You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@spamassassin.apache.org by fe...@apache.org on 2004/08/28 03:02:46 UTC
svn commit: rev 37136 - in spamassassin/trunk: . lib/Mail
Author: felicity
Date: Fri Aug 27 18:02:45 2004
New Revision: 37136
Modified:
spamassassin/trunk/Changes
spamassassin/trunk/lib/Mail/SpamAssassin.pm
Log:
prepare for 3.0.0-rc2 release
Modified: spamassassin/trunk/Changes
==============================================================================
--- spamassassin/trunk/Changes (original)
+++ spamassassin/trunk/Changes Fri Aug 27 18:02:45 2004
@@ -1,4 +1,208 @@
------------------------------------------------------------------------
+r37130 | jm | 2004-08-27 20:16:39 +0000 (Fri, 27 Aug 2004) | 1 line
+
+oops, I'd forgotten to check in the SAME fixes from minotaur. here's some parts of what I did that are still worthwhile with Theo's changes
+------------------------------------------------------------------------
+r37129 | felicity | 2004-08-27 18:15:59 +0000 (Fri, 27 Aug 2004) | 1 line
+
+updates to the build process -- md5sum and sha1sum aren't installed on minotaur, so use quick n' dirty perl versions I wrote up. /bin/bash doesn't exist on minotaur either, so change to /usr/local/bin/bash ...
+------------------------------------------------------------------------
+r37097 | felicity | 2004-08-26 21:37:01 +0000 (Thu, 26 Aug 2004) | 1 line
+
+bug 3731: there were some situations where umask would be changed and then not reset back to the original value
+------------------------------------------------------------------------
+r37092 | quinlan | 2004-08-26 18:12:49 +0000 (Thu, 26 Aug 2004) | 2 lines
+
+bug 3730: make DNS unavailable unless versions are high enough
+
+------------------------------------------------------------------------
+r37091 | felicity | 2004-08-26 17:13:13 +0000 (Thu, 26 Aug 2004) | 1 line
+
+bug 3728: check_languages() wasn't able to find the languages hint file so it didn't run.
+------------------------------------------------------------------------
+r37041 | quinlan | 2004-08-25 19:43:35 +0000 (Wed, 25 Aug 2004) | 2 lines
+
+use Constants.pm
+
+------------------------------------------------------------------------
+r36852 | felicity | 2004-08-25 13:11:49 +0000 (Wed, 25 Aug 2004) | 1 line
+
+bug 3719: ignore X-Bogosity header from potential upstream Bogofilter run
+------------------------------------------------------------------------
+r36851 | felicity | 2004-08-25 13:09:08 +0000 (Wed, 25 Aug 2004) | 1 line
+
+bug 3705: add a regression test for relative scores
+------------------------------------------------------------------------
+r36836 | jm | 2004-08-25 04:23:15 +0000 (Wed, 25 Aug 2004) | 1 line
+
+bug 3715: Conf::Parser double-counts certain errors in --lint, fixed. bug 3699: --lint should warn if a description is empty. also, some doc changes to clarify Conf::Parser operation
+------------------------------------------------------------------------
+r36835 | jm | 2004-08-25 04:23:03 +0000 (Wed, 25 Aug 2004) | 1 line
+
+bug 3715: Conf::Parser double-counts certain errors in --lint, fixed. bug 3699: --lint should warn if a description is empty. also, some doc changes to clarify Conf::Parser operation
+------------------------------------------------------------------------
+r36826 | mss | 2004-08-24 22:48:18 +0000 (Tue, 24 Aug 2004) | 2 lines
+
+Removed two obsolete commented-out lines. It's now documented better in INSTALL.
+
+------------------------------------------------------------------------
+r36823 | mss | 2004-08-24 20:35:39 +0000 (Tue, 24 Aug 2004) | 2 lines
+
+Also strip the directory from the converted spamc docs.
+
+------------------------------------------------------------------------
+r36822 | mss | 2004-08-24 20:12:24 +0000 (Tue, 24 Aug 2004) | 2 lines
+
+Fixed up some PODs about which convert_pods_to_doc complained.
+
+------------------------------------------------------------------------
+r36816 | mss | 2004-08-24 17:22:00 +0000 (Tue, 24 Aug 2004) | 8 lines
+
+bug 3717: Loading PlugIns from outside @INC didn't work because do() looks in @INC, too, for relative paths. Fixed by always using absolute paths.
+
+also:
+ * Added Justin's regression test for this feature.
+ * Marked some tests which missed the bit as executable.
+ * Sorted MANIFEST alphabetically (again).
+
+
+------------------------------------------------------------------------
+r36815 | felicity | 2004-08-24 16:07:22 +0000 (Tue, 24 Aug 2004) | 1 line
+
+bug 3705: relative scores worked when all 4 scoresets were specified, but not when only 1 number was given. fix that, and fix debug output to use the correct variable. also, instead of 2 sections for 'score rule # # # #' and 'score rule #', always assume all 4 scoresets are specified and deal with the 1 score=>4 scores bit appropriately.
+------------------------------------------------------------------------
+r36792 | felicity | 2004-08-24 00:19:29 +0000 (Tue, 24 Aug 2004) | 1 line
+
+bug 3722: document the --folders option
+------------------------------------------------------------------------
+r36791 | felicity | 2004-08-24 00:02:33 +0000 (Tue, 24 Aug 2004) | 1 line
+
+bug 3725: added documentation for a bunch of PerMsgStatus methods, including get_uri_list()
+------------------------------------------------------------------------
+r36786 | quinlan | 2004-08-23 21:17:37 +0000 (Mon, 23 Aug 2004) | 2 lines
+
+clarify Net::DNS version requirements
+
+------------------------------------------------------------------------
+r36784 | mss | 2004-08-23 21:01:15 +0000 (Mon, 23 Aug 2004) | 2 lines
+
+Net::DNS 0.46 or higher is needed for DNS tests under Windows. Thanks to Fred Tarasevicius for debugging this issue.
+
+------------------------------------------------------------------------
+r36781 | felicity | 2004-08-23 20:20:53 +0000 (Mon, 23 Aug 2004) | 1 line
+
+bug 2437: make the code to figure out the autolearn score be in a single function and not scattered around PMS. Also, properly ignore noautolearn, userconf, and learn rules instead of just noautolearn. also, don't shift the global scoreset when figuring out whether to autolearn, just use a local scoreset copy.
+------------------------------------------------------------------------
+r36766 | quinlan | 2004-08-23 18:08:44 +0000 (Mon, 23 Aug 2004) | 2 lines
+
+add conditional around URIDNSBL scores
+
+------------------------------------------------------------------------
+r36758 | jm | 2004-08-23 05:48:40 +0000 (Mon, 23 Aug 2004) | 1 line
+
+doco fix: simplify sample plugin text a little
+------------------------------------------------------------------------
+r36757 | jm | 2004-08-23 05:38:20 +0000 (Mon, 23 Aug 2004) | 1 line
+
+trivial: typo
+------------------------------------------------------------------------
+r36701 | parker | 2004-08-22 18:11:21 +0000 (Sun, 22 Aug 2004) | 1 line
+
+Bug 3713: Little clarify to docs on _TABLE_ value for custom sql query
+------------------------------------------------------------------------
+r36698 | parker | 2004-08-22 16:58:15 +0000 (Sun, 22 Aug 2004) | 1 line
+
+Bug 3713: UPGRADE needs to mention that user_scores_sql_table is no longer supported
+------------------------------------------------------------------------
+r36695 | mss | 2004-08-22 14:13:29 +0000 (Sun, 22 Aug 2004) | 2 lines
+
+Added note about the proper use of DROPPRIVS in site-wide procmailrcs.
+
+------------------------------------------------------------------------
+r36690 | quinlan | 2004-08-22 06:53:49 +0000 (Sun, 22 Aug 2004) | 2 lines
+
+bug 3702: bayes version mismatch should warn() not dbg()
+
+------------------------------------------------------------------------
+r36689 | felicity | 2004-08-22 06:32:01 +0000 (Sun, 22 Aug 2004) | 1 line
+
+clean up the 'score' doco a little
+------------------------------------------------------------------------
+r36688 | felicity | 2004-08-22 06:18:37 +0000 (Sun, 22 Aug 2004) | 1 line
+
+make the tflag/autolearn/awl interactions more clear. actually document what rules the awl ignores when figuring out the rules to use.
+------------------------------------------------------------------------
+r36686 | felicity | 2004-08-22 05:50:06 +0000 (Sun, 22 Aug 2004) | 1 line
+
+make some of the debug output more clear
+------------------------------------------------------------------------
+r36685 | felicity | 2004-08-22 04:01:34 +0000 (Sun, 22 Aug 2004) | 1 line
+
+bug 3707: added build requirements into spec file, updated required perl version to 5.6.1
+------------------------------------------------------------------------
+r36673 | quinlan | 2004-08-20 22:27:59 +0000 (Fri, 20 Aug 2004) | 2 lines
+
+die whitespace, die!
+
+------------------------------------------------------------------------
+r36638 | felicity | 2004-08-19 22:11:42 +0000 (Thu, 19 Aug 2004) | 1 line
+
+another doco update wrt autolearning
+------------------------------------------------------------------------
+r36625 | felicity | 2004-08-19 19:03:27 +0000 (Thu, 19 Aug 2004) | 1 line
+
+make the documentation a little more explicit about what tflags disable autolearn control
+------------------------------------------------------------------------
+r36540 | felicity | 2004-08-17 21:01:11 +0000 (Tue, 17 Aug 2004) | 1 line
+
+bug 3692: DCC eval code was trying to check for DCC headers already in the message indicating bulk status, but wasn't actually looking at the right header(s).
+------------------------------------------------------------------------
+r36538 | mss | 2004-08-17 20:39:12 +0000 (Tue, 17 Aug 2004) | 5 lines
+
+Some docu tweaks:
+* Split the documentation of the spamd/spamc protocol into an extra file
+* Some little cleanup in the README, added FIXMEs to outdated docu (I have
+ currently no time to wordsmith the actual updates)
+
+------------------------------------------------------------------------
+r36536 | quinlan | 2004-08-17 20:35:37 +0000 (Tue, 17 Aug 2004) | 2 lines
+
+trivial documentation tweak
+
+------------------------------------------------------------------------
+r36535 | quinlan | 2004-08-17 20:32:05 +0000 (Tue, 17 Aug 2004) | 2 lines
+
+clarify the URIBL blurb to make it easier to understand
+
+------------------------------------------------------------------------
+r36516 | mss | 2004-08-17 11:59:22 +0000 (Tue, 17 Aug 2004) | 2 lines
+
+bug 3690: skip SPF regression tests if RUN_NET_TESTS is false
+
+------------------------------------------------------------------------
+r36490 | quinlan | 2004-08-17 03:54:11 +0000 (Tue, 17 Aug 2004) | 2 lines
+
+use explicit return for clarity
+
+------------------------------------------------------------------------
+r36456 | quinlan | 2004-08-16 06:44:13 +0000 (Mon, 16 Aug 2004) | 3 lines
+
+revise the release highlights, note: removing hashcash - sorry, it
+doesn't make the top list of features people care about
+
+------------------------------------------------------------------------
+r36452 | jm | 2004-08-16 06:14:53 +0000 (Mon, 16 Aug 2004) | 1 line
+
+going back into non-release-build mode
+------------------------------------------------------------------------
+r36451 | jm | 2004-08-16 06:14:23 +0000 (Mon, 16 Aug 2004) | 1 line
+
+3.0.0-rc1 RELEASED
+------------------------------------------------------------------------
+r36449 | jm | 2004-08-16 06:11:18 +0000 (Mon, 16 Aug 2004) | 1 line
+
+preparing to release 3.0.0-rc1
+------------------------------------------------------------------------
r36447 | parker | 2004-08-16 05:29:08 +0000 (Mon, 16 Aug 2004) | 1 line
Bug 3331: Adds hooks to Plugin API for bayes_scan, bayes_learn and bayes_forget that passes in hashed and raw token values, amoung other things. Useful for creating the hashed to raw token value map that people may want.
@@ -902,4 +1106,6846 @@
Move spamassassin from the incubator to top level.
------------------------------------------------------------------------
+r22094 | jm | 2004-06-25 03:02:51 +0000 (Fri, 25 Jun 2004) | 1 line
+
+forgot hashcash in the change announcement mail
+------------------------------------------------------------------------
+r22081 | felicity | 2004-06-24 19:53:19 +0000 (Thu, 24 Jun 2004) | 1 line
+
+bug 3537: the hk TLD has a few more second-level TLD parts (gov, edu, idv,) so add them into our list.
+------------------------------------------------------------------------
+r22079 | parker | 2004-06-24 19:12:58 +0000 (Thu, 24 Jun 2004) | 1 line
+
+Bug 3526: modify description of --import switch to be clearer
+------------------------------------------------------------------------
+r21615 | mss | 2004-06-24 00:24:15 +0000 (Thu, 24 Jun 2004) | 2 lines
+
+sorted alphabetically
+
+------------------------------------------------------------------------
+r21525 | jm | 2004-06-21 03:43:29 +0000 (Mon, 21 Jun 2004) | 1 line
+
+documentation: corrected attribution for the 30_text_de.cf translations
+------------------------------------------------------------------------
+r21514 | parker | 2004-06-20 18:35:06 +0000 (Sun, 20 Jun 2004) | 1 line
+
+Bug 3477: do not exit child when there is an error looking up SQL prefs
+------------------------------------------------------------------------
+r21481 | quinlan | 2004-06-20 02:10:52 +0000 (Sun, 20 Jun 2004) | 2 lines
+
+nit
+
+------------------------------------------------------------------------
+r21480 | quinlan | 2004-06-20 02:09:32 +0000 (Sun, 20 Jun 2004) | 2 lines
+
+more tweaking, duncf made me do it ;-)
+
+------------------------------------------------------------------------
+r21479 | quinlan | 2004-06-20 02:03:22 +0000 (Sun, 20 Jun 2004) | 2 lines
+
+more tweaks based on comments from duncf
+
+------------------------------------------------------------------------
+r21477 | quinlan | 2004-06-20 01:43:30 +0000 (Sun, 20 Jun 2004) | 2 lines
+
+spelling
+
+------------------------------------------------------------------------
+r21476 | quinlan | 2004-06-20 01:41:35 +0000 (Sun, 20 Jun 2004) | 2 lines
+
+some last changes to the summary before the 3.0.0 pre-release
+
+------------------------------------------------------------------------
+r21475 | jm | 2004-06-20 01:19:17 +0000 (Sun, 20 Jun 2004) | 1 line
+
+3.0.0 devel cycle started
+------------------------------------------------------------------------
+r21473 | jm | 2004-06-20 01:17:31 +0000 (Sun, 20 Jun 2004) | 1 line
+
+er, that should be: 3.0.0-pre1 RELEASED
+------------------------------------------------------------------------
+r21472 | jm | 2004-06-20 01:17:15 +0000 (Sun, 20 Jun 2004) | 1 line
+
+3.0.0 RELEASED
+------------------------------------------------------------------------
+r21471 | jm | 2004-06-20 01:16:32 +0000 (Sun, 20 Jun 2004) | 1 line
+
+updated hashes to match built distribution
+------------------------------------------------------------------------
+r21470 | jm | 2004-06-20 01:12:23 +0000 (Sun, 20 Jun 2004) | 1 line
+
+doco and build script fixes
+------------------------------------------------------------------------
+r21468 | jm | 2004-06-20 01:02:54 +0000 (Sun, 20 Jun 2004) | 1 line
+
+translations
+------------------------------------------------------------------------
+r21467 | parker | 2004-06-20 01:00:25 +0000 (Sun, 20 Jun 2004) | 1 line
+
+added comment to SQL changes
+------------------------------------------------------------------------
+r21466 | quinlan | 2004-06-20 01:00:06 +0000 (Sun, 20 Jun 2004) | 2 lines
+
+a few tweaks
+
+------------------------------------------------------------------------
+r21465 | jm | 2004-06-20 00:48:55 +0000 (Sun, 20 Jun 2004) | 1 line
+
+updating change summary with a couple of overlooked items
+------------------------------------------------------------------------
+r21464 | quinlan | 2004-06-20 00:37:03 +0000 (Sun, 20 Jun 2004) | 4 lines
+
+it's been a very long time since Matt and Craig committed code and they
+don't have ASF committer access, so list them separately like other ASF
+projects do; spell my name "Daniel"
+
+------------------------------------------------------------------------
+r21462 | jm | 2004-06-20 00:33:23 +0000 (Sun, 20 Jun 2004) | 1 line
+
+added t/README to MANIFEST
+------------------------------------------------------------------------
+r21461 | jm | 2004-06-20 00:31:06 +0000 (Sun, 20 Jun 2004) | 1 line
+
+added change summary mail text
+------------------------------------------------------------------------
+r21459 | jm | 2004-06-20 00:26:44 +0000 (Sun, 20 Jun 2004) | 1 line
+
+And Matt Y
+------------------------------------------------------------------------
+r21458 | jm | 2004-06-20 00:24:18 +0000 (Sun, 20 Jun 2004) | 1 line
+
+outdated stuff in doco
+------------------------------------------------------------------------
+r21457 | jm | 2004-06-20 00:21:11 +0000 (Sun, 20 Jun 2004) | 1 line
+
+Sidney wasn't in the CREDITS file, oops ;)
+------------------------------------------------------------------------
+r21455 | jm | 2004-06-20 00:08:23 +0000 (Sun, 20 Jun 2004) | 1 line
+
+note that --virtual-config was dropped
+------------------------------------------------------------------------
+r21454 | sidney | 2004-06-20 00:06:05 +0000 (Sun, 20 Jun 2004) | 1 line
+
+bug #3513: document make test in a README
+------------------------------------------------------------------------
+r21453 | jm | 2004-06-19 23:47:18 +0000 (Sat, 19 Jun 2004) | 1 line
+
+UPGRADE documentation fixes: moved around some pieces, noted rewrite_subject is now dropped, and reformatted to be more consistent with INSTALL and release announcement mails
+------------------------------------------------------------------------
+r21452 | jm | 2004-06-19 23:31:52 +0000 (Sat, 19 Jun 2004) | 1 line
+
+added masses/config to MANIFEST.SKIP
+------------------------------------------------------------------------
+r21451 | jm | 2004-06-19 23:30:17 +0000 (Sat, 19 Jun 2004) | 1 line
+
+bug 3522: 'make disttest' fails due to removed FORGED_*_WHITELIST rules, fixed
+------------------------------------------------------------------------
+r21450 | parker | 2004-06-19 23:25:16 +0000 (Sat, 19 Jun 2004) | 1 line
+
+Doco updates
+------------------------------------------------------------------------
+r21449 | jm | 2004-06-19 23:12:51 +0000 (Sat, 19 Jun 2004) | 1 line
+
+more build doco updates
+------------------------------------------------------------------------
+r21448 | jm | 2004-06-19 22:59:29 +0000 (Sat, 19 Jun 2004) | 1 line
+
+bug 3520: disable t/bayesdbm_flock.t when running test suite on Windows
+------------------------------------------------------------------------
+r21447 | jm | 2004-06-19 22:43:49 +0000 (Sat, 19 Jun 2004) | 1 line
+
+updated build doco
+------------------------------------------------------------------------
+r21446 | jm | 2004-06-19 22:40:21 +0000 (Sat, 19 Jun 2004) | 1 line
+
+preparing for 3.0.0-pre1 release
+------------------------------------------------------------------------
+r21445 | jm | 2004-06-19 22:08:19 +0000 (Sat, 19 Jun 2004) | 1 line
+
+minor doco tweaks
+------------------------------------------------------------------------
+r21444 | sidney | 2004-06-19 21:40:20 +0000 (Sat, 19 Jun 2004) | 1 line
+
+bug #3521: replace use of platform dependent system call with perl version
+------------------------------------------------------------------------
+r21443 | sidney | 2004-06-19 08:22:41 +0000 (Sat, 19 Jun 2004) | 1 line
+
+bug #3513: fix spamc tests under Windows
+------------------------------------------------------------------------
+r21439 | parker | 2004-06-19 04:38:53 +0000 (Sat, 19 Jun 2004) | 1 line
+
+Bug 3503: config needs to be copied for sql/ldap configs as well
+------------------------------------------------------------------------
+r21433 | jm | 2004-06-19 02:03:13 +0000 (Sat, 19 Jun 2004) | 1 line
+
+allow_user_rules fixes: block user-defined SenderBase rules; sanitise arguments to many regexp rules; run helper apps as uid=user euid=user; ensure regexps are valid before they are used in the ruleset. fix a couple of rules to work with these changes.
+------------------------------------------------------------------------
+r21432 | parker | 2004-06-19 01:33:19 +0000 (Sat, 19 Jun 2004) | 1 line
+
+Bug 3334: create sed_path cache to be used by sed_path and clear the cache in compile_now after running the test msg
+------------------------------------------------------------------------
+r21429 | parker | 2004-06-19 00:47:12 +0000 (Sat, 19 Jun 2004) | 1 line
+
+Bug 3519: add documentation about failures with postgresql less than 7.4.2
+------------------------------------------------------------------------
+r21409 | jm | 2004-06-18 06:01:35 +0000 (Fri, 18 Jun 2004) | 1 line
+
+script was not executable, fixed
+------------------------------------------------------------------------
+r21408 | jm | 2004-06-18 05:56:28 +0000 (Fri, 18 Jun 2004) | 1 line
+
+bug 3509: make building spamc easier for packagers with new version.h.pl script
+------------------------------------------------------------------------
+r21406 | jm | 2004-06-18 05:50:28 +0000 (Fri, 18 Jun 2004) | 1 line
+
+bug 3511: add support for multi-valued RHSBLs; add multi.surbl.org as well
+------------------------------------------------------------------------
+r21389 | parker | 2004-06-17 20:21:39 +0000 (Thu, 17 Jun 2004) | 1 line
+
+Bug 3518: change char(200) to char(5) to bring it in line with the current database revision
+------------------------------------------------------------------------
+r21353 | mss | 2004-06-16 18:33:35 +0000 (Wed, 16 Jun 2004) | 3 lines
+
+* Removed some obsolete entries from the ignore list.
+* Prepared for Module::Build :)
+
+------------------------------------------------------------------------
+r21338 | jm | 2004-06-16 02:41:16 +0000 (Wed, 16 Jun 2004) | 1 line
+
+bug 3467: uri_to_domain should only return domains with valid TLDs
+------------------------------------------------------------------------
+r21335 | quinlan | 2004-06-16 01:48:57 +0000 (Wed, 16 Jun 2004) | 3 lines
+
+bug 3495: remove DNS_FROM_RFC_WHOIS_B due to high score and poor FP rate,
+rename DNS_FROM_RFC_WHOIS_A to DNS_FROM_RFC_WHOIS
+
+------------------------------------------------------------------------
+r21322 | parker | 2004-06-15 23:00:26 +0000 (Tue, 15 Jun 2004) | 1 line
+
+Bug 3508: Type in variable name
+------------------------------------------------------------------------
+r21320 | jm | 2004-06-15 22:24:25 +0000 (Tue, 15 Jun 2004) | 1 line
+
+test update: ensure 2 hashcash tokens in multiple headers work sensibly
+------------------------------------------------------------------------
+r21294 | quinlan | 2004-06-15 16:14:56 +0000 (Tue, 15 Jun 2004) | 3 lines
+
+move some network tests out into the network test section and set their
+non-net score set scores to 0
+
+------------------------------------------------------------------------
+r21257 | quinlan | 2004-06-15 05:38:36 +0000 (Tue, 15 Jun 2004) | 2 lines
+
+bug 3447: fix test failures
+
+------------------------------------------------------------------------
+r21256 | quinlan | 2004-06-15 05:38:03 +0000 (Tue, 15 Jun 2004) | 15 lines
+
+bug 3447
+- change TCR lambda to 50 everywhere and add lambda value to output
+- add -t option to perceptron for tuning FP/FN ratio (workable solution for
+ now)
+- add -h option to perceptron
+- remove extra newline from rewrite-cf-with-new-scores (noticible when you
+ run it four times in a row)
+- update runGA to be more or less what I used
+ - revert 50_scores.cf before runs: very important
+ - iterate over thresholds (relatively cheap) to find optimal FN/FP point
+ (used: 4.2 for set0, 4.8 for set1 and set2, 4.9 for set3)
+ - also rewrite and validate scores on each iteration
+- masses/mk-baseline-results: fix some option names
+- some other minor changes
+
+------------------------------------------------------------------------
+r21255 | quinlan | 2004-06-15 05:37:36 +0000 (Tue, 15 Jun 2004) | 5 lines
+
+bug 3447
+- change SPF_PASS and SPF_HELO_PASS to be userconf rules so perceptron does
+ not try to assign them a score
+- new scores!
+
+------------------------------------------------------------------------
+r21251 | quinlan | 2004-06-15 02:33:54 +0000 (Tue, 15 Jun 2004) | 3 lines
+
+remove svn:mime-type flag from all these files so I can svn diff them
+(they are not binary files, so the flag isn't really needed)
+
+------------------------------------------------------------------------
+r21195 | quinlan | 2004-06-14 09:40:00 +0000 (Mon, 14 Jun 2004) | 2 lines
+
+documentation update, remove ancient statistics
+
+------------------------------------------------------------------------
+r21065 | quinlan | 2004-06-11 04:43:19 +0000 (Fri, 11 Jun 2004) | 2 lines
+
+now you only have to change SCORESET once
+
+------------------------------------------------------------------------
+r21042 | quinlan | 2004-06-11 00:40:23 +0000 (Fri, 11 Jun 2004) | 2 lines
+
+add "\n"
+
+------------------------------------------------------------------------
+r21041 | quinlan | 2004-06-11 00:30:15 +0000 (Fri, 11 Jun 2004) | 2 lines
+
+more clean-up of output
+
+------------------------------------------------------------------------
+r21040 | quinlan | 2004-06-10 23:41:14 +0000 (Thu, 10 Jun 2004) | 3 lines
+
+use consistent naming for ham vs. spam
+make output of fp-fn-statistics non-confusing
+
+------------------------------------------------------------------------
+r21039 | quinlan | 2004-06-10 23:38:03 +0000 (Thu, 10 Jun 2004) | 4 lines
+
+turn on score ranges code
+picks initial weights between [range_lo..range_hi)
+change output to be non-confusing
+
+------------------------------------------------------------------------
+r21038 | quinlan | 2004-06-10 23:10:31 +0000 (Thu, 10 Jun 2004) | 2 lines
+
+Call the files ham or call them nonspam, but don't call them both!!!!!!!!!!
+------------------------------------------------------------------------
+r21036 | parker | 2004-06-10 19:45:47 +0000 (Thu, 10 Jun 2004) | 1 line
+
+Bug 3498 - Update documentation
+------------------------------------------------------------------------
+r20962 | parker | 2004-06-09 21:01:47 +0000 (Wed, 09 Jun 2004) | 1 line
+
+Bug 3471: add sanity check so too many tokens do not get expired
+------------------------------------------------------------------------
+r20961 | jm | 2004-06-09 20:13:34 +0000 (Wed, 09 Jun 2004) | 1 line
+
+bug 3442: renamed the .lock file to .mutex for the flock-locking case, since it is really a mutual-exclusion semaphore, not a lock. in other words, it sticks around after unlocking.
+------------------------------------------------------------------------
+r20960 | jm | 2004-06-09 20:11:23 +0000 (Wed, 09 Jun 2004) | 1 line
+
+reinstating accidentally-deleted line from last patch
+------------------------------------------------------------------------
+r20959 | felicity | 2004-06-09 20:02:05 +0000 (Wed, 09 Jun 2004) | 1 line
+
+bug 3482: having the children reset the signal handling removed perl-internal signal handles as well, such as __WARN__, which broke the debug output. we now explicitly only reset the signals that we changed in the parent via 2 new functions in spamd where all the signal handling should happen.
+------------------------------------------------------------------------
+r20958 | felicity | 2004-06-09 19:58:37 +0000 (Wed, 09 Jun 2004) | 1 line
+
+bug 3490: spamd wasn't warning if the pidfile or socket was failing to be removed during a restart
+------------------------------------------------------------------------
+r20956 | felicity | 2004-06-09 16:48:36 +0000 (Wed, 09 Jun 2004) | 1 line
+
+wordsmithing the file lock POD
+------------------------------------------------------------------------
+r20922 | felicity | 2004-06-08 21:08:32 +0000 (Tue, 08 Jun 2004) | 1 line
+
+fix broken spf test -- expected __SPF_PASS and __SPF_HELO_PASS which were renamed the other day
+------------------------------------------------------------------------
+r20921 | felicity | 2004-06-08 20:58:51 +0000 (Tue, 08 Jun 2004) | 1 line
+
+bug 3295: FORGED_DEF_WHITELIST and FORGED_IN_WHITELIST will fail in non-forged situations like when mails are received via mailing list. reverting to test mode for now.
+------------------------------------------------------------------------
+r20920 | felicity | 2004-06-08 20:49:51 +0000 (Tue, 08 Jun 2004) | 1 line
+
+bug 3485: missing meta dependencies... __OUTLOOK_MUA was renamed __ANY_OUTLOOK_MUA, and __DRUGS_ERECTILE7 doesn't exist. fixed the first, removed the second.
+------------------------------------------------------------------------
+r20919 | felicity | 2004-06-08 20:42:47 +0000 (Tue, 08 Jun 2004) | 1 line
+
+bug 3483: the fix to 3345 caused the journal to not sync to the DB in the case where scan wasn't possible. we now allow the sync to occur if the db is accessible only, which fixes both 3345 and 3483.
+------------------------------------------------------------------------
+r20897 | quinlan | 2004-06-08 05:47:34 +0000 (Tue, 08 Jun 2004) | 2 lines
+
+minor comment change
+
+------------------------------------------------------------------------
+r20896 | jm | 2004-06-08 03:39:49 +0000 (Tue, 08 Jun 2004) | 1 line
+
+bug 3391: CPAN familiarity required during INSTALL, doco fixes
+------------------------------------------------------------------------
+r20892 | jm | 2004-06-08 02:34:06 +0000 (Tue, 08 Jun 2004) | 1 line
+
+bug 3479: reinstate SPF_PASS, SPF_HELO_PASS, but with very small scores
+------------------------------------------------------------------------
+r20885 | felicity | 2004-06-07 20:41:15 +0000 (Mon, 07 Jun 2004) | 1 line
+
+small typo
+------------------------------------------------------------------------
+r20884 | felicity | 2004-06-07 20:25:46 +0000 (Mon, 07 Jun 2004) | 1 line
+
+remove more metachars before doing diff
+------------------------------------------------------------------------
+r20883 | quinlan | 2004-06-07 19:28:18 +0000 (Mon, 07 Jun 2004) | 2 lines
+
+minor fix
+
+------------------------------------------------------------------------
+r20844 | jm | 2004-06-06 02:50:42 +0000 (Sun, 06 Jun 2004) | 1 line
+
+doco fixes: a few 'its' cleared up
+------------------------------------------------------------------------
+r20843 | felicity | 2004-06-06 02:05:01 +0000 (Sun, 06 Jun 2004) | 1 line
+
+bug 3474: spamd was doing getpwnam() instead of getgrnam() for some reason. doh...
+------------------------------------------------------------------------
+r20839 | felicity | 2004-06-05 19:06:33 +0000 (Sat, 05 Jun 2004) | 1 line
+
+bug 3446: the spamd children would inherit the parent's signal handler which caused lockups among other issues. also, if max-conn-per-child is 1, don't bother copying the configuration around since each child will only process 1 message.
+------------------------------------------------------------------------
+r20836 | parker | 2004-06-05 15:50:10 +0000 (Sat, 05 Jun 2004) | 1 line
+
+Bug 3476 change -u to -U for psql command
+------------------------------------------------------------------------
+r20826 | jm | 2004-06-05 01:16:17 +0000 (Sat, 05 Jun 2004) | 1 line
+
+bug 3438: conflicting API documentation in PerMsgLearner. fixed
+------------------------------------------------------------------------
+r20825 | quinlan | 2004-06-04 21:30:56 +0000 (Fri, 04 Jun 2004) | 3 lines
+
+promote T_INCOMPLETE_TEMPLATE_CX to UNRESOLVED_TEMPLATE (also removed the
+/s modifier, it's not needed since there's no "." in the expression)
+
+------------------------------------------------------------------------
+r20797 | quinlan | 2004-06-03 21:57:56 +0000 (Thu, 03 Jun 2004) | 2 lines
+
+almost finished with T_INCOMPLETE_TEMPLATE*
+
+------------------------------------------------------------------------
+r20793 | felicity | 2004-06-03 19:19:51 +0000 (Thu, 03 Jun 2004) | 1 line
+
+bug 3465: typo in bayessql test got my() repeated warnings
+------------------------------------------------------------------------
+r20790 | parker | 2004-06-03 18:06:04 +0000 (Thu, 03 Jun 2004) | 1 line
+
+Bug 3415: add a warning about old versions of DBD::Pg
+------------------------------------------------------------------------
+r20771 | parker | 2004-06-03 05:38:27 +0000 (Thu, 03 Jun 2004) | 1 line
+
+Bug 3457: Remove subselect in test in bayessql.t, start using new clear_database call, also add a test for clear_database
+------------------------------------------------------------------------
+r20768 | jm | 2004-06-03 03:22:26 +0000 (Thu, 03 Jun 2004) | 1 line
+
+bug 3462: fix 'make distcheck'
+------------------------------------------------------------------------
+r20764 | quinlan | 2004-06-02 23:18:35 +0000 (Wed, 02 Jun 2004) | 4 lines
+
+promote RATWARE_RCVD_PF and RATWARE_RCVD_AT with no changes
+remove T_MANY_MX_5 and T_MANY_MX_10
+add last batch of T_INCOMPLETE_TEMPLATE* variations to finish this rule off
+
+------------------------------------------------------------------------
+r20760 | felicity | 2004-06-02 19:19:16 +0000 (Wed, 02 Jun 2004) | 1 line
+
+ADDRESS_IN_SUBJECT description was misleading (address can appear anywhere, not just at the front)
+------------------------------------------------------------------------
+r20743 | felicity | 2004-06-01 20:47:42 +0000 (Tue, 01 Jun 2004) | 1 line
+
+bug 3453: add more platform support for masses/cpucount
+------------------------------------------------------------------------
+r20723 | sidney | 2004-06-01 00:50:39 +0000 (Tue, 01 Jun 2004) | 1 line
+
+bug #3452: more correct fix using proper C idiom to suppress unreferenced formal parameter warning
+------------------------------------------------------------------------
+r20719 | sidney | 2004-05-31 22:44:55 +0000 (Mon, 31 May 2004) | 1 line
+
+Bug 3433: remove some compiler warnings from spamc build and compile with warnings on
+------------------------------------------------------------------------
+r20716 | sidney | 2004-05-31 21:41:06 +0000 (Mon, 31 May 2004) | 1 line
+
+Bug 3433: remove some compiler warnings from spamc build
+------------------------------------------------------------------------
+r20715 | mss | 2004-05-31 21:15:21 +0000 (Mon, 31 May 2004) | 2 lines
+
+bug 3433: Added a README about the script
+
+------------------------------------------------------------------------
+r20714 | mss | 2004-05-31 21:03:54 +0000 (Mon, 31 May 2004) | 2 lines
+
+bug 3433: to make it really obvious, rename it :)
+
+------------------------------------------------------------------------
+r20713 | mss | 2004-05-31 21:02:23 +0000 (Mon, 31 May 2004) | 2 lines
+
+bug 3433: wave the clusestick for people who don't read the docu
+
+------------------------------------------------------------------------
+r20712 | felicity | 2004-05-31 19:59:28 +0000 (Mon, 31 May 2004) | 1 line
+
+bug 3450: spamd shouldn't be started by default upon installation of the RPM. leave that up to the admin.
+------------------------------------------------------------------------
+r20711 | felicity | 2004-05-31 19:55:50 +0000 (Mon, 31 May 2004) | 1 line
+
+bug 3347: change the process name to make it easier to see which spamd is parent and which spamd is child
+------------------------------------------------------------------------
+r20710 | parker | 2004-05-31 19:48:15 +0000 (Mon, 31 May 2004) | 1 line
+
+bug 3381: restore did not like a spam_count or ham_count == 0
+------------------------------------------------------------------------
+r20709 | mss | 2004-05-31 19:23:01 +0000 (Mon, 31 May 2004) | 2 lines
+
+nothing fancy, just a newline
+
+------------------------------------------------------------------------
+r20708 | parker | 2004-05-31 19:17:48 +0000 (Mon, 31 May 2004) | 1 line
+
+bug 3378: add some logic to nspam_nham_change to only update spam or ham count when necessary
+------------------------------------------------------------------------
+r20705 | parker | 2004-05-31 19:10:08 +0000 (Mon, 31 May 2004) | 1 line
+
+bug 3414: fixed up schmea for postgresql
+------------------------------------------------------------------------
+r20695 | mss | 2004-05-31 16:09:35 +0000 (Mon, 31 May 2004) | 2 lines
+
+bug 3451: the LOG_* stuff is an evil hack for Windows, should be cleanup up one day
+
+------------------------------------------------------------------------
+r20694 | mss | 2004-05-31 15:32:36 +0000 (Mon, 31 May 2004) | 2 lines
+
+Not the most elegant way to add -Wall but should work for everybody with a GNU toolchain. If anybody uses something else which doesn't support -Wall, please supply a patch ;-)
+
+------------------------------------------------------------------------
+r20693 | mss | 2004-05-31 15:23:19 +0000 (Mon, 31 May 2004) | 2 lines
+
+I need to get -Wall into the default Makefile :-/
+
+------------------------------------------------------------------------
+r20692 | mss | 2004-05-31 15:21:43 +0000 (Mon, 31 May 2004) | 2 lines
+
+the fail: target is actually needed under Windows, too
+
+------------------------------------------------------------------------
+r20688 | mss | 2004-05-31 13:47:30 +0000 (Mon, 31 May 2004) | 2 lines
+
+svn:ignore blibdirs
+
+------------------------------------------------------------------------
+r20687 | mss | 2004-05-31 13:41:22 +0000 (Mon, 31 May 2004) | 2 lines
+
+bug 3431: fourth part, finally :) a conflict held this and number three back
+
+------------------------------------------------------------------------
+r20685 | quinlan | 2004-05-31 06:15:16 +0000 (Mon, 31 May 2004) | 2 lines
+
+change SPF_PASS and SPF_HELO_PASS to be prerequisite rules
+
+------------------------------------------------------------------------
+r20683 | quinlan | 2004-05-31 05:55:49 +0000 (Mon, 31 May 2004) | 2 lines
+
+removing T_SPF_PASS_NO_SBL - all of my hits are really spam
+
+------------------------------------------------------------------------
+r20682 | quinlan | 2004-05-31 05:11:02 +0000 (Mon, 31 May 2004) | 2 lines
+
+T_MANY_MX* variations not as good as I thought they would be
+
+------------------------------------------------------------------------
+r20681 | quinlan | 2004-05-31 05:08:39 +0000 (Mon, 31 May 2004) | 2 lines
+
+fix T_RATWARE_RCVD_PF_1
+
+------------------------------------------------------------------------
+r20679 | quinlan | 2004-05-31 04:46:03 +0000 (Mon, 31 May 2004) | 4 lines
+
+remove T_REDIR_AOL_CLICKTHRU and T_SAVE_BUCKS
+add T_MANY_MX_* for EnvelopeFrom instead of From:addr
+fix description for T_DNS_FROM_SECURITYSAGE
+
+------------------------------------------------------------------------
+r20678 | quinlan | 2004-05-31 04:32:23 +0000 (Mon, 31 May 2004) | 2 lines
+
+typo
+
+------------------------------------------------------------------------
+r20677 | quinlan | 2004-05-31 04:30:14 +0000 (Mon, 31 May 2004) | 2 lines
+
+one of the SenderBase rules not working great in wider testing, so remove it
+
+------------------------------------------------------------------------
+r20676 | quinlan | 2004-05-31 04:13:41 +0000 (Mon, 31 May 2004) | 2 lines
+
+remove -firsttrusted test rules
+
+------------------------------------------------------------------------
+r20675 | quinlan | 2004-05-31 04:08:34 +0000 (Mon, 31 May 2004) | 2 lines
+
+some descriptions
+
+------------------------------------------------------------------------
+r20672 | quinlan | 2004-05-31 02:08:39 +0000 (Mon, 31 May 2004) | 4 lines
+
+switch RCVD_IN_XBL, RCVD_IN_RFC_IPWHOIS, and RCVD_IN_DSBL to use
+ -notfirsthop instead of -firsttrusted since it preserves the original hit
+ rate much better with most or all of the FP benefit of -firstttrusted
+
+------------------------------------------------------------------------
+r20664 | sidney | 2004-05-30 23:50:00 +0000 (Sun, 30 May 2004) | 1 line
+
+bug #3443: fix race conditions that left spamd zombies durin make test
+------------------------------------------------------------------------
+r20662 | quinlan | 2004-05-30 21:21:15 +0000 (Sun, 30 May 2004) | 2 lines
+
+fix error that killed RCVD_IN_XBL hits for the weekly run
+
+------------------------------------------------------------------------
+r20659 | quinlan | 2004-05-30 19:46:46 +0000 (Sun, 30 May 2004) | 2 lines
+
+remove T_RCVD_IN_RFC_IPWHOIS - already got promoted, not needed
+
+------------------------------------------------------------------------
+r20658 | quinlan | 2004-05-30 19:45:46 +0000 (Sun, 30 May 2004) | 2 lines
+
+remove T_RCVD_IN_XBL and T_RCVD_IN_DSBL - not needed now
+
+------------------------------------------------------------------------
+r20619 | duncf | 2004-05-30 01:14:35 +0000 (Sun, 30 May 2004) | 1 line
+
+add missing files to manifest
+------------------------------------------------------------------------
+r20598 | mss | 2004-05-29 14:37:07 +0000 (Sat, 29 May 2004) | 6 lines
+
+Cleaned up the handling of networking related command line switches a bit:
+* Invalid IP-addresses/hostnames weren't handled before -- spamd was always falling back to 127.0.0.1 silently.
