailtrain backend	Johannes Berg	2008-02-02	1	-9/+38
\|
*	use -Wextra	Johannes Berg	2007-10-15	1	-2/+3
\|
*	make settings per-plugin	Johannes Berg	2007-10-14	1	-9/+9
\|
*	use 'enum classification' instead of 'bool from_spam'	Johannes Berg	2007-10-01	1	-8/+17
\|
*	add mailtrain backend and docs	Johannes Berg	2007-10-01	1	-0/+294