+* Even though the code looks different, hostnames for -i never worked. Those also fell back to 127.0.0.1 silently. Now the first available IP-address is used (or spamd dies if none is available).
+* Added feature: ports can now be specified as names, too.
+* Added feature: giving just the -i switch without a value will make spamd listen on all interfaces (0.0.0.0). That's still limited by the host filter (given by -A).
+
+------------------------------------------------------------------------
+r20580 | quinlan | 2004-05-29 06:15:56 +0000 (Sat, 29 May 2004) | 2 lines
+
+renames
+
+------------------------------------------------------------------------
+r20579 | quinlan | 2004-05-29 06:13:47 +0000 (Sat, 29 May 2004) | 2 lines
+
+renames
+
+------------------------------------------------------------------------
+r20578 | quinlan | 2004-05-29 06:12:57 +0000 (Sat, 29 May 2004) | 2 lines
+
+fix compile error
+
+------------------------------------------------------------------------
+r20574 | jm | 2004-05-29 04:04:41 +0000 (Sat, 29 May 2004) | 1 line
+
+promoted T_FAKE_HELO_MAIL_COM_DOM
+------------------------------------------------------------------------
+r20570 | quinlan | 2004-05-29 03:05:59 +0000 (Sat, 29 May 2004) | 3 lines
+
+promote Subject test for sexually-explicit
+revise header template rule
+
+------------------------------------------------------------------------
+r20567 | quinlan | 2004-05-29 02:03:29 +0000 (Sat, 29 May 2004) | 2 lines
+
+remove most of the T_MANY_MX* rules
+
+------------------------------------------------------------------------
+r20566 | quinlan | 2004-05-29 01:59:12 +0000 (Sat, 29 May 2004) | 2 lines
+
+remove T_DNS_FROM_AHBL_RHSBL - no real improvement with EnvelopeFrom
+
+------------------------------------------------------------------------
+r20565 | quinlan | 2004-05-29 01:57:39 +0000 (Sat, 29 May 2004) | 5 lines
+
+change the three DNSBLS that improved the most via -firsttrusted logic
+ to use it: RCVD_IN_XBL, RCVD_IN_DSBL
+promote all of the RFC-Ignorant EnvelopeFrom rules, delete the old
+ RFC-Ignorant From: rules
+
+------------------------------------------------------------------------
+r20564 | jm | 2004-05-29 01:22:13 +0000 (Sat, 29 May 2004) | 1 line
+
+'spamtrap' commands are no longer supported; remove them from the Conf.pm parsing code
+------------------------------------------------------------------------
+r20563 | jm | 2004-05-29 01:18:17 +0000 (Sat, 29 May 2004) | 1 line
+
+bug 2987: finish reordering Conf items to put frequently-used user options at top. also took the opportunity to fix up some of the text, too
+------------------------------------------------------------------------
+r20562 | jm | 2004-05-29 00:50:22 +0000 (Sat, 29 May 2004) | 1 line
+
+why was this using these modules? that doesn't work
+------------------------------------------------------------------------
+r20560 | jm | 2004-05-29 00:42:21 +0000 (Sat, 29 May 2004) | 1 line
+
+bug 3431: other third of the patch applied ;)
+------------------------------------------------------------------------
+r20559 | quinlan | 2004-05-28 23:52:51 +0000 (Fri, 28 May 2004) | 4 lines
+
+switch NJABL dynamic test to use new zone
+rename RCVD_IN_NJABL_DIALUP to RCVD_IN_NJABL_DUL to match other rules
+(maybe they should all be named *DYN instead, though)
+
+------------------------------------------------------------------------
+r20555 | mss | 2004-05-28 22:27:43 +0000 (Fri, 28 May 2004) | 2 lines
+
+bug 3431: added tags _REMOTEHOSTNAME_ and _REMOTEHOSTADDR_
+
+------------------------------------------------------------------------
+r20554 | mss | 2004-05-28 22:20:51 +0000 (Fri, 28 May 2004) | 2 lines
+
+bug 3441: completing the fix (note to myself: vi doesn't always DWIM)
+
+------------------------------------------------------------------------
+r20552 | mss | 2004-05-28 21:47:40 +0000 (Fri, 28 May 2004) | 2 lines
+
+bug 3441: make sure the string we free was allocated by us.
+
+------------------------------------------------------------------------
+r20551 | felicity | 2004-05-28 20:12:05 +0000 (Fri, 28 May 2004) | 1 line
+
+3.0.0 is really 3.000000, so fix it in the spec file
+------------------------------------------------------------------------
+r20550 | felicity | 2004-05-28 20:02:01 +0000 (Fri, 28 May 2004) | 1 line
+
+bug 3340: new conf code has a multi-level hash which has CODE references. skip that in the config copy function.
+------------------------------------------------------------------------
+r20549 | quinlan | 2004-05-28 19:40:58 +0000 (Fri, 28 May 2004) | 2 lines
+
+update SenderBase tests to use revised subtest format
+
+------------------------------------------------------------------------
+r20548 | quinlan | 2004-05-28 19:39:04 +0000 (Fri, 28 May 2004) | 6 lines
+
+revise and update the DNSBL documentation
+change SenderBase tests to be prefixed with "sb:" instead of relying on
+ the set name to distinguish them
+lower magnitude cut-off for SB_NSP_VOLUME_SPIKE based on test rule results
+ and remove those test rules
+
+------------------------------------------------------------------------
+r20537 | jm | 2004-05-28 07:26:43 +0000 (Fri, 28 May 2004) | 1 line
+
+oops; test failure
+------------------------------------------------------------------------
+r20536 | jm | 2004-05-28 07:25:45 +0000 (Fri, 28 May 2004) | 1 line
+
+moved another method, and some documentation, over to new Parser class. also had to convert some constants back into vars for the move; possible TODO: move them to Constants.pm
+------------------------------------------------------------------------
+r20529 | jm | 2004-05-28 07:06:55 +0000 (Fri, 28 May 2004) | 1 line
+
+Conf reorg continued; broke out parsing code into a separate parsing class. also fixed some bugs that 'make test' hadn't shown up; however 'make test' is failing in t/spamd_allow_user_rules.t due to spamc coredumping when the -u switch is used ;)
+------------------------------------------------------------------------
+r20528 | jm | 2004-05-28 07:06:33 +0000 (Fri, 28 May 2004) | 1 line
+
+Conf reorg continued; broke out parsing code into a separate parsing class. also fixed some bugs that 'make test' hadn't shown up; however 'make test' is failing in t/spamd_allow_user_rules.t due to spamc coredumping when the -u switch is used ;)
+------------------------------------------------------------------------
+r20527 | jm | 2004-05-28 07:00:20 +0000 (Fri, 28 May 2004) | 1 line
+
+removed some not-so-hot ratware rules
+------------------------------------------------------------------------
+r20526 | jm | 2004-05-28 05:36:53 +0000 (Fri, 28 May 2004) | 1 line
+
+bug 2987: first part of Conf reorg. adopt new model which allows us to move conf items around. also faster: 0m23.763s for 100 --lint runs, compared with 0m26.861s for the old style
+------------------------------------------------------------------------
+r20522 | mss | 2004-05-27 23:00:45 +0000 (Thu, 27 May 2004) | 2 lines
+
+ANSI comment :)
+
+------------------------------------------------------------------------
+r20521 | mss | 2004-05-27 22:42:09 +0000 (Thu, 27 May 2004) | 2 lines
+
+EX_NOTSPAM == EX_OK *grmbl*
+
+------------------------------------------------------------------------
+r20520 | mss | 2004-05-27 22:36:33 +0000 (Thu, 27 May 2004) | 3 lines
+
+* 'goto finish' is the way to go because else we miss WSACleanup() on Windows
+* Removed a FIXME added by me, added a descriptive comment instead.
+
+------------------------------------------------------------------------
+r20519 | mss | 2004-05-27 22:24:43 +0000 (Thu, 27 May 2004) | 4 lines
+
+* Refactored the username determination into its own subroutine. This actually fixed a bug on Windows as WSACleanup() is now called on failure.
+* Made the memory for username allocated dynamically. Why always use 255 bytes with some weird buffer handling when often a few bytes are enough?
+* Got rid of an ifdef-bug on Windows I introduced before.
+
+------------------------------------------------------------------------
+r20518 | mss | 2004-05-27 21:22:45 +0000 (Thu, 27 May 2004) | 5 lines
+
+Some code cleanup:
+* Mostly fixed indentations and spacing.
+* Replaced two simple cases with guard clauses instead of if() or #ifdef
+* Replaced the double WSACleanup() with a goto. This is a case where gotos aren't evil :)
+
+------------------------------------------------------------------------
+r20515 | mss | 2004-05-27 19:12:53 +0000 (Thu, 27 May 2004) | 2 lines
+
+Some more help/usage/versioning cleanup.
+
+------------------------------------------------------------------------
+r20503 | mss | 2004-05-27 12:48:55 +0000 (Thu, 27 May 2004) | 2 lines
+
+Added some maintenance stuff I forgot for the spamc versioning.
+
+------------------------------------------------------------------------
+r20500 | mss | 2004-05-27 12:35:47 +0000 (Thu, 27 May 2004) | 2 lines
+
+Added often-requested -V switch to spamc. Finally one can check out which version of spamc is installed on his box :)
+
+------------------------------------------------------------------------
+r20497 | mss | 2004-05-27 11:07:12 +0000 (Thu, 27 May 2004) | 4 lines
+
+* spamc now returns success if it was called with -h (like any good tool).
+* In main, the return value of read_args is honored.
+* Added a missing newline in the help I accidently removed before.
+
+------------------------------------------------------------------------
+r20489 | mss | 2004-05-26 21:19:27 +0000 (Wed, 26 May 2004) | 4 lines
+
+Rearranged the help (again?), make it break before 80 chars.
+
+Also added a few empty lines and a FALLTHROUGH to make the code easier to read.
+
+------------------------------------------------------------------------
+r20471 | quinlan | 2004-05-26 12:09:17 +0000 (Wed, 26 May 2004) | 3 lines
+
+remove low TTL domains
+add some new domains
+
+------------------------------------------------------------------------
+r20452 | duncf | 2004-05-26 04:04:29 +0000 (Wed, 26 May 2004) | 1 line
+
+Added UPGRADE to MANIFEST. Damn MANIFEST
+------------------------------------------------------------------------
+r20451 | jm | 2004-05-26 01:44:24 +0000 (Wed, 26 May 2004) | 1 line
+
+comment about TTLs
+------------------------------------------------------------------------
+r20450 | jm | 2004-05-26 01:36:01 +0000 (Wed, 26 May 2004) | 1 line
+
+bug 3419: use NS queries instead of MX to lessen load
+------------------------------------------------------------------------
+r20449 | jm | 2004-05-26 01:25:12 +0000 (Wed, 26 May 2004) | 1 line
+
+bug 3320: insecure setuid eval possibly worked around using a similar fix to bug 3325
+------------------------------------------------------------------------
+r20444 | quinlan | 2004-05-25 19:39:19 +0000 (Tue, 25 May 2004) | 5 lines
+
+do some checking for T_HAS_TEMPLATE* rules since we're not catching all
+ the ham yet
+remove T_SPECIAL_BY_CHOICE* because it sucks
+add some more T_MANY_MX variants
+
+------------------------------------------------------------------------
+r20428 | mss | 2004-05-25 11:18:43 +0000 (Tue, 25 May 2004) | 2 lines
+
+Tiny docu update.
+
+------------------------------------------------------------------------
+r20372 | jm | 2004-05-24 19:22:26 +0000 (Mon, 24 May 2004) | 1 line
+
+some code from the weekend; fix UTF-8 warnings due to UTF-8-tainted data from HTML-Parser, and some interesting new ratware Received header rules
+------------------------------------------------------------------------
+r20371 | jm | 2004-05-24 19:14:14 +0000 (Mon, 24 May 2004) | 1 line
+
+clarified Duncan's clarification
+------------------------------------------------------------------------
+r20364 | quinlan | 2004-05-24 17:26:23 +0000 (Mon, 24 May 2004) | 2 lines
+
+add new attempt at template rule
+
+------------------------------------------------------------------------
+r20283 | duncf | 2004-05-24 03:42:04 +0000 (Mon, 24 May 2004) | 5 lines
+
+More documentation changes. Created "UPGRADE" to deal with upgrade
+notes rather than sticking them in INSTALL. This is useful for Debian
+since they've already installed SA, so we don't need to show them how;
+yet upgrade issues are still relavent.
+
+------------------------------------------------------------------------
+r20282 | duncf | 2004-05-24 03:32:33 +0000 (Mon, 24 May 2004) | 2 lines
+
+Clarified user_prefs stuff. Was a debian patch
+
+------------------------------------------------------------------------
+r20280 | duncf | 2004-05-24 03:23:05 +0000 (Mon, 24 May 2004) | 3 lines
+
+Added note about anti-From munging thingy in procmailrc.example
+I'd had this as a debian specific patch before, but that's unnecessary!
+
+------------------------------------------------------------------------
+r20274 | duncf | 2004-05-24 01:51:38 +0000 (Mon, 24 May 2004) | 5 lines
+
+Made some updates to the README -- removed some stuff referring to old
+code. Also added Apache Incubator notice (I think we're required to
+under ASF policy, someone correct me if I'm wrong. Also, I'm assuming
+Sponsor = Apache BOD)
+
+------------------------------------------------------------------------
+r20272 | felicity | 2004-05-23 23:15:44 +0000 (Sun, 23 May 2004) | 1 line
+
+bug 3422: fix FPs for RCVD_ILLEGAL_IP, only use untrusted relays, use the parsed data, not the raw received string...
+------------------------------------------------------------------------
+r20236 | quinlan | 2004-05-23 06:35:54 +0000 (Sun, 23 May 2004) | 2 lines
+
+this needs to be case-insensitive
+
+------------------------------------------------------------------------
+r20232 | quinlan | 2004-05-23 04:47:30 +0000 (Sun, 23 May 2004) | 3 lines
+
+be less restrictive about score ranges and mutability (mutability is
+probably what makes the most difference)
+
+------------------------------------------------------------------------
+r20228 | quinlan | 2004-05-22 21:05:32 +0000 (Sat, 22 May 2004) | 2 lines
+
+alphabetize
+
+------------------------------------------------------------------------
+r20227 | quinlan | 2004-05-22 21:04:29 +0000 (Sat, 22 May 2004) | 2 lines
+
+formatting
+
+------------------------------------------------------------------------
+r20226 | quinlan | 2004-05-22 21:04:04 +0000 (Sat, 22 May 2004) | 2 lines
+
+remove 58/8 and 59/8 - no longer reserved
+
+------------------------------------------------------------------------
+r20225 | quinlan | 2004-05-22 20:53:21 +0000 (Sat, 22 May 2004) | 2 lines
+
+update list of well-connected domains (see bug 3419)
+
+------------------------------------------------------------------------
+r20217 | felicity | 2004-05-22 16:56:36 +0000 (Sat, 22 May 2004) | 1 line
+
+bug 3345: sometimes for a bayes scan we can't tie the db, so we will skip the scan but still try the opportunistic calls which expect a tied db... we will now only do the opportunistic calls if the scan completes sucessfully.
+------------------------------------------------------------------------
+r20213 | parker | 2004-05-22 14:55:32 +0000 (Sat, 22 May 2004) | 1 line
+
+Bug 3377: Remove trailing comma in table create
+------------------------------------------------------------------------
+r20206 | quinlan | 2004-05-22 05:17:41 +0000 (Sat, 22 May 2004) | 2 lines
+
+try using -firsttrusted for most DNSBL tests
+
+------------------------------------------------------------------------
+r20205 | quinlan | 2004-05-22 05:16:53 +0000 (Sat, 22 May 2004) | 2 lines
+
+remove OPM and an old comment
+
+------------------------------------------------------------------------
+r20204 | quinlan | 2004-05-22 03:57:21 +0000 (Sat, 22 May 2004) | 3 lines
+
+add rule for sexually-explicit tag
+add new version of XBL test that should FP less
+
+------------------------------------------------------------------------
+r20203 | quinlan | 2004-05-22 01:53:29 +0000 (Sat, 22 May 2004) | 2 lines
+
+performance tweaks (21% faster)
+
+------------------------------------------------------------------------
+r20202 | felicity | 2004-05-21 23:52:43 +0000 (Fri, 21 May 2004) | 1 line
+
+rh bug 123739: -P is still documented, even though it's been removed from the parameter parsing section.
+------------------------------------------------------------------------
+r20201 | quinlan | 2004-05-21 21:43:10 +0000 (Fri, 21 May 2004) | 5 lines
+
+rename CT_BOUND_DDNUM to MIME_BOUND_DD_DIGITS per convention
+remove some poor test rules
+revise T_OFFER_BY_CHOICE* to T_SPECIAL_BY_CHOICE*
+promote MSGID_YAHOO_CAPS and MSGID_SPAM_CAPS (bugs 3395 and 3399)
+
+------------------------------------------------------------------------
+r20200 | mss | 2004-05-21 19:08:20 +0000 (Fri, 21 May 2004) | 2 lines
+
+Added ->finish() calls to SYNOPSIS
+
+------------------------------------------------------------------------
+r20199 | mss | 2004-05-21 19:01:32 +0000 (Fri, 21 May 2004) | 2 lines
+
+Removed some code duplication, now Message.pm is used to parse mail headers. Also made it possible to free some mem after the mail is parsed.
+
+------------------------------------------------------------------------
+r20186 | felicity | 2004-05-21 06:47:30 +0000 (Fri, 21 May 2004) | 1 line
+
+fix long descriptions
+------------------------------------------------------------------------
+r20185 | felicity | 2004-05-21 06:44:37 +0000 (Fri, 21 May 2004) | 1 line
+
+put in a default score for CT_BOUND_DDNUM ...
+------------------------------------------------------------------------
+r20178 | felicity | 2004-05-21 05:58:40 +0000 (Fri, 21 May 2004) | 1 line
+
+promote CT_BOUND_DDNUM ... exceptional rule, at least for now...
+------------------------------------------------------------------------
+r20177 | duncf | 2004-05-21 04:20:54 +0000 (Fri, 21 May 2004) | 1 line
+
+Apparently my wishlist is a 404 -- if you want to buy me stuff, just send me an e-mail :-)
+------------------------------------------------------------------------
+r20175 | quinlan | 2004-05-21 02:47:26 +0000 (Fri, 21 May 2004) | 2 lines
+
+some T_MANY_MX* tests
+
+------------------------------------------------------------------------
+r20174 | yackley | 2004-05-21 02:10:51 +0000 (Fri, 21 May 2004) | 1 line
+
+Updated fraud tests, updated many_mx testes, removed tests for bug: 3349
+------------------------------------------------------------------------
+r20173 | quinlan | 2004-05-21 01:49:30 +0000 (Fri, 21 May 2004) | 2 lines
+
+bug 3410: add envelope-from testing for RFC-ignorant (much better results!)
+
+------------------------------------------------------------------------
+r20172 | quinlan | 2004-05-21 01:48:36 +0000 (Fri, 21 May 2004) | 2 lines
+
+no -rhs needed since the non-RHS blacklist was dropped
+
+------------------------------------------------------------------------
+r20171 | quinlan | 2004-05-20 19:15:51 +0000 (Thu, 20 May 2004) | 2 lines
+
+some code formatting clean-up
+
+------------------------------------------------------------------------
+r20170 | quinlan | 2004-05-20 19:13:03 +0000 (Thu, 20 May 2004) | 4 lines
+
+add check_rbl_envfrom for RFCI DSN test (bug 3410) and add
+ T_DNS_FROM_RFCI_DSN to tests it
+some code formatting clean-up
+
+------------------------------------------------------------------------
+r20169 | felicity | 2004-05-20 18:21:50 +0000 (Thu, 20 May 2004) | 1 line
+
+whowhere.com seems to be everyone else's FPs for the MSGID_CAPS rules, so add them to the negative look-ahead ...
+------------------------------------------------------------------------
+r20168 | quinlan | 2004-05-20 18:11:29 +0000 (Thu, 20 May 2004) | 2 lines
+
+new version of OBSCURED_EMAIL
+
+------------------------------------------------------------------------
+r20167 | quinlan | 2004-05-20 18:01:35 +0000 (Thu, 20 May 2004) | 3 lines
+
+add some T_MSGID_CAPS* variants (bug 3399)
+remove T_HEXOCTDWORD* due to poor results (bug 3403)
+
+------------------------------------------------------------------------
+r10719 | felicity | 2004-05-19 17:24:08 +0000 (Wed, 19 May 2004) | 1 line
+
+circular meta dependencies ought to trigger a --lint warning. added a --lint generic 'something is wrong' message. instead of exiting with # of errors, exit 1 if errors, 0 if not. (don't overflow the exit code...)
+------------------------------------------------------------------------
+r10714 | yackley | 2004-05-19 03:33:13 +0000 (Wed, 19 May 2004) | 1 line
+
+Fixed test fraud rule typo
+------------------------------------------------------------------------
+r10713 | yackley | 2004-05-19 03:29:13 +0000 (Wed, 19 May 2004) | 1 line
+
+Updated test fraud rules
+------------------------------------------------------------------------
+r10712 | quinlan | 2004-05-18 09:00:10 +0000 (Tue, 18 May 2004) | 3 lines
+
+add option to only read first N lines (useful for doing comparisons
+between old mass-check logs and new ones if you're impatient)
+
+------------------------------------------------------------------------
+r10711 | quinlan | 2004-05-18 08:58:33 +0000 (Tue, 18 May 2004) | 2 lines
+
+default score set should be 0
+
+------------------------------------------------------------------------
+r10710 | quinlan | 2004-05-18 08:54:19 +0000 (Tue, 18 May 2004) | 3 lines
+
+some test rule churn
+bug 3403: rule for detecting encoded IP addresses
+
+------------------------------------------------------------------------
+r10709 | quinlan | 2004-05-18 08:53:22 +0000 (Tue, 18 May 2004) | 2 lines
+
+lower DOMAIN_RATIO threshold a bit
+
+------------------------------------------------------------------------
+r10708 | mss | 2004-05-18 07:47:00 +0000 (Tue, 18 May 2004) | 2 lines
+
+Consistently prefix all internal function with an underscore.
+
+------------------------------------------------------------------------
+r10707 | yackley | 2004-05-18 05:51:06 +0000 (Tue, 18 May 2004) | 1 line
+
+Updated test fraud rules
+------------------------------------------------------------------------
+r10706 | mss | 2004-05-17 22:58:23 +0000 (Mon, 17 May 2004) | 2 lines
+
+bug 3353: this is the correct fix, hopefully
+
+------------------------------------------------------------------------
+r10705 | mss | 2004-05-17 22:35:00 +0000 (Mon, 17 May 2004) | 2 lines
+
+bug 3354: the local[0-9] syslog facilities didn't work because [0-9] are non-word characters (?)
+
+------------------------------------------------------------------------
+r10703 | felicity | 2004-05-16 23:01:51 +0000 (Sun, 16 May 2004) | 1 line
+
+bug 3399: more suggested message-id based rules
+------------------------------------------------------------------------
+r10702 | yackley | 2004-05-16 21:36:15 +0000 (Sun, 16 May 2004) | 1 line
+
+Bug 3374: Promoted DRUG_DOSAGE to 20_drugs.cf
+------------------------------------------------------------------------
+r10701 | quinlan | 2004-05-16 20:25:17 +0000 (Sun, 16 May 2004) | 2 lines
+
+get most of the tenpass stuff to the point where it runs
+
+------------------------------------------------------------------------
+r10700 | quinlan | 2004-05-16 09:51:41 +0000 (Sun, 16 May 2004) | 2 lines
+
+promote a few of the SenderBase rules that have been in testing
+
+------------------------------------------------------------------------
+r10699 | quinlan | 2004-05-16 09:10:58 +0000 (Sun, 16 May 2004) | 2 lines
+
+add <th> and <dt> as whitespace tags (reduces BACKHAIR_* hits some)
+
+------------------------------------------------------------------------
+r10692 | felicity | 2004-05-16 08:30:36 +0000 (Sun, 16 May 2004) | 1 line
+
+bug 3396: new rule suggestion, look for content-type boundary format
+------------------------------------------------------------------------
+r10691 | felicity | 2004-05-16 08:25:04 +0000 (Sun, 16 May 2004) | 1 line
+
+bug 3395: suggested rule T_YAHOO_MSGID_CAPS
+------------------------------------------------------------------------
+r10689 | felicity | 2004-05-16 08:12:54 +0000 (Sun, 16 May 2004) | 1 line
+
+update INSTALL doc to talk a little about CPAN, and how we don't support third party modules...
+------------------------------------------------------------------------
+r10687 | quinlan | 2004-05-16 06:47:23 +0000 (Sun, 16 May 2004) | 4 lines
+
+bug 3372: HTML parser confused by blank lines in paragraph
+ buffer up text until a new whitespace tag is encountered,
+ when printing the buffer, collapse whitespace reasonably correctly
+
+------------------------------------------------------------------------
+r10686 | yackley | 2004-05-16 04:52:12 +0000 (Sun, 16 May 2004) | 1 line
+
+Updated test rules for PER_DOSE and fraud
+------------------------------------------------------------------------
+r10685 | quinlan | 2004-05-15 19:27:00 +0000 (Sat, 15 May 2004) | 2 lines
+
+formatting, don't use T_ prefix for prerequisite tests
+
+------------------------------------------------------------------------
+r10684 | quinlan | 2004-05-15 19:26:05 +0000 (Sat, 15 May 2004) | 2 lines
+
+switch to using new flag for HTML_MESSAGE
+
+------------------------------------------------------------------------
+r10683 | felicity | 2004-05-15 18:52:29 +0000 (Sat, 15 May 2004) | 1 line
+
+skip using a variable for a constant in mbx code, also change 'foo ;' to just 'foo;'
+------------------------------------------------------------------------
+r10681 | mss | 2004-05-15 15:43:43 +0000 (Sat, 15 May 2004) | 2 lines
+
+This makes more sense.
+
+------------------------------------------------------------------------
+r10680 | mss | 2004-05-15 13:19:29 +0000 (Sat, 15 May 2004) | 2 lines
+
+bug 3383: added support for folded Message-Ids
+
+------------------------------------------------------------------------
+r10679 | quinlan | 2004-05-15 11:10:06 +0000 (Sat, 15 May 2004) | 3 lines
+
+remove some old testing code
+add html_message flag for HTML_MESSAGE
+
+------------------------------------------------------------------------
+r10678 | jm | 2004-05-15 08:24:40 +0000 (Sat, 15 May 2004) | 1 line
+
+oops, leftover test code
+------------------------------------------------------------------------
+r10672 | jm | 2004-05-15 04:49:27 +0000 (Sat, 15 May 2004) | 1 line
+
+bug 3325: Insecure dependency in Dns.pm line 214: fixed, here at least. seems to be tickling a perl bug...
+------------------------------------------------------------------------
+r10669 | yackley | 2004-05-15 04:03:51 +0000 (Sat, 15 May 2004) | 1 line
+
+Fixed typo in rule name T__FRAUD_BARRISTR
+------------------------------------------------------------------------
+r10668 | jm | 2004-05-15 03:46:10 +0000 (Sat, 15 May 2004) | 1 line
+
+oops, MIME::Base64 listed twice
+------------------------------------------------------------------------
+r10667 | jm | 2004-05-15 02:37:55 +0000 (Sat, 15 May 2004) | 1 line
+
+bug 3322: fixed LDAP driver problems when there's no username/password, thanks to Turbo Fredriksson. also updated ldap/README.testing for Debian unstable.
+------------------------------------------------------------------------
+r10666 | jm | 2004-05-15 01:38:42 +0000 (Sat, 15 May 2004) | 1 line
+
+bug 3368: URL parsing assumed mailto if scheme was missing in a href. Fixed
+------------------------------------------------------------------------
+r10665 | jm | 2004-05-15 01:33:37 +0000 (Sat, 15 May 2004) | 1 line
+
+bug 2372: added support for the parsing of mbx mailboxes (the mailbox format used within the University of Washington's IMAP implementation -- http://www.washington.edu/imap/). Thanks to John Newman for this patch
+------------------------------------------------------------------------
+r10649 | felicity | 2004-05-14 16:56:59 +0000 (Fri, 14 May 2004) | 1 line
+
+bug 3384: if a message came in with a doctype set, but no other html tags, we could get undefined/division by 0 errors when rendering/setting HTML results.
+------------------------------------------------------------------------
+r10633 | yackley | 2004-05-14 01:54:38 +0000 (Fri, 14 May 2004) | 1 line
+
+Added my fraud rule meta set
+------------------------------------------------------------------------
+r10630 | mss | 2004-05-14 00:00:01 +0000 (Fri, 14 May 2004) | 2 lines
+
+Fix up the SYNOPSIS (parse() was missing the data)
+
+------------------------------------------------------------------------
+r10629 | quinlan | 2004-05-13 23:02:15 +0000 (Thu, 13 May 2004) | 5 lines
+
+clean up anchor code - no additional ham hits, spam hits changed as follows:
+ HTML_LINK_CLICK_CAPS increased from 3.14% to 3.82%
+ HTML_LINK_CLICK_HERE increased from 15.16% to 19.71%
+ HTML_LINK_PUSH_HERE increased from 2.02% to 2.23%
+
+------------------------------------------------------------------------
+r10614 | jm | 2004-05-13 04:28:40 +0000 (Thu, 13 May 2004) | 1 line
+
+bug 3197: Dutch translation, thanks to Jesse Houwing
+------------------------------------------------------------------------
+r10613 | quinlan | 2004-05-13 03:36:39 +0000 (Thu, 13 May 2004) | 3 lines
+
+some test rules (some really added due to other changes forthcoming in
+whitespace handling, but these are safe to check in now)
+
+------------------------------------------------------------------------
+r10612 | yackley | 2004-05-13 03:34:21 +0000 (Thu, 13 May 2004) | 1 line
+
+Submitted rules for bugs 3374 & 3349, added a few more MANY_MX rules
+------------------------------------------------------------------------
+r10611 | sidney | 2004-05-13 02:25:44 +0000 (Thu, 13 May 2004) | 1 line
+
+Bug #3039 further fixes for WIN32 and remove use of %m left over from calls to syslog
+------------------------------------------------------------------------
+r10600 | jm | 2004-05-12 07:16:59 +0000 (Wed, 12 May 2004) | 1 line
+
+bug 3039: added log-to-stderr support for spamc, -l switch
+------------------------------------------------------------------------
+r10599 | jm | 2004-05-12 06:38:54 +0000 (Wed, 12 May 2004) | 1 line
+
+bug in Received; double-unpacking of IP addrs
+------------------------------------------------------------------------
+r10598 | yackley | 2004-05-12 04:17:00 +0000 (Wed, 12 May 2004) | 1 line
+
+Fixed slight typo in T_MANY_MX_4.
+------------------------------------------------------------------------
+r10597 | yackley | 2004-05-12 03:10:26 +0000 (Wed, 12 May 2004) | 1 line
+
+Added rules to check From addresses for high numbered MX hosts.
+------------------------------------------------------------------------
+r10592 | quinlan | 2004-05-11 10:02:07 +0000 (Tue, 11 May 2004) | 2 lines
+
+fix annoying warning
+
+------------------------------------------------------------------------
+r10591 | jm | 2004-05-11 07:19:44 +0000 (Tue, 11 May 2004) | 1 line
+
+added a diagnostic method to --lint, which lists the versions of various installed modules for debugging
+------------------------------------------------------------------------
+r10589 | felicity | 2004-05-11 04:46:05 +0000 (Tue, 11 May 2004) | 1 line
+
+specify which version of the Storable module is found
+------------------------------------------------------------------------
+r10583 | jm | 2004-05-10 17:54:29 +0000 (Mon, 10 May 2004) | 1 line
+
+need DB_File for this test
+------------------------------------------------------------------------
+r10577 | quinlan | 2004-05-10 08:56:14 +0000 (Mon, 10 May 2004) | 2 lines
+
+some test rules resurrected from the past
+
+------------------------------------------------------------------------
+r10574 | felicity | 2004-05-10 05:10:10 +0000 (Mon, 10 May 2004) | 1 line
+
+bug 3366: a from address of 'foo@bar...' was causing issues with Net::DNS. 1) don't bother doing an A or MX lookup on 'bar...' -- trim it to 'bar.', then notice that it's not at made up of at least 2 parts and skip. 2) in all_from_addrs, unique the list of addresses, and also deal with the 'bar...' names -- squish to 'bar.'.
+------------------------------------------------------------------------
+r10572 | quinlan | 2004-05-10 00:54:06 +0000 (Mon, 10 May 2004) | 2 lines
+
+update of SenderBase rules
+
+------------------------------------------------------------------------
+r10571 | jm | 2004-05-09 22:46:56 +0000 (Sun, 09 May 2004) | 1 line
+
+bug 3327: fix 'undefined' warning in spamd_utf8.t
+------------------------------------------------------------------------
+r10570 | jm | 2004-05-09 22:45:16 +0000 (Sun, 09 May 2004) | 1 line
+
+a couple of things that people say should never be in legit mail
+------------------------------------------------------------------------
+r10569 | quinlan | 2004-05-09 20:17:04 +0000 (Sun, 09 May 2004) | 5 lines
+
+update of corpus tools
+- more retries and reliability stuff for corpus-nightly
+- more complete crontab
+- updates to summary script
+
+------------------------------------------------------------------------
+r10565 | quinlan | 2004-05-08 08:41:24 +0000 (Sat, 08 May 2004) | 2 lines
+
+add SenderBase tests
+
+------------------------------------------------------------------------
+r10564 | jm | 2004-05-08 00:51:32 +0000 (Sat, 08 May 2004) | 1 line
+
+bug 3327: spamd_utf8.t test fails if the locale is not installed
+------------------------------------------------------------------------
+r10563 | felicity | 2004-05-07 19:43:30 +0000 (Fri, 07 May 2004) | 4 lines
+
+replace code for decode_ulong_to_ip. in a test, 100_000 iterations
+took 0.81 seconds, whereas the new version, same 100_000 iterations took
+0.55 seconds.
+
+------------------------------------------------------------------------
+r10553 | jm | 2004-05-07 05:16:17 +0000 (Fri, 07 May 2004) | 1 line
+
+bug 3009: Matt Kettler's antidrug.cf ruleset taken out of testing and added to core ruleset
+------------------------------------------------------------------------
+r10552 | jm | 2004-05-07 04:36:42 +0000 (Fri, 07 May 2004) | 1 line
+
+spam spotted in wild evading URIBL, so deal with several URI obfuscations: http://0x425c45de/, http://66.92.0x45.221/, http://1113343455/, http://slashdot.org@1113343456/ in get_uri_list
+------------------------------------------------------------------------
+r10551 | quinlan | 2004-05-07 02:20:36 +0000 (Fri, 07 May 2004) | 2 lines
+
+SA doesn't "block", so say "label" instead
+
+------------------------------------------------------------------------
+r10550 | quinlan | 2004-05-07 01:32:46 +0000 (Fri, 07 May 2004) | 3 lines
+
+replace FROM_NO_LOWER with T_FROM_NO_LOWER_3
+promote T_ADVERT_CODE to ENGLISH_UCE_SUBJECT
+
+------------------------------------------------------------------------
+r10543 | quinlan | 2004-05-06 08:56:48 +0000 (Thu, 06 May 2004) | 2 lines
+
+translations
+
+------------------------------------------------------------------------
+r10542 | quinlan | 2004-05-06 08:34:27 +0000 (Thu, 06 May 2004) | 2 lines
+
+bug 2419: merging a bunch of Bayes ranges
+
+------------------------------------------------------------------------
+r10541 | jm | 2004-05-06 05:47:53 +0000 (Thu, 06 May 2004) | 1 line
+
+bug 3241: added an implementation of an NFS-safe unlocker. Thanks to Alexis Rosen for reminding me on this. also tweaked the test scripts to avoid frequent test failures I'm getting on linux 2.6.x kernels due to un-shut-down spamd
+------------------------------------------------------------------------
+r10540 | quinlan | 2004-05-06 05:40:46 +0000 (Thu, 06 May 2004) | 2 lines
+
+so long, goodbye
+
+------------------------------------------------------------------------
+r10538 | jm | 2004-05-06 05:00:30 +0000 (Thu, 06 May 2004) | 1 line
+
+bug 3046: added --socketowner, --socketgroup, --socketmode parameters to spamd
+------------------------------------------------------------------------
+r10537 | jm | 2004-05-06 04:11:20 +0000 (Thu, 06 May 2004) | 1 line
+
+merged in newer version of antidrug.cf: v0.65
+------------------------------------------------------------------------
+r10536 | quinlan | 2004-05-06 02:36:40 +0000 (Thu, 06 May 2004) | 3 lines
+
+remove T_FROM_DELPHI
+add T_FROM_NO_LOWER_4 variant
+
+------------------------------------------------------------------------
+r10535 | quinlan | 2004-05-05 22:20:28 +0000 (Wed, 05 May 2004) | 2 lines
+
+minor typo (no effect)
+
+------------------------------------------------------------------------
+r10534 | quinlan | 2004-05-05 22:09:54 +0000 (Wed, 05 May 2004) | 3 lines
+
+need a normalization factor since number of ranks will differ between
+wanted and unwanted
+
+------------------------------------------------------------------------
+r10533 | quinlan | 2004-05-05 21:34:19 +0000 (Wed, 05 May 2004) | 3 lines
+
+new wanted/unwanted ranking system
+add -i option for IG ranking system
+
+------------------------------------------------------------------------
+r10532 | quinlan | 2004-05-05 08:59:13 +0000 (Wed, 05 May 2004) | 2 lines
+
+promote HTML size tests
+
+------------------------------------------------------------------------
+r10529 | quinlan | 2004-05-05 06:19:45 +0000 (Wed, 05 May 2004) | 2 lines
+
+bunch of rule stuff
+
+------------------------------------------------------------------------
+r10528 | quinlan | 2004-05-05 05:46:36 +0000 (Wed, 05 May 2004) | 2 lines
+
+fix SPF bug where one-word HELO hostnames were being checked
+
+------------------------------------------------------------------------
+r10527 | jm | 2004-05-05 04:53:07 +0000 (Wed, 05 May 2004) | 1 line
+
+attempt to placate Dan
+------------------------------------------------------------------------
+r10526 | quinlan | 2004-05-05 04:34:17 +0000 (Wed, 05 May 2004) | 5 lines
+
+as predicted, spammers are already using SPF in good numbers
+last week of email:
+ 5.063 2.1322 6.6538 0.243 1.00 -0.00 SPF_HELO_PASS
+ 0.625 0.5330 0.6750 0.441 0.01 -0.00 SPF_PASS
+
+------------------------------------------------------------------------
+r10521 | quinlan | 2004-05-04 23:13:40 +0000 (Tue, 04 May 2004) | 2 lines
+
+bug 3351: possible FROM_NO_LOWER replacements
+
+------------------------------------------------------------------------
+r10518 | quinlan | 2004-05-04 13:53:48 +0000 (Tue, 04 May 2004) | 2 lines
+
+typo
+
+------------------------------------------------------------------------
+r10517 | jm | 2004-05-04 02:52:52 +0000 (Tue, 04 May 2004) | 1 line
+
+flock() Locker code was using a non-blocking method where it should have been using blocking
+------------------------------------------------------------------------
+r10516 | felicity | 2004-05-03 22:55:54 +0000 (Mon, 03 May 2004) | 1 line
+
+do more error checking during header parsing. don't allow undef header values, etc. also, get_header() wasn't checking for an undef return value from header() or raw_header(), which happens if a query is made for a non-existant header.
+------------------------------------------------------------------------
+r10504 | quinlan | 2004-05-03 04:31:19 +0000 (Mon, 03 May 2004) | 2 lines
+
+more possible Senderbase rules
+
+------------------------------------------------------------------------
+r10502 | parker | 2004-05-03 02:05:12 +0000 (Mon, 03 May 2004) | 1 line
+
+Implement tok_touch_all to update useful tokens in one call
+------------------------------------------------------------------------
+r10501 | quinlan | 2004-05-03 00:18:27 +0000 (Mon, 03 May 2004) | 3 lines
+
+some rule promotions
+remove some SenderBase rules
+
+------------------------------------------------------------------------
+r10500 | quinlan | 2004-05-03 00:15:12 +0000 (Mon, 03 May 2004) | 2 lines
+
+replace X_MESSAGE_INFO with less-narrow version
+
+------------------------------------------------------------------------
+r10499 | quinlan | 2004-05-03 00:14:10 +0000 (Mon, 03 May 2004) | 2 lines
+
+replace RCVD_NUMERIC_HELO with better eval
+
+------------------------------------------------------------------------
+r10498 | quinlan | 2004-05-03 00:13:47 +0000 (Mon, 03 May 2004) | 3 lines
+
+promote HTML_TEXT_AFTER_BODY
+promote HTML_TEXT_AFTER_HTML
+
+------------------------------------------------------------------------
+r10497 | quinlan | 2004-05-02 21:51:16 +0000 (Sun, 02 May 2004) | 2 lines
+
+always start with a reasonable default configuration
+
+------------------------------------------------------------------------
+r10496 | felicity | 2004-05-02 21:34:43 +0000 (Sun, 02 May 2004) | 1 line
+
+M::SA::Received is really part of Metadata, so move it to the appropriate area. undo some metadata reference kluges. be more complete in finish() calls -- instead of just deleting certain keys, go ahead and delete them all. also get rid of the useless "$self={}" statements ("%{$self} = ()" is what was meant...)
+------------------------------------------------------------------------
+r10495 | parker | 2004-05-02 20:40:49 +0000 (Sun, 02 May 2004) | 1 line
+
+Bug 3344: added a comment
+------------------------------------------------------------------------
+r10494 | parker | 2004-05-02 20:27:44 +0000 (Sun, 02 May 2004) | 1 line
+
+Bug 3344: Avoid calling sqrt on negative number
+------------------------------------------------------------------------
+r10493 | quinlan | 2004-05-02 19:32:11 +0000 (Sun, 02 May 2004) | 2 lines
+
+T_PERCENT_RANDOM*
+
+------------------------------------------------------------------------
+r10488 | felicity | 2004-05-02 17:30:37 +0000 (Sun, 02 May 2004) | 1 line
+
+two things: move receive_date from Bayes and ArchiveIterator and put it in Util instead -- two copies of the same logic code makes no sense. add a --before option to mass-check, works in the same was as --after.
+------------------------------------------------------------------------
+r10486 | felicity | 2004-05-02 16:01:35 +0000 (Sun, 02 May 2004) | 1 line
+
+bug 3343: rewrote most of split_domain() to be more efficient and do better handling of exception cases
+------------------------------------------------------------------------
+r10485 | felicity | 2004-05-01 22:19:32 +0000 (Sat, 01 May 2004) | 1 line
+
+bug 3341: added documentation that IP::Country::Fast is required if using the optional pluging RelayCountry.
+------------------------------------------------------------------------
+r10479 | quinlan | 2004-05-01 08:42:07 +0000 (Sat, 01 May 2004) | 2 lines
+
+separate out prefix characters
+
+------------------------------------------------------------------------
+r10478 | quinlan | 2004-05-01 07:25:01 +0000 (Sat, 01 May 2004) | 2 lines
+
+add perl version and archname
+
+------------------------------------------------------------------------
+r10475 | quinlan | 2004-05-01 06:32:13 +0000 (Sat, 01 May 2004) | 2 lines
+
+just wait for the results :-)
+
+------------------------------------------------------------------------
+r10474 | felicity | 2004-05-01 06:01:35 +0000 (Sat, 01 May 2004) | 1 line
+
+found that get_uri_list() can get called before check() which is what sets PerMsgStatus->{html}, which get_uri_list() was using to get the HTML parsed uris. modified get_uri_list() to use the metadata directly, which is set. also added a test to look for this type of thing. also moved the scope of a variable in html_uri to where it was used, not the whole function.
+------------------------------------------------------------------------
+r10473 | quinlan | 2004-05-01 05:27:21 +0000 (Sat, 01 May 2004) | 2 lines
+
+add parsing for a HTTP format
+
+------------------------------------------------------------------------
+r10472 | felicity | 2004-05-01 03:49:28 +0000 (Sat, 01 May 2004) | 1 line
+
+urls don't have leading/trailing whitespace
+------------------------------------------------------------------------
+r10471 | felicity | 2004-05-01 03:41:01 +0000 (Sat, 01 May 2004) | 1 line
+
+uri_to_domain was letting a bunch of stuff through that it really shouldn't -- terms with whitespace, single 'word' domains (www), etc.
+------------------------------------------------------------------------
+r10468 | jm | 2004-05-01 02:11:41 +0000 (Sat, 01 May 2004) | 1 line
+
+more comprehensive handling of the (sick, sick, sick) rules for TLD delegation -- now deals with 3 and 4 level TLDs correctly
+------------------------------------------------------------------------
+r10462 | quinlan | 2004-05-01 00:41:01 +0000 (Sat, 01 May 2004) | 2 lines
+
+add T_X_MESSAGE_INFO just in case a + is not always there
+
+------------------------------------------------------------------------
+r10456 | quinlan | 2004-04-30 21:59:07 +0000 (Fri, 30 Apr 2004) | 3 lines
+
+T_HTML_TEXT_AFTER_HTML and T_HTML_TEXT_AFTER_HTML test rules, remove
+code and rules later if it doesn't hit enough spam to matter
+
+------------------------------------------------------------------------
+r10453 | quinlan | 2004-04-30 19:27:59 +0000 (Fri, 30 Apr 2004) | 2 lines
+
+add version requirement -- doesn't seem to trigger with t/spf.t though
+
+------------------------------------------------------------------------
+r10449 | quinlan | 2004-04-30 08:46:16 +0000 (Fri, 30 Apr 2004) | 2 lines
+
+this hurts so bad
+
+------------------------------------------------------------------------
+r10448 | quinlan | 2004-04-30 08:45:01 +0000 (Fri, 30 Apr 2004) | 2 lines
+
+removing RCVD_IN_AHBL_*
+
+------------------------------------------------------------------------
+r10447 | quinlan | 2004-04-30 07:28:52 +0000 (Fri, 30 Apr 2004) | 3 lines
+
+promote T_X_MESSAGE_INFO to X_MESSAGE_INFO
+promote T_REDIRECTOR to URI_REDIRECTOR
+
+------------------------------------------------------------------------
+r10446 | quinlan | 2004-04-30 07:22:23 +0000 (Fri, 30 Apr 2004) | 2 lines
+
+just say no
+
+------------------------------------------------------------------------
+r10439 | quinlan | 2004-04-30 03:59:37 +0000 (Fri, 30 Apr 2004) | 2 lines
+
+might as well do rawbody too
+
+------------------------------------------------------------------------
+r10438 | quinlan | 2004-04-30 02:28:40 +0000 (Fri, 30 Apr 2004) | 2 lines
+
+revise RND rules
+
+------------------------------------------------------------------------
+r10434 | felicity | 2004-04-29 21:36:51 +0000 (Thu, 29 Apr 2004) | 1 line
+
+backup more per-user state when running spamd w/ setuid. also, set setuid_to_user more appropriately for what the code actually does.
+------------------------------------------------------------------------
+r10425 | felicity | 2004-04-29 16:25:06 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 3329: modify debug output to be clearer that there are multiple queries per domain -- we're not querying the same thing multiple times
+------------------------------------------------------------------------
+r10423 | felicity | 2004-04-29 07:19:10 +0000 (Thu, 29 Apr 2004) | 1 line
+
+for some extra debugging information, let us know what version of Net::DNS is installed if we find it installed. :)
+------------------------------------------------------------------------
+r10422 | felicity | 2004-04-29 07:12:25 +0000 (Thu, 29 Apr 2004) | 1 line
+
+some conf variables can be undef, so don't try to back those up when doing compile_now() stuff...
+------------------------------------------------------------------------
+r10420 | jm | 2004-04-29 06:57:50 +0000 (Thu, 29 Apr 2004) | 1 line
+
+some leftovers of dev code
+------------------------------------------------------------------------
+r10417 | felicity | 2004-04-29 06:33:13 +0000 (Thu, 29 Apr 2004) | 1 line
+
+when doing DB upgrades, we need to make sure the lock doesn't go stale, so periodically (once every 1000 tokens), update the lock.
+------------------------------------------------------------------------
+r10413 | felicity | 2004-04-29 05:55:15 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 3064: added 'use_bayes_rules' config option to disable the bayes rules in an easy-to-use option. lets people disable checking while still allowing learning/db repair tools/etc.
+------------------------------------------------------------------------
+r10412 | felicity | 2004-04-29 05:29:15 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 3296: lower the default number of connections per child to 200
+------------------------------------------------------------------------
+r10410 | quinlan | 2004-04-29 05:18:21 +0000 (Thu, 29 Apr 2004) | 2 lines
+
+use standard naming
+
+------------------------------------------------------------------------
+r10409 | quinlan | 2004-04-29 05:17:16 +0000 (Thu, 29 Apr 2004) | 3 lines
+
+promote RCVD_IN_AHBL_RHSBL
+change __RCVD_NUMERIC_HELO to use an eval ignoring private address spaces
+
+------------------------------------------------------------------------
+r10408 | felicity | 2004-04-29 05:06:14 +0000 (Thu, 29 Apr 2004) | 1 line
+
+fix spamc_E test -- spamd-related tests should only stop spamd once, and spamd will get stopped if spamc returns an error, so put the failure one at the end.
+------------------------------------------------------------------------
+r10407 | felicity | 2004-04-29 04:59:04 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 2762: if spamd is restarted, and a file needs to get written (AWL or Bayes for instance) which is configured to be outside the default area (ie: awl_path /var/foo/bar/blah), the file will be created by the user launching spamd (root in most cases) which could end up being unaccessible by children processes. in compile_now(), we backup the default configuration values which include __userstate__, reset the defaults after init(), then after the whole check() completes, reset to the init() values. this lets spamd (and any other caller, optionally,) put all the startup files in a temp area.
+------------------------------------------------------------------------
+r10405 | jm | 2004-04-29 04:40:47 +0000 (Thu, 29 Apr 2004) | 1 line
+
+Rank replaced by information gain, defined as 'the average reduction in the entropy of C (classification) given the value of X (the rule)'.
+------------------------------------------------------------------------
+r10404 | quinlan | 2004-04-29 04:39:45 +0000 (Thu, 29 Apr 2004) | 2 lines
+
+some HTML rules
+
+------------------------------------------------------------------------
+r10401 | felicity | 2004-04-29 03:43:44 +0000 (Thu, 29 Apr 2004) | 1 line
+
+update comment
+------------------------------------------------------------------------
+r10400 | jm | 2004-04-29 03:15:28 +0000 (Thu, 29 Apr 2004) | 1 line
+
+calling unlock multiple times is OK, allow that without a warning
+------------------------------------------------------------------------
+r10399 | jm | 2004-04-29 02:56:01 +0000 (Thu, 29 Apr 2004) | 1 line
+
+--deencap didn't deal with the case where a local scanner overwrote the remote scanner's X-Spam-Checker-Version header. fix by grovelling through the body parts :(
+------------------------------------------------------------------------
+r10398 | quinlan | 2004-04-29 01:54:46 +0000 (Thu, 29 Apr 2004) | 2 lines
+
+keyword space, function no space
+
+------------------------------------------------------------------------
+r10397 | felicity | 2004-04-29 01:40:07 +0000 (Thu, 29 Apr 2004) | 1 line
+
+untie the DBs after we preload everything, prespawn...
+------------------------------------------------------------------------
+r10396 | quinlan | 2004-04-29 01:28:51 +0000 (Thu, 29 Apr 2004) | 4 lines
+
+bug 3291: update references to activating auto-whitelist with -a in the docs
+also move some upgrade notes from README to INSTALL (needs more work and
+cleaning)
+
+------------------------------------------------------------------------
+r10395 | jm | 2004-04-29 01:18:32 +0000 (Thu, 29 Apr 2004) | 1 line
+
+test for spamc -E
+------------------------------------------------------------------------
+r10394 | parker | 2004-04-29 01:02:59 +0000 (Thu, 29 Apr 2004) | 26 lines
+
+Bug 3225: Bayes Optimization
+
+Many optimizations for bayes storage.
+
+1) Implement a tok_get_all that fetches multiple tokens at one time, instead of one at a time.
+
+2) Modify SQL storage so that it keeps a running tally for total
+tokens, oldest/newest token age instead using expensive select
+count(*)/min(atime)/max(atime) calls.
+
+3) In SQL, clean up tokens whose spam_count and ham_count both equal
+0.
+
+4) Added some basic caching in SQL code to avoid doing some SQL
+operations.
+
+5) Switched SQL backend to use userids instead of username for keys in
+all of it's tables.
+
+6) Removed some dead code.
+
+7) We now store tokens as the lower 40 bits of a SHA1 hash.
+
+8) Updated to backup/restore code to support the new database format.
+
+
+------------------------------------------------------------------------
+r10393 | jm | 2004-04-29 00:51:55 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 2855: added spamc -E switch, to return exit code based on spam/nonspam status in normal use
+------------------------------------------------------------------------
+r10392 | jm | 2004-04-29 00:39:48 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 2641 'fix' reverted: I was wrong, virtual-config-dir was doing exactly what it's supposed to do.
+------------------------------------------------------------------------
+r10391 | jm | 2004-04-29 00:26:41 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 2776: make test fails in de_DE locale
+------------------------------------------------------------------------
+r10390 | jm | 2004-04-29 00:11:22 +0000 (Thu, 29 Apr 2004) | 1 line
+
+bug 3017: fix dccifd bug when message is empty
+------------------------------------------------------------------------
+r10389 | jm | 2004-04-28 23:51:45 +0000 (Wed, 28 Apr 2004) | 1 line
+
+removed terse-report from all intl files; added a lock_method line to demo local.cf file
+------------------------------------------------------------------------
+r10388 | jm | 2004-04-28 23:40:25 +0000 (Wed, 28 Apr 2004) | 1 line
+
+tiny tweak
+------------------------------------------------------------------------
+r10387 | felicity | 2004-04-28 23:35:55 +0000 (Wed, 28 Apr 2004) | 1 line
+
+if the first bayes tie succeeds, but the second fails, we don't untie the first. also, extra cleanup in permsglearner...
+------------------------------------------------------------------------
+r10386 | jm | 2004-04-28 23:34:07 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 2641: spamd was using wrong path for virtual-config-dir user_prefs file
+------------------------------------------------------------------------
+r10385 | jm | 2004-04-28 23:17:25 +0000 (Wed, 28 Apr 2004) | 1 line
+
+refactored and fixed spamd syslog code. the flag 'ndelay' is now used, which should speed us up some; redundant calls to openlog() removed for efficiency; the 'try with a different switch' message is now only output once.
+------------------------------------------------------------------------
+r10384 | jm | 2004-04-28 23:00:55 +0000 (Wed, 28 Apr 2004) | 1 line
+
+avoid cases where msgid could contain ','
+------------------------------------------------------------------------
+r10383 | jm | 2004-04-28 22:49:17 +0000 (Wed, 28 Apr 2004) | 1 line
+
+rewrote flock() sleep code to use a single SIGALRM-based timeout
+------------------------------------------------------------------------
+r10382 | jm | 2004-04-28 22:34:31 +0000 (Wed, 28 Apr 2004) | 1 line
+
+updated comments
+------------------------------------------------------------------------
+r10381 | jm | 2004-04-28 22:32:17 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 3242: NFS-unsafe locker added, using flock(). currently it's polling (attempt lock,sleep,attempt lock,sleep). controlled using the 'lock_method' config parameter.
+------------------------------------------------------------------------
+r10380 | felicity | 2004-04-28 22:28:35 +0000 (Wed, 28 Apr 2004) | 1 line
+
+sa-learn wasn't cleaning up the parsed message after it was finished...
+------------------------------------------------------------------------
+r10377 | felicity | 2004-04-28 21:35:08 +0000 (Wed, 28 Apr 2004) | 1 line
+
+added log msg when spamd spawns
+------------------------------------------------------------------------
+r10370 | jm | 2004-04-28 19:12:32 +0000 (Wed, 28 Apr 2004) | 1 line
+
+I've been getting a number of FPs on FORGED_DEF_WHITELIST, due to older mailing list software. Given this, lowering score to avoid FPs
+------------------------------------------------------------------------
+r10367 | quinlan | 2004-04-28 18:10:12 +0000 (Wed, 28 Apr 2004) | 2 lines
+
+add tflags net
+
+------------------------------------------------------------------------
+r10366 | quinlan | 2004-04-28 18:08:54 +0000 (Wed, 28 Apr 2004) | 3 lines
+
+new SenderBase rules
+promote new timezone rules
+
+------------------------------------------------------------------------
+r10351 | quinlan | 2004-04-28 06:02:43 +0000 (Wed, 28 Apr 2004) | 2 lines
+
+use a bit more complicated message
+
+------------------------------------------------------------------------
+r10348 | felicity | 2004-04-28 05:00:32 +0000 (Wed, 28 Apr 2004) | 1 line
+
+streamline the body modification test: just pass the file glob, don't bother pre-reading the message into memory.
+------------------------------------------------------------------------
+r10346 | jm | 2004-04-28 04:28:30 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 3057: added Return-path compression for exim
+------------------------------------------------------------------------
+r10345 | jm | 2004-04-28 04:27:06 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 3126: added test to ensure that body text is not modified during scanning
+------------------------------------------------------------------------
+r10344 | jm | 2004-04-28 04:04:05 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 2798: added code to log message-id, resent-message-id, the tests hit, autolearn status, and several other things in a mass-check compatible format to syslog. Would appreciate comments / additions...
+------------------------------------------------------------------------
+r10340 | jm | 2004-04-28 03:28:14 +0000 (Wed, 28 Apr 2004) | 1 line
+
+refactored spamd eval { } block into its own method, for clarity
+------------------------------------------------------------------------
+r10339 | jm | 2004-04-28 03:19:48 +0000 (Wed, 28 Apr 2004) | 1 line
+
+ensure $client gets closed
+------------------------------------------------------------------------
+r10338 | jm | 2004-04-28 02:28:25 +0000 (Wed, 28 Apr 2004) | 1 line
+
+avoid extra tok_get() call
+------------------------------------------------------------------------
+r10337 | jm | 2004-04-28 02:11:46 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 2769: message_filter() reportedly returned the wrong error codes
+------------------------------------------------------------------------
+r10336 | jm | 2004-04-28 02:06:23 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 2266: add 'add_header' tags to display information about the message's Bayesian classification and the tokens used; patch contributed by David Koppelman
+------------------------------------------------------------------------
+r10335 | felicity | 2004-04-28 01:55:36 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 3318: added a test for encoded redirects
+------------------------------------------------------------------------
+r10334 | felicity | 2004-04-28 01:46:38 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 3318: fully unencode URIs that are found.
+------------------------------------------------------------------------
+r10333 | jm | 2004-04-28 01:22:34 +0000 (Wed, 28 Apr 2004) | 1 line
+
+test for include
+------------------------------------------------------------------------
+r10332 | jm | 2004-04-28 01:12:19 +0000 (Wed, 28 Apr 2004) | 1 line
+
+bug 1244: 'include' facility added to user_prefs. also fixed a bug where 'loadplugin' with a relative path was broken
+------------------------------------------------------------------------
+r10326 | felicity | 2004-04-27 21:01:17 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 3308: more work on get_uri_list ...
+------------------------------------------------------------------------
+r10324 | felicity | 2004-04-27 20:28:45 +0000 (Tue, 27 Apr 2004) | 1 line
+
+updated comment info
+------------------------------------------------------------------------
+r10323 | felicity | 2004-04-27 20:25:18 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 3317: copy_config() needs to deal with the one-off object copying (in this case the trusted networks object array) _before_ dealing with standard reference copying.
+------------------------------------------------------------------------
+r10321 | felicity | 2004-04-27 19:33:57 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 3308: turns out that a lot of browsers (IE, Mozilla, Thunderbird, Opera, etc) will allow non-valid URLs such as 'http:www', and 'http:/www', in addition to 'http://www'. I modified the appropriate parts of the code to handle those. Moved a section of get_uri_list into its own Util function, and added tests to make sure we can canonify uris properly.
+------------------------------------------------------------------------
+r10320 | parker | 2004-04-27 19:09:02 +0000 (Tue, 27 Apr 2004) | 1 line
+
+Minor tweaks to bayes restore
+------------------------------------------------------------------------
+r10312 | parker | 2004-04-27 14:29:07 +0000 (Tue, 27 Apr 2004) | 1 line
+
+Bug 3049: Bayes database backup/restore
+------------------------------------------------------------------------
+r10311 | parker | 2004-04-27 14:27:56 +0000 (Tue, 27 Apr 2004) | 1 line
+
+Bug 3049: Bayes database backup/restore
+------------------------------------------------------------------------
+r10305 | jm | 2004-04-27 07:10:18 +0000 (Tue, 27 Apr 2004) | 1 line
+
+revise to handle a legit die() case
+------------------------------------------------------------------------
+r10304 | jm | 2004-04-27 07:05:05 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 1832: avoid nmap crashing spamd by trapping die()s from all socket operations
+------------------------------------------------------------------------
+r10302 | quinlan | 2004-04-27 03:51:06 +0000 (Tue, 27 Apr 2004) | 2 lines
+
+add net flag
+
+------------------------------------------------------------------------
+r10300 | jm | 2004-04-27 02:03:13 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 3122: USER_IN_BLACKLIST penalty should not be added to AWL, as this makes it harder to de-blacklist
+------------------------------------------------------------------------
+r10299 | jm | 2004-04-27 01:47:24 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 2198: spamd tests now do not log to syslog at all, except for one new one -- spamd_syslog.t
+------------------------------------------------------------------------
+r10298 | jm | 2004-04-27 01:36:47 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 2419: expand Bayes score breakdown near the extremes
+------------------------------------------------------------------------
+r10297 | felicity | 2004-04-27 00:32:37 +0000 (Tue, 27 Apr 2004) | 1 line
+
+digging around, SIGCHLD=IGNORE doesn't have guaranteed behavior on most platforms. so change the restart and kill handlers appropriately -- kill sets to default then dies, restart sends kills and does a wait for each process.
+------------------------------------------------------------------------
+r10296 | felicity | 2004-04-27 00:28:54 +0000 (Tue, 27 Apr 2004) | 1 line
+
+bug 3312: world-writable path taint check was invalid. the mode&2 == 2 not 1. gah!
+------------------------------------------------------------------------
+r10294 | felicity | 2004-04-26 22:49:44 +0000 (Mon, 26 Apr 2004) | 1 line
+
+add a test for bug 2292, parsing from addresses out of the comment section of the From header...
+------------------------------------------------------------------------
+r10293 | felicity | 2004-04-26 22:39:36 +0000 (Mon, 26 Apr 2004) | 1 line
+
+bug 2292: whitelist_from would have issues when the From header had a comment with an email address in it which was whitelisted. changed to using the get('...:addr') code instead which knows how to ignore comments and the like.
+------------------------------------------------------------------------
+r10292 | felicity | 2004-04-26 21:02:29 +0000 (Mon, 26 Apr 2004) | 1 line
+
+bug 3304: turns out that if you set a constant to qr/.../, you need to do =~CONSTANT, not =~/CONSTANT/ ...
+------------------------------------------------------------------------
+r10291 | felicity | 2004-04-26 20:38:01 +0000 (Mon, 26 Apr 2004) | 1 line
+
+bug 3290: whitelist_addrs.t was failing because we ignore squirrelmail headers in 3.0. changed the IP in the received header prior to the squirrelmail header to match the t/data/nice/002 mail so the test succeeds. the other tests didn't break when doing this, so ... :)
+------------------------------------------------------------------------
+r10290 | quinlan | 2004-04-26 20:22:32 +0000 (Mon, 26 Apr 2004) | 2 lines
+
+BondedSender data in SenderBase appears to be stale or incomplete
+
+------------------------------------------------------------------------
+r10289 | quinlan | 2004-04-26 19:05:00 +0000 (Mon, 26 Apr 2004) | 3 lines
+
+revise T_BAD_TIMEZONE*
+add T_RCVD_IN_BSP_* to see if SenderBase BSP is working now
+
+------------------------------------------------------------------------
+r10288 | felicity | 2004-04-26 18:57:25 +0000 (Mon, 26 Apr 2004) | 1 line
+
+update spamd pod/usage...
+------------------------------------------------------------------------
+r10286 | felicity | 2004-04-26 18:41:54 +0000 (Mon, 26 Apr 2004) | 1 line
+
+bug 3308: redirector handling didn't deal with 'http:/www ...', while things like the yahoo redirector does. so handle that, also fix uris that come through with only one forward-slash. also, add debugging output for list of URIs found.
+------------------------------------------------------------------------
+r10276 | quinlan | 2004-04-26 06:47:54 +0000 (Mon, 26 Apr 2004) | 2 lines
+
+Cuba
+
+------------------------------------------------------------------------
+r10275 | quinlan | 2004-04-26 06:40:50 +0000 (Mon, 26 Apr 2004) | 2 lines
+
+lint
+
+------------------------------------------------------------------------
+r10274 | quinlan | 2004-04-26 05:31:53 +0000 (Mon, 26 Apr 2004) | 3 lines
+
+remove T_HTML_FONT_SIZE_BIG
+remove HTML_FONT_FACE_ODD
+
+------------------------------------------------------------------------
+r10273 | quinlan | 2004-04-26 03:39:42 +0000 (Mon, 26 Apr 2004) | 4 lines
+
+remove some test rules
+replace LOTS_OF_STUFF with T_LOTS_OF_STUFF_13
+add T_ENTITY_HEX_ANY
+
+------------------------------------------------------------------------
+r10268 | quinlan | 2004-04-26 00:39:29 +0000 (Mon, 26 Apr 2004) | 3 lines
+
+revise T_ENTITY*
+tweak T_INVALID_RCVD_TZ_ABSURD_A
+
+------------------------------------------------------------------------
+r10267 | quinlan | 2004-04-25 22:51:54 +0000 (Sun, 25 Apr 2004) | 2 lines
+
+some test rules
+
+------------------------------------------------------------------------
+r10266 | felicity | 2004-04-25 22:46:45 +0000 (Sun, 25 Apr 2004) | 1 line
+
+fix debug output issue in bayes expiry
+------------------------------------------------------------------------
+r10265 | felicity | 2004-04-25 22:38:14 +0000 (Sun, 25 Apr 2004) | 1 line
+
+make expiry percent, period, and max exponent config variables instead of strictly hardcoded values
+------------------------------------------------------------------------
+r10264 | quinlan | 2004-04-25 22:05:12 +0000 (Sun, 25 Apr 2004) | 2 lines
+
+some time zone rules
+
+------------------------------------------------------------------------
+r10263 | quinlan | 2004-04-25 21:48:33 +0000 (Sun, 25 Apr 2004) | 2 lines
+
+tweak T_BAD_TIMEZONE_CST*
+
+------------------------------------------------------------------------
+r10262 | felicity | 2004-04-25 20:58:40 +0000 (Sun, 25 Apr 2004) | 1 line
+
+move the euid and config change-back to after the client connection is closed in the child...
+------------------------------------------------------------------------
+r10256 | quinlan | 2004-04-25 10:22:12 +0000 (Sun, 25 Apr 2004) | 2 lines
+
+remove now unnecessary backhair code
+
+------------------------------------------------------------------------
+r10255 | quinlan | 2004-04-25 10:21:43 +0000 (Sun, 25 Apr 2004) | 6 lines
+
+promote T_URIBL_WS_SURBL to URIBL_WS_SURBL
+promote T_URIBL_BE_SURBL to URIBL_BE_SURBL
+promote SENDERBASE_NEW_BULK (hits about 8% of spam real-time)
+promote HTML_BACKHAIR*
+settle on a single T_BAD_TIMEZONE* version
+
+------------------------------------------------------------------------
+r10246 | parker | 2004-04-25 06:44:07 +0000 (Sun, 25 Apr 2004) | 1 line
+
+remove warning
+------------------------------------------------------------------------
+r10219 | quinlan | 2004-04-24 07:53:53 +0000 (Sat, 24 Apr 2004) | 3 lines
+
+revise T_ENTITY_* rules (rawbody is only type really working, test other
+decimal codes, also experiment with leading zeroes)
+
+------------------------------------------------------------------------
+r10218 | quinlan | 2004-04-24 04:25:58 +0000 (Sat, 24 Apr 2004) | 2 lines
+
+adjust threshold for SUSPICIOUS_RECIPS
+
+------------------------------------------------------------------------
+r10217 | quinlan | 2004-04-24 04:22:36 +0000 (Sat, 24 Apr 2004) | 5 lines
+
+remove TO_HAS_SPACES and T_TO_HAS_SPACES*
+promote T_LONG_HOSTNAME_25_A to URI_LONG_HOSTNAME
+remove T_GEVALIACOFFEE
+add some T_LOTS_OF_STUFF* variations
+
+------------------------------------------------------------------------
+r10216 | quinlan | 2004-04-24 04:02:04 +0000 (Sat, 24 Apr 2004) | 2 lines
+
+a few variations
+
+------------------------------------------------------------------------
+r10215 | quinlan | 2004-04-24 03:33:05 +0000 (Sat, 24 Apr 2004) | 3 lines
+
+add backhair_count rules
+add T_BAD_TIMEZONE* rules
+
+------------------------------------------------------------------------
+r10214 | quinlan | 2004-04-24 03:32:27 +0000 (Sat, 24 Apr 2004) | 2 lines
+
+add backhair_count
+
+------------------------------------------------------------------------
+r10213 | quinlan | 2004-04-24 03:31:53 +0000 (Sat, 24 Apr 2004) | 2 lines
+
+some revisions to the time zone code
+
+------------------------------------------------------------------------
+r10212 | quinlan | 2004-04-24 03:29:13 +0000 (Sat, 24 Apr 2004) | 2 lines
+
+remove redundant orbitz entry
+
+------------------------------------------------------------------------
+r10211 | quinlan | 2004-04-24 03:28:25 +0000 (Sat, 24 Apr 2004) | 2 lines
+
+add a few whitelist entries
+
+------------------------------------------------------------------------
+r10210 | quinlan | 2004-04-24 02:14:12 +0000 (Sat, 24 Apr 2004) | 2 lines
+
+whitespace only
+
+------------------------------------------------------------------------
+r10201 | felicity | 2004-04-23 21:23:53 +0000 (Fri, 23 Apr 2004) | 1 line
+
+bug 3288: kh and ni are subdelegated cctld ...
+------------------------------------------------------------------------
+r10198 | felicity | 2004-04-23 20:18:45 +0000 (Fri, 23 Apr 2004) | 1 line
+
+fix rcvd_parser.t errors
+------------------------------------------------------------------------
+r10190 | jm | 2004-04-23 01:05:55 +0000 (Fri, 23 Apr 2004) | 1 line
+
+added T_SA_MARKED_UP to track how many encapsulated mails get through to the scanner
+------------------------------------------------------------------------
+r10189 | jm | 2004-04-23 01:04:16 +0000 (Fri, 23 Apr 2004) | 1 line
+
+added support for --deencap switch: de-encapsulate only mails that were marked up on a specific host
+------------------------------------------------------------------------
+r10188 | felicity | 2004-04-22 20:48:35 +0000 (Thu, 22 Apr 2004) | 1 line
+
+adding some comments for get_uri_list()
+------------------------------------------------------------------------
+r10187 | felicity | 2004-04-22 20:25:22 +0000 (Thu, 22 Apr 2004) | 1 line
+
+bug 3287: URI parsing should ignore/strip leading/trailing whitespace.
+------------------------------------------------------------------------
+r10186 | felicity | 2004-04-22 20:15:17 +0000 (Thu, 22 Apr 2004) | 1 line
+
+let GC handle the temp array
+------------------------------------------------------------------------
+r10185 | felicity | 2004-04-22 19:47:55 +0000 (Thu, 22 Apr 2004) | 1 line
+
+during expiry, handle the situation where the newest tokens is in the future by resetting the newest token stamp to current time, and changing any token atimes > newest atime to newest atime.
+------------------------------------------------------------------------
+r10184 | felicity | 2004-04-22 19:12:34 +0000 (Thu, 22 Apr 2004) | 1 line
+
+add prelim docs for upgrading to 3.0
+------------------------------------------------------------------------
+r10183 | quinlan | 2004-04-22 18:31:35 +0000 (Thu, 22 Apr 2004) | 2 lines
+
+add T_URIBL_BE_SURBL
+
+------------------------------------------------------------------------
+r10182 | quinlan | 2004-04-22 18:24:10 +0000 (Thu, 22 Apr 2004) | 2 lines
+
+need to ignore case when ignoring repeated addresses
+
+------------------------------------------------------------------------
+r10169 | quinlan | 2004-04-22 07:42:15 +0000 (Thu, 22 Apr 2004) | 2 lines
+
+drill down a bit on SenderBase rules
+
+------------------------------------------------------------------------
+r10168 | quinlan | 2004-04-22 06:52:16 +0000 (Thu, 22 Apr 2004) | 4 lines
+
+merge VERY_SUSP_RECIPS into SUSPICIOUS_RECIPS (hit rate for
+ VERY_SUSP_RECIPS is too low and scores were close to each other anyway)
+test different thresholds for new SUSPICIOUS_RECIPS code
+
+------------------------------------------------------------------------
+r10167 | parker | 2004-04-22 06:08:32 +0000 (Thu, 22 Apr 2004) | 1 line
+
+another SHA1 cleanup
+------------------------------------------------------------------------
+r10166 | quinlan | 2004-04-22 04:59:49 +0000 (Thu, 22 Apr 2004) | 2 lines
+
+merge AOL redirect rules
+
+------------------------------------------------------------------------
+r10165 | quinlan | 2004-04-22 04:48:44 +0000 (Thu, 22 Apr 2004) | 2 lines
+
+remove extra step
+
+------------------------------------------------------------------------
+r10164 | quinlan | 2004-04-22 04:21:00 +0000 (Thu, 22 Apr 2004) | 2 lines
+
+oops, forgot to check these in too
+
+------------------------------------------------------------------------
+r10163 | sidney | 2004-04-22 04:18:14 +0000 (Thu, 22 Apr 2004) | 1 line
+
+Change sha1.t test to match changed function names in SHA1.pm
+------------------------------------------------------------------------
+r10162 | quinlan | 2004-04-22 03:57:11 +0000 (Thu, 22 Apr 2004) | 3 lines
+
+use same naming as Digest::SHA1 for exported functions
+rename SHA1 to perl_sha1 for better clarity
+
+------------------------------------------------------------------------
+r10159 | sidney | 2004-04-22 00:19:53 +0000 (Thu, 22 Apr 2004) | 1 line
+
+Add sha1bin function for binary form of SHA1 using Digest::SHA1 if available or else slow pure perl. Update test cases for the changes
+------------------------------------------------------------------------
+r10146 | quinlan | 2004-04-21 06:06:58 +0000 (Wed, 21 Apr 2004) | 2 lines
+
+bug 3285: SUSPICIOUS_RECIPS triggers on "user@dom.ain" <us...@dom.ain>
+
+------------------------------------------------------------------------
+r10145 | quinlan | 2004-04-21 06:02:33 +0000 (Wed, 21 Apr 2004) | 2 lines
+
+URI scheme removal should be case-insensitive
+
+------------------------------------------------------------------------
+r10144 | jm | 2004-04-21 03:05:34 +0000 (Wed, 21 Apr 2004) | 1 line
+
+bug 3201: current Mailman not detected by __KNOWN_MAILING_LIST
+------------------------------------------------------------------------
+r10143 | jm | 2004-04-21 01:59:00 +0000 (Wed, 21 Apr 2004) | 1 line
+
+bug 3236: ignore SquirrelMail injection Received headers
+------------------------------------------------------------------------
+r10142 | quinlan | 2004-04-21 01:25:29 +0000 (Wed, 21 Apr 2004) | 2 lines
+
+fix undef warning for malformed URLs
+
+------------------------------------------------------------------------
+r10141 | quinlan | 2004-04-21 01:08:19 +0000 (Wed, 21 Apr 2004) | 2 lines
+
+fixes for T_REDIRECTOR
+
+------------------------------------------------------------------------
+r10139 | jm | 2004-04-20 23:55:16 +0000 (Tue, 20 Apr 2004) | 1 line
+
+bug 3009: antidrug.cf, several good anti-drug-spam rules, added. thanks to Matt Kettler
+------------------------------------------------------------------------
+r10138 | jm | 2004-04-20 23:43:59 +0000 (Tue, 20 Apr 2004) | 1 line
+
+bug 2539: updated RAND rules from Fred T
+------------------------------------------------------------------------
+r10137 | jm | 2004-04-20 23:41:28 +0000 (Tue, 20 Apr 2004) | 1 line
+
+oops, didn't mean to check that in
+------------------------------------------------------------------------
+r10136 | jm | 2004-04-20 23:38:19 +0000 (Tue, 20 Apr 2004) | 1 line
+
+bug 2539: added more rules to catch RAND signs thanks to Fred T
+------------------------------------------------------------------------
+r10134 | quinlan | 2004-04-20 21:07:30 +0000 (Tue, 20 Apr 2004) | 2 lines
+
+remove test rule
+
+------------------------------------------------------------------------
+r10133 | quinlan | 2004-04-20 20:07:34 +0000 (Tue, 20 Apr 2004) | 2 lines
+
+remove old score
+
+------------------------------------------------------------------------
+r10132 | quinlan | 2004-04-20 20:06:58 +0000 (Tue, 20 Apr 2004) | 2 lines
+
+remove old scores
+
+------------------------------------------------------------------------
+r10131 | quinlan | 2004-04-20 20:06:04 +0000 (Tue, 20 Apr 2004) | 6 lines
+
+replace MSGID_FROM_MTA_SHORT with MSGID_FROM_MTA_ID
+remove MSGID_FROM_MTA_LATER and MSGID_FROM_MTA_BACKUP
+remove URL shortening rules
+add T_LOTS_OF_STUFF*
+add T_TO_HAS_SPACES*
+
+------------------------------------------------------------------------
+r10122 | jm | 2004-04-20 06:26:31 +0000 (Tue, 20 Apr 2004) | 1 line
+
+added 'id=' to rcvd_parser.t test cases -- still a lot of failures however
+------------------------------------------------------------------------
+r10121 | jm | 2004-04-20 06:24:47 +0000 (Tue, 20 Apr 2004) | 1 line
+
+added 'envelope_sender_header' config setting for sites to override EnvelopeFrom heuristics
+------------------------------------------------------------------------
+r10120 | jm | 2004-04-20 06:05:51 +0000 (Tue, 20 Apr 2004) | 1 line
+
+bug 2321: list of headers checked by whitelist_to and whitelist_from was incomplete
+------------------------------------------------------------------------
+r10115 | felicity | 2004-04-20 03:08:50 +0000 (Tue, 20 Apr 2004) | 1 line
+
+spamd would complain about a pidfile, even if it wasn't supposed to make one ...
+------------------------------------------------------------------------
+r10114 | felicity | 2004-04-20 03:03:37 +0000 (Tue, 20 Apr 2004) | 1 line
+
+log_socket may not be defined when spamd ends (when usage() runs, for instance...)
+------------------------------------------------------------------------
+r10113 | quinlan | 2004-04-20 02:20:10 +0000 (Tue, 20 Apr 2004) | 2 lines
+
+add cleaner version of MSGID_FROM_MTA*
+
+------------------------------------------------------------------------
+r10112 | felicity | 2004-04-20 02:13:32 +0000 (Tue, 20 Apr 2004) | 1 line
+
+bug 3283: had some issues with umask settings
+------------------------------------------------------------------------
+r10111 | felicity | 2004-04-19 23:33:53 +0000 (Mon, 19 Apr 2004) | 1 line
+
+bug 3268: get_uri_list() was using decoded_body to find URIs, then looking at the HTML parser results for more uris. however, the decoded_body has the HTML in it, so we were looking at the HTML twice, sometimes grabbing newline-separated uris. so disable passing in a text array, and use the rendered text for RE matching, let the HTML parsing do the right thing there. also, revert the redirector code.
+------------------------------------------------------------------------
+r10110 | felicity | 2004-04-19 23:20:33 +0000 (Mon, 19 Apr 2004) | 1 line
+
+bug 3268: rule to look for generic HTTP redirects. ignore redirects that are to the same domain.
+------------------------------------------------------------------------
+r10109 | jm | 2004-04-19 17:52:49 +0000 (Mon, 19 Apr 2004) | 1 line
+
+no need to use $1
+------------------------------------------------------------------------
+r10108 | jm | 2004-04-19 17:49:06 +0000 (Mon, 19 Apr 2004) | 1 line
+
+fix regression: more than 1 layer of redirection not caught
+------------------------------------------------------------------------
+r10107 | quinlan | 2004-04-19 17:31:45 +0000 (Mon, 19 Apr 2004) | 2 lines
+
+another one
+
+------------------------------------------------------------------------
+r10106 | quinlan | 2004-04-19 17:26:02 +0000 (Mon, 19 Apr 2004) | 2 lines
+
+add rules for URL obfuscation services
+
+------------------------------------------------------------------------
+r10105 | felicity | 2004-04-19 16:29:16 +0000 (Mon, 19 Apr 2004) | 1 line
+
+minor uri-based cleanup
+------------------------------------------------------------------------
+r10104 | felicity | 2004-04-19 16:00:42 +0000 (Mon, 19 Apr 2004) | 1 line
+
+fix redirect handler again
+------------------------------------------------------------------------
+r10103 | quinlan | 2004-04-19 15:37:13 +0000 (Mon, 19 Apr 2004) | 3 lines
+
+Nevermind, looks okay (too early in the morning).
+I think this method is still going to miss encoded redirects, though.
+
+------------------------------------------------------------------------
+r10102 | quinlan | 2004-04-19 15:34:26 +0000 (Mon, 19 Apr 2004) | 2 lines
+
+comment out infinite loop
+
+------------------------------------------------------------------------
+r10095 | felicity | 2004-04-19 05:13:10 +0000 (Mon, 19 Apr 2004) | 1 line
+
+fix up uri parsing code -- we have been catching URIs on multiple lines and embedding newlines, so strip those out. do better version of unencoding. deal with redirectors by adding in all of the redirection urls as well. it made all my results better/more accurate, so plus plus. :)
+------------------------------------------------------------------------
+r10067 | quinlan | 2004-04-16 23:17:15 +0000 (Fri, 16 Apr 2004) | 2 lines
+
+change from Fisher-Yates shuffle to SHA1 shuffle
+
+------------------------------------------------------------------------
+r10054 | jm | 2004-04-16 20:51:22 +0000 (Fri, 16 Apr 2004) | 1 line
+
+BIZ_TLD_2 -> BIZ_TLD
+------------------------------------------------------------------------
+r10053 | quinlan | 2004-04-16 20:39:21 +0000 (Fri, 16 Apr 2004) | 2 lines
+
+TLD rules
+
+------------------------------------------------------------------------
+r10052 | quinlan | 2004-04-16 20:34:21 +0000 (Fri, 16 Apr 2004) | 2 lines
+
+remove MIME_ODD_CASE - last spam hit in Jan 2003
+
+------------------------------------------------------------------------
+r10051 | felicity | 2004-04-16 20:22:08 +0000 (Fri, 16 Apr 2004) | 1 line
+
+bug 3272: logic in copy_config for checking source + dest for hash reference was wrong. oops.
+------------------------------------------------------------------------
+r10049 | felicity | 2004-04-16 19:08:45 +0000 (Fri, 16 Apr 2004) | 1 line
+
+bug 3276: spamd's -m didn't work right wrt max-connects per child. spun it off into --max-conn-per-child, fixed docs and such.
+------------------------------------------------------------------------
+r10043 | quinlan | 2004-04-16 05:44:40 +0000 (Fri, 16 Apr 2004) | 2 lines
+
+try some looser T_INFO_TLD rules
+
+------------------------------------------------------------------------
+r10042 | felicity | 2004-04-16 05:09:53 +0000 (Fri, 16 Apr 2004) | 1 line
+
+bug 3240: __ANY_QUALCOMM_MUA got removed in 2.61, but it's a subrule for FORGED_QUALCOMM_TAGS ... oops.
+------------------------------------------------------------------------
+r10041 | felicity | 2004-04-16 04:59:34 +0000 (Fri, 16 Apr 2004) | 1 line
+
+bug 3269: message parser needs to return mime boundary states for the MIME_MISSING_BOUNDARY rule to figure out if boundaries are missing
+------------------------------------------------------------------------
+r10040 | felicity | 2004-04-16 04:34:02 +0000 (Fri, 16 Apr 2004) | 1 line
+
+bug 3269: mime header information was expected to be lowercased, but wasn't, causing issues.
+------------------------------------------------------------------------
+r10039 | felicity | 2004-04-16 04:20:35 +0000 (Fri, 16 Apr 2004) | 1 line
+
+bug 3275: ':addr' was a little broken as it was trying to remove multiple addresses without checking for quotes and such first.
+------------------------------------------------------------------------
+r10038 | felicity | 2004-04-16 04:06:21 +0000 (Fri, 16 Apr 2004) | 1 line
+
+better rule output in debug mode
+------------------------------------------------------------------------
+r10037 | felicity | 2004-04-16 04:00:22 +0000 (Fri, 16 Apr 2004) | 1 line
+
+more pre/post-daemonize stuff
+------------------------------------------------------------------------
+r10035 | felicity | 2004-04-15 23:26:19 +0000 (Thu, 15 Apr 2004) | 1 line
+
+better deal with pidfile and daemonizing and such
+------------------------------------------------------------------------
+r10034 | felicity | 2004-04-15 23:13:07 +0000 (Thu, 15 Apr 2004) | 1 line
+
+move parts around in spamd so that errors are displayed before becoming a daemon (like Storable is needed but not available ...)
+------------------------------------------------------------------------
+r10033 | felicity | 2004-04-15 20:53:50 +0000 (Thu, 15 Apr 2004) | 1 line
+
+bug 3272: modify how the configuration is backed up for spamd. added new copy_config() API method to allow third party folks to backup/restore the config.
+------------------------------------------------------------------------
+r10030 | felicity | 2004-04-15 04:11:37 +0000 (Thu, 15 Apr 2004) | 1 line
+
+bug 3262: attempt to improve NO_COST
+------------------------------------------------------------------------
+r10026 | felicity | 2004-04-15 03:54:12 +0000 (Thu, 15 Apr 2004) | 1 line
+
+updated docs to point at pyzor website and remove repeated pointers for razor and dcc
+------------------------------------------------------------------------
+r10025 | felicity | 2004-04-15 03:41:02 +0000 (Thu, 15 Apr 2004) | 1 line
+
+reimplement ADVERT_CODE* rules
+------------------------------------------------------------------------
+r10023 | jm | 2004-04-15 02:34:22 +0000 (Thu, 15 Apr 2004) | 1 line
+
+some updates to plugin doco and a bugfix in URIDNSBL
+------------------------------------------------------------------------
+r10022 | felicity | 2004-04-15 02:17:18 +0000 (Thu, 15 Apr 2004) | 1 line
+
+bug 3269: found that _check_attachments() was always looking at the message header, and not the MIME part headers. :(
+------------------------------------------------------------------------
+r10006 | felicity | 2004-04-14 21:21:24 +0000 (Wed, 14 Apr 2004) | 1 line
+
+found a few other holdouts for 'required_hits' -- should be 'required_score' now...
+------------------------------------------------------------------------
+r10004 | felicity | 2004-04-14 20:21:50 +0000 (Wed, 14 Apr 2004) | 1 line
+
+bug 3270: have spamassassin and spamd check to make sure they're the same version as the modules that they load. this will avoid an issue which has come up where people have, for instance, the 'spamassassin' script from 2.31 running with modules from 2.63, and then complaining there are issues.
+------------------------------------------------------------------------
+r10000 | felicity | 2004-04-14 03:23:56 +0000 (Wed, 14 Apr 2004) | 1 line
+
+bug 3097: switch spamd from doing 'fork per message' to a 'prefork' model
+------------------------------------------------------------------------
+r9999 | jm | 2004-04-14 02:06:35 +0000 (Wed, 14 Apr 2004) | 1 line
+
+attempting to remove the non-deterministic element from 'make test'
+------------------------------------------------------------------------
+r9998 | felicity | 2004-04-14 01:49:16 +0000 (Wed, 14 Apr 2004) | 1 line
+
+bayes magic dump should display in unsigned mode...
+------------------------------------------------------------------------
+r9997 | quinlan | 2004-04-14 01:40:38 +0000 (Wed, 14 Apr 2004) | 2 lines
+
+add T_BIZ_TLD*
+
+------------------------------------------------------------------------
+r9996 | quinlan | 2004-04-13 17:02:41 +0000 (Tue, 13 Apr 2004) | 2 lines
+
+name change
+
+------------------------------------------------------------------------
+r9990 | quinlan | 2004-04-13 04:16:33 +0000 (Tue, 13 Apr 2004) | 2 lines
+
+remove HTML_RELAYING_FRAME (just about the worst HTML rule we have)
+
+------------------------------------------------------------------------
+r9989 | quinlan | 2004-04-13 04:11:02 +0000 (Tue, 13 Apr 2004) | 2 lines
+
+remove HTML_IMAGE_AREA* rules
+
+------------------------------------------------------------------------
+r9988 | quinlan | 2004-04-13 04:01:00 +0000 (Tue, 13 Apr 2004) | 2 lines
+
+remove HTML color rules
+
+------------------------------------------------------------------------
+r9987 | quinlan | 2004-04-13 03:38:07 +0000 (Tue, 13 Apr 2004) | 2 lines
+
+guess size if we don't know it (should help HTML_IMAGE_RATIO a bit)
+
+------------------------------------------------------------------------
+r9979 | quinlan | 2004-04-12 20:42:28 +0000 (Mon, 12 Apr 2004) | 3 lines
+
+remove T_URIBL_DSBL
+add T_URIBL_SA_SURBL
+
+------------------------------------------------------------------------
+r9964 | quinlan | 2004-04-11 08:31:10 +0000 (Sun, 11 Apr 2004) | 4 lines
+
+bug 3252: remove redundant HTML_FONTCOLOR_NAME
+clean up HTML color tests (test both fg and bg, simplify naming)
+remove low performing HTML_WITH_BGCOLOR
+
+------------------------------------------------------------------------
+r9960 | felicity | 2004-04-10 17:23:59 +0000 (Sat, 10 Apr 2004) | 1 line
+
+one of the bayesstore 'not implemented' messages was wrong. we were also caching the token magic_re when we didn't need to (it's only used in 1 place).
+------------------------------------------------------------------------
+r9952 | felicity | 2004-04-09 19:19:00 +0000 (Fri, 09 Apr 2004) | 1 line
+
+move ConfSource*.pm to Conf/*.pm, again for perl-OO-y-ish-ness...
+------------------------------------------------------------------------
+r9949 | felicity | 2004-04-09 18:54:54 +0000 (Fri, 09 Apr 2004) | 1 line
+
+make *Locker.pm into Locker/*.pm to be more perl-OO-y. also, make set_running_expire_token refresh the bayes db lock.
+------------------------------------------------------------------------
+r9940 | felicity | 2004-04-08 16:14:57 +0000 (Thu, 08 Apr 2004) | 1 line
+
+bug 3251: for things like sa-learn, the bayes db gets locked, then a bunch of learning occurs, then the db is unlocked. however, if the learning takes more than 600 seconds (10 mins), another process would see the lock as stale and break it, even though the original process is still going. now, everytime a lock is requested (which is still per message), the lock file will be 'refreshed' via utime().
+------------------------------------------------------------------------
+r9882 | quinlan | 2004-04-06 11:43:19 +0000 (Tue, 06 Apr 2004) | 3 lines
+
+promote T_URIBL_SC_SURBL to URIBL_SC_SURBL
+demote URIBL_DSBL to T_URIBL_DSBL
+
+------------------------------------------------------------------------
+r9881 | jm | 2004-04-06 02:49:33 +0000 (Tue, 06 Apr 2004) | 1 line
+
+added RHSBL lookup capability to URIBL plugin
+------------------------------------------------------------------------
+r9879 | jm | 2004-04-05 22:18:23 +0000 (Mon, 05 Apr 2004) | 1 line
+
+oops typo, and added T_X_MESSAGE_INFO
+------------------------------------------------------------------------
+r9878 | jm | 2004-04-05 21:48:49 +0000 (Mon, 05 Apr 2004) | 1 line
+
+promoted a stack of T_ rules
+------------------------------------------------------------------------
+r9875 | jm | 2004-04-05 18:49:11 +0000 (Mon, 05 Apr 2004) | 1 line
+
+promoted T_FORGED_MUA_THEBAT_BOUN; added a couple of RND-tag rules
+------------------------------------------------------------------------
+r9866 | quinlan | 2004-04-05 03:55:14 +0000 (Mon, 05 Apr 2004) | 2 lines
+
+fix taint error
+
+------------------------------------------------------------------------
+r9865 | quinlan | 2004-04-05 02:34:02 +0000 (Mon, 05 Apr 2004) | 2 lines
+
+fix scores for HTML_IMAGE_ONLY_*
+
+------------------------------------------------------------------------
+r9864 | felicity | 2004-04-05 02:23:38 +0000 (Mon, 05 Apr 2004) | 1 line
+
+bug 3212: we may sometimes attempt to free the same memory segment twice. do better checking to make sure we don't do that.
+------------------------------------------------------------------------
+r9859 | quinlan | 2004-04-04 08:57:28 +0000 (Sun, 04 Apr 2004) | 3 lines
+
+promote T_MISSING_SUBJECT to MISSING_SUBJECT
+rename DATE_MISSING to MISSING_DATE
+
+------------------------------------------------------------------------
+r9858 | quinlan | 2004-04-04 08:46:16 +0000 (Sun, 04 Apr 2004) | 5 lines
+
+replace HTML_IMAGE_ONLY* with new rules
+promote new HTML_SHORT_* rules
+remove T_HTML_TAGS_* for poor performance
+remove T_HTML_SHORT_IMG_* since it is the same as HTML_IMAGE_ONLY
+
+------------------------------------------------------------------------
+r9855 | quinlan | 2004-04-03 22:01:55 +0000 (Sat, 03 Apr 2004) | 2 lines
+
+comment
+
+------------------------------------------------------------------------
+r9851 | quinlan | 2004-04-03 08:48:26 +0000 (Sat, 03 Apr 2004) | 2 lines
+
+XBL changed A record format too! @%$#*&!
+
+------------------------------------------------------------------------
+r9850 | quinlan | 2004-04-03 08:35:33 +0000 (Sat, 03 Apr 2004) | 2 lines
+
+bug 3234: workaround XBL format change by reverting to using A queries only
+
+------------------------------------------------------------------------
+r9849 | quinlan | 2004-04-03 01:13:59 +0000 (Sat, 03 Apr 2004) | 3 lines
+
+revisions of the new HTML rules
+remove the T_LONGWORDS tests
+
+------------------------------------------------------------------------
+r9840 | felicity | 2004-04-01 23:47:48 +0000 (Thu, 01 Apr 2004) | 1 line
+
+reworked the mpart_alt_diff code a little bit, added more comments, added code to handle html w/ no words and only images...
+------------------------------------------------------------------------
+r9836 | quinlan | 2004-04-01 20:01:37 +0000 (Thu, 01 Apr 2004) | 2 lines
+
+add some potential HTML rules for short HTML spam
+
+------------------------------------------------------------------------
+r9832 | felicity | 2004-03-31 02:17:22 +0000 (Wed, 31 Mar 2004) | 1 line
+
+bug 3209: added MISSING_SUBJECT rule
+------------------------------------------------------------------------
+r9831 | quinlan | 2004-03-31 01:09:32 +0000 (Wed, 31 Mar 2004) | 2 lines
+
+minor change
+
+------------------------------------------------------------------------
+r9830 | felicity | 2004-03-30 23:53:21 +0000 (Tue, 30 Mar 2004) | 1 line
+
+bug 3227: 'To: display-name: foo@bar ;' wasn't being properly handled (see RFC2822 for 'group'). I modified the code to trim out the address/name in the group situation, as well as cleaning off the excess whitespace in the address/name situations.
+------------------------------------------------------------------------
+r9829 | felicity | 2004-03-30 17:24:32 +0000 (Tue, 30 Mar 2004) | 1 line
+
+bug 3230: don't allow 0-length tokens out of the bayes tokenizer
+------------------------------------------------------------------------
+r9812 | quinlan | 2004-03-29 21:15:28 +0000 (Mon, 29 Mar 2004) | 3 lines
+
+make sure we're getting good incremental value from the LONGWORDS sub-rules
+also aim for lower FP rate if possible
+
+------------------------------------------------------------------------
+r9810 | quinlan | 2004-03-29 20:52:13 +0000 (Mon, 29 Mar 2004) | 3 lines
+
+update of IADB rules
+still need to add http://www.isipp.com/iadb_query_sign_up1.php
+
+------------------------------------------------------------------------
+r9809 | felicity | 2004-03-29 17:59:15 +0000 (Mon, 29 Mar 2004) | 1 line
+
+bug 3217: let init_learner() call init() so that callers can do a rebuild_learner_caches() call without doing a learn() or check() first, and leave init() as a non-public function.
+------------------------------------------------------------------------
+r9808 | jm | 2004-03-29 07:07:32 +0000 (Mon, 29 Mar 2004) | 1 line
+
+rule_names.t again
+------------------------------------------------------------------------
+r9807 | jm | 2004-03-29 06:28:19 +0000 (Mon, 29 Mar 2004) | 1 line
+
+rule_names.t again. this time *removed* STOP_OFFERS, it's a rule with a .72 hitrate, with what can be a 128-character lookahead -- so terrible speed impact too
+------------------------------------------------------------------------
+r9806 | jm | 2004-03-29 05:24:13 +0000 (Mon, 29 Mar 2004) | 1 line
+
+bug 2415: avoid Bat forgery FPs due to MailMan adding MIME boundaries
+------------------------------------------------------------------------
+r9805 | jm | 2004-03-29 05:18:25 +0000 (Mon, 29 Mar 2004) | 1 line
+
+a little bit of cleanup on trusted_networks doco
+------------------------------------------------------------------------
+r9804 | quinlan | 2004-03-28 23:50:08 +0000 (Sun, 28 Mar 2004) | 2 lines
+
+add T_URIBL_SC_SURBL
+
+------------------------------------------------------------------------
+r9803 | quinlan | 2004-03-28 23:05:12 +0000 (Sun, 28 Mar 2004) | 4 lines
+
+revert combined backhair rules - I'm not really happy with the method
+revision SenderBase domain age/volume rules - test per-IP instead of
+ per-message
+
+------------------------------------------------------------------------
+r9802 | felicity | 2004-03-28 16:31:55 +0000 (Sun, 28 Mar 2004) | 1 line
+
+add the SA version to the debug output... I can't believe we didn't do that before.
+------------------------------------------------------------------------
+r9801 | felicity | 2004-03-28 16:26:04 +0000 (Sun, 28 Mar 2004) | 1 line
+
+allow sa-learn --force-expire or sa-learn --expire to be used...
+------------------------------------------------------------------------
+r9800 | zelgadis | 2004-03-28 09:03:51 +0000 (Sun, 28 Mar 2004) | 2 lines
+
+Bug #3223: RelayCountry: IP::Country::Fast can return undef
+
+------------------------------------------------------------------------
+r9799 | quinlan | 2004-03-28 06:17:46 +0000 (Sun, 28 Mar 2004) | 6 lines
+
+merge T_FROM_EUDORA and FAKE_FROM_ALTAVISTA into FROM_NONSENDING_DOMAIN
+delete FAKE_REPTO_ALTAVISTA and FAKE_ENVFROM_ALTAVISTA since
+ FROM_NONSENDING_DOMAIN hits everything they do
+delete T_FROM_DELPHI_BASIC - some FPs and it's a real company!
+delete extra font rules not needed
+
+------------------------------------------------------------------------
+r9798 | felicity | 2004-03-28 04:16:40 +0000 (Sun, 28 Mar 2004) | 1 line
+
+added a little more commenting on the default to ham mode...
+------------------------------------------------------------------------
+r9797 | felicity | 2004-03-28 04:13:24 +0000 (Sun, 28 Mar 2004) | 1 line
+
+3.0 requires perl 5.6.1, so remove more code that is kluging around 5.005, etc. also, found that ArchiveIterator was unnecessarily figuring out the class letter (h vs s) more often than was required...
+------------------------------------------------------------------------
+r9796 | quinlan | 2004-03-28 02:53:11 +0000 (Sun, 28 Mar 2004) | 2 lines
+
+keep wheedling down backhair
+
+------------------------------------------------------------------------
+r9795 | quinlan | 2004-03-28 00:15:25 +0000 (Sun, 28 Mar 2004) | 3 lines
+
+promote T_LONGWORDS_* to LONGWORDS
+delete T_BAYESBUSTER_LINE_* because it overlaps LONGWORDS almost completely
+
+------------------------------------------------------------------------
+r9794 | sidney | 2004-03-27 22:31:27 +0000 (Sat, 27 Mar 2004) | 4 lines
+
+bug #3192: anonymous sub spawned in spamd was sometimes not returning a value,
+causing a cryptic undef warning in spawn's exit call. Fixed by always
+returning some value.
+
+------------------------------------------------------------------------
+r9789 | quinlan | 2004-03-27 07:53:38 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+inf, undef, whatever
+
+------------------------------------------------------------------------
+r9788 | quinlan | 2004-03-27 07:52:31 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+typo
+
+------------------------------------------------------------------------
+r9787 | quinlan | 2004-03-27 07:44:34 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+HTML tests
+
+------------------------------------------------------------------------
+r9786 | quinlan | 2004-03-27 07:43:52 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+cosmetic
+
+------------------------------------------------------------------------
+r9785 | parker | 2004-03-27 07:38:56 +0000 (Sat, 27 Mar 2004) | 7 lines
+
+Bug 3214 - Spamd not using per-user Bayes settings
+
+When using -q/--sql-config or --ldap-config spamd was not switching to
+the proper user. Added call to signal_user_changed when those are
+set.
+
+
+------------------------------------------------------------------------
+r9784 | quinlan | 2004-03-27 07:22:33 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+iterate on T_BAYESBUSTER_LINE_*
+
+------------------------------------------------------------------------
+r9783 | quinlan | 2004-03-27 07:11:30 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+fix a rule_names.t failure
+
+------------------------------------------------------------------------
+r9782 | quinlan | 2004-03-27 07:03:38 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+remove descriptions for deleted rules
+
+------------------------------------------------------------------------
+r9781 | quinlan | 2004-03-27 07:02:41 +0000 (Sat, 27 Mar 2004) | 5 lines
+
+remove HTML_COMMENT_EMAIL (low performer)
+remove HTML_COMMENT_8BITS (locale-specific, 0.5 S/O)
+promote T_HTML_COMMENT_TINY6 to HTML_COMMENT_SHORT (T_HTML_TINY_COMMENTS had
+ better accuracy, but seemed too specific/fragile)
+
+------------------------------------------------------------------------
+r9780 | quinlan | 2004-03-27 06:47:51 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+finally trim down the LONGWORDS rules
+
+------------------------------------------------------------------------
+r9779 | quinlan | 2004-03-27 06:35:22 +0000 (Sat, 27 Mar 2004) | 3 lines
+
+remove extra low contrast tests
+more backhair stuff
+
+------------------------------------------------------------------------
+r9778 | quinlan | 2004-03-27 05:03:29 +0000 (Sat, 27 Mar 2004) | 3 lines
+
+fix undef readline bug
+cosmetic changes
+
+------------------------------------------------------------------------
+r9777 | quinlan | 2004-03-27 05:02:36 +0000 (Sat, 27 Mar 2004) | 2 lines
+
+get count right
+
+------------------------------------------------------------------------
+r9744 | jm | 2004-03-25 18:16:44 +0000 (Thu, 25 Mar 2004) | 1 line
+
+number of tests increased
+------------------------------------------------------------------------
+r9740 | felicity | 2004-03-24 20:18:58 +0000 (Wed, 24 Mar 2004) | 1 line
+
+remove some unneeded anon subroutines in get_tag() code.
+------------------------------------------------------------------------
+r9739 | felicity | 2004-03-24 19:54:31 +0000 (Wed, 24 Mar 2004) | 1 line
+
+due to cpan's 'issues' with x.y.z version numbers, our internal numbering scheme will be based on perl's (x.yyyzzz), and convert to x.y.z for display... also added coded, commented out for now, to allow 'require_version 3.0' to work for all 3.0.z versions.
+------------------------------------------------------------------------
+r9737 | quinlan | 2004-03-24 07:02:11 +0000 (Wed, 24 Mar 2004) | 2 lines
+
+restoring tabs mysteriously removed
+
+------------------------------------------------------------------------
+r9736 | quinlan | 2004-03-24 06:54:55 +0000 (Wed, 24 Mar 2004) | 2 lines
+
+some SenderBase rules for testing this weekend
+
+------------------------------------------------------------------------
+r9735 | quinlan | 2004-03-24 06:10:31 +0000 (Wed, 24 Mar 2004) | 2 lines
+
+don't warn if a SenderBase field is undefined
+
+------------------------------------------------------------------------
+r9734 | jm | 2004-03-24 05:57:19 +0000 (Wed, 24 Mar 2004) | 1 line
+
+bug 2326: add support for IPv4-in-v6 addrs and couriertcpd
+------------------------------------------------------------------------
+r9733 | jm | 2004-03-24 05:48:18 +0000 (Wed, 24 Mar 2004) | 1 line
+
+bug 3173: add invisible tokens with special prefix
+------------------------------------------------------------------------
+r9732 | jm | 2004-03-24 05:31:09 +0000 (Wed, 24 Mar 2004) | 1 line
+
+updated BUGGY_CGI for newer NMS FormMail releases, thx to Jonathan Stowe
+------------------------------------------------------------------------
+r9731 | jm | 2004-03-24 04:12:14 +0000 (Wed, 24 Mar 2004) | 1 line
+
+bug 2342: remove vestigial descriptions for removed rules
+------------------------------------------------------------------------
+r9730 | quinlan | 2004-03-24 03:45:19 +0000 (Wed, 24 Mar 2004) | 2 lines
+
+bug 3153: promote T_HTML_PLAIN_TEXT to HTML_MISSING_CTYPE
+
+------------------------------------------------------------------------
+r9729 | quinlan | 2004-03-24 03:42:40 +0000 (Wed, 24 Mar 2004) | 4 lines
+
+promote T_RCVD_HELO_NE_IP1 to RCVD_HELO_IP_MISMATCH
+ 9.887 12.1696 0.0051 1.000 0.96 0.01 T_RCVD_HELO_NE_IP1
+more tiny comment rules
+
+------------------------------------------------------------------------
+r9728 | jm | 2004-03-24 03:38:47 +0000 (Wed, 24 Mar 2004) | 1 line
+
+fixed up last rules_names.t failures; also ditched a whole load of rules that were no longer hitting anything useful -- with particular emphasis on obsolete ratware rules and low-accuracy, hit-runtime phrases.
+------------------------------------------------------------------------
+r9721 | felicity | 2004-03-23 15:55:53 +0000 (Tue, 23 Mar 2004) | 1 line
+
+make sure that bayes journal entries have the token in them appropriately ...
+------------------------------------------------------------------------
+r9696 | quinlan | 2004-03-23 01:31:00 +0000 (Tue, 23 Mar 2004) | 2 lines
+
+somewhat less confusing logic (and maybe more correct)
+
+------------------------------------------------------------------------
+r9686 | jm | 2004-03-22 20:59:10 +0000 (Mon, 22 Mar 2004) | 1 line
+
+updated MANIFEST
+------------------------------------------------------------------------
+r9685 | jm | 2004-03-22 20:56:47 +0000 (Mon, 22 Mar 2004) | 1 line
+
+oops. thanks Mike, caught it again
+------------------------------------------------------------------------
+r9684 | parker | 2004-03-22 19:33:38 +0000 (Mon, 22 Mar 2004) | 1 line
+
+bump number of tests
+------------------------------------------------------------------------
+r9660 | jm | 2004-03-22 08:16:09 +0000 (Mon, 22 Mar 2004) | 1 line
+
+SPF now providing no FPs or FNs for me that are bug-related; so, promoting to main ruleset. (if you're getting bug-related FPs/FNs, post them please.) also, moved into a plugin and added a test
+------------------------------------------------------------------------
+r9659 | jm | 2004-03-21 02:50:26 +0000 (Sun, 21 Mar 2004) | 1 line
+
+added new spamsign: T_SUBJ_2_NUM_PARENS
+------------------------------------------------------------------------
+r9634 | quinlan | 2004-03-19 20:59:38 +0000 (Fri, 19 Mar 2004) | 2 lines
+
+remove T_HTML_WEB_BUGS
+
+------------------------------------------------------------------------
+r9612 | quinlan | 2004-03-19 08:21:35 +0000 (Fri, 19 Mar 2004) | 2 lines
+
+bug 2996: HTML attribute testing
+
+------------------------------------------------------------------------
+r9611 | quinlan | 2004-03-19 08:06:57 +0000 (Fri, 19 Mar 2004) | 2 lines
+
+promote T_HTML_OBFUSCATE5* to HTML_OBFUSCATE* starting with 5% instead of 0%
+
+------------------------------------------------------------------------
+r9610 | quinlan | 2004-03-19 07:46:31 +0000 (Fri, 19 Mar 2004) | 4 lines
+
+promote T_DOMAIN_RATIO_026 to DOMAIN_RATIO
+at least cache URIs in get_uri_list()
+tidy some code
+
+------------------------------------------------------------------------
+r9606 | quinlan | 2004-03-19 05:24:02 +0000 (Fri, 19 Mar 2004) | 2 lines
+
+bug 3194: constants can't be used directly in RE
+
+------------------------------------------------------------------------
+r9598 | felicity | 2004-03-19 01:22:28 +0000 (Fri, 19 Mar 2004) | 1 line
+
+bug 3193: find_parts() wasn't going deep enough due to a bug introduced when find_parts() and _find_parts() got split... :(
+------------------------------------------------------------------------
+r9595 | quinlan | 2004-03-19 00:32:51 +0000 (Fri, 19 Mar 2004) | 2 lines
+
+add new tests to check for when a HELO IP address is wrong
+
+------------------------------------------------------------------------
+r9593 | quinlan | 2004-03-18 19:58:26 +0000 (Thu, 18 Mar 2004) | 3 lines
+
+remove HTML_COMMENT_RATIO
+fix T_HTML_WEB_BUGS regular expression
+
+------------------------------------------------------------------------
+r9592 | felicity | 2004-03-18 18:16:35 +0000 (Thu, 18 Mar 2004) | 1 line
+
+bug 3189: report_safe 1 and rewrite_headers weren't working together... fixed.
+------------------------------------------------------------------------
+r9591 | parker | 2004-03-18 16:51:03 +0000 (Thu, 18 Mar 2004) | 1 line
+
+upped number of tests
+------------------------------------------------------------------------
+r9590 | parker | 2004-03-18 07:24:05 +0000 (Thu, 18 Mar 2004) | 8 lines
+
+
+Bug 3187 - trim_rules broke
+
+trim_rules was not aware of the new rule priorities. Created helper
+methods in Conf.pm to hide the whole priority thing. Also moved
+trim_rules and add_meta_depends into Conf.pm.
+
+
+------------------------------------------------------------------------
+r9589 | quinlan | 2004-03-18 07:21:25 +0000 (Thu, 18 Mar 2004) | 2 lines
+
+delete score
+
+------------------------------------------------------------------------
+r9588 | quinlan | 2004-03-18 07:10:19 +0000 (Thu, 18 Mar 2004) | 2 lines
+
+invert weird negation
+
+------------------------------------------------------------------------
+r9587 | quinlan | 2004-03-18 06:59:41 +0000 (Thu, 18 Mar 2004) | 3 lines
+
+fix test failure - I don't really think we need to do a ham test here
+anyway
+
+------------------------------------------------------------------------
+r9586 | quinlan | 2004-03-18 06:34:00 +0000 (Thu, 18 Mar 2004) | 4 lines
+
+check_for_mime_text_only is not needed yet
+promote T_BILL_1618 to BILL_1618, remove UNDER_BILL_1618
+T_DIGEST_MULTIPLE is a net test
+
+------------------------------------------------------------------------
+r9582 | jm | 2004-03-18 04:58:24 +0000 (Thu, 18 Mar 2004) | 1 line
+
+bug 3048: noted sleep(1) libdb bug
+------------------------------------------------------------------------
+r9581 | jm | 2004-03-18 04:50:42 +0000 (Thu, 18 Mar 2004) | 1 line
+
+bug 3183: Received line with extra space misparsed
+------------------------------------------------------------------------
+r9580 | jm | 2004-03-18 04:48:08 +0000 (Thu, 18 Mar 2004) | 1 line
+
+oops, typo and too-long desc
+------------------------------------------------------------------------
+r9578 | jm | 2004-03-18 03:33:01 +0000 (Thu, 18 Mar 2004) | 1 line
+
+killed T_REPLY_WITHOUT_HDRS, made JS charcode one harder
+------------------------------------------------------------------------
+r9577 | jm | 2004-03-18 03:27:01 +0000 (Thu, 18 Mar 2004) | 1 line
+
+promoted remaining HELO_DYNAMIC rules, working well
+------------------------------------------------------------------------
+r9576 | jm | 2004-03-18 02:52:38 +0000 (Thu, 18 Mar 2004) | 1 line
+
+bug 3176: inviz text that should be ignored, was not; bug 3137: BAYES rules were modifying body, fixed; bug 3173: invisible text investigation, still ongoing
+------------------------------------------------------------------------
+r9575 | quinlan | 2004-03-18 00:11:22 +0000 (Thu, 18 Mar 2004) | 3 lines
+
+one last try before we bail on HTML_COMMENT_RATIO
+
+
+------------------------------------------------------------------------
+r9573 | quinlan | 2004-03-17 22:06:01 +0000 (Wed, 17 Mar 2004) | 2 lines
+
+try variation of HTML_WEB_BUGS instead since the other rules weren't good
+
+------------------------------------------------------------------------
+r9572 | sidney | 2004-03-17 17:42:36 +0000 (Wed, 17 Mar 2004) | 2 lines
+
+Bug 3021: Use whitelist_from_rcvd for a new forgery rule
+
+------------------------------------------------------------------------
+r9571 | parker | 2004-03-17 17:32:54 +0000 (Wed, 17 Mar 2004) | 10 lines
+
+Bug 2912 - Allow test order to be controlled by "priority" setting
+
+Added ability to assign priority to tests (ignored for RBL tests).
+Meta tests must have a priority of at least 500. Moved all (except
+starting of RBL queries) test methods into priority loop. Optimized
+test methods to make up for speed lost by running methods multiple
+times for each message. Set AWL test to priority 1000 to make sure
+that it runs after all other tests have been run.
+
+
+------------------------------------------------------------------------
+r9570 | sidney | 2004-03-17 15:43:15 +0000 (Wed, 17 Mar 2004) | 2 lines
+
+Add comment to rule_names test output so people will stop filing bug reports about it
+
+------------------------------------------------------------------------
+r9565 | quinlan | 2004-03-17 08:29:37 +0000 (Wed, 17 Mar 2004) | 2 lines
+
+oops
+
+------------------------------------------------------------------------
+r9564 | quinlan | 2004-03-17 08:26:20 +0000 (Wed, 17 Mar 2004) | 3 lines
+
+fix comment ratio code
+test new thresholds for HTML_COMMENT_RATIO test
+
+------------------------------------------------------------------------
+r9557 | quinlan | 2004-03-17 08:03:45 +0000 (Wed, 17 Mar 2004) | 2 lines
+
+minor change to UNWANTED_LANGUAGE_BODY
+
+------------------------------------------------------------------------
+r9556 | quinlan | 2004-03-17 08:00:05 +0000 (Wed, 17 Mar 2004) | 2 lines
+
+add img_src text array for tests
+
+------------------------------------------------------------------------
+r9547 | jm | 2004-03-17 05:56:23 +0000 (Wed, 17 Mar 2004) | 1 line
+
+oops, typo
+------------------------------------------------------------------------
+r9546 | jm | 2004-03-17 05:48:33 +0000 (Wed, 17 Mar 2004) | 1 line
+
+bug 2017: add rule for IMGs ending in /
+------------------------------------------------------------------------
+r9541 | jm | 2004-03-17 03:47:47 +0000 (Wed, 17 Mar 2004) | 1 line
+
+promoted the 1.0-S/O HELO_DYNAMIC_* rules; removed some FAKE_HELO rules with lousy scores.
+------------------------------------------------------------------------
+r9540 | jm | 2004-03-17 03:23:22 +0000 (Wed, 17 Mar 2004) | 1 line
+
+oops, didn't mean to check those in
+------------------------------------------------------------------------
+r9539 | jm | 2004-03-17 03:22:05 +0000 (Wed, 17 Mar 2004) | 1 line
+
+fix T_HELO_DYNAMIC fps; add a load more patterns too
+------------------------------------------------------------------------
+r9536 | quinlan | 2004-03-16 20:14:01 +0000 (Tue, 16 Mar 2004) | 3 lines
+
+add T_HTML_PLAIN_TEXT2 for kicks
+revisions of the drug rules
+
+------------------------------------------------------------------------
+r9535 | quinlan | 2004-03-16 20:08:13 +0000 (Tue, 16 Mar 2004) | 2 lines
+
+just provide the fragile HABEAS_SWE rule as a commented-out example
+
+------------------------------------------------------------------------
+r9531 | felicity | 2004-03-16 15:01:02 +0000 (Tue, 16 Mar 2004) | 1 line
+
+updates to the accessdb code -- some versions of makemap/postmap add nulls to the end of values and keys so look for that, accept \d+ as an error code (postfix).
+------------------------------------------------------------------------
+r9530 | felicity | 2004-03-16 14:58:43 +0000 (Tue, 16 Mar 2004) | 1 line
+
+bug 3180: if matching headers with RE, make sure to use ':[ \t]' and not ':\s' since the \s matches \n, which works great on blank headers, leaving the rest of the RE likely to explode in your face.
+------------------------------------------------------------------------
+r9523 | jm | 2004-03-16 06:34:32 +0000 (Tue, 16 Mar 2004) | 1 line
+
+plugin speedup: cache the list of plugins and method refs for each subroutine name
+------------------------------------------------------------------------
+r9520 | jm | 2004-03-16 06:23:50 +0000 (Tue, 16 Mar 2004) | 1 line
+
+fix basic_lint.t failure; multiple plugins registered meant that return value from one was overwritten by the next
+------------------------------------------------------------------------
+r9519 | jm | 2004-03-16 05:56:35 +0000 (Tue, 16 Mar 2004) | 1 line
+
+oops, test failure in hashcash.t
+------------------------------------------------------------------------
+r9515 | jm | 2004-03-16 03:58:17 +0000 (Tue, 16 Mar 2004) | 1 line
+
+bug 3027: bayes_dump_to_trusted_networks missing prereq warning
+------------------------------------------------------------------------
+r9509 | jm | 2004-03-16 03:02:50 +0000 (Tue, 16 Mar 2004) | 1 line
+
+dammit! conditional code due to plugin use and --lint behaviour for scores with no rules
+------------------------------------------------------------------------
+r9508 | jm | 2004-03-16 03:01:06 +0000 (Tue, 16 Mar 2004) | 1 line
+
+oops, score in the rules file
+------------------------------------------------------------------------
+r9507 | jm | 2004-03-16 03:00:58 +0000 (Tue, 16 Mar 2004) | 1 line
+
+oops, score in the rules file
+------------------------------------------------------------------------
+r9506 | jm | 2004-03-16 02:41:24 +0000 (Tue, 16 Mar 2004) | 1 line
+
+Hashcash moved to main ruleset. Scores may need manual tweaking
+------------------------------------------------------------------------
+r9505 | jm | 2004-03-16 02:40:33 +0000 (Tue, 16 Mar 2004) | 1 line
+
+look for init.pre in system rules dir if not in /etc/mail; used to support running from build dir
+------------------------------------------------------------------------
+r9499 | jm | 2004-03-15 23:09:38 +0000 (Mon, 15 Mar 2004) | 1 line
+
+oops, thx Mike
+------------------------------------------------------------------------
+r9491 | jm | 2004-03-15 08:37:39 +0000 (Mon, 15 Mar 2004) | 1 line
+
+Theo, pls review. think you meant this; avoids undef errors
+------------------------------------------------------------------------
+r9490 | jm | 2004-03-15 08:34:25 +0000 (Mon, 15 Mar 2004) | 1 line
+
+moved hashcash code into a plugin (bug 3170)
+------------------------------------------------------------------------
+r9489 | jm | 2004-03-15 08:34:04 +0000 (Mon, 15 Mar 2004) | 1 line
+
+moved hashcash code into a plugin (bug 3170)
+------------------------------------------------------------------------
+r9485 | jm | 2004-03-15 06:07:57 +0000 (Mon, 15 Mar 2004) | 1 line
+
+added init.pre file. NOTE: filename will probably change based on review, so get making suggestions ;)
+------------------------------------------------------------------------
+r9482 | jm | 2004-03-15 05:45:03 +0000 (Mon, 15 Mar 2004) | 1 line
+
+allow user override of plugins through a specially-treated file, '/etc/mail/spamassassin/init.pre'
+------------------------------------------------------------------------
+r9481 | jm | 2004-03-15 05:22:17 +0000 (Mon, 15 Mar 2004) | 1 line
+
+adding test for bug 2906 - Courier received header format
+------------------------------------------------------------------------
+r9480 | jm | 2004-03-15 05:16:39 +0000 (Mon, 15 Mar 2004) | 1 line
+
+bug 1923: multicast and other 'very reserved' IANA addresses should not appear in Received headers; they're forgeries. also noted some failures to catch those in Constants.pm
+------------------------------------------------------------------------
+r9460 | jm | 2004-03-14 05:49:08 +0000 (Sun, 14 Mar 2004) | 1 line
+
+bug 2129: bayes tweak -- use only visible text for Bayesian classification and learning, in other words ignore invisible or low-contrast text.
+------------------------------------------------------------------------
+r9448 | jm | 2004-03-14 02:05:14 +0000 (Sun, 14 Mar 2004) | 1 line
+
+22-char limit
+------------------------------------------------------------------------
+r9447 | jm | 2004-03-14 01:16:48 +0000 (Sun, 14 Mar 2004) | 1 line
+
+bug 2129: results of new Bayes tweaking -- numeric token decomposition no longer useful, dropped; decomposed text tokens (Foo? is duped as foo? Foo foo) added; dual header-mapping tweaks added for Message-Id, In-Reply-To, From/To/Cc, X-Mailer and User-Agent headers; and Henry's fix for bug 3118 for dealing with unbalanced corpora.
+------------------------------------------------------------------------
+r9446 | jm | 2004-03-14 01:03:48 +0000 (Sun, 14 Mar 2004) | 1 line
+
+updated bayes-testing with corrected details on how to run a 10-fold cross-validation run for testing Bayes. Go forth and tweak -- please...
+------------------------------------------------------------------------
+r9445 | jm | 2004-03-14 00:52:59 +0000 (Sun, 14 Mar 2004) | 1 line
+
+added test case for bug 3004
+------------------------------------------------------------------------
+r9433 | felicity | 2004-03-13 17:52:39 +0000 (Sat, 13 Mar 2004) | 1 line
+
+bug 3167: a few more places needed to be aware that bayes_score can be undef now via bug 2279
+------------------------------------------------------------------------
+r9402 | quinlan | 2004-03-13 06:58:39 +0000 (Sat, 13 Mar 2004) | 2 lines
+
+add IADB trusted
+
+------------------------------------------------------------------------
+r9397 | quinlan | 2004-03-13 05:27:43 +0000 (Sat, 13 Mar 2004) | 2 lines
+
+bug 3128: change AUTHOR to project
+
+------------------------------------------------------------------------
+r9396 | felicity | 2004-03-13 05:08:10 +0000 (Sat, 13 Mar 2004) | 1 line
+
+Making BayesStoreSQL and BayesStoreDBM to BayesStore/SQL and BayesStore/DBM ... It's a bit cleaner this way. :)
+------------------------------------------------------------------------
+r9394 | quinlan | 2004-03-13 02:18:27 +0000 (Sat, 13 Mar 2004) | 2 lines
+
+more HTML obfuscation
+
+------------------------------------------------------------------------
+r9392 | quinlan | 2004-03-13 00:04:55 +0000 (Sat, 13 Mar 2004) | 2 lines
+
+assume this paren is wanted
+
+------------------------------------------------------------------------
+r9391 | felicity | 2004-03-12 21:07:34 +0000 (Fri, 12 Mar 2004) | 1 line
+
+update sa-learn docs. rename (but keep backwards compat support for) --rebuild and --no-rebuild to be the more accurate --sync and --nosync since the option specifies that a journal sync should/should not occur... there's really no rebuilding involved.
+------------------------------------------------------------------------
+r9390 | felicity | 2004-03-12 21:06:07 +0000 (Fri, 12 Mar 2004) | 1 line
+
+bug 3164: 2.6x allowed old databaes to be used for check(), dump(), etc (anything that didn't write to the db). for 3.0 we're dropping that support -- databases must now be upgraded before they're usable.
+------------------------------------------------------------------------
+r9389 | quinlan | 2004-03-12 20:46:12 +0000 (Fri, 12 Mar 2004) | 2 lines
+
+more work on HTML obfuscation rules
+
+------------------------------------------------------------------------
+r9388 | felicity | 2004-03-12 20:31:17 +0000 (Fri, 12 Mar 2004) | 1 line
+
+no more PORN_4, so kill the score.
+------------------------------------------------------------------------
+r9387 | quinlan | 2004-03-12 19:00:18 +0000 (Fri, 12 Mar 2004) | 2 lines
+
+bug 3163: probable improvement of HTML obfuscation test
+
+------------------------------------------------------------------------
+r9386 | felicity | 2004-03-12 16:44:31 +0000 (Fri, 12 Mar 2004) | 1 line
+
+bug 3155: parts which claim to be message/rfc822 but have no body content would cause the parser to wait on STDIN... if there is no body content, don't bother parsing into a subtree.
+------------------------------------------------------------------------
+r9385 | parker | 2004-03-12 16:15:05 +0000 (Fri, 12 Mar 2004) | 1 line
+
+Bug 3162: document noautolearn tflag
+------------------------------------------------------------------------
+r9384 | climent | 2004-03-12 13:12:16 +0000 (Fri, 12 Mar 2004) | 2 lines
+
+Uploading Spanish ruls...
+
+------------------------------------------------------------------------
+r9381 | quinlan | 2004-03-12 07:07:13 +0000 (Fri, 12 Mar 2004) | 2 lines
+
+some drug rules
+
+------------------------------------------------------------------------
+r9374 | jm | 2004-03-12 02:18:43 +0000 (Fri, 12 Mar 2004) | 1 line
+
+moved PORN URI rules to 20_porn.cf, and shortened descs
+------------------------------------------------------------------------
+r9373 | jm | 2004-03-12 01:21:47 +0000 (Fri, 12 Mar 2004) | 1 line
+
+bug 3150: removed PORN_4 and split up into the worthwhile sub-rule parts; dropped the 'teen' rule and a few long-dead spammer domains
+------------------------------------------------------------------------
+r9372 | jm | 2004-03-12 00:56:42 +0000 (Fri, 12 Mar 2004) | 1 line
+
+bug 2713: removed code from Dustin Sallings, since we never got a CLA on-paper and it's holding up ASF migration. This means the --virtual-config switch is now gone; but --virtual-config-dir is still present, so users can migrate to that
+------------------------------------------------------------------------
+r9371 | zelgadis | 2004-03-12 00:56:12 +0000 (Fri, 12 Mar 2004) | 4 lines
+
+I modified this intead of /etc/mail/spamassassin/local.cf; I feel so stupid.
+
+*whacks self with a clue-stick*
+
+------------------------------------------------------------------------
+r9364 | felicity | 2004-03-11 23:26:30 +0000 (Thu, 11 Mar 2004) | 1 line
+
+bug 1843: try less complex S.1618 matching
+------------------------------------------------------------------------
+r9362 | felicity | 2004-03-11 22:42:59 +0000 (Thu, 11 Mar 2004) | 1 line
+
+bug 2168: look for malformed/non-existent mozilla user-agent ...
+------------------------------------------------------------------------
+r9361 | felicity | 2004-03-11 22:30:11 +0000 (Thu, 11 Mar 2004) | 1 line
+
+bug 2279: bayes shouldn't use 0.5 as an error condition. leave that up to the calling functions. we now return undef when a probability can't be calculated. the BAYES_* rules can now include 0.5 as that's a valid probabality now.
+------------------------------------------------------------------------
+r9360 | felicity | 2004-03-11 18:52:14 +0000 (Thu, 11 Mar 2004) | 1 line
+
+bug 3112: if at least one bayes token was found in the db, but they were all discounted due to their overall strength, we'd potentially continue and get division by zero errors when calculating probability. so exit out if there are no 'strong' tokens available.
+------------------------------------------------------------------------
+r9359 | felicity | 2004-03-11 18:40:30 +0000 (Thu, 11 Mar 2004) | 1 line
+
+bug 3152: ldap/README had attr and scope in the wrong order in the ldap url. corrected to be compliant with rfc 2255.
+------------------------------------------------------------------------
+r9348 | zelgadis | 2004-03-11 05:07:36 +0000 (Thu, 11 Mar 2004) | 2 lines
+
+Bug 3153: Detect HTML pretending to be text
+
+------------------------------------------------------------------------
+r9347 | felicity | 2004-03-11 04:22:46 +0000 (Thu, 11 Mar 2004) | 1 line
+
+get_magic_re can just use db_version ...
+------------------------------------------------------------------------
+r9346 | felicity | 2004-03-11 03:28:15 +0000 (Thu, 11 Mar 2004) | 1 line
+
+tokenize_line was making lots of subroutine calls to get_magic_re in its tight loop of token checking. let's just make the call once and cache it.
+------------------------------------------------------------------------
+r9345 | quinlan | 2004-03-10 23:26:55 +0000 (Wed, 10 Mar 2004) | 2 lines
+
+even tinier
+
+------------------------------------------------------------------------
+r9344 | quinlan | 2004-03-10 23:25:05 +0000 (Wed, 10 Mar 2004) | 2 lines
+
+very tiny speed-up
+
+------------------------------------------------------------------------
+r9340 | quinlan | 2004-03-10 08:25:46 +0000 (Wed, 10 Mar 2004) | 2 lines
+
+revise Habeas rules
+
+------------------------------------------------------------------------
+r9337 | jm | 2004-03-10 05:58:37 +0000 (Wed, 10 Mar 2004) | 1 line
+
+bug 3121: added test suite for Received-header parser
+------------------------------------------------------------------------
+r9330 | felicity | 2004-03-10 03:52:57 +0000 (Wed, 10 Mar 2004) | 1 line
+
+move the receive_date code after the probability check in case we exit before using the atime value
+------------------------------------------------------------------------
+r9329 | jm | 2004-03-10 03:44:17 +0000 (Wed, 10 Mar 2004) | 1 line
+
+bug 2811: Polish translations from Jerzy Szczudlowski and radek at alter dot pl
+------------------------------------------------------------------------
+r9328 | jm | 2004-03-10 03:31:40 +0000 (Wed, 10 Mar 2004) | 1 line
+
+URIBL: stop querying against XBL, it provides 0 hits
+------------------------------------------------------------------------
+r9327 | felicity | 2004-03-10 03:17:43 +0000 (Wed, 10 Mar 2004) | 1 line
+
+fix a typo and make dump more efficient (use while/each instead of foreach).
+------------------------------------------------------------------------
+r9321 | felicity | 2004-03-10 00:53:49 +0000 (Wed, 10 Mar 2004) | 1 line
+
+fix up a few small issues related to the previous bayes changes
+------------------------------------------------------------------------
+r9320 | felicity | 2004-03-10 00:48:54 +0000 (Wed, 10 Mar 2004) | 1 line
+
+make the bayes scan/tokenize code work a bit more straight forward. for instance, have tokenize() uniq the list of tokens, not each function which calls it...
+------------------------------------------------------------------------
+r9319 | felicity | 2004-03-09 23:57:12 +0000 (Tue, 09 Mar 2004) | 1 line
+
+fix up a comment...
+------------------------------------------------------------------------
+r9316 | felicity | 2004-03-09 23:34:23 +0000 (Tue, 09 Mar 2004) | 1 line
+
+put the html results in metadata!
+------------------------------------------------------------------------
+r9315 | quinlan | 2004-03-09 20:39:16 +0000 (Tue, 09 Mar 2004) | 2 lines
+
+promote new double IP rules
+
+------------------------------------------------------------------------
+r9314 | quinlan | 2004-03-09 20:30:05 +0000 (Tue, 09 Mar 2004) | 2 lines
+
+add net tflag
+
+------------------------------------------------------------------------
+r9309 | felicity | 2004-03-09 07:38:25 +0000 (Tue, 09 Mar 2004) | 1 line
+
+made the --mbox docs a little more verbose
+------------------------------------------------------------------------
+r9307 | felicity | 2004-03-09 06:04:09 +0000 (Tue, 09 Mar 2004) | 1 line
+
+bug 3077: more work to speed up remove_spamassassin_markup... 'report_safe 0' now saves a pristine copy of the header to X-Spam-Prev-Header, and remove_sa_markup properly moves it back. it also keep backward compatibility for pre-3.0's rewrite headers, which unfortunately means the config is loaded. but now we only load the config when necessary (ie: if the message(s) are report_safe encapsulated, no config is read. ;) made a few other efficiency improvements, and fixed the stripmarkup test to deal with the new way the markup is removed.
+------------------------------------------------------------------------
+r9306 | quinlan | 2004-03-09 06:01:56 +0000 (Tue, 09 Mar 2004) | 2 lines
+
+bug 3146: add test rules for the ISIPP Accreditation Database
+
+------------------------------------------------------------------------
+r7059 | jm | 2004-03-09 03:14:36 +0000 (Tue, 09 Mar 2004) | 1 line
+
+fixed a constants-related test failure. also reimplemented code that needed reimplementation in Constants.pm, which I think will fix the ALL_IN_TRUSTED fps
+------------------------------------------------------------------------
+r7058 | jm | 2004-03-09 02:59:49 +0000 (Tue, 09 Mar 2004) | 1 line
+
+took Michael's advice and made the constants into REAL constants
+------------------------------------------------------------------------
+r7057 | quinlan | 2004-03-09 00:26:06 +0000 (Tue, 09 Mar 2004) | 2 lines
+
+a few changes to double IP rules
+
+------------------------------------------------------------------------
+r7056 | felicity | 2004-03-09 00:23:49 +0000 (Tue, 09 Mar 2004) | 1 line
+
+have ArchiveIterator -j1 --restart do a forking thing...
+------------------------------------------------------------------------
+r7055 | felicity | 2004-03-08 22:58:48 +0000 (Mon, 08 Mar 2004) | 1 line
+
+the rendered and decoded generation code is now in Message instead of Metadata
+------------------------------------------------------------------------
+r7054 | felicity | 2004-03-08 22:48:07 +0000 (Mon, 08 Mar 2004) | 1 line
+
+running in test mode caused confusion...
+------------------------------------------------------------------------
+r7053 | felicity | 2004-03-08 21:30:26 +0000 (Mon, 08 Mar 2004) | 1 line
+
+bug 3144: don't subparse message/* types if the whole message claims to be message/* ... also, make sure we parse the decoded version of the part...
+------------------------------------------------------------------------
+r7052 | quinlan | 2004-03-08 08:05:22 +0000 (Mon, 08 Mar 2004) | 3 lines
+
+a backhair variant
+revision of double IP rules
+
+------------------------------------------------------------------------
+r7051 | jm | 2004-03-08 07:10:19 +0000 (Mon, 08 Mar 2004) | 1 line
+
+updated MANIFEST
+------------------------------------------------------------------------
+r7050 | quinlan | 2004-03-08 06:53:57 +0000 (Mon, 08 Mar 2004) | 2 lines
+
+add backhair rules based on idea from Jennifer Wheeler and Adam Lopresto
+
+------------------------------------------------------------------------
+r7049 | jm | 2004-03-08 06:17:50 +0000 (Mon, 08 Mar 2004) | 1 line
+
+removed comment I meant to remove before ;)
+------------------------------------------------------------------------
+r7048 | jm | 2004-03-08 05:58:04 +0000 (Mon, 08 Mar 2004) | 1 line
+
+a whole load of rules promoted, several removed, including OPT_HEADER (which was giving a .64 S/O)
+------------------------------------------------------------------------
+r7047 | jm | 2004-03-08 04:58:59 +0000 (Mon, 08 Mar 2004) | 1 line
+
+trimmed down desc lengths
+------------------------------------------------------------------------
+r7046 | jm | 2004-03-08 04:53:19 +0000 (Mon, 08 Mar 2004) | 1 line
+
+added rule to catch 'String.fromCharCode(chinner[encroach] ^ despairs[encroach % Rumford]);' type spams
+------------------------------------------------------------------------
+r7045 | quinlan | 2004-03-08 04:42:19 +0000 (Mon, 08 Mar 2004) | 2 lines
+
+revision of drug rules
+
+------------------------------------------------------------------------
+r7044 | jm | 2004-03-08 04:31:11 +0000 (Mon, 08 Mar 2004) | 1 line
+
+super-easy recent pattern
+------------------------------------------------------------------------
+r7043 | duncf | 2004-03-07 22:28:21 +0000 (Sun, 07 Mar 2004) | 1 line
+
+Bug 2988: Patch from Bob Apthorpe for sa-stats.pl
+------------------------------------------------------------------------
+r7042 | felicity | 2004-03-07 08:51:33 +0000 (Sun, 07 Mar 2004) | 1 line
+
+bug 2632: if running spamassassin in test mode, force disable things like autowhitelist and bayes autolearn (things that have a memory...)
+------------------------------------------------------------------------
+r7041 | felicity | 2004-03-07 08:43:10 +0000 (Sun, 07 Mar 2004) | 1 line
+
+bug 2220: added test rule to check if more than 1 network digest check considers the mail spam
+------------------------------------------------------------------------
+r7040 | felicity | 2004-03-07 08:40:01 +0000 (Sun, 07 Mar 2004) | 1 line
+
+bug 2201: allow relative scores to be put in the configuration file
+------------------------------------------------------------------------
+r7039 | felicity | 2004-03-07 06:10:14 +0000 (Sun, 07 Mar 2004) | 1 line
+
+oops. wrong paren thing.
+------------------------------------------------------------------------
+r7038 | felicity | 2004-03-07 06:08:01 +0000 (Sun, 07 Mar 2004) | 1 line
+
+fix up a few small issues with the message/* parsing patch ...
+------------------------------------------------------------------------
+r7037 | felicity | 2004-03-07 06:05:09 +0000 (Sun, 07 Mar 2004) | 1 line
+
+bug 3069: parse message/* parts into a subtree of the message/* part. makes get_rendered/get_decoded only look for text parts (will go into the message tree), etc.
+------------------------------------------------------------------------
+r7036 | quinlan | 2004-03-07 05:31:02 +0000 (Sun, 07 Mar 2004) | 3 lines
+
+add some more AHBL subtests
+rename T_RCVD_IN_AHBL to __RCVD_IN_AHBL
+
+------------------------------------------------------------------------
+r7035 | quinlan | 2004-03-07 05:18:57 +0000 (Sun, 07 Mar 2004) | 2 lines
+
+bug 3129: only load a plugin once, fix from Matthew Cline
+
+------------------------------------------------------------------------
+r7034 | quinlan | 2004-03-06 22:51:26 +0000 (Sat, 06 Mar 2004) | 3 lines
+
+remove backup flag for rsync
+chmod after rsync
+
+------------------------------------------------------------------------
+r7033 | quinlan | 2004-03-06 20:19:01 +0000 (Sat, 06 Mar 2004) | 2 lines
+
+bug 3132: some more double IP rules to test
+
+------------------------------------------------------------------------
+r7032 | felicity | 2004-03-06 19:36:28 +0000 (Sat, 06 Mar 2004) | 1 line
+
+make the decoded/rendered body part functions a bit more efficient and streamlined.
+------------------------------------------------------------------------
+r7030 | felicity | 2004-03-06 06:05:50 +0000 (Sat, 06 Mar 2004) | 1 line
+
+let spamassassin report the number of messages examined as long as the examination doesn't produce output of its own.
+------------------------------------------------------------------------
+r7029 | quinlan | 2004-03-06 04:58:22 +0000 (Sat, 06 Mar 2004) | 2 lines
+
+fix another t/rule_names.t failure
+
+------------------------------------------------------------------------
+r7025 | felicity | 2004-03-06 00:47:28 +0000 (Sat, 06 Mar 2004) | 1 line
+
+bug 3127: autolearn now has a 'disabled' state which occurs if autolearning is, well, disabled.
+------------------------------------------------------------------------
+r7024 | felicity | 2004-03-05 23:47:47 +0000 (Fri, 05 Mar 2004) | 1 line
+
+fix small pod oopsie
+------------------------------------------------------------------------
+r7023 | felicity | 2004-03-05 23:45:03 +0000 (Fri, 05 Mar 2004) | 1 line
+
+fix more rule hitting itself issues...
+------------------------------------------------------------------------
+r7022 | felicity | 2004-03-05 23:15:46 +0000 (Fri, 05 Mar 2004) | 1 line
+
+bug 1890: spamassassin now uses ArchiveIterator. This lets us do things like 'spamassassin -d --mbox file1' and have the functionality go over the entire mbox file. same for checks, adding to white/black-lists, etc.
+------------------------------------------------------------------------
+r7021 | felicity | 2004-03-05 22:20:59 +0000 (Fri, 05 Mar 2004) | 1 line
+
+if we're looking at messages that aren't specified as spam or ham, message_array gets confused. so default to ham.
+------------------------------------------------------------------------
+r7020 | quinlan | 2004-03-05 21:18:12 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+fix a t/rule_names.t failure
+
+------------------------------------------------------------------------
+r7019 | quinlan | 2004-03-05 20:54:09 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+don't change @$body since it affects other tests!
+
+------------------------------------------------------------------------
+r7014 | parker | 2004-03-05 16:43:25 +0000 (Fri, 05 Mar 2004) | 7 lines
+
+PR3006 SQL user-prefs stopped working latest trunk
+ When parsing the Conf object expected to be able call methods off of
+ the main SA object. The problem is that we delete the object in
+ init so it's space can be re-used. This fix sets the main SA object
+ in conf before a call to parse the scores, and then deletes it
+ afterward.
+
+------------------------------------------------------------------------
+r7013 | parker | 2004-03-05 16:32:14 +0000 (Fri, 05 Mar 2004) | 1 line
+
+removed extra debug statement
+------------------------------------------------------------------------
+r7012 | quinlan | 2004-03-05 08:57:45 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+Net::DNS 0.34 or higher is required to use Mail::SPF::Query
+
+------------------------------------------------------------------------
+r7011 | quinlan | 2004-03-05 05:31:30 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+fix another t/rule_names.t failure
+
+------------------------------------------------------------------------
+r7010 | quinlan | 2004-03-05 05:24:11 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+fix another t/rule_names.t failure by changing OFFERS_ETC to SPECIAL_OFFERS
+
+------------------------------------------------------------------------
+r7009 | quinlan | 2004-03-05 05:19:38 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+just a bit of trimming since I'm stalled on the CSS stuff
+
+------------------------------------------------------------------------
+r7008 | parker | 2004-03-05 05:16:17 +0000 (Fri, 05 Mar 2004) | 15 lines
+
+PR1222: die on error connecting or querying the SQL database
+ If we are unable to connect to the SQL database or if there is an
+ error while performing the query, then we will return a service
+ unavailable error to spamc.
+
+PR1602/1079: Configurable SQL query/per domain defaults
+ Allow customizable SQL queries to retrieve user preferences.
+
+ This allows for all sorts of things, such as domain based overrides,
+ integration with other systems (ie Horde/IMP).
+
+Also removed use of GLOBAL username as a global override in
+ConfSourceSQL.
+
+
+------------------------------------------------------------------------
+r7005 | quinlan | 2004-03-05 03:54:45 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+some rule renaming to fix t/rule_names.t failures
+
+------------------------------------------------------------------------
+r7004 | quinlan | 2004-03-05 03:08:19 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+Subject: drug rules
+
+------------------------------------------------------------------------
+r7003 | quinlan | 2004-03-05 03:05:31 +0000 (Fri, 05 Mar 2004) | 2 lines
+
+replace VIAGRA with DRUG_ED_CAPS
+
+------------------------------------------------------------------------
+r7002 | jm | 2004-03-05 02:07:24 +0000 (Fri, 05 Mar 2004) | 23 lines
+
+a few changes from last night I hadn't checked in yet:
+
+- Fixed fd leak in URIBL code, which was pretty serious after a big mass-check
+ run. my bad!
+
+- URIBL code now correctly disables itself if mass-check --rules is used and it
+ is not selected; previously it did not.
+
+- added more paranoia to envelope-from retrieval, to deal with the case where
+ an intermediate relay added an X-Envelope-From header (for example) then
+ relayed into a mailing list, which in turn used a new env-from. old code
+ would have picked up X-Envelope-From, even though that was no longer
+ appropriate. New code checks for Received headers before that hdr, and
+ ignores it if they exist.
+
+- SPF now working well; only FPs I now get are from ISPs who've set up over-
+ permissive SPF records (*cough BlueYonder.co.uk cough*) and from a couple
+ of domains with buggy SPF records.
+
+- added another (infrequent) Received line format (I think it's TIS -- now NAI
+ -- fwtk).
+
+
+------------------------------------------------------------------------
+r7001 | quinlan | 2004-03-05 01:47:17 +0000 (Fri, 05 Mar 2004) | 5 lines
+
+never say never...
+- add domain spoofing test (getting good results now)
+- move uri_to_domain to Util.pm
+- drop #fragments from URIs
+
+------------------------------------------------------------------------
+r7000 | jm | 2004-03-04 07:21:38 +0000 (Thu, 04 Mar 2004) | 1 line
+
+bug 2603: stop-list for numeric Received tokens not working too well
+------------------------------------------------------------------------
+r6999 | jm | 2004-03-04 07:19:35 +0000 (Thu, 04 Mar 2004) | 1 line
+
+bug 2189: add bayes compression tokens for X-Spam-Relays-Trusted, X-Spam-Relays-Untrusted
+------------------------------------------------------------------------
+r6998 | jm | 2004-03-04 07:14:58 +0000 (Thu, 04 Mar 2004) | 1 line
+
+bug 3016: avoid serious delays from mails with non-resolvable domains
+------------------------------------------------------------------------
+r6997 | jm | 2004-03-04 06:46:55 +0000 (Thu, 04 Mar 2004) | 1 line
+
+added some header-order rules
+------------------------------------------------------------------------
+r6994 | quinlan | 2004-03-04 06:26:04 +0000 (Thu, 04 Mar 2004) | 2 lines
+
+drop the /i
+
+------------------------------------------------------------------------
+r6992 | jm | 2004-03-04 05:56:34 +0000 (Thu, 04 Mar 2004) | 1 line
+
+bug 2876: catch faked reply messages
+------------------------------------------------------------------------
+r6990 | jm | 2004-03-04 05:45:07 +0000 (Thu, 04 Mar 2004) | 1 line
+
+bug 2983: a mistimed timeout results in spurious errors in spamc
+------------------------------------------------------------------------
+r6989 | jm | 2004-03-04 05:35:36 +0000 (Thu, 04 Mar 2004) | 1 line
+
+bug 2809: added rule to work around PRIORITY_NO_NAME fps
+------------------------------------------------------------------------
+r6988 | jm | 2004-03-04 05:25:07 +0000 (Thu, 04 Mar 2004) | 1 line
+
+bug 2952: document :raw suffix
+------------------------------------------------------------------------
+r6987 | jm | 2004-03-04 05:16:31 +0000 (Thu, 04 Mar 2004) | 1 line
+
+added a rule for long hostnames in URIs
+------------------------------------------------------------------------
+r6984 | quinlan | 2004-03-04 02:50:36 +0000 (Thu, 04 Mar 2004) | 2 lines
+
+variations of some drug rules that FP a bit too much
+
+------------------------------------------------------------------------
+r6983 | quinlan | 2004-03-04 02:45:23 +0000 (Thu, 04 Mar 2004) | 2 lines
+
+promote T_RCVD_BY_IP_2 to RCVD_BY_IP
+
+------------------------------------------------------------------------
+r6982 | quinlan | 2004-03-04 02:38:23 +0000 (Thu, 04 Mar 2004) | 2 lines
+
+some drug rules
+
+------------------------------------------------------------------------
+r6981 | quinlan | 2004-03-04 02:21:50 +0000 (Thu, 04 Mar 2004) | 2 lines
+
+separate drug rules (enough for a separate file to make sense)
+
+------------------------------------------------------------------------
+r6980 | quinlan | 2004-03-04 02:14:52 +0000 (Thu, 04 Mar 2004) | 2 lines
+
+fix CELL_PHONE_BOOST self-hits
+
+------------------------------------------------------------------------
+r6979 | quinlan | 2004-03-04 01:01:21 +0000 (Thu, 04 Mar 2004) | 2 lines
+
+bug 887: t script to test that rules don't trigger on their own names
+
+------------------------------------------------------------------------
+r6978 | parker | 2004-03-03 20:57:00 +0000 (Wed, 03 Mar 2004) | 1 line
+
+opps, obviously wrong column in username index
+------------------------------------------------------------------------
+r6977 | quinlan | 2004-03-03 06:43:17 +0000 (Wed, 03 Mar 2004) | 3 lines
+
+promote T_FAKED_AUTHWARN_2 to X_AUTH_WARN_FAKED
+revisions of T_RCVD_BY_IP
+
+------------------------------------------------------------------------
+r6976 | quinlan | 2004-03-03 04:05:44 +0000 (Wed, 03 Mar 2004) | 2 lines
+
+remove some unnecessary translations and scores
+
+------------------------------------------------------------------------
+r6975 | quinlan | 2004-03-03 04:03:56 +0000 (Wed, 03 Mar 2004) | 7 lines
+
+remove some really low eval tests, 7day results:
+ 0.000 0.0000 0.0000 0.500 0.12 0.00 HTML_TAG_BALANCE_FONT
+ 0.013 0.0125 0.0129 0.493 0.11 2.90 FAKE_HELO_BIGFOOT
+ 0.009 0.0085 0.0103 0.453 0.09 0.07 HTML_TAG_BALANCE_A
+ 0.003 0.0015 0.0103 0.129 0.00 2.80 FAKE_HELO_USA_NET
+ 0.007 0.0015 0.0283 0.051 0.00 0.00 FAKE_HELO_NETZERO
+
+------------------------------------------------------------------------
+r6974 | quinlan | 2004-03-03 03:50:58 +0000 (Wed, 03 Mar 2004) | 2 lines
+
+remove some old spam signature rules
+
+------------------------------------------------------------------------
+r6973 | quinlan | 2004-03-03 03:32:47 +0000 (Wed, 03 Mar 2004) | 7 lines
+
+share same eval test between HTML comments and HTML titles
+promote new RCVD_DOUBLE_IP rules
+promote T_MIME_BOUND_DIGITS15 to MIME_BOUND_DIGITS_15
+demote T_RCVD_BY_IP
+demote T_FAKED_IP_IN_RCVD
+remove HTML_COMMENT_SKY
+
+------------------------------------------------------------------------
+r6972 | felicity | 2004-03-03 01:00:44 +0000 (Wed, 03 Mar 2004) | 1 line
+
+another update
+------------------------------------------------------------------------
+r6971 | felicity | 2004-03-02 22:08:28 +0000 (Tue, 02 Mar 2004) | 1 line
+
+update the update_devel script a little bit -- clean up more after itself, etc.
+------------------------------------------------------------------------
+r6970 | felicity | 2004-03-02 19:45:42 +0000 (Tue, 02 Mar 2004) | 1 line
+
+bug 2012: move anything that's not a check() out from check(). in this case, the report generation code -- moved to get_report().
+------------------------------------------------------------------------
+r6963 | jm | 2004-03-02 04:31:18 +0000 (Tue, 02 Mar 2004) | 1 line
+
+fixed MANIFEST bug
+------------------------------------------------------------------------
+r6962 | felicity | 2004-03-02 04:26:16 +0000 (Tue, 02 Mar 2004) | 1 line
+
+added some more meat to report_safe. specifically, look for report_safe 0 to kill 'x-spam-status: no' in the original message
+------------------------------------------------------------------------
+r6961 | felicity | 2004-03-02 03:16:31 +0000 (Tue, 02 Mar 2004) | 1 line
+
+bug 3117: merged sa-learn and CmdLearn together. set --dump to work even if use_bayes is 0. gave untie_db a return value. some bayes functions were returning 1 for error, fixed that around too...
+------------------------------------------------------------------------
+r6960 | jm | 2004-03-02 02:45:58 +0000 (Tue, 02 Mar 2004) | 1 line
+
+updated MANIFEST for other people's additions ;)
+------------------------------------------------------------------------
+r6959 | jm | 2004-03-02 02:40:51 +0000 (Tue, 02 Mar 2004) | 1 line
+
+added test to ensure various semantics of our rule types are working
+------------------------------------------------------------------------
+r6958 | quinlan | 2004-03-02 02:04:57 +0000 (Tue, 02 Mar 2004) | 2 lines
+
+remove a few rules
+
+------------------------------------------------------------------------
+r6957 | quinlan | 2004-03-02 01:06:14 +0000 (Tue, 02 Mar 2004) | 2 lines
+
+some basic tests for report_safe option
+
+------------------------------------------------------------------------
+r6956 | felicity | 2004-03-01 22:57:28 +0000 (Mon, 01 Mar 2004) | 1 line
+
+bug 3113: for some reason 'report_safe 0' wasn't removing the original X-Spam- headers before adding the new ones. doh!
+------------------------------------------------------------------------
+r6955 | quinlan | 2004-03-01 22:24:03 +0000 (Mon, 01 Mar 2004) | 2 lines
+
+bug fix
+
+------------------------------------------------------------------------
+r6954 | quinlan | 2004-03-01 22:22:28 +0000 (Mon, 01 Mar 2004) | 2 lines
+
+T_DOUBLE_IP_* evolves
+
+------------------------------------------------------------------------
+r6953 | quinlan | 2004-03-01 18:51:41 +0000 (Mon, 01 Mar 2004) | 3 lines
+
+bug 3114: add T_HTML_COMMENT_TAG
+add T_FAKED_AUTHWARN_2 to avoid using meta rule for this
+
+------------------------------------------------------------------------
+r6952 | quinlan | 2004-03-01 09:31:42 +0000 (Mon, 01 Mar 2004) | 5 lines
+
+fix test (seems to be working although trusted_networks inference
+seems to be having problems, 212.17.35.14 should be queried twice
+and get back two 127.0.0.1 hits, one query for DNSBL_TEST_WHITELIST
+and one for DNSBL_TEST_TOP)
+
+------------------------------------------------------------------------
+r6945 | jm | 2004-03-01 08:05:06 +0000 (Mon, 01 Mar 2004) | 1 line
+
+try to reduce impact of really slow DNS servers on SA's lookups
+------------------------------------------------------------------------
+r6944 | jm | 2004-03-01 07:07:09 +0000 (Mon, 01 Mar 2004) | 1 line
+
+fixed SPF failures; Received header parser was not working with new message object get_header() method's semantics
+------------------------------------------------------------------------
+r6943 | felicity | 2004-03-01 05:11:43 +0000 (Mon, 01 Mar 2004) | 1 line
+
+add more pod docs
+------------------------------------------------------------------------
+r6942 | parker | 2004-03-01 05:09:39 +0000 (Mon, 01 Mar 2004) | 1 line
+
+PR2457: Updated userpref create def, also re added the PHP admin stuff with a corrected URL
+------------------------------------------------------------------------
+r6941 | parker | 2004-03-01 04:07:29 +0000 (Mon, 01 Mar 2004) | 1 line
+
+PR3095: Update sql documentation to include PostgreSQL examples and instructions, parts from Feargal Reilly
+------------------------------------------------------------------------
+r6940 | parker | 2004-03-01 03:20:27 +0000 (Mon, 01 Mar 2004) | 1 line
+
+PR3091: Do not allow spam/ham counts to go negative on bayes tokens
+------------------------------------------------------------------------
+r6938 | jm | 2004-03-01 00:52:57 +0000 (Mon, 01 Mar 2004) | 1 line
+
+added T_FAKED_AUTHWARN rule
+------------------------------------------------------------------------
+r6937 | quinlan | 2004-02-29 23:42:29 +0000 (Sun, 29 Feb 2004) | 2 lines
+
+a few minor bug fixes
+
+------------------------------------------------------------------------
+r6936 | felicity | 2004-02-29 22:12:03 +0000 (Sun, 29 Feb 2004) | 1 line
+
+doing a bunch of code touchups... Made MsgNode and MsgMetadata Message::Node and Message::Metadata respectively. merged MsgParser into Message. made M::SA::parse() just call Message::new(). added/reworked some docuementation. found a few places where Message->finish() wasn't being called. etc.
+------------------------------------------------------------------------
+r6935 | felicity | 2004-02-29 20:06:03 +0000 (Sun, 29 Feb 2004) | 1 line
+
+bug 3100: if the fs fills up, we'll get a partial write failure to the bayes journal. we used to loop and keep attempting to write out the data, but this would crush systems... we'll now only try to write out the journal 5 times before failing out, we'll also truncate the journal upon write failure so we don't do a partial journal commit.
+------------------------------------------------------------------------
+r6934 | quinlan | 2004-02-29 06:29:13 +0000 (Sun, 29 Feb 2004) | 2 lines
+
+remember to add back the T_ prefix
+
+------------------------------------------------------------------------
+r6933 | quinlan | 2004-02-29 06:25:35 +0000 (Sun, 29 Feb 2004) | 2 lines
+
+a different spin on the DOUBLE_IP rules
+
+------------------------------------------------------------------------
+r6932 | quinlan | 2004-02-29 05:49:04 +0000 (Sun, 29 Feb 2004) | 2 lines
+
+die() if a meta test depends on itself
+
+------------------------------------------------------------------------
+r6931 | felicity | 2004-02-29 05:09:10 +0000 (Sun, 29 Feb 2004) | 1 line
+
+we prefer /(foo)?/ instead of /(foo|)/ ...
+------------------------------------------------------------------------
+r6930 | jm | 2004-02-29 02:50:06 +0000 (Sun, 29 Feb 2004) | 1 line
+
+oh yeah, LOCALHOST too
+------------------------------------------------------------------------
+r6929 | jm | 2004-02-29 02:36:50 +0000 (Sun, 29 Feb 2004) | 1 line
+
+bug 1399: cannot tell window-blurring and widget-defocusing apart, so dropping HTML_WIN_BLUR rule
+------------------------------------------------------------------------
+r6928 | jm | 2004-02-29 02:32:29 +0000 (Sun, 29 Feb 2004) | 1 line
+
+bug 2793: IP_ADDRESS re was matching IPv4-in-IPv6 addrs wrongly
+------------------------------------------------------------------------
+r6927 | quinlan | 2004-02-29 02:08:47 +0000 (Sun, 29 Feb 2004) | 2 lines
+
+some updates and fixes
+
+------------------------------------------------------------------------
+r6926 | felicity | 2004-02-29 02:01:44 +0000 (Sun, 29 Feb 2004) | 1 line
+
+bug 3108: make the mbox seperator not a header (take it out of pristine_headers, etc.) remove the 'look for the mbox seperator and remove it' code in a few sections of the code. added different report_safe remove markup tests to strip2.t, etc.
+------------------------------------------------------------------------
+r6925 | quinlan | 2004-02-29 01:37:35 +0000 (Sun, 29 Feb 2004) | 2 lines
+
+comments
+
+------------------------------------------------------------------------
+r6924 | jm | 2004-02-29 00:18:31 +0000 (Sun, 29 Feb 2004) | 1 line
+
+bug 3087: order of loadplugin conf lines was wrong
+------------------------------------------------------------------------
+r6923 | jm | 2004-02-29 00:04:38 +0000 (Sun, 29 Feb 2004) | 1 line
+
+bug 2973: possible replacement for OPT_HEADER to avoid fps
+------------------------------------------------------------------------
+r6922 | jm | 2004-02-28 23:32:26 +0000 (Sat, 28 Feb 2004) | 1 line
+
+bug 2525: workaround for IMS sending HTML-only mail
+------------------------------------------------------------------------
+r6921 | jm | 2004-02-28 22:57:32 +0000 (Sat, 28 Feb 2004) | 1 line
+
+bug 2164: catch pipe- and apostrophe obfuscation, thx to David Skoll
+------------------------------------------------------------------------
+r6918 | quinlan | 2004-02-28 03:11:33 +0000 (Sat, 28 Feb 2004) | 3 lines
+
+only process HTML elements
+stub subroutine for CSS styles
+
+------------------------------------------------------------------------
+r6916 | quinlan | 2004-02-27 22:08:06 +0000 (Fri, 27 Feb 2004) | 2 lines
+
+merge stuff into incubator status page
+
+------------------------------------------------------------------------
+r6914 | parker | 2004-02-27 20:19:13 +0000 (Fri, 27 Feb 2004) | 1 line
+
+Account for change in name of dns.t to dnsbl.t
+------------------------------------------------------------------------
+r6913 | jm | 2004-02-27 17:44:40 +0000 (Fri, 27 Feb 2004) | 1 line
+
+fix undef warnings
+------------------------------------------------------------------------
+r6912 | quinlan | 2004-02-27 09:21:46 +0000 (Fri, 27 Feb 2004) | 2 lines
+
+some revisions
+
+------------------------------------------------------------------------
+r6911 | quinlan | 2004-02-27 08:11:58 +0000 (Fri, 27 Feb 2004) | 2 lines
+
+move to new name
+
+------------------------------------------------------------------------
+r6910 | quinlan | 2004-02-27 08:11:40 +0000 (Fri, 27 Feb 2004) | 2 lines
+
+add new DNS tests for better coverage (some are currently failing)
+
+------------------------------------------------------------------------
+r6909 | parker | 2004-02-27 07:48:41 +0000 (Fri, 27 Feb 2004) | 1 line
+
+Fix double not
+------------------------------------------------------------------------
+r6908 | felicity | 2004-02-27 07:39:32 +0000 (Fri, 27 Feb 2004) | 1 line
+
+bug 2657: default body-based rule descriptions wouldn't specify between body, rawbody, full, and uri as expected. I merged got_body_pattern_hit and got_uri_pattern_hit into a single got_pattern_hit, which accepts a description prefix. now using BODY, RAW, FULL, and URI as appropriate.
+------------------------------------------------------------------------
+r6907 | felicity | 2004-02-27 06:49:09 +0000 (Fri, 27 Feb 2004) | 1 line
+
+bug 2836: scan_directories() would require that there be no whitespace in filenames, but spaces are valid in file names, so fix the RE to ignore what we don't want (starts with . or ,) and allow anything else.
+------------------------------------------------------------------------
+r6906 | felicity | 2004-02-27 06:37:34 +0000 (Fri, 27 Feb 2004) | 1 line
+
+fix description to fit within standard convention
+------------------------------------------------------------------------
+r6899 | jm | 2004-02-27 05:22:40 +0000 (Fri, 27 Feb 2004) | 2 lines
+
+updated MANIFEST
+
+------------------------------------------------------------------------
+r6898 | jm | 2004-02-27 05:07:57 +0000 (Fri, 27 Feb 2004) | 1 line
+
+bug 1375: do DNSBL lookups on URLs, implemented as plugin Mail::SpamAssassin::Plugin::URIDNSBL
+------------------------------------------------------------------------
+r6897 | jm | 2004-02-27 04:49:48 +0000 (Fri, 27 Feb 2004) | 1 line
+
+added documentation for the newly-added asynch-event-loop support plugin hooks
+------------------------------------------------------------------------
+r6896 | quinlan | 2004-02-27 04:43:49 +0000 (Fri, 27 Feb 2004) | 2 lines
+
+add line break for <div> tag
+
+------------------------------------------------------------------------
+r6895 | jm | 2004-02-27 04:42:47 +0000 (Fri, 27 Feb 2004) | 1 line
+
+bug 1375: do DNSBL lookups on URLs, implemented as plugin Mail::SpamAssassin::Plugin::URIDNSBL
+------------------------------------------------------------------------
+r6894 | felicity | 2004-02-27 02:42:12 +0000 (Fri, 27 Feb 2004) | 1 line
+
+remove scores for rules that don't exist, shorten some rule names to fit convention...
+------------------------------------------------------------------------
+r6893 | felicity | 2004-02-27 00:11:12 +0000 (Fri, 27 Feb 2004) | 1 line
+
+bug 3093: add lint tests for rule names, descriptions, scores, etc. fix some documentation...
+------------------------------------------------------------------------
+r6890 | parker | 2004-02-26 22:49:15 +0000 (Thu, 26 Feb 2004) | 1 line
+
+Add Makefile.old to the ignore list
+------------------------------------------------------------------------
+r6889 | parker | 2004-02-26 22:43:28 +0000 (Thu, 26 Feb 2004) | 1 line
+
+when reading a file specified with -f it was adding more than one target, one of which was bogus
+------------------------------------------------------------------------
+r6887 | parker | 2004-02-26 22:21:02 +0000 (Thu, 26 Feb 2004) | 1 line
+
+PR3079: Moved actual connect to tie_db_writable because when using spamd there are a couple of cases where a new BayesStoreSQL object are not created and we tried to use the database handle from before the fork
+------------------------------------------------------------------------
+r6880 | parker | 2004-02-26 16:14:24 +0000 (Thu, 26 Feb 2004) | 1 line
+
+If the DBI module is not found, do not offer the bayes/awl sql tests
+------------------------------------------------------------------------
+r6879 | quinlan | 2004-02-26 11:01:30 +0000 (Thu, 26 Feb 2004) | 2 lines
+
+trim down T_RM_BPT_LONGWORDS a bit
+
+------------------------------------------------------------------------
+r6861 | quinlan | 2004-02-26 04:25:25 +0000 (Thu, 26 Feb 2004) | 3 lines
+
+remove low performing rawbody and body rules based on nightly corpus
+results, not including userconf rules
+
+------------------------------------------------------------------------
+r6860 | quinlan | 2004-02-26 03:41:27 +0000 (Thu, 26 Feb 2004) | 3 lines
+
+remove T_SPRINTF_5X
+pretty up SPF rules
+
+------------------------------------------------------------------------
+r6859 | quinlan | 2004-02-26 03:22:56 +0000 (Thu, 26 Feb 2004) | 2 lines
+
+nuking some more
+
+------------------------------------------------------------------------
+r6858 | quinlan | 2004-02-26 03:10:52 +0000 (Thu, 26 Feb 2004) | 2 lines
+
+delete some poor rules
+
+------------------------------------------------------------------------
+r6857 | quinlan | 2004-02-26 00:35:55 +0000 (Thu, 26 Feb 2004) | 2 lines
+
+rule promotions
+
+------------------------------------------------------------------------
+r6856 | quinlan | 2004-02-26 00:24:00 +0000 (Thu, 26 Feb 2004) | 3 lines
+
+promote UNIQUE_WORDS test
+tweak variations on a theme of double IP
+
+------------------------------------------------------------------------
+r6855 | duncf | 2004-02-25 14:57:41 +0000 (Wed, 25 Feb 2004) | 1 line
+
+mass-check should be more verbose when dying if Time::ParseDate not installed
+------------------------------------------------------------------------
+r6854 | duncf | 2004-02-25 14:55:57 +0000 (Wed, 25 Feb 2004) | 1 line
+
+ArchiveIterator should pass the class of the message to wanted -- this is needed for proper support of -f with sa-learn
+------------------------------------------------------------------------
+r6846 | parker | 2004-02-24 07:12:27 +0000 (Tue, 24 Feb 2004) | 1 line
+
+Bring msgid test up to speed with code changes
+------------------------------------------------------------------------
+r6845 | quinlan | 2004-02-24 06:48:19 +0000 (Tue, 24 Feb 2004) | 2 lines
+
+add some T_RCVD_DOUBLE_IP variants
+
+------------------------------------------------------------------------
+r6844 | parker | 2004-02-24 06:40:22 +0000 (Tue, 24 Feb 2004) | 1 line
+
+Add missing # on comment
+------------------------------------------------------------------------
+r6843 | quinlan | 2004-02-24 06:14:49 +0000 (Tue, 24 Feb 2004) | 2 lines
+
+some test rules for bug 2901
+
+------------------------------------------------------------------------
+r6841 | quinlan | 2004-02-24 04:57:38 +0000 (Tue, 24 Feb 2004) | 2 lines
+
+another rule that doesn't hit any more -- too long since last release
+
+------------------------------------------------------------------------
+r6840 | quinlan | 2004-02-24 04:49:52 +0000 (Tue, 24 Feb 2004) | 2 lines
+
+promote, delete, promote, delete
+
+------------------------------------------------------------------------
+r6839 | quinlan | 2004-02-24 04:19:17 +0000 (Tue, 24 Feb 2004) | 2 lines
+
+some work on test rules
+
+------------------------------------------------------------------------
+r6838 | felicity | 2004-02-23 23:07:51 +0000 (Mon, 23 Feb 2004) | 1 line
+
+bug 3074: when rewriting headers, the code assumed 'name:<spaces>something' which isn't always going to happen as 'name:<tab>something' is valid in a message. fixed the RE to account for this.
+------------------------------------------------------------------------
+r6837 | quinlan | 2004-02-23 18:46:59 +0000 (Mon, 23 Feb 2004) | 3 lines
+
+okay, I think the function-based method is better, so let's go with it
+and explore that space a bit more precisely
+
+------------------------------------------------------------------------
+r6836 | quinlan | 2004-02-23 15:55:17 +0000 (Mon, 23 Feb 2004) | 2 lines
+
+increase timeout (60 seconds was timing out)
+
+------------------------------------------------------------------------
+r6835 | quinlan | 2004-02-23 15:54:59 +0000 (Mon, 23 Feb 2004) | 2 lines
+
+add way to ignore T_ tests
+
+------------------------------------------------------------------------
+r6834 | quinlan | 2004-02-23 08:18:56 +0000 (Mon, 23 Feb 2004) | 3 lines
+
+HTML title code update
+remove some unnecessary double quotes
+
+------------------------------------------------------------------------
+r6833 | quinlan | 2004-02-23 08:07:35 +0000 (Mon, 23 Feb 2004) | 2 lines
+
+some test rule work
+
+------------------------------------------------------------------------
+r6832 | quinlan | 2004-02-23 07:25:51 +0000 (Mon, 23 Feb 2004) | 2 lines
+
+best rule was _59A so I think we should explore the space a bit more
+
+------------------------------------------------------------------------
+r6831 | quinlan | 2004-02-23 07:10:41 +0000 (Mon, 23 Feb 2004) | 3 lines
+
+consolidate Yahoo! Message-ID forgery rules into one final candidate
+split out new Yahoo! redirector rules to figure out the FPs
+
+------------------------------------------------------------------------
+r6830 | quinlan | 2004-02-23 05:33:18 +0000 (Mon, 23 Feb 2004) | 2 lines
+
+a probably better slice of the unique words test
+
+------------------------------------------------------------------------
+r6829 | duncf | 2004-02-23 04:42:47 +0000 (Mon, 23 Feb 2004) | 1 line
+
+Fixes removal of rewritten heTo/From headers (bug 3075)
+------------------------------------------------------------------------
+r6820 | quinlan | 2004-02-22 07:02:16 +0000 (Sun, 22 Feb 2004) | 3 lines
+
+bug 3023: detecting random garbage in emails
+also tweak tests for illegal attributes
+
+------------------------------------------------------------------------
+r6819 | quinlan | 2004-02-22 01:56:21 +0000 (Sun, 22 Feb 2004) | 2 lines
+
+test out higher range on low contrast text
+
+------------------------------------------------------------------------
+r6818 | quinlan | 2004-02-21 23:58:29 +0000 (Sat, 21 Feb 2004) | 2 lines
+
+fix some typos
+
+------------------------------------------------------------------------
+r6817 | quinlan | 2004-02-21 23:56:43 +0000 (Sat, 21 Feb 2004) | 3 lines
+
+remove HTML_TAG_EXISTS_BASE
+promote T_HTML_TAG_EXISTS_MARQUEE to HTML_TAG_EXISTS_MARQUEE
+
+------------------------------------------------------------------------
+r6816 | quinlan | 2004-02-21 23:52:32 +0000 (Sat, 21 Feb 2004) | 2 lines
+
+fix selection of logs for weekly on Saturdays
+
+------------------------------------------------------------------------
+r6806 | quinlan | 2004-02-21 03:21:07 +0000 (Sat, 21 Feb 2004) | 3 lines
+
+promote T_MSGID_EVIL_32 to MSGID_RANDY
+new Yahoo! redirector apparently
+
+------------------------------------------------------------------------
+r6792 | quinlan | 2004-02-20 08:32:07 +0000 (Fri, 20 Feb 2004) | 2 lines
+
+some rules to try for Yahoo! forgeries
+
+------------------------------------------------------------------------
+r6789 | felicity | 2004-02-20 05:33:39 +0000 (Fri, 20 Feb 2004) | 1 line
+
+make report_as_spam and revoke_as_spam deal with specificially the message passed in... make sure when we call those functions that the message has SA markup removed. do more calls to finish() to remove noticable memory leaks(), specifically when we do remove_spamassassin_markup...
+------------------------------------------------------------------------
+r6788 | felicity | 2004-02-20 05:13:18 +0000 (Fri, 20 Feb 2004) | 1 line
+
+we don't need already_parsed since we have toparse. implemented MsgNode::finish() and have things call it to be a good housekeeper.
+------------------------------------------------------------------------
+r6787 | felicity | 2004-02-20 04:34:47 +0000 (Fri, 20 Feb 2004) | 1 line
+
+deal with already_parsed a little better
+------------------------------------------------------------------------
+r6786 | jm | 2004-02-20 03:35:49 +0000 (Fri, 20 Feb 2004) | 1 line
+
+try out the AHBL blocklists
+------------------------------------------------------------------------
+r6785 | quinlan | 2004-02-20 01:45:47 +0000 (Fri, 20 Feb 2004) | 2 lines
+
+return empty matches if message isn't long enough for language guessing
+
+------------------------------------------------------------------------
+r6784 | jm | 2004-02-19 23:47:10 +0000 (Thu, 19 Feb 2004) | 1 line
+
+split up MsgContainer into two new classes - Message, for top-level node, and MsgNode for sub-nodes. Message is a subclass of MsgNode with additional features to hold metadata, pristine message representation and top-level message-wide flags
+------------------------------------------------------------------------
+r6783 | jm | 2004-02-19 22:59:35 +0000 (Thu, 19 Feb 2004) | 1 line
+
+moved constants to new module
+------------------------------------------------------------------------
+r6782 | jm | 2004-02-19 22:56:34 +0000 (Thu, 19 Feb 2004) | 1 line
+
+did I check in?
+------------------------------------------------------------------------
+r6781 | jm | 2004-02-19 18:46:48 +0000 (Thu, 19 Feb 2004) | 1 line
+
+moved finish_metadata() call in PMS to finish()
+------------------------------------------------------------------------
+r6780 | jm | 2004-02-19 18:45:24 +0000 (Thu, 19 Feb 2004) | 1 line
+
+oops. make sure finish_metadata can be called multiple times
+------------------------------------------------------------------------
+r6779 | jm | 2004-02-19 18:41:47 +0000 (Thu, 19 Feb 2004) | 1 line
+
+some cleanup of last night's metadata code
+------------------------------------------------------------------------
+r6776 | quinlan | 2004-02-19 10:05:49 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+change two of the lesser subrules for evil message-id
+
+------------------------------------------------------------------------
+r6774 | jm | 2004-02-19 08:22:37 +0000 (Thu, 19 Feb 2004) | 1 line
+
+rule_tests passes again
+------------------------------------------------------------------------
+r6773 | jm | 2004-02-19 08:11:54 +0000 (Thu, 19 Feb 2004) | 1 line
+
+fixed 'make test' failures
+------------------------------------------------------------------------
+r6772 | jm | 2004-02-19 08:03:08 +0000 (Thu, 19 Feb 2004) | 1 line
+
+sorted MANIFEST
+------------------------------------------------------------------------
+r6771 | jm | 2004-02-19 08:02:32 +0000 (Thu, 19 Feb 2004) | 1 line
+
+added plugin to look up IP address of relays with IP::Country (bug 2575)
+------------------------------------------------------------------------
+r6768 | jm | 2004-02-19 07:34:09 +0000 (Thu, 19 Feb 2004) | 1 line
+
+metadata support; sa-learn can now learn languages and trusted/untrusted relays, so these can now be Bayes tokens
+------------------------------------------------------------------------
+r6767 | quinlan | 2004-02-19 05:44:32 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+one variant
+
+------------------------------------------------------------------------
+r6766 | felicity | 2004-02-19 05:29:36 +0000 (Thu, 19 Feb 2004) | 1 line
+
+deal with unneccessary hex uri encoding in addition to the decimal ones...
+------------------------------------------------------------------------
+r6765 | jm | 2004-02-19 04:58:18 +0000 (Thu, 19 Feb 2004) | 1 line
+
+added T_SILDENAFIL in response to bug 2572
+------------------------------------------------------------------------
+r6764 | quinlan | 2004-02-19 04:42:17 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+regular expression mistake
+
+------------------------------------------------------------------------
+r6763 | quinlan | 2004-02-19 04:19:35 +0000 (Thu, 19 Feb 2004) | 3 lines
+
+some teen variants
+also use tabs
+
+------------------------------------------------------------------------
+r6762 | felicity | 2004-02-19 04:10:31 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2897: 'sa-learn --rebuild' would always set --showdots due to passing a reference then an actual value...
+------------------------------------------------------------------------
+r6761 | quinlan | 2004-02-19 04:04:43 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+bug 2996: HTML attribute testing
+
+------------------------------------------------------------------------
+r6760 | felicity | 2004-02-19 03:59:16 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2816: redhat rc script 'status' didn't set the return value.
+------------------------------------------------------------------------
+r6759 | felicity | 2004-02-19 03:48:19 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2784: allow NICELEVEL to be specified for the redhat rc script to start spamd at a certain nice level
+------------------------------------------------------------------------
+r6758 | quinlan | 2004-02-19 03:07:17 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+allow negative numbers
+
+------------------------------------------------------------------------
+r6757 | felicity | 2004-02-19 02:58:57 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2414: allow the hostname (_HOSTNAME_) used in reports/header rewriting to be set via a config option since the actual system hostname may not be the hostname most people look for (ie: mail CNAME hostfoo, we should report mail not hostfoo ...)
+------------------------------------------------------------------------
+r6756 | jm | 2004-02-19 02:55:42 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2273: bayes_ignore_from, bayes_ignore_to added, thanks to David Koppelman
+------------------------------------------------------------------------
+r6755 | jm | 2004-02-19 02:52:03 +0000 (Thu, 19 Feb 2004) | 1 line
+
+possible Red Hat bug with smrsh noted, thanks to Jim Severino for the heads-up (bug 2393)
+------------------------------------------------------------------------
+r6754 | jm | 2004-02-19 02:45:11 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2935: PENIS_ENLARGE rule name rephrased more diplomatically, fix by Rich Puhek
+------------------------------------------------------------------------
+r6753 | felicity | 2004-02-19 02:33:17 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2401: apparently perl on Windows doesn't seem to know what SIGHUP is, so we need to protect our HUP handler calls.
+------------------------------------------------------------------------
+r6752 | jm | 2004-02-19 02:16:12 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2496: spamc usage should go to stderr
+------------------------------------------------------------------------
+r6751 | felicity | 2004-02-19 02:12:40 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2261: change the 'db not initialised' message to be more meaningful -- none of the message's tokens were found in the db.
+------------------------------------------------------------------------
+r6750 | mss | 2004-02-19 02:05:20 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+*grmbl*
+
+------------------------------------------------------------------------
+r6749 | mss | 2004-02-19 02:02:37 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+Better rounding "fix" plus a bit refactoring.
+
+------------------------------------------------------------------------
+r6748 | mss | 2004-02-19 01:56:39 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+Make the syslog-SIGPIPE-handler more local and make sure its always removed again (to avoid it being called if pipes to helper apps break).
+
+------------------------------------------------------------------------
+r6747 | duncf | 2004-02-19 01:32:10 +0000 (Thu, 19 Feb 2004) | 1 line
+
+Fix a few instances of _HITS_. Also, fix despamassassinification of _SCORE_ in spam with report_safe 0
+------------------------------------------------------------------------
+r6746 | duncf | 2004-02-19 01:24:48 +0000 (Thu, 19 Feb 2004) | 1 line
+
+Bug 1332: replace hits with points (or score) internally and externally... Some variable names have changed, notably $self->{hits} is now $self->{score}. Backwards compatibility is maintained where possible
+------------------------------------------------------------------------
+r6745 | felicity | 2004-02-19 01:03:15 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 3014: allow dbpath for everything, not just dump and import
+------------------------------------------------------------------------
+r6744 | jm | 2004-02-19 00:51:37 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 2377: T_URI_HTTP_TO_HEX_IP added
+------------------------------------------------------------------------
+r6743 | jm | 2004-02-19 00:47:36 +0000 (Thu, 19 Feb 2004) | 1 line
+
+added note about FDs re bug 2352
+------------------------------------------------------------------------
+r6742 | jm | 2004-02-19 00:45:08 +0000 (Thu, 19 Feb 2004) | 1 line
+
+added note about FDs re bug 2352
+------------------------------------------------------------------------
+r6741 | felicity | 2004-02-19 00:35:47 +0000 (Thu, 19 Feb 2004) | 1 line
+
+full tests get the raw header (no decoding). also fix the comment about that
+------------------------------------------------------------------------
+r6740 | mss | 2004-02-19 00:30:14 +0000 (Thu, 19 Feb 2004) | 2 lines
+
+bug 2607: Always round down to make the 5.0!=5.0-phenomenon go away (patch by Dan Kohn).
+
+------------------------------------------------------------------------
+r6739 | duncf | 2004-02-19 00:06:02 +0000 (Thu, 19 Feb 2004) | 1 line
+
+Added support for ? : operator in meta tests (Bug 3038)
+------------------------------------------------------------------------
+r6738 | felicity | 2004-02-19 00:04:32 +0000 (Thu, 19 Feb 2004) | 1 line
+
+bug 3061: if bayes_learn_to_journal was set (sa-learn --norebuild does this too), and no bayes DB already existed, the learn would fail because tie_db_readonly() wouldn't create a db if it didn't already exist. so we now failback to tieing r/w if r/o fails only if learn_to_journal is set. cleaned up a bit of bayesdbm.t as well.
+------------------------------------------------------------------------
+r6737 | quinlan | 2004-02-18 23:56:28 +0000 (Wed, 18 Feb 2004) | 2 lines
+
+bug 2950: various redirector rules
+
+------------------------------------------------------------------------
+r6736 | jm | 2004-02-18 23:36:45 +0000 (Wed, 18 Feb 2004) | 1 line
+
+added T_RATWARE_FAKED_AOL_UA from bug 2299
+------------------------------------------------------------------------
+r6735 | jm | 2004-02-18 23:34:13 +0000 (Wed, 18 Feb 2004) | 1 line
+
+bug 2300: avoid FPs on Exchange messages sent with Priority set to High
+------------------------------------------------------------------------
+r6734 | jm | 2004-02-18 23:29:28 +0000 (Wed, 18 Feb 2004) | 1 line
+
+bug 2542: translated X-Spam-Report header contains unencoded 8-bit chars
+------------------------------------------------------------------------
+r6733 | felicity | 2004-02-18 23:26:01 +0000 (Wed, 18 Feb 2004) | 1 line
+
+bug 3055: spammers are using the same message id to get around bayes being able to learn different messages. make the hash message-id the default now, but be backwards compatible with the seen db.
+------------------------------------------------------------------------
+r6729 | jm | 2004-02-18 23:07:49 +0000 (Wed, 18 Feb 2004) | 1 line
+
+bug 3044: IT_PATTERN for find_all_addrs_in_line() is too restrictive
+------------------------------------------------------------------------
+r6728 | quinlan | 2004-02-18 22:49:36 +0000 (Wed, 18 Feb 2004) | 2 lines
+
+bug 1780: fix for PORN_4
+
+------------------------------------------------------------------------
+r6727 | jm | 2004-02-18 22:09:20 +0000 (Wed, 18 Feb 2004) | 1 line
+
+use 127.0.0.1 IP addr instead of 'localhost' to avoid test failure reported in bug 2749
+------------------------------------------------------------------------
+r6726 | jm | 2004-02-18 22:03:22 +0000 (Wed, 18 Feb 2004) | 1 line
+
+bug 1677: generalise request header parsing in spamd for future new methods
+------------------------------------------------------------------------
+r6725 | jm | 2004-02-18 22:02:45 +0000 (Wed, 18 Feb 2004) | 1 line
+
+bonus-space-in-Received line rule spotted last week
+------------------------------------------------------------------------
+r6724 | quinlan | 2004-02-18 21:52:14 +0000 (Wed, 18 Feb 2004) | 2 lines
+
+bug 2891: Insert qpsmtpd parsing into Received.pm
+
+------------------------------------------------------------------------
+r6723 | felicity | 2004-02-18 21:05:38 +0000 (Wed, 18 Feb 2004) | 1 line
+
+MsgContainer->rendered() was calling the HTML renderer twice per part for no apparent reason.
+------------------------------------------------------------------------
+r6722 | mss | 2004-02-18 21:01:35 +0000 (Wed, 18 Feb 2004) | 2 lines
+
+bug 2927: Added possibility to skip messages if RELAYCLIENT is set (configure check to follow). Also get rid of some vars.
+
+------------------------------------------------------------------------
+r6721 | quinlan | 2004-02-18 20:51:19 +0000 (Wed, 18 Feb 2004) | 2 lines
+
+clean up hue sub, add white and black "hues"
+
+------------------------------------------------------------------------
+r6720 | quinlan | 2004-02-18 20:44:56 +0000 (Wed, 18 Feb 2004) | 3 lines
+
+add minimal support for <marquee> and a marquee rule
+turn hue code into a sub
+
+------------------------------------------------------------------------
+r6719 | quinlan | 2004-02-18 07:51:15 +0000 (Wed, 18 Feb 2004) | 2 lines
+
+remove some old URI code not needed now
+
+------------------------------------------------------------------------
+r6718 | quinlan | 2004-02-18 07:12:02 +0000 (Wed, 18 Feb 2004) | 3 lines
+
+tweak to guess whether getting a file or a file.offset
+(maybe this behavior should have a flag?)
+
+------------------------------------------------------------------------
+r6717 | quinlan | 2004-02-18 07:02:45 +0000 (Wed, 18 Feb 2004) | 2 lines
+
+more evil message-id rules to test
+
+------------------------------------------------------------------------
+r6716 | felicity | 2004-02-18 06:47:19 +0000 (Wed, 18 Feb 2004) | 1 line
+
+make M::SA::parse() not do the full recursive mime tree generation. let M::SA::MsgContainer::find_parts() do it as necessary. this makes check() run at about the same speed, but things like 'spamassassin -d' should run a little faster due to the reduced amount of work that needs to occur.
+------------------------------------------------------------------------
+r6715 | quinlan | 2004-02-18 05:52:52 +0000 (Wed, 18 Feb 2004) | 7 lines
+
+add min/max font size support using new font code
+add <basefont> support
+add remaining CSS3 colors to avoid missing invisible or low contrast text
+encode HTML and CSS3 colors using hex (saves about 2k of memory)
+merge tr/td handling into one routine
+fix some minor bugs with new font code
+
+------------------------------------------------------------------------
+r6714 | felicity | 2004-02-18 03:47:44 +0000 (Wed, 18 Feb 2004) | 1 line
+
+MsgContainer::is_root() is a misnomer in a way, it checks whether or not the tree node is a leaf. 'root' sounds like the top of the tree, where it actually means a root of the tree or any subtree there-under. I just renamed it is_leaf to make it more correct.
+------------------------------------------------------------------------
+r6713 | felicity | 2004-02-18 02:41:34 +0000 (Wed, 18 Feb 2004) | 1 line
+
+API CHANGE! Moving from MsgParser->parse() to M::SA->parse(), so future changes to how we parse do not mean a change in the backend. Everything else is the same for now.
+------------------------------------------------------------------------
+r6711 | felicity | 2004-02-17 21:25:31 +0000 (Tue, 17 Feb 2004) | 1 line
+
+rewrite_mail() now just returns a scalar with the rewritten message instead of a fully parsed object
+------------------------------------------------------------------------
+r6710 | quinlan | 2004-02-17 20:55:39 +0000 (Tue, 17 Feb 2004) | 4 lines
+
+rewrite font attribute handling:
+ - fix invisibility FPs in Theo's corpus
+ - add handling of text attributes in addition to bgcolor and fgcolor
+
+------------------------------------------------------------------------
+r6703 | felicity | 2004-02-17 05:11:48 +0000 (Tue, 17 Feb 2004) | 1 line
+
+bug 3053: the strip2.t test was severely broken. the output redirection would fail, so it would end up comparing a blank file to a blank file, which of course always works ...
+------------------------------------------------------------------------
+r6702 | felicity | 2004-02-17 04:26:57 +0000 (Tue, 17 Feb 2004) | 1 line
+
+if a complete blank message is passed to parse(), pristine_headers would be undef which causes lots of error checking to have to go in... just default it to ''. also, _parse_body() was doing a CRLF->LF conversion on the whole message body, and then again for each multipart part, which is a waste since we did it all the first time. got rid of that.
+------------------------------------------------------------------------
+r6701 | felicity | 2004-02-17 03:42:35 +0000 (Tue, 17 Feb 2004) | 1 line
+
+get rid of MsgContainer::ignore() which nothing calls ... this also lets us get rid of the noexit setting code which is only used in ignore...
+------------------------------------------------------------------------
+r6700 | felicity | 2004-02-17 03:16:51 +0000 (Tue, 17 Feb 2004) | 1 line
+
+bug 1971: 'spamassassin -d -t' used to only remove the markup of the input mail, but will now remove the markup and run through -t as usual. also got rid of the calls to MsgContainer->ignore() from spamassassin and replaced with a simple exit() command. also found that -d wasn't working correctly due to the new API so that's fixed. :)
+------------------------------------------------------------------------
+r6699 | felicity | 2004-02-17 02:58:16 +0000 (Tue, 17 Feb 2004) | 1 line
+
+bug 1985: new version of urgent_biz for testing
+------------------------------------------------------------------------
+r6698 | felicity | 2004-02-17 02:10:58 +0000 (Tue, 17 Feb 2004) | 1 line
+
+bug 3050: 'spamassassin -D' would only show the rewritten message being parsed, and not the original due to the debug code being setup after the parse had occured.
+------------------------------------------------------------------------
+r6697 | quinlan | 2004-02-16 23:19:29 +0000 (Mon, 16 Feb 2004) | 2 lines
+
+typo
+
+------------------------------------------------------------------------
+r6696 | quinlan | 2004-02-16 23:18:40 +0000 (Mon, 16 Feb 2004) | 2 lines
+
+promote T_HTML_FONT_TINY
+
+------------------------------------------------------------------------
+r6695 | quinlan | 2004-02-16 20:12:11 +0000 (Mon, 16 Feb 2004) | 2 lines
+
+new URI tests
+
+------------------------------------------------------------------------
+r6691 | mss | 2004-02-16 12:02:02 +0000 (Mon, 16 Feb 2004) | 2 lines
+
+Got rid of warnings, made it fail a bit more gracefully.
+
+------------------------------------------------------------------------
+r6690 | quinlan | 2004-02-16 09:18:40 +0000 (Mon, 16 Feb 2004) | 2 lines
+
+rewrite most of the URI handling code
+
+------------------------------------------------------------------------
+r6689 | quinlan | 2004-02-16 09:11:09 +0000 (Mon, 16 Feb 2004) | 2 lines
+
+some more HTML test rules
+
+------------------------------------------------------------------------
+r6687 | quinlan | 2004-02-16 04:33:15 +0000 (Mon, 16 Feb 2004) | 2 lines
+
+typo
+
+------------------------------------------------------------------------
+r6686 | quinlan | 2004-02-16 04:21:18 +0000 (Mon, 16 Feb 2004) | 3 lines
+
+bump up score for MPART_ALT_DIFF
+add another evil message-id rule to try
+
+------------------------------------------------------------------------
+r6685 | felicity | 2004-02-16 04:10:12 +0000 (Mon, 16 Feb 2004) | 1 line
+
+bug 3028: FAKED_HOTMAIL_DAV wasn't allowing uppercase hotmail.com.
+------------------------------------------------------------------------
+r6672 | felicity | 2004-02-16 01:12:42 +0000 (Mon, 16 Feb 2004) | 1 line
+
+bug 2635: if the bayes_journal file can't be opened for some reason, spit out the error string from the system call.
+------------------------------------------------------------------------
+r6671 | felicity | 2004-02-16 01:07:10 +0000 (Mon, 16 Feb 2004) | 1 line
+
+bug 2840: if spamc was passed '-d non.existant.host', the message wouldn't be sent back out to STDOUT as it does with a host that exists but is not able to be contacted.
+------------------------------------------------------------------------
+r6670 | felicity | 2004-02-15 23:29:45 +0000 (Sun, 15 Feb 2004) | 1 line
+
+bug 3026: update the redhat rc script to be correct via their specifications
+------------------------------------------------------------------------
+r6669 | felicity | 2004-02-15 23:25:09 +0000 (Sun, 15 Feb 2004) | 1 line
+
+bug 3024: trying to forget() a message that hasn't been learned caused a fatal error to be reported incorrectly. also added in code to learn() to check the result of the forget() call (if it was called) and propagate up fatal errors appropriately.
+------------------------------------------------------------------------
+r6664 | felicity | 2004-02-15 22:22:25 +0000 (Sun, 15 Feb 2004) | 1 line
+
+3.0.0 isn't numeric... stupid perl. <G>
+------------------------------------------------------------------------
+r6660 | felicity | 2004-02-15 21:56:11 +0000 (Sun, 15 Feb 2004) | 1 line
+
+renumber the release from 2.70 to 3.0.0
+------------------------------------------------------------------------
+r6659 | felicity | 2004-02-15 21:46:14 +0000 (Sun, 15 Feb 2004) | 1 line
+
+split up the meta token RE to be a bit more readable
+------------------------------------------------------------------------
+r6658 | quinlan | 2004-02-15 19:53:39 +0000 (Sun, 15 Feb 2004) | 3 lines
+
+add major overhaul of URI parsing to HTML parser
+also remove HTML order code and unclickable anchor test code
+
+------------------------------------------------------------------------
+r6657 | quinlan | 2004-02-15 12:41:53 +0000 (Sun, 15 Feb 2004) | 2 lines
+
+remove HTML order tests
+
+------------------------------------------------------------------------
+r6656 | quinlan | 2004-02-15 12:40:48 +0000 (Sun, 15 Feb 2004) | 4 lines
+
+replace HTML_TITLE_EMPTY with T_HTML_TITLE_EMPTY
+remove T_HTML_LINK_UNCLICKABLE
+remove HTML order rules
+
+------------------------------------------------------------------------
+r6653 | quinlan | 2004-02-15 01:53:37 +0000 (Sun, 15 Feb 2004) | 2 lines
+
+spell mutable correctly
+
+------------------------------------------------------------------------
+r6652 | quinlan | 2004-02-14 21:19:50 +0000 (Sat, 14 Feb 2004) | 4 lines
+
+fix T_RCVD_IN_XBL_NFH rule (-notfirsthop should really be an argument to
+make it clearer that only check_rbl and check_rbl_txt support notfirsthop
+mode)
+
+------------------------------------------------------------------------
+r6651 | quinlan | 2004-02-14 08:42:53 +0000 (Sat, 14 Feb 2004) | 2 lines
+
+revisions of the evil spam Message-ID rules
+
+------------------------------------------------------------------------
+r6646 | mss | 2004-02-14 01:06:10 +0000 (Sat, 14 Feb 2004) | 2 lines
+
+Two more typos which I didn't see.
+
+------------------------------------------------------------------------
+r6645 | mss | 2004-02-14 00:04:02 +0000 (Sat, 14 Feb 2004) | 2 lines
+
+Some more sorting.
+
+------------------------------------------------------------------------
+r6644 | mss | 2004-02-13 23:38:59 +0000 (Fri, 13 Feb 2004) | 2 lines
+
+Forgot a newline.
+
+------------------------------------------------------------------------
+r6643 | mss | 2004-02-13 23:35:05 +0000 (Fri, 13 Feb 2004) | 2 lines
+
+Cleaned up the usage a bit.
+
+------------------------------------------------------------------------
+r6642 | quinlan | 2004-02-13 21:05:09 +0000 (Fri, 13 Feb 2004) | 3 lines
+
+promote T_MSGID_DOLLARS_1 to T_MSGID_DOLLARS
+remove other T_MSGID_DOLLARS_* variants (equal or worse accuracy)
+
+------------------------------------------------------------------------
+r6641 | quinlan | 2004-02-13 20:57:38 +0000 (Fri, 13 Feb 2004) | 2 lines
+
+remove T_HTML_ALT_LONGEST_* rules, poor results for theo and rodbegbie
+
+------------------------------------------------------------------------
+r6640 | mss | 2004-02-13 20:06:22 +0000 (Fri, 13 Feb 2004) | 2 lines
+
+Just a small change, feels better.
+
+------------------------------------------------------------------------
+r6639 | mss | 2004-02-13 20:03:38 +0000 (Fri, 13 Feb 2004) | 2 lines
+
+Updated license.
+
+------------------------------------------------------------------------
+r6638 | mss | 2004-02-13 13:37:51 +0000 (Fri, 13 Feb 2004) | 2 lines
+
+bug 3041: Some Windows-related fixes sumbitted by Sidney.
+
+------------------------------------------------------------------------
+r6637 | quinlan | 2004-02-13 09:51:08 +0000 (Fri, 13 Feb 2004) | 2 lines
+
+more evil message-id stuff
+
+------------------------------------------------------------------------
+r6636 | quinlan | 2004-02-13 07:59:45 +0000 (Fri, 13 Feb 2004) | 3 lines
+
+stop rendering ALT text since most mailers don't
+add a test for excessively long ALT text (total length did not work well)
+
+------------------------------------------------------------------------
+r6635 | quinlan | 2004-02-12 22:24:48 +0000 (Thu, 12 Feb 2004) | 2 lines
+
+fix rewriting
+
+------------------------------------------------------------------------
+r6634 | mss | 2004-02-12 13:40:07 +0000 (Thu, 12 Feb 2004) | 2 lines
+
+bug 3022: Applied Michael Parker's patch to skip the DBI/DB_File tests on platforms where they're not supported.
+
+------------------------------------------------------------------------
+r6633 | quinlan | 2004-02-12 10:31:35 +0000 (Thu, 12 Feb 2004) | 2 lines
+
+a few more T_XORIGIP_INVALID variants
+
+------------------------------------------------------------------------
+r6629 | quinlan | 2004-02-12 07:16:17 +0000 (Thu, 12 Feb 2004) | 9 lines
+
+promote T_MPART_ALT_DIFF_99 to MPART_ALT_DIFF (lowest FNs by far)
+promote T_HTML_BADTAGS_* to HTML_BADTAGS_*
+promote T_HTML_BADTAGS_U_* to HTML_NONELEMENT_*
+promote T_HTML_MESSAGE_1 to HTML_MESSAGE
+add T_HTML_LINK_UNCLICKABLE (anchor with no image or text)
+add HTML_OBFUSCATION_* descriptions
+make MsgContainer post-rendering texts somewhat more readable
+move RATWARE_HASH_DASH to 20_ratware.cf
+
+------------------------------------------------------------------------
+r6628 | quinlan | 2004-02-12 06:26:46 +0000 (Thu, 12 Feb 2004) | 4 lines
+
+bug 3030: nightly corpus DETAILS to include temporal data
+don't bother if nothing has been uploaded (reduce load on my poor machine)
+minor speedup for sorting routine
+
+------------------------------------------------------------------------
+r6625 | jm | 2004-02-11 20:41:19 +0000 (Wed, 11 Feb 2004) | 1 line
+
+added reply to FIXME
+------------------------------------------------------------------------
+r6624 | mss | 2004-02-11 19:42:46 +0000 (Wed, 11 Feb 2004) | 4 lines
+
+bug 1360: Add support for logging to a file instead of syslog.
+
+There might be a bug in this code, I'm not sure how the logfile should be shared; see line 618.
+
+------------------------------------------------------------------------
+r6617 | jm | 2004-02-11 07:25:23 +0000 (Wed, 11 Feb 2004) | 1 line
+
+missing entire hostnames
+------------------------------------------------------------------------
+r6616 | jm | 2004-02-11 07:18:42 +0000 (Wed, 11 Feb 2004) | 1 line
+
+bug 2895: fixed description for 'rawbody' rules
+------------------------------------------------------------------------
+r6615 | jm | 2004-02-11 07:17:40 +0000 (Wed, 11 Feb 2004) | 1 line
+
+QP decoding should deal with lowercase QP as well
+------------------------------------------------------------------------
+r6612 | quinlan | 2004-02-11 06:21:51 +0000 (Wed, 11 Feb 2004) | 2 lines
+
+trivial spacing change
+
+------------------------------------------------------------------------
+r6611 | quinlan | 2004-02-11 06:21:29 +0000 (Wed, 11 Feb 2004) | 2 lines
+
+revise T_HTML_MESSAGE_*
+
+------------------------------------------------------------------------
+r6610 | quinlan | 2004-02-11 05:02:26 +0000 (Wed, 11 Feb 2004) | 2 lines
+
+add HTML order rules and some possible HTML_MESSAGE replacements
+
+------------------------------------------------------------------------
+r6609 | jm | 2004-02-11 03:16:34 +0000 (Wed, 11 Feb 2004) | 1 line
+
+wantonly updated Changes from svn log
+------------------------------------------------------------------------
+r6608 | jm | 2004-02-11 03:16:03 +0000 (Wed, 11 Feb 2004) | 1 line
+
+fixed T_REGSOFT, added T_DATE_COMMA
+------------------------------------------------------------------------
+r6607 | quinlan | 2004-02-11 00:20:09 +0000 (Wed, 11 Feb 2004) | 2 lines
+
+move 70_cvs_rules_under_test.cf to rules/70_testing.cf
+
+------------------------------------------------------------------------
+r6606 | quinlan | 2004-02-11 00:04:07 +0000 (Wed, 11 Feb 2004) | 3 lines
+
+test XBL as -notfirsthop to compare against current rule before changing
+current rule
+
+------------------------------------------------------------------------
+r6605 | quinlan | 2004-02-10 21:14:16 +0000 (Tue, 10 Feb 2004) | 2 lines
+
+XBL is supposed to be a DUL list according to Steve Linford
+
+------------------------------------------------------------------------
+r6603 | felicity | 2004-02-10 17:37:06 +0000 (Tue, 10 Feb 2004) | 1 line
+
+PerMsgStatus shouldn't go digging into the internal message format. bayes wants the mbox seperator if it exists. and let's not use future atimes in the bayes db since that should never happen.
+------------------------------------------------------------------------
+r6601 | jm | 2004-02-10 01:13:10 +0000 (Tue, 10 Feb 2004) | 1 line
+
+added T_REGSOFT ratware-forgery sign
+------------------------------------------------------------------------
+r6586 | jm | 2004-02-08 05:13:39 +0000 (Sun, 08 Feb 2004) | 1 line
+
+couple more patterns
+------------------------------------------------------------------------
+r6585 | jm | 2004-02-08 05:12:39 +0000 (Sun, 08 Feb 2004) | 1 line
+
+couple more patterns
+------------------------------------------------------------------------
+r6584 | jm | 2004-02-08 05:10:31 +0000 (Sun, 08 Feb 2004) | 1 line
+
+couple more patterns
+------------------------------------------------------------------------
+r6583 | jm | 2004-02-08 04:56:16 +0000 (Sun, 08 Feb 2004) | 1 line
+
+redo T_HELO_DYNAMIC rules using last-untrusted semantics
+------------------------------------------------------------------------
+r6582 | jm | 2004-02-08 04:51:57 +0000 (Sun, 08 Feb 2004) | 1 line
+
+oops, forgot to shorten the ip
+------------------------------------------------------------------------
+r6579 | jm | 2004-02-08 04:40:12 +0000 (Sun, 08 Feb 2004) | 1 line
+
+use Net::CIDR::Lite to consolidate into ranges where possible
+------------------------------------------------------------------------
+r6578 | jm | 2004-02-08 04:19:52 +0000 (Sun, 08 Feb 2004) | 1 line
+
+fixed bug: bayes was picking up non-IP tokens as IPs
+------------------------------------------------------------------------
+r6577 | jm | 2004-02-08 02:44:20 +0000 (Sun, 08 Feb 2004) | 1 line
+
+sorted MANIFEST
+------------------------------------------------------------------------
+r6576 | jm | 2004-02-08 02:44:03 +0000 (Sun, 08 Feb 2004) | 1 line
+
+added tool
+------------------------------------------------------------------------
+r6575 | jm | 2004-02-08 02:15:13 +0000 (Sun, 08 Feb 2004) | 1 line
+
+added tool to generate 'trusted_networks' lines from a Bayes db dump
+------------------------------------------------------------------------
+r6574 | jm | 2004-02-08 01:45:16 +0000 (Sun, 08 Feb 2004) | 1 line
+
+bayes: track IP addresses from Received for trusted_networks mining
+------------------------------------------------------------------------
+r6573 | jm | 2004-02-08 01:44:52 +0000 (Sun, 08 Feb 2004) | 16 lines
+
+Plugins:
+
+ - allow loading from @INC
+
+ - allow loading with relative paths
+
+ - added 'ifplugin' and 'endif' conditional-loading keywords for Conf
+
+ - added a Test plugin class
+
+ - added test for above (and for plugins in general)
+
+ - cleaned up a few more cases where Conf uses regexps instead of
+ $key and $value
+
+
+------------------------------------------------------------------------
+r6572 | jm | 2004-02-08 00:45:27 +0000 (Sun, 08 Feb 2004) | 1 line
+
+added relative-path support to loadplugin
+------------------------------------------------------------------------
+r6571 | jm | 2004-02-07 23:55:22 +0000 (Sat, 07 Feb 2004) | 1 line
+
+bug 2556: added 'auto_whitelist_db_modules' to provide way for user to specify db module preference so that old AWL files can be used
+------------------------------------------------------------------------
+r6570 | jm | 2004-02-07 23:47:08 +0000 (Sat, 07 Feb 2004) | 1 line
+
+bug 2898: GTUBE mails should not affect the AWL
+------------------------------------------------------------------------
+r6569 | jm | 2004-02-07 23:28:52 +0000 (Sat, 07 Feb 2004) | 1 line
+
+added POD documentation and --clean switch to tools/check_whitelist, allowing AWL to be cleaned of single-hit old entries. fixes bug 1883: record number of emails received in AWL; bug 2103: AWL needs a way to be cleaned out
+------------------------------------------------------------------------
+r6565 | jm | 2004-02-07 22:36:16 +0000 (Sat, 07 Feb 2004) | 1 line
+
+bug 3006: SQL user-prefs stop working due to use of self->{main} in Conf.pm. Fixed by setting {main} in Conf ctor, and providing a finish() method on the main object as we should have been doing all along
+------------------------------------------------------------------------
+r6564 | jm | 2004-02-07 22:18:45 +0000 (Sat, 07 Feb 2004) | 1 line
+
+affiliate_id rule, suggestion from sa-users list
+------------------------------------------------------------------------
+r6558 | jm | 2004-02-07 03:54:20 +0000 (Sat, 07 Feb 2004) | 1 line
+
+some new rules for testing
+------------------------------------------------------------------------
+r6553 | jm | 2004-02-06 20:25:59 +0000 (Fri, 06 Feb 2004) | 1 line
+
+rule development helper script: scan a corpus of mail for one or more named headers, producing easilly-greppable output, to spot correlations between header patterns
+------------------------------------------------------------------------
+r6550 | jm | 2004-02-06 17:55:57 +0000 (Fri, 06 Feb 2004) | 1 line
+
+more simple drugs-in-subject rules for testing
+------------------------------------------------------------------------
+r6549 | jm | 2004-02-06 17:52:14 +0000 (Fri, 06 Feb 2004) | 1 line
+
+simple xanax-in-subject rule
+------------------------------------------------------------------------
+r6548 | jm | 2004-02-06 17:39:06 +0000 (Fri, 06 Feb 2004) | 1 line
+
+some errors in MANIFEST discovered; all files with 'o' in the name may have been skipped
+------------------------------------------------------------------------
+r6539 | jm | 2004-02-06 07:00:31 +0000 (Fri, 06 Feb 2004) | 1 line
+
+bug 2205: LDAP support, from Kristian Koehntopp <kk(at)netuse.de>. spamd can now read configs from LDAP
+------------------------------------------------------------------------
+r6536 | jm | 2004-02-06 06:49:55 +0000 (Fri, 06 Feb 2004) | 1 line
+
+bug 2205: LDAP support, from Kristian Koehntopp <kk(at)netuse.de>. spamd can now read configs from LDAP
+------------------------------------------------------------------------
+r6532 | felicity | 2004-02-06 04:30:44 +0000 (Fri, 06 Feb 2004) | 1 line
+
+updated sa-learn doco a little
+------------------------------------------------------------------------
+r6531 | jm | 2004-02-06 04:04:28 +0000 (Fri, 06 Feb 2004) | 1 line
+
+added a whole load more plugin APIs, for scanning start/end events, autolearning, and when the user is changed
+------------------------------------------------------------------------
+r6529 | jm | 2004-02-06 03:36:06 +0000 (Fri, 06 Feb 2004) | 1 line
+
+some error-avoidance in plugin support
+------------------------------------------------------------------------
+r6528 | jm | 2004-02-06 03:35:41 +0000 (Fri, 06 Feb 2004) | 1 line
+
+trying out some new rules to detect spamware which attempts to evade rDNS/HELO crosschecks by using correct rDNS for HELO; however, if we see HELOs from dynamic IPs using their rDNS names, I bet it's a good rule. idea from CL
+------------------------------------------------------------------------
+r6524 | quinlan | 2004-02-05 23:10:44 +0000 (Thu, 05 Feb 2004) | 7 lines
+
+DNSBL changes
+- rename RCVD_IN_DYNABLOCK to RCVD_IN_SORBS_DUL
+- change multi-RBLs into prerequisite tests for better stability
+- promote RCVD_IN_XBL, switch to new version of SBL+XBL rule
+- fix T_RCVD_IN_SORBS_NOMAIL and T_RCVD_IN_SORBS_BADCONF to query
+ the SORBS RHSBL
+
+------------------------------------------------------------------------
+r6523 | quinlan | 2004-02-05 23:08:02 +0000 (Thu, 05 Feb 2004) | 2 lines
+
+remove some debugging accidentally committed
+
+------------------------------------------------------------------------
+r6522 | felicity | 2004-02-05 22:48:10 +0000 (Thu, 05 Feb 2004) | 1 line
+
+reverted the last score checkin, forced the 0 scores to 0.001 to get the same result as I was going for...
+------------------------------------------------------------------------
+r6520 | felicity | 2004-02-05 21:10:30 +0000 (Thu, 05 Feb 2004) | 4 lines
+
+since we had a bunch of scores still at 0, weren't not testing all of the
+rules ... killed the generated scores so everything will default to 1.
+enabled net rules that were disabled with 0.001 score.
+
+------------------------------------------------------------------------
+r6519 | felicity | 2004-02-05 20:40:15 +0000 (Thu, 05 Feb 2004) | 1 line
+
+bug 2869: sa-learn would fail if run with --mbox and reading from STDIN in certain circumstances. this commit lets sa-learn make a temp file to store the contents from STDIN, runs the appropriate code over the temp file, then removes it.
+------------------------------------------------------------------------
+r6518 | felicity | 2004-02-05 20:26:15 +0000 (Thu, 05 Feb 2004) | 1 line
+
+the loop should deal with the file name, not the folder name ...
+------------------------------------------------------------------------
+r6515 | felicity | 2004-02-05 18:03:42 +0000 (Thu, 05 Feb 2004) | 1 line
+
+more mime parsing tests
+------------------------------------------------------------------------
+r6514 | felicity | 2004-02-05 17:50:11 +0000 (Thu, 05 Feb 2004) | 1 line
+
+found spam trying to use blank mime boundaries, in violation of the rfc. it's not clear if we should accept or ignore the blank boundary (muas are all over the board wrt how they handle it), so this commit will have us accept the blank boundary. we can easily change it to ignore the blank boundary with a 2-line comment change if we later decide to do things differently.
+------------------------------------------------------------------------
+r6513 | felicity | 2004-02-05 16:03:25 +0000 (Thu, 05 Feb 2004) | 1 line
+
+solve a div by zero error
+------------------------------------------------------------------------
+r6509 | quinlan | 2004-02-05 07:34:27 +0000 (Thu, 05 Feb 2004) | 3 lines
+
+promote HTML_CONVERTED
+add a new test rule (not usable yet)
+
+------------------------------------------------------------------------
+r6504 | felicity | 2004-02-05 04:08:59 +0000 (Thu, 05 Feb 2004) | 1 line
+
+remove MICROSOFT_EXECUTABLE
+------------------------------------------------------------------------
+r6503 | felicity | 2004-02-05 03:19:20 +0000 (Thu, 05 Feb 2004) | 1 line
+
+don't bother looking for uuencoded parts, since we don't use that knowledge for anything
+------------------------------------------------------------------------
+r6502 | felicity | 2004-02-04 22:28:34 +0000 (Wed, 04 Feb 2004) | 3 lines
+
+bug 3000: make check_attachments() use the mime tree instead of trying
+to figure out what is going on by itself.
+
+------------------------------------------------------------------------
+r6501 | felicity | 2004-02-04 21:11:54 +0000 (Wed, 04 Feb 2004) | 1 line
+
+fixed no-self issue...
+------------------------------------------------------------------------
+r6500 | felicity | 2004-02-04 20:27:53 +0000 (Wed, 04 Feb 2004) | 1 line
+
+allow access to raw data if necessary
+------------------------------------------------------------------------
+r6499 | felicity | 2004-02-04 20:24:58 +0000 (Wed, 04 Feb 2004) | 1 line
+
+parse more things out of content-type
+------------------------------------------------------------------------
+r6494 | quinlan | 2004-02-04 08:12:19 +0000 (Wed, 04 Feb 2004) | 2 lines
+
+remove HTML bad tag count rules - not as good as ratios
+
+------------------------------------------------------------------------
+r6493 | quinlan | 2004-02-04 08:09:34 +0000 (Wed, 04 Feb 2004) | 4 lines
+
+Promote one of the parser-based HTML obfuscation rule sets
+(not quite the best results, but seems least easily evaded).
+This blows OBFUSCATING_COMMENT out of the water, by the way.
+
+------------------------------------------------------------------------
+r6466 | quinlan | 2004-02-04 04:35:49 +0000 (Wed, 04 Feb 2004) | 2 lines
+
+remove in-memory cache of probabilities since code was not being used
+
+------------------------------------------------------------------------
+r6463 | quinlan | 2004-02-04 00:29:53 +0000 (Wed, 04 Feb 2004) | 4 lines
+
+add HTML parser-based version of T_HTML_CONVERTED
+add xbl+sbl support
+a few other new test rules
+
+------------------------------------------------------------------------
+r6441 | quinlan | 2004-02-03 06:02:22 +0000 (Tue, 03 Feb 2004) | 2 lines
+
+allow HTML comment rules to be specified in html_eval, convert current rules
+
+------------------------------------------------------------------------
+r6440 | quinlan | 2004-02-03 05:54:48 +0000 (Tue, 03 Feb 2004) | 2 lines
+
+add parser-based HTML obfuscation tests and a few tweaks to HTML rendering
+
+------------------------------------------------------------------------
+r6436 | felicity | 2004-02-03 00:30:30 +0000 (Tue, 03 Feb 2004) | 1 line
+
+don't need MsgContainer::as_string() anymore. instead of rendering as HTML any text* attachment, only do 'text' and 'text/plain'. fixes some FPs for MPART_ALT_DIFF, etc.
+------------------------------------------------------------------------
+r6435 | felicity | 2004-02-02 21:36:50 +0000 (Mon, 02 Feb 2004) | 1 line
+
+spamd wasn't putting the mbox seperator back ... <grrr>
+------------------------------------------------------------------------
+r6434 | quinlan | 2004-02-02 21:34:33 +0000 (Mon, 02 Feb 2004) | 2 lines
+
+add some Message-ID rules for test
+
+------------------------------------------------------------------------
+r6433 | felicity | 2004-02-02 15:57:07 +0000 (Mon, 02 Feb 2004) | 1 line
+
+if the raw size is 0, there's nothing to render, so don't bother. (solves a div by 0 error too)
+------------------------------------------------------------------------
+r6432 | quinlan | 2004-02-02 06:39:05 +0000 (Mon, 02 Feb 2004) | 3 lines
+
+testing moving some comment rules from code to here
+bug 2992: a few proposed new rules by Martin Radford
+
+------------------------------------------------------------------------
+r6431 | quinlan | 2004-02-02 06:38:03 +0000 (Mon, 02 Feb 2004) | 2 lines
+
+allow HTML comment tests to be specified in html_eval rules
+
+------------------------------------------------------------------------
+r6430 | felicity | 2004-02-02 02:50:25 +0000 (Mon, 02 Feb 2004) | 1 line
+
+deal with the mbox seperator ...
+------------------------------------------------------------------------
+r6421 | felicity | 2004-02-01 21:14:39 +0000 (Sun, 01 Feb 2004) | 1 line
+
+use 3 different states for ArchiveIterator, 1 for sa-learn (do everything in memory), 2 for mass-check: tempfile to keep memory use down, 1 process (no forking/etc), 2+ processes (forking, IPC, etc.)
+------------------------------------------------------------------------
+r6400 | duncf | 2004-02-01 05:41:34 +0000 (Sun, 01 Feb 2004) | 1 line
+
+Fixed credits to avoid belittling some contributions... removed references to bugzilla and dates (we really dont need to be so specific...)
+------------------------------------------------------------------------
+r6399 | felicity | 2004-02-01 05:12:49 +0000 (Sun, 01 Feb 2004) | 1 line
+
+previous change broke sa-learn -- we now only use external files for message passing if -j > 1, also just stick # of messages at top of temp file, skip the whole IO child/parent thing ...
+------------------------------------------------------------------------
+r6398 | jm | 2004-02-01 04:37:29 +0000 (Sun, 01 Feb 2004) | 2 lines
+
+Michael Parker: Fixing up the license in several files, also moves over to the new key/value
+Conf stuff and makes it so BayesStoreSQL no longer returns undef in new.
+------------------------------------------------------------------------
+r6393 | mss | 2004-02-01 01:46:46 +0000 (Sun, 01 Feb 2004) | 4 lines
+
+Tried to find out why 'make test' fails with
+ t/bayesdbm....ok 32/43Can't use an undefined value as a symbol reference at ../lib/Mail/SpamAssassin/ArchiveIterator.pm line 296.
+Anybody an idea?
+
+------------------------------------------------------------------------
+r6390 | mss | 2004-02-01 01:17:14 +0000 (Sun, 01 Feb 2004) | 1 line
+
+"clean-room" re-implementation of the spamc $(EXE_EXT) stuff
+------------------------------------------------------------------------
+r6385 | felicity | 2004-01-31 21:40:05 +0000 (Sat, 31 Jan 2004) | 1 line
+
+reduce memory usage during mass-check -- for me I go from 110-250 to 50-60MB
+------------------------------------------------------------------------
+r6383 | felicity | 2004-01-31 14:35:51 +0000 (Sat, 31 Jan 2004) | 1 line
+
+use nightly/weekly as appropriate, not just nightly ... <grrr>
+------------------------------------------------------------------------
+r6381 | jm | 2004-01-31 05:26:58 +0000 (Sat, 31 Jan 2004) | 1 line
+
+reverted patch from bug 1031: removed EXE_EXT use from code to save having to get a written CLA from Reini Urban
+------------------------------------------------------------------------
+r6380 | jm | 2004-01-31 05:21:25 +0000 (Sat, 31 Jan 2004) | 1 line
+
+bug 2949: add wiki link to README
+------------------------------------------------------------------------
+r6379 | jm | 2004-01-31 05:08:39 +0000 (Sat, 31 Jan 2004) | 1 line
+
+updated CREDITS list to be comprehensive and up-to-date
+------------------------------------------------------------------------
+r6378 | jm | 2004-01-31 04:43:04 +0000 (Sat, 31 Jan 2004) | 1 line
+
+removed copyright notices from all files, apart from the ASF one. Moved authorship and credits information into a new top-level file, CREDITS
+------------------------------------------------------------------------
+r6377 | jm | 2004-01-31 04:18:08 +0000 (Sat, 31 Jan 2004) | 1 line
+
+ran indent -i4 -npsl -di0 -br -nce -d0 -cli0 -npcs -nfc1 (the Apache C style, http://www.apache.org/dev/styleguide.html) over the .c and .h files in spamc
+------------------------------------------------------------------------
+r6376 | jm | 2004-01-31 04:12:48 +0000 (Sat, 31 Jan 2004) | 1 line
+
+noted some warnings about implicit declarations and unused vars; cleaned
+------------------------------------------------------------------------
+r6375 | jm | 2004-01-31 04:10:17 +0000 (Sat, 31 Jan 2004) | 1 line
+
+bug 2959: spamc -d causes seg fault when hostname used does not resolve
+------------------------------------------------------------------------
+r6374 | jm | 2004-01-31 04:00:18 +0000 (Sat, 31 Jan 2004) | 1 line
+
+bug 1642: remove use of AnyDBM_File, since it pollutes the global namespace
+------------------------------------------------------------------------
+r6371 | quinlan | 2004-01-31 03:28:18 +0000 (Sat, 31 Jan 2004) | 4 lines
+
+store list of messages in a temporary file
+restore offset as byte offset of message starting at "From "
+move secure_tmpfile to Util.pm
+
+------------------------------------------------------------------------
+r6370 | quinlan | 2004-01-31 03:26:41 +0000 (Sat, 31 Jan 2004) | 2 lines
+
+fix a warning from new test code
+
+------------------------------------------------------------------------
+r6365 | felicity | 2004-01-30 21:03:54 +0000 (Fri, 30 Jan 2004) | 1 line
+
+quinlan didn't like my sha1.t modifications, so I'm backing out the patch since it's not a hot issue for me.
+------------------------------------------------------------------------
+r6364 | felicity | 2004-01-30 18:49:59 +0000 (Fri, 30 Jan 2004) | 1 line
+
+when bayes learns a message, put the atime of that message in the debug output
+------------------------------------------------------------------------
+r6363 | felicity | 2004-01-30 18:45:19 +0000 (Fri, 30 Jan 2004) | 1 line
+
+if sa-learn is run but the configuration disables bayes, tell the user!
+------------------------------------------------------------------------
+r6361 | jm | 2004-01-30 06:12:57 +0000 (Fri, 30 Jan 2004) | 1 line
+
+T_HTTP_77 now a uri rule
+------------------------------------------------------------------------
+r6360 | jm | 2004-01-30 05:54:01 +0000 (Fri, 30 Jan 2004) | 1 line
+
+oops, typo in T_HTTP_77
+------------------------------------------------------------------------
+r6359 | felicity | 2004-01-30 04:15:38 +0000 (Fri, 30 Jan 2004) | 2 lines
+
+faster version of scan_mailbox. brings the time for me from 55s to 44s.
+
+------------------------------------------------------------------------
+r6358 | felicity | 2004-01-30 03:43:03 +0000 (Fri, 30 Jan 2004) | 2 lines
+
+more removal of deprecated options and such
+
+------------------------------------------------------------------------
+r6357 | jm | 2004-01-30 03:09:51 +0000 (Fri, 30 Jan 2004) | 1 line
+
+minor doco update
+------------------------------------------------------------------------
+r6356 | jm | 2004-01-30 03:04:54 +0000 (Fri, 30 Jan 2004) | 1 line
+
+catch encoded 'www' in URLs
+------------------------------------------------------------------------
+r6355 | jm | 2004-01-30 03:04:47 +0000 (Fri, 30 Jan 2004) | 1 line
+
+catch encoded 'www' in URLs
+------------------------------------------------------------------------
+r6354 | jm | 2004-01-30 03:04:17 +0000 (Fri, 30 Jan 2004) | 1 line
+
+this went in with last change. highlighting independently
+------------------------------------------------------------------------
+r6353 | jm | 2004-01-30 02:54:33 +0000 (Fri, 30 Jan 2004) | 1 line
+
+bug 195: SQL support for AWL and Bayes storage, thanks to Michael Parker
+------------------------------------------------------------------------
+r6352 | felicity | 2004-01-29 19:43:38 +0000 (Thu, 29 Jan 2004) | 1 line
+
+more housekeeping -- check_bayes_db no longer exists, and modified some of the sha1 test code (more text blobs ...)
+------------------------------------------------------------------------
+r6351 | felicity | 2004-01-29 19:27:12 +0000 (Thu, 29 Jan 2004) | 1 line
+
+first pass at removing deprecated options
+------------------------------------------------------------------------
+r6350 | felicity | 2004-01-29 19:09:45 +0000 (Thu, 29 Jan 2004) | 1 line
+
+fixed up some doco test bits
+------------------------------------------------------------------------
+r6349 | felicity | 2004-01-29 18:47:23 +0000 (Thu, 29 Jan 2004) | 1 line
+
+here's cpucount ...
+------------------------------------------------------------------------
+r6348 | quinlan | 2004-01-29 18:44:56 +0000 (Thu, 29 Jan 2004) | 2 lines
+
+find number of processors without cpucount
+
+------------------------------------------------------------------------
+r6347 | felicity | 2004-01-29 14:18:15 +0000 (Thu, 29 Jan 2004) | 1 line
+
+another stupid typo
+------------------------------------------------------------------------
+r6346 | quinlan | 2004-01-29 01:57:58 +0000 (Thu, 29 Jan 2004) | 2 lines
+
+more SVN updates
+
+------------------------------------------------------------------------
+r6343 | felicity | 2004-01-28 22:29:01 +0000 (Wed, 28 Jan 2004) | 1 line
+
+remove out-of-date tools that aren't used anyway
+------------------------------------------------------------------------
+r6342 | felicity | 2004-01-28 22:27:18 +0000 (Wed, 28 Jan 2004) | 1 line
+
+added license text to calling scripts
+------------------------------------------------------------------------
+r6341 | felicity | 2004-01-28 22:24:05 +0000 (Wed, 28 Jan 2004) | 1 line
+
+need to kill it from MANIFEST too... <sigh>
+------------------------------------------------------------------------
+r6340 | felicity | 2004-01-28 22:23:21 +0000 (Wed, 28 Jan 2004) | 1 line
+
+people won't run it if it's not there ... ;)
+------------------------------------------------------------------------
+r6336 | felicity | 2004-01-28 16:02:39 +0000 (Wed, 28 Jan 2004) | 1 line
+
+fixed svnversion issue
+------------------------------------------------------------------------
+r6335 | quinlan | 2004-01-28 06:40:25 +0000 (Wed, 28 Jan 2004) | 2 lines
+
+switch to SVN version (does anyone actually use the tag of mass-check itself?)
+
+------------------------------------------------------------------------
+r6334 | quinlan | 2004-01-28 06:34:10 +0000 (Wed, 28 Jan 2004) | 2 lines
+
+add $Rev$
+
+------------------------------------------------------------------------
+r6333 | jm | 2004-01-28 06:13:17 +0000 (Wed, 28 Jan 2004) | 1 line
+
+updating instructions for svn
+------------------------------------------------------------------------
+r6331 | felicity | 2004-01-28 00:57:39 +0000 (Wed, 28 Jan 2004) | 1 line
+
+updates for the corpora run scripts
+------------------------------------------------------------------------
+r6330 | quinlan | 2004-01-28 00:37:10 +0000 (Wed, 28 Jan 2004) | 2 lines
+
+update scripts for SVN
+
+------------------------------------------------------------------------
+r6329 | quinlan | 2004-01-27 22:50:11 +0000 (Tue, 27 Jan 2004) | 2 lines
+
+bug 2211: tests to detect invalid HTML tags
+
+------------------------------------------------------------------------
+r6328 | quinlan | 2004-01-27 22:35:08 +0000 (Tue, 27 Jan 2004) | 2 lines
+
+naming convention for html_results that are still in testing
+
+------------------------------------------------------------------------
+r6327 | felicity | 2004-01-27 21:26:56 +0000 (Tue, 27 Jan 2004) | 1 line
+
+finish removing this stuff
+------------------------------------------------------------------------
+r6324 | felicity | 2004-01-27 21:00:17 +0000 (Tue, 27 Jan 2004) | 1 line
+
+removing old changes file
+------------------------------------------------------------------------
+r6323 | felicity | 2004-01-27 20:58:39 +0000 (Tue, 27 Jan 2004) | 1 line
+
+require razor 2.40 or later
+------------------------------------------------------------------------
+r6322 | jm | 2004-01-27 20:11:42 +0000 (Tue, 27 Jan 2004) | 1 line
+
+oops. thanks Theo for spotting that. fixed a bug
+------------------------------------------------------------------------
+r6321 | felicity | 2004-01-27 19:19:10 +0000 (Tue, 27 Jan 2004) | 2 lines
+
+bug 2910: looks like these weren't "svn add"ed ... oops!
+
+------------------------------------------------------------------------
+r6319 | jm | 2004-01-27 08:24:29 +0000 (Tue, 27 Jan 2004) | 1 line
+
+bug 2910: replaced evolver with fast perceptron score generation tool by Henry Stern; also cleaned up a few old scripts that aren't used anymore
+------------------------------------------------------------------------
+r6318 | jm | 2004-01-27 08:04:58 +0000 (Tue, 27 Jan 2004) | 1 line
+
+metadata's not going away that easily ;)
+------------------------------------------------------------------------
+r6317 | jm | 2004-01-27 08:01:27 +0000 (Tue, 27 Jan 2004) | 1 line
+
+bug 2970: Bob Menschel's 'longwords' rules
+------------------------------------------------------------------------
+r6316 | jm | 2004-01-27 07:55:25 +0000 (Tue, 27 Jan 2004) | 1 line
+
+bug 2965: hit-frequencies -x header was wrong
+------------------------------------------------------------------------
+r6312 | felicity | 2004-01-27 03:29:54 +0000 (Tue, 27 Jan 2004) | 1 line
+
+bug 2968: sa-learn failed because there's no more 'get' function...
+------------------------------------------------------------------------
+r6305 | felicity | 2004-01-26 23:27:57 +0000 (Mon, 26 Jan 2004) | 1 line
+
+mime parser test code and related data files
+------------------------------------------------------------------------
+r6304 | jm | 2004-01-26 22:20:21 +0000 (Mon, 26 Jan 2004) | 1 line
+
+added bogusmx DNSBL test
+------------------------------------------------------------------------
+r6303 | felicity | 2004-01-26 22:09:05 +0000 (Mon, 26 Jan 2004) | 30 lines
+
+- bug 2267-related: I found that M::SA::Conf::_parse() was sucking
+ up a lot of the time of a single message run. I modified the
+ majority of configuration options to do string equality instead of
+ REs and got an improvement. It probably still needs some tweaks,
+ but it's working for me. :)
+
+- started removing backwards compatibility options -- for instance,
+ bayes_auto_learn is now the required version, auto_learn alone won't
+ work.
+
+- the previous Conf code would set the default scores whenever a
+ rule was defined in the config files. unfortunately, since we
+ typically define the rule, and then the score, this means we're
+ setting the scores to the default, then again when the real scores
+ are read in. I moved the default score setting to the end of the
+ parsing routine, so all the real scores are set, then anything
+ that's left without a score gets the appropriate default.
+
+- more documentation bits around the MsgContainer and MsgParser code
+
+- got rid of some module code not required in PerMsgStatus
+
+- added a quick n' dirty useless entity modification to the URL list
+
+- added license jargon to MsgParser
+
+- fixed dns.t to exit if it actually shouldn't test anything
+
+- killed useless Data::Dumper from rule_tests.t
+
+------------------------------------------------------------------------
+r6299 | felicity | 2004-01-26 04:41:00 +0000 (Mon, 26 Jan 2004) | 1 line
+
+bug 2071: gevalia coffee rule
+------------------------------------------------------------------------
+r6298 | felicity | 2004-01-26 04:12:26 +0000 (Mon, 26 Jan 2004) | 1 line
+
+bug 2844: web browsers will correctly parse 'http://www.foo.biz?id=3' to be 'http://www.foo.biz/?id=3', so we should too...
+------------------------------------------------------------------------
+r6297 | felicity | 2004-01-26 03:27:09 +0000 (Mon, 26 Jan 2004) | 1 line
+
+bug 2963: make rule_tests work with the new mime parser. also make sure the tests run against the test configs and don't read the site or user prefs files.
+------------------------------------------------------------------------
+r6295 | felicity | 2004-01-25 22:24:35 +0000 (Sun, 25 Jan 2004) | 1 line
+
+stop running body rules a little faster ...
+------------------------------------------------------------------------
+r6282 | felicity | 2004-01-25 20:06:33 +0000 (Sun, 25 Jan 2004) | 1 line
+
+potentially mark when a bad uri encoding occurs.
+------------------------------------------------------------------------
+r6265 | felicity | 2004-01-25 03:17:01 +0000 (Sun, 25 Jan 2004) | 1 line
+
+merge the new html rendering with the old html rules ...
+------------------------------------------------------------------------
+r6263 | mss | 2004-01-24 20:11:43 +0000 (Sat, 24 Jan 2004) | 1 line
+
+Some more standardized copyrights.
+------------------------------------------------------------------------
+r6262 | mss | 2004-01-24 20:05:23 +0000 (Sat, 24 Jan 2004) | 1 line
+
+Added some license/copyright headers
+------------------------------------------------------------------------
+r6261 | quinlan | 2004-01-24 19:54:44 +0000 (Sat, 24 Jan 2004) | 2 lines
+
+use standard copyright format
+
+------------------------------------------------------------------------
+r6260 | felicity | 2004-01-24 18:36:47 +0000 (Sat, 24 Jan 2004) | 1 line
+
+bug 2956: uri tests weren't matching where they should if uris were encoded improperly. we now reencode the uris we found correctly and check them too.
+------------------------------------------------------------------------
+r6259 | felicity | 2004-01-24 18:10:34 +0000 (Sat, 24 Jan 2004) | 1 line
+
+bug 2957: add INFO_TLD, same as BIZ_TLD
+------------------------------------------------------------------------
+r6258 | felicity | 2004-01-24 18:07:38 +0000 (Sat, 24 Jan 2004) | 4 lines
+
+bug 2858: add a "PING" protocol command to spamd which allows for remote
+monitoring to see if spamd is alive without it complaining there was no
+message sent.
+
+------------------------------------------------------------------------
+r6257 | felicity | 2004-01-24 05:33:24 +0000 (Sat, 24 Jan 2004) | 1 line
+
+more parser work, replaced the get_*body functions in PerMsgStatus, etc, etc.
+------------------------------------------------------------------------
+r6256 | quinlan | 2004-01-24 01:15:36 +0000 (Sat, 24 Jan 2004) | 2 lines
+
+update to Apache License Version 2.0
+
+------------------------------------------------------------------------
+r6255 | quinlan | 2004-01-24 01:11:57 +0000 (Sat, 24 Jan 2004) | 2 lines
+
+minor fix
+
+------------------------------------------------------------------------
+r6250 | jm | 2004-01-23 06:01:21 +0000 (Fri, 23 Jan 2004) | 1 line
+
+plugin support, using new loadplugin configuration command
+------------------------------------------------------------------------
+r6249 | felicity | 2004-01-21 21:49:00 +0000 (Wed, 21 Jan 2004) | 1 line
+
+work towards merging new parser into other areas ...
+------------------------------------------------------------------------
+r6248 | mss | 2004-01-21 19:21:42 +0000 (Wed, 21 Jan 2004) | 2 lines
+
+I wonder why this file is suddenly executabel again...
+
+------------------------------------------------------------------------
+r6247 | felicity | 2004-01-20 22:33:37 +0000 (Tue, 20 Jan 2004) | 1 line
+
+bug 2939: initial work to remove Mail::Audit code
+------------------------------------------------------------------------
+r6245 | jm | 2004-01-20 21:37:01 +0000 (Tue, 20 Jan 2004) | 1 line
+
+added details of spamtrap machine etc.
+------------------------------------------------------------------------
+r6244 | jm | 2004-01-20 20:59:39 +0000 (Tue, 20 Jan 2004) | 1 line
+
+minor status updates
+------------------------------------------------------------------------
+r6240 | felicity | 2004-01-20 20:32:27 +0000 (Tue, 20 Jan 2004) | 2 lines
+
+bug 2291: routines to process body text could be run multiple times due to lack of caching of results.
+
+------------------------------------------------------------------------
+r6235 | jm | 2004-01-20 07:57:55 +0000 (Tue, 20 Jan 2004) | 1 line
+
+bug 2791: qmail Received headers may still be misinterpreted, fix from Michel Bouissou
+------------------------------------------------------------------------
+r6234 | jm | 2004-01-20 07:55:55 +0000 (Tue, 20 Jan 2004) | 1 line
+
+more SPF work; added test for numeric HELOs as the non-first Received line; added test to find images/hrefs to terra.es, which seems to have a serious problem with hosting spammers
+------------------------------------------------------------------------
+r6233 | jm | 2004-01-20 07:54:11 +0000 (Tue, 20 Jan 2004) | 1 line
+
+ok_locales now supports two new locales -- zh.big5 and zh.gb2312, to limit charsets more accurately; bug 2846
+------------------------------------------------------------------------
+r6228 | mss | 2004-01-19 11:46:23 +0000 (Mon, 19 Jan 2004) | 1 line
+
+vim likes tabs...
+------------------------------------------------------------------------
+r6227 | mss | 2004-01-19 11:36:57 +0000 (Mon, 19 Jan 2004) | 2 lines
+
+More magic for Windows.
+
+------------------------------------------------------------------------
+r6226 | mss | 2004-01-19 11:27:04 +0000 (Mon, 19 Jan 2004) | 2 lines
+
+English lesson: We're RUNNING_ON_WINDOWS, *not* IN (stupid me).
+
+------------------------------------------------------------------------
+r6216 | mss | 2004-01-18 22:44:05 +0000 (Sun, 18 Jan 2004) | 2 lines
+
+Some refactoring of our test environment to make it work smoother with Windows.
+
+------------------------------------------------------------------------
+r6215 | mss | 2004-01-18 22:43:24 +0000 (Sun, 18 Jan 2004) | 2 lines
+
+Corrected a changed filename.
+
+------------------------------------------------------------------------
+r6212 | felicity | 2004-01-18 18:58:37 +0000 (Sun, 18 Jan 2004) | 1 line
+
+get rid of the last 'delivery' code (log-to-mbox), move pristine header into the mime parser.
+------------------------------------------------------------------------
+r6211 | mss | 2004-01-18 18:36:44 +0000 (Sun, 18 Jan 2004) | 2 lines
+
+Add t/SATest.pl which I forgot, sorted this bloody file alphabetically while I was around.
+
+------------------------------------------------------------------------
+r6209 | mss | 2004-01-18 18:21:14 +0000 (Sun, 18 Jan 2004) | 2 lines
+
+Add workaround for systems/shells (esp. Windows) which do not support calls a la 'cmd >file 2>&1 &'. Hope this really works under Windows, will test it in afew hours.
+
+------------------------------------------------------------------------
+r6206 | mss | 2004-01-18 16:56:24 +0000 (Sun, 18 Jan 2004) | 2 lines
+
+Don't knock over the slashes under Windows -- the apps have to handle it themselves.
+
+------------------------------------------------------------------------
+r6200 | felicity | 2004-01-18 06:28:02 +0000 (Sun, 18 Jan 2004) | 4 lines
+
+starting to get rid of the Mail::Audit stuff
+mime parser will now accept parse data that's already an array or a
+file descriptor.
+
+------------------------------------------------------------------------
+r6199 | felicity | 2004-01-18 05:55:08 +0000 (Sun, 18 Jan 2004) | 1 line
+
+updated spec file, removed unkempt polish translations, etc.
+------------------------------------------------------------------------
+r6197 | felicity | 2004-01-18 05:06:21 +0000 (Sun, 18 Jan 2004) | 8 lines
+
+More parser work ...
+- be a little smarter when parsing the boundary. pay a little more attention to the rfc.
+- do decode/render on-first-request. this'll save time during parsing
+ and memory until the need arises for a decoded/rendered part.
+- generalized the text v. html difference rule to not know about internal data formats
+- the parser will now deal with all of the message headers, not just mime-related ones
+- decoding and rendering are now done in the object, not in the parser
+
+------------------------------------------------------------------------
+r6189 | felicity | 2004-01-17 22:33:54 +0000 (Sat, 17 Jan 2004) | 1 line
+
+add warning to build doc
+------------------------------------------------------------------------
+r6188 | felicity | 2004-01-17 22:30:58 +0000 (Sat, 17 Jan 2004) | 1 line
+
+update extra_versions and docs for svn
+------------------------------------------------------------------------
+r6182 | mss | 2004-01-17 17:56:19 +0000 (Sat, 17 Jan 2004) | 2 lines
+
+Set svn:ignore values for spamc.
+
+------------------------------------------------------------------------
+r6181 | mss | 2004-01-17 17:50:25 +0000 (Sat, 17 Jan 2004) | 2 lines
+
+The spamd directory was getting too crowded and became a mess of EU::MM and autoconf generated stuff. So all spamc stuff (including qmail-spamc) was moved to its own directory. 'make disttest' finished successfully both on Linux and Windows but YMMV so please test. Especially packagers might want to check their scripts.
+
+------------------------------------------------------------------------
+r6180 | mss | 2004-01-16 21:33:33 +0000 (Fri, 16 Jan 2004) | 1 line
+
+remove some obsolete entries
+------------------------------------------------------------------------
+r6177 | mss | 2004-01-16 01:19:47 +0000 (Fri, 16 Jan 2004) | 1 line
+
+Set some svn:keywords properties.
+------------------------------------------------------------------------
+r6171 | mss | 2004-01-15 19:26:43 +0000 (Thu, 15 Jan 2004) | 11 lines
+
+bug 2928: avoid syslog-hickups on weird Message-Ids:
+* check(): be more specific when parsing the id, replace all control and high chars with question marks
+* logmsg(): also replace all control chars with underscores
+
+also made the "processing" log message easier to parse. its format is now:
+ logmsg = ("processing" | "checking")
+ WSP "message" (("<" msgid ">") | "(unknown)")
+ [ WSP "aka" "<" resent-msgid ">" ]
+ WSP "for" (uname | "(unknown)") ":" uid
+ "."
+
+------------------------------------------------------------------------
+r6170 | mss | 2004-01-15 00:34:30 +0000 (Thu, 15 Jan 2004) | 2 lines
+
+bug 2924: Don't create a malformed ("from" Domain Address-literal) Received header.
+
+------------------------------------------------------------------------
+r6169 | felicity | 2004-01-14 07:44:26 +0000 (Wed, 14 Jan 2004) | 10 lines
+
+More parser work. Cleaned up debug statements some, got rid of
+some more code that is unneeded now that we do a tree as opposed to
+a pseudo-tree, resolved a potential recursion issue with leaf nodes
+pointing to themselves, simplified some code to use -1 for the last
+element in an array instead of "scalar @array - 1", find_parts() will
+now look at the tree structure to find children instead of just looking
+for children under multipart/* nodes (it's the same thing, but now it's
+not tied to the content-type, just the fact there are children nodes...)
+
+
+------------------------------------------------------------------------
+r6167 | jm | 2004-01-14 07:41:56 +0000 (Wed, 14 Jan 2004) | 1 line
+
+bug 2787: Received header misunderstood; bug 2896: debug code to report unparseable received header incorrect
+------------------------------------------------------------------------
+r6166 | felicity | 2004-01-14 06:16:30 +0000 (Wed, 14 Jan 2004) | 4 lines
+
+Added new rules to determine the % difference between text/html and text/*
+attachments in a multipart/alternative mail. It's got amazing hit-rate
+and accuracy on my corpus even given the simple algorithm I used.
+
+------------------------------------------------------------------------
+r6165 | felicity | 2004-01-14 06:12:43 +0000 (Wed, 14 Jan 2004) | 10 lines
+
+More parser work. The parser now handles all of my test messages
+including rather annoying corner cases. I added some comment/pod
+documentation so the process is a little less cloudy (I'm still not quite
+sure where tree nodes are created ...) I got rid of a bunch of duplicated
+code and made some of the parser more generic. The tree that's made is
+now actually a full tree instead of a sorta-tree (root nodes had leaf
+data, but no actual leaves ...) Renamed private functions to have an
+underscore in front of their names. Implemented a find_parts() function
+to search the tree and return leaves/sub-trees that match. Phew!
+
+------------------------------------------------------------------------
+r6164 | felicity | 2004-01-14 05:54:21 +0000 (Wed, 14 Jan 2004) | 4 lines
+
+bug 2569: MSGID_FROM_MTA_SHORT was FPing on certain postfix received
+headers due to its reliance on what old Postfix versions did. updated to
+be more accurate although the hit rate goes down.
+
+------------------------------------------------------------------------
+r6161 | felicity | 2004-01-14 05:46:04 +0000 (Wed, 14 Jan 2004) | 3 lines
+
+per bug 2822, added a section around the bayes expiry debug statements to completely ignore expiry
+histogram output if debug is disabled. this will cause a for() loop to be skipped. :)
+
+------------------------------------------------------------------------
+r6160 | felicity | 2004-01-14 05:39:22 +0000 (Wed, 14 Jan 2004) | 2 lines
+
+bug 2871: fixes for FORGED_YAHOO_RCVD fps
+
+------------------------------------------------------------------------
+r6157 | mss | 2004-01-13 23:56:20 +0000 (Tue, 13 Jan 2004) | 2 lines
+
+bug 2900: log also an existing Resent-Message-Id to ease log-analysis.
+
+------------------------------------------------------------------------
+r6156 | mss | 2004-01-13 23:05:15 +0000 (Tue, 13 Jan 2004) | 2 lines
+
+Skip spamd regression tests on Windows (based on patches by Sidney Markowitz).
+
+------------------------------------------------------------------------
+r6155 | felicity | 2004-01-13 18:14:35 +0000 (Tue, 13 Jan 2004) | 3 lines
+
+make sure content-type is valid, parse all multipart/* types the same
+(don't make exception for alternative)
+
+------------------------------------------------------------------------
+r6154 | jm | 2004-01-13 08:03:18 +0000 (Tue, 13 Jan 2004) | 1 line
+
+spamd -m default was not taking effect; fixed
+------------------------------------------------------------------------
+r6152 | jm | 2004-01-13 07:45:46 +0000 (Tue, 13 Jan 2004) | 1 line
+
+bug 2919: bayes expiry may fail if a previous expiry was killed
+------------------------------------------------------------------------
+r6148 | felicity | 2004-01-13 03:12:53 +0000 (Tue, 13 Jan 2004) | 2 lines
+
+I've found that at least HTML::Parser 3.31 will do something stupid with and render it into "?".
+
+------------------------------------------------------------------------
+r6147 | felicity | 2004-01-13 01:39:29 +0000 (Tue, 13 Jan 2004) | 2 lines
+
+more fixing of the parser
+
+------------------------------------------------------------------------
+r6146 | jm | 2004-01-13 01:22:52 +0000 (Tue, 13 Jan 2004) | 1 line
+
+now seeing this in spam/scam mails
+------------------------------------------------------------------------
+r6145 | felicity | 2004-01-12 19:32:44 +0000 (Mon, 12 Jan 2004) | 7 lines
+
+the parser still had a few bugs ... headers from previous same-level attachments would carry over (ie:
+content-type, content-type-encoding, etc.) we wouldn't actually get a tree, but a flat parse of the
+message, so things like mult/mixed(mult/alt(text/plain,text/html),application/msword) would become
+text/plain, text/html, application/msword ... we now get objectization in the parse tree so that the full
+message is multipart/mixed which has a multipart/alternative and application/msword part, and the
+multipart/alternative part has a text/plain and text/html part.
+
+------------------------------------------------------------------------
+r6143 | felicity | 2004-01-12 03:37:40 +0000 (Mon, 12 Jan 2004) | 4 lines
+
+debug statements shouldn't put \n at the end. fixed a slight bug with my last patch where m@^text/html$@
+would be rendered, but it would ignore valid "text/html; ....."
+
+
+------------------------------------------------------------------------
+r6142 | felicity | 2004-01-11 23:54:08 +0000 (Sun, 11 Jan 2004) | 2 lines
+
+we need to render text/html or text/* that looks like HTML ...
+
+------------------------------------------------------------------------
+r6141 | felicity | 2004-01-11 19:21:28 +0000 (Sun, 11 Jan 2004) | 3 lines
+
+perl was complaining about the use of qw{} and text with a comma to
+figure out last modified date of SA.pm. changed to split instead.
+
+------------------------------------------------------------------------
+r6140 | felicity | 2004-01-11 18:19:04 +0000 (Sun, 11 Jan 2004) | 4 lines
+
+bug 2917: Failure to remove /tmp/spamd-$$-init for certain settings of
+auto_whitelist_path (and bayes_path)
+
+
+------------------------------------------------------------------------
+r6137 | mss | 2004-01-10 19:53:16 +0000 (Sat, 10 Jan 2004) | 2 lines
+
+Fix some bugs Sidney pointed out in the Native Windows Build stuff.
+
+------------------------------------------------------------------------
+r6136 | mss | 2004-01-09 22:40:18 +0000 (Fri, 09 Jan 2004) | 2 lines
+
+Set the correct mime-types on all files. Hopefully the next version of my IDE will pick up those SVN props and use them accordingly :)
+
+------------------------------------------------------------------------
+r6135 | mss | 2004-01-09 21:29:57 +0000 (Fri, 09 Jan 2004) | 2 lines
+
+Adapted the versioning code to SVN.
+
+------------------------------------------------------------------------
+r6134 | jm | 2004-01-09 19:26:41 +0000 (Fri, 09 Jan 2004) | 1 line
+
+bug 2915: added rule to detect Message-ID headers that don't match Received
+------------------------------------------------------------------------
+r6133 | jm | 2004-01-09 06:41:18 +0000 (Fri, 09 Jan 2004) | 1 line
+
+added replacement for RCVD_NUMERIC_HELO using new Received parser; added 'tflags nice' for T_ALL_TRUSTED
+------------------------------------------------------------------------
+r6132 | jm | 2004-01-09 06:26:51 +0000 (Fri, 09 Jan 2004) | 1 line
+
+added test for tiny font sizes
+------------------------------------------------------------------------
+r6127 | jm | 2004-01-09 01:37:40 +0000 (Fri, 09 Jan 2004) | 1 line
+
+removed CBL again, since Spamhaus XBL includes all that database
+------------------------------------------------------------------------
+r6126 | felicity | 2004-01-08 23:56:37 +0000 (Thu, 08 Jan 2004) | 3 lines
+
+bug 2608: HABEAS_SWE wasn't trigerring if the X-Habeas-SWE headers are
+out of order in the message header (some MUAs do this).
+
+------------------------------------------------------------------------
+r6123 | felicity | 2004-01-08 23:45:15 +0000 (Thu, 08 Jan 2004) | 2 lines
+
+bug 2856: 'report_safe_copy_headers Received' would reverse the order of the Received headers
+
+------------------------------------------------------------------------
+r6120 | jm | 2004-01-08 23:28:38 +0000 (Thu, 08 Jan 2004) | 1 line
+
+use of UNUSED_VARIABLE cannot be interspersed between variable declarations with older compilers
+------------------------------------------------------------------------
+r6119 | jm | 2004-01-08 22:59:08 +0000 (Thu, 08 Jan 2004) | 1 line
+
+added rule to catch the 'Re: UPPERCASE, word word word' spams by Subject
+------------------------------------------------------------------------
+r6117 | jm | 2004-01-08 07:06:31 +0000 (Thu, 08 Jan 2004) | 1 line
+
+added ASL 1.1 text to all source, rules and many scripts
+------------------------------------------------------------------------
+r6116 | jm | 2004-01-08 06:59:30 +0000 (Thu, 08 Jan 2004) | 1 line
+
+added license-replacing helper script
+------------------------------------------------------------------------
+r6115 | jm | 2004-01-08 06:42:45 +0000 (Thu, 08 Jan 2004) | 1 line
+
+preparing to insert license text using an automated tool
+------------------------------------------------------------------------
+r6114 | jm | 2004-01-08 06:40:11 +0000 (Thu, 08 Jan 2004) | 1 line
+
+some files showing up even though they were removed on sf.net
+------------------------------------------------------------------------
+r6112 | jm | 2004-01-08 04:40:47 +0000 (Thu, 08 Jan 2004) | 1 line
+
+updating distro to use ASL instead of GPL/PAL; preparing for massive license fixup
+------------------------------------------------------------------------
+r6110 | duncf | 2004-01-08 03:06:36 +0000 (Thu, 08 Jan 2004) | 4 lines
+
+Fixes backwards compat of rewrite_subject -> rewrite_header which I
+screwed up the first time... Also removes any reference to
+rewrite_subject in the docs. (Closes: #2905)
+
+------------------------------------------------------------------------
+r6109 | jm | 2004-01-08 02:57:34 +0000 (Thu, 08 Jan 2004) | 1 line
+
+bug 2842: exiscan spamc passes empty User: on to spamd, fix support for this
+------------------------------------------------------------------------
+r6108 | jm | 2004-01-08 02:51:57 +0000 (Thu, 08 Jan 2004) | 1 line
+
+bug 2842: empty User header breaks exiscan
+------------------------------------------------------------------------
+r6106 | jm | 2004-01-08 02:44:13 +0000 (Thu, 08 Jan 2004) | 2 lines
+
+bug 2787: Received header parser not handling one type of header
+
+------------------------------------------------------------------------
+r6104 | jm | 2004-01-08 02:20:24 +0000 (Thu, 08 Jan 2004) | 2 lines
+
+removed old GPL/PAL contrib cert
+
+------------------------------------------------------------------------
+r6103 | jm | 2004-01-08 02:19:28 +0000 (Thu, 08 Jan 2004) | 2 lines
+
+Added STATUS file, Apache-style
+
+------------------------------------------------------------------------
+r6102 | mss | 2004-01-06 21:23:31 +0000 (Tue, 06 Jan 2004) | 1 line
+
+Forgot to remove some debugging code.
+------------------------------------------------------------------------
+r6101 | mss | 2004-01-06 21:21:15 +0000 (Tue, 06 Jan 2004) | 2 lines
+
+Some Native Windows fixes.
+
+------------------------------------------------------------------------
+r6100 | mss | 2004-01-06 14:19:46 +0000 (Tue, 06 Jan 2004) | 2 lines
+
+Fixed a stupid bug which always enabled SSL on Windows.
+
+------------------------------------------------------------------------
+r6099 | mss | 2004-01-06 02:36:44 +0000 (Tue, 06 Jan 2004) | 2 lines
+
+Added a modified version of Sidney's Native Windows Patch(tm). build/configure is now enabled.
+
+------------------------------------------------------------------------
+r6098 | mss | 2004-01-06 02:29:49 +0000 (Tue, 06 Jan 2004) | 1 line
+
+Oops. Stupid typo. It's late.
+------------------------------------------------------------------------
+r6097 | mss | 2004-01-06 02:11:38 +0000 (Tue, 06 Jan 2004) | 2 lines
+
+Added some autoconf wrapper/replacement to make it easier to incorporate Sidney's Native Windows Patches(tm) -- not yet enabled.
+
+------------------------------------------------------------------------
+r6096 | mss | 2004-01-06 02:08:14 +0000 (Tue, 06 Jan 2004) | 3 lines
+
+* Make -Mvars work with uninitialized variables.
+* Don't fiddle around with STDIN/-OUT. This is probably not really necessary but I changed it while hunting another bug and it won't hurt.
+
+------------------------------------------------------------------------
+r6095 | jm | 2004-01-05 22:57:49 +0000 (Mon, 05 Jan 2004) | 2 lines
+
+added comment attribution
+
+------------------------------------------------------------------------
+r6094 | jm | 2004-01-05 22:55:57 +0000 (Mon, 05 Jan 2004) | 2 lines
+
+added MSN redirector test rule
+
+------------------------------------------------------------------------
+r6093 | jm | 2004-01-05 22:52:25 +0000 (Mon, 05 Jan 2004) | 2 lines
+
+added XBL test
+
+------------------------------------------------------------------------
+r6092 | felicity | 2004-01-05 16:28:02 +0000 (Mon, 05 Jan 2004) | 3 lines
+
+bug 2864: RPM building on newer RPM versions would cause complaints for
+files installed but not packaged, etc. quell the warnings.
+
+------------------------------------------------------------------------
+r6090 | jm | 2004-01-05 08:20:43 +0000 (Mon, 05 Jan 2004) | 4 lines
+
+SPF tweak; avoid fetchmail-incurred FPs, and use envelope-from data from
+Received headers where available.
+
+
+------------------------------------------------------------------------
+r6089 | jm | 2004-01-05 07:57:49 +0000 (Mon, 05 Jan 2004) | 3 lines
+
+Added some more test rules that seem good against the "DFKGJDFGJ, the blah etc"
+spammer
+
+------------------------------------------------------------------------
+r6088 | mss | 2004-01-04 23:07:22 +0000 (Sun, 04 Jan 2004) | 2 lines
+
+(Finally) corrected some executable bits
+
+------------------------------------------------------------------------
+r6087 | duncf | 2004-01-02 01:32:59 +0000 (Fri, 02 Jan 2004) | 4 lines
+
+Updated sa-stats.pl from Bob Apthorpe (Bug 2870) Also included a hack
+to fix Parse::Syslog brokenness and allow the script to work with log
+files not in current year. (I.e. after new years)
+
+------------------------------------------------------------------------
+r6086 | mss | 2003-12-30 23:20:57 +0000 (Tue, 30 Dec 2003) | 2 lines
+
+Add SVN files .r[0-9] and .mine to list of ignored files, remove obsolete windows_install.
+
+------------------------------------------------------------------------
+r6085 | duncf | 2003-12-30 06:10:04 +0000 (Tue, 30 Dec 2003) | 11 lines
+
+Adds rewrite_header option to allow tagging of Subject, To and From
+headers. For To and From headers, markup will be in the form of an RFC
+2822 Comment (in parentheses separated by a tab from the
+address). (Closes bug 478)
+
+Also, fixes Subject tagging to allow for all tags to be used, noting
+that remobval of markup will only work if report_safe is > 0, or only
+_HITS_ and _REQD_ are used. (Closes 2550)
+
+Thanks to Malte for some of the regexps I stole from him ;-)
+
+------------------------------------------------------------------------
+r6084 | duncf | 2003-12-29 19:22:09 +0000 (Mon, 29 Dec 2003) | 3 lines
+
+Remove .cvsignore files and setting svn:ignore as suggested by Michael
+Parker
+
+------------------------------------------------------------------------
+r6083 | duncf | 2003-12-29 19:06:08 +0000 (Mon, 29 Dec 2003) | 1 line
+
+Have makefile.pl ignore .svn directory when building blib
+------------------------------------------------------------------------
+r6082 | duncf | 2003-12-29 18:14:12 +0000 (Mon, 29 Dec 2003) | 2 lines
+
+Updating MAINFEST to deal with file removals (why do we always forget this?)
+
+------------------------------------------------------------------------
+r6081 | felicity | 2003-12-27 06:52:04 +0000 (Sat, 27 Dec 2003) | 5 lines
+
+bug 2837: -w is a left over from early versions of SpamAssassin, but
+the program is now designed to only filter mail, so it shouldn't try
+responding with a bounce. disabling/removing "spamassassin -w" code.
+
+
+------------------------------------------------------------------------
+r6080 | felicity | 2003-12-27 06:29:56 +0000 (Sat, 27 Dec 2003) | 6 lines
+
+bug 2827: sometimes there is a "Argument "" isn't numeric in numeric lt
+(<)" warning that gets printed from the Bayes code. it turns out that
+where the DB module ought to be returning undef, it sometimes returns
+"", so add a check for that possibility.
+
+
+------------------------------------------------------------------------
+r6079 | felicity | 2003-12-27 06:18:44 +0000 (Sat, 27 Dec 2003) | 5 lines
+
+bug 2851: when calling spamassassin -r and -k (report and revoke),
+Bayes will see the modified message (ie: SA markup included), which is
+a bad thing. patched the functions to strip the markup pre-learn.
+
+
+------------------------------------------------------------------------
+r6077 | felicity | 2003-12-26 04:13:36 +0000 (Fri, 26 Dec 2003) | 3 lines
+
+bug 2847: Let CmdLearn know whether or not the learning attempt completed
+or failed. If it failed, throw an error and die.
+
+------------------------------------------------------------------------
+r6040 | felicity | 2003-12-17 15:06:29 +0000 (Wed, 17 Dec 2003) | 4 lines
+
+bug 2847: added an "unavailable" and "failed" state for _AUTOLEARN_
+which describes when bayes isn't available (use_bayes 0, or no DB_File,
+etc,) and when learning fails (see debug output) respectively.
+
+------------------------------------------------------------------------
+r6039 | jmason | 2003-12-17 08:09:00 +0000 (Wed, 17 Dec 2003) | 2 lines
+
+SPF update
+
+------------------------------------------------------------------------
+r6038 | jmason | 2003-12-17 07:08:44 +0000 (Wed, 17 Dec 2003) | 2 lines
+
+SPF code updated for latest Mail::SPF::Query release
+
+------------------------------------------------------------------------
+r6037 | jmason | 2003-12-17 06:23:48 +0000 (Wed, 17 Dec 2003) | 2 lines
+
+interim checkin of SPF changes
+
+------------------------------------------------------------------------
+r6036 | jmason | 2003-12-17 06:18:25 +0000 (Wed, 17 Dec 2003) | 2 lines
+
+promoted RCVD_IN_RSL and RCVD_IN_CBL, gave them approximate scores based on similar GA tests from http://www.pathname.com/~corpus/NET.7day
+
+------------------------------------------------------------------------
+r6035 | jmason | 2003-12-17 06:17:42 +0000 (Wed, 17 Dec 2003) | 2 lines
+
+promoted RCVD_IN_RSL and RCVD_IN_CBL, gave them approximate scores based on similar GA tests from http://www.pathname.com/~corpus/NET.7day
+
+------------------------------------------------------------------------
+r6034 | jmason | 2003-12-17 06:14:52 +0000 (Wed, 17 Dec 2003) | 2 lines
+
+promoted RCVD_IN_RSL and RCVD_IN_CBL, gave them approximate scores based on similar GA tests from http://www.pathname.com/~corpus/NET.7day
+
+------------------------------------------------------------------------
+r6033 | jmason | 2003-12-17 05:51:56 +0000 (Wed, 17 Dec 2003) | 2 lines
+
+FINALLY fixed mass-check to have a usage msg
+
+------------------------------------------------------------------------
+r6032 | felicity | 2003-12-16 22:55:36 +0000 (Tue, 16 Dec 2003) | 2 lines
+
+bug 2848: DB_File detection wasn't working properly in the Bayes tie_db_* functions.
+
+------------------------------------------------------------------------
+r6031 | jmason | 2003-12-16 20:14:11 +0000 (Tue, 16 Dec 2003) | 2 lines
+
+added anti-Altavista-forgery rules
+
+------------------------------------------------------------------------
+r6030 | felicity | 2003-12-16 05:11:00 +0000 (Tue, 16 Dec 2003) | 2 lines
+
+for debug information, display which configuration files are being read.
+
+------------------------------------------------------------------------
+r6029 | jmason | 2003-12-15 08:21:21 +0000 (Mon, 15 Dec 2003) | 2 lines
+
+bug 1041: add a method to determine intended recipient's address; bug 796: add hashcash support; also a few doco cleanups
+
+------------------------------------------------------------------------
+r6028 | felicity | 2003-12-10 18:58:01 +0000 (Wed, 10 Dec 2003) | 3 lines
+
+bug 2822: added a listing of atime deltas and reduction counts to help
+people understand why expiry isn't occuring.
+
+------------------------------------------------------------------------
+r6027 | jmason | 2003-12-10 07:41:48 +0000 (Wed, 10 Dec 2003) | 2 lines
+
+bug 2661: dropped rules from contributor Wolfgang Weisselberg due to missing CLA
+
+------------------------------------------------------------------------
+r6026 | jmason | 2003-12-10 07:26:04 +0000 (Wed, 10 Dec 2003) | 2 lines
+
+bug 2688: removed contributor Radoslaw Stachowiak due to lack of CLA
+
+------------------------------------------------------------------------
+r6025 | jmason | 2003-12-10 07:25:02 +0000 (Wed, 10 Dec 2003) | 2 lines
+
+some vestiges of code from Thomas Hurst removed; no CLA received
+
+------------------------------------------------------------------------
+r6024 | jmason | 2003-12-10 07:22:23 +0000 (Wed, 10 Dec 2003) | 2 lines
+
+terse-report vestiges removed
+
+------------------------------------------------------------------------
+r6023 | jmason | 2003-12-10 07:18:03 +0000 (Wed, 10 Dec 2003) | 2 lines
+
+clean-roomed pyzor_options
+
+------------------------------------------------------------------------
+r6022 | jmason | 2003-12-10 07:13:15 +0000 (Wed, 10 Dec 2003) | 2 lines
+
+removed always_add_report, pyzor_options; didn't get CLA from Steve Abatangle
+
+------------------------------------------------------------------------
+r6021 | jmason | 2003-12-10 07:03:36 +0000 (Wed, 10 Dec 2003) | 2 lines
+
+removed contributor David Barroso; no CLA received
+
+------------------------------------------------------------------------
+r6013 | jmason | 2003-12-09 06:15:53 +0000 (Tue, 09 Dec 2003) | 2 lines
+
+bug 2787: Received line misparsed due to ordering problem
+
+------------------------------------------------------------------------
+r6012 | jmason | 2003-12-04 21:20:30 +0000 (Thu, 04 Dec 2003) | 2 lines
+
+bug 2625: spamd -m5 now default
+
+------------------------------------------------------------------------
+r6011 | jmason | 2003-12-04 21:19:27 +0000 (Thu, 04 Dec 2003) | 2 lines
+
+oops, typo
+
+------------------------------------------------------------------------
+r6009 | jmason | 2003-12-04 17:32:09 +0000 (Thu, 04 Dec 2003) | 2 lines
+
+minor updates to SQL doco, note use of spamd -q switch
+
+------------------------------------------------------------------------
+r6008 | felicity | 2003-12-03 22:43:34 +0000 (Wed, 03 Dec 2003) | 2 lines
+
+fixed documentation for bayes_journal_max_size. 0 disables opportunistic sync, not force once a day.
+
+------------------------------------------------------------------------
+r6006 | jmason | 2003-12-03 02:36:34 +0000 (Wed, 03 Dec 2003) | 2 lines
+
+bug 2676: John Levine patch reinstated, received CLA
+
+------------------------------------------------------------------------
+r6005 | jmason | 2003-12-03 02:33:34 +0000 (Wed, 03 Dec 2003) | 2 lines
+
+bug 2670, bug 2666: got CLAs for Brad Rathbun and Bob Apthorpe so can readd tools/sa-stats.pl
+
+------------------------------------------------------------------------
+r6004 | jmason | 2003-12-03 02:31:43 +0000 (Wed, 03 Dec 2003) | 2 lines
+
+bug 2710: Anirvan Chatterjee rule reinstated on CLA
+
+------------------------------------------------------------------------
+r6003 | jmason | 2003-12-03 02:27:44 +0000 (Wed, 03 Dec 2003) | 2 lines
+
+bug 2705: Dave Lugo's contribution reinstated
+
+------------------------------------------------------------------------
+r6002 | jmason | 2003-12-03 02:25:37 +0000 (Wed, 03 Dec 2003) | 2 lines
+
+bug 2671: Chris Eykamp rules reinstated
+
+------------------------------------------------------------------------
+r6001 | jmason | 2003-12-03 02:22:07 +0000 (Wed, 03 Dec 2003) | 2 lines
+
+bug 2662: Alexander Kourakos CLA received, reinstating patch
+
+------------------------------------------------------------------------
+r6000 | jmason | 2003-12-02 08:20:03 +0000 (Tue, 02 Dec 2003) | 2 lines
+
+updated for new Dynablock DNSBL location; re-added PDL lookup
+
+------------------------------------------------------------------------
+r5999 | jmason | 2003-12-02 08:05:15 +0000 (Tue, 02 Dec 2003) | 2 lines
+
+couple of new SORBS subrules to test
+
+------------------------------------------------------------------------
+r5997 | jmason | 2003-12-02 07:45:32 +0000 (Tue, 02 Dec 2003) | 2 lines
+
+bug 2758: doco issue for _STARS_ capping
+
+------------------------------------------------------------------------
+r5996 | duncf | 2003-11-25 15:49:35 +0000 (Tue, 25 Nov 2003) | 10 lines
+
+Removing the debian/ directory -- it's not elegant to begin with as I
+have to fudge some stuff to get it to build from cvs anyways, and I'd
+rather start an alioth site or something to collabratively maintain
+the packaging. When I actually do this, I'll write a README on how to
+build your own packages.
+
+Also, since debian/rules was deleted(!), this is all useless anyways,
+and I'm not going to reimplement the minor patches (like putting
+quotes around $(CFLAGS))
+
+------------------------------------------------------------------------
+r5995 | jmason | 2003-11-25 07:47:29 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+file removed
+
+------------------------------------------------------------------------
+r5994 | jmason | 2003-11-25 07:33:42 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+bug 2685: removed contributor Michael Moose Dinn
+
+------------------------------------------------------------------------
+r5993 | jmason | 2003-11-25 07:19:01 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contributions from Matt Simerson
+
+------------------------------------------------------------------------
+r5992 | jmason | 2003-11-25 07:09:11 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contributor Marek Huda'k
+
+------------------------------------------------------------------------
+r5991 | jmason | 2003-11-25 07:03:38 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+oops, broke the rules files
+
+------------------------------------------------------------------------
+r5990 | jmason | 2003-11-25 07:01:52 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+bug 2674: removed contributions from David Kewley
+
+------------------------------------------------------------------------
+r5989 | jmason | 2003-11-25 03:02:49 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+dropped contributor Chris Eykamp
+
+------------------------------------------------------------------------
+r5988 | jmason | 2003-11-25 03:01:22 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contributor Mike Nolan
+
+------------------------------------------------------------------------
+r5987 | jmason | 2003-11-25 02:57:27 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+dropped contributor Chris Eykamp
+
+------------------------------------------------------------------------
+r5986 | jmason | 2003-11-25 02:12:45 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contributor Radoslaw Zielinski
+
+------------------------------------------------------------------------
+r5985 | jmason | 2003-11-25 02:08:59 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contributor Robert James Kaes
+
+------------------------------------------------------------------------
+r5984 | jmason | 2003-11-25 02:05:47 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+bug 2672: removed 30_text_de.cf due to a lack of CLAs from some of the contributors
+
+------------------------------------------------------------------------
+r5983 | jmason | 2003-11-25 02:01:34 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contributor Bob Apthorpe due to lack of CLA
+
+------------------------------------------------------------------------
+r5982 | jmason | 2003-11-25 01:50:23 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contrib files that we still have no CLA for
+
+------------------------------------------------------------------------
+r5981 | jmason | 2003-11-25 01:45:17 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+removed contributor Ian Murdock (bug 2694)
+
+------------------------------------------------------------------------
+r5980 | jmason | 2003-11-25 01:43:27 +0000 (Tue, 25 Nov 2003) | 2 lines
+
+bug 2695: removed contributor Tino Keitel
+
+------------------------------------------------------------------------
+r5979 | jmason | 2003-11-24 05:18:33 +0000 (Mon, 24 Nov 2003) | 2 lines
+
+bug 2774: undef warnings fixed in qmail Received-header parsing
+
+------------------------------------------------------------------------
+r5978 | jmason | 2003-11-20 19:35:06 +0000 (Thu, 20 Nov 2003) | 2 lines
+
+added more debugging to RBL code; found bug in interim list of reserved-IPs, fixed pending their clean-room reimplementation
+
+------------------------------------------------------------------------
+r5975 | jmason | 2003-11-20 07:22:48 +0000 (Thu, 20 Nov 2003) | 2 lines
+
+bug 2759: Received parser can't deal with a variety of qmail formats
+
+------------------------------------------------------------------------
+r5974 | jmason | 2003-11-19 20:25:27 +0000 (Wed, 19 Nov 2003) | 2 lines
+
+bug 2742: updated French translations, Michel Bouissou
+
+------------------------------------------------------------------------
+r5973 | jmason | 2003-11-19 02:37:50 +0000 (Wed, 19 Nov 2003) | 2 lines
+
+bug 2742: new set of French translations from Michel Bouissou
+
+------------------------------------------------------------------------
+r5972 | duncf | 2003-11-18 17:04:49 +0000 (Tue, 18 Nov 2003) | 2 lines
+
+Doh. $current_user was already declared... oops
+
+------------------------------------------------------------------------
+r5971 | duncf | 2003-11-18 16:49:17 +0000 (Tue, 18 Nov 2003) | 2 lines
+
+Yes this is silly. Reimplementing set current_user to value in $1. I cant test this, so I hope for the best ;-)
+
+------------------------------------------------------------------------
+r5970 | jmason | 2003-11-18 07:15:08 +0000 (Tue, 18 Nov 2003) | 2 lines
+
+bug 2692: removed contributor Tim Bell. CVS HEAD IS NOW BROKEN -- SOMEONE ELSE PLEASE REIMPLEMENT ASAP
+
+------------------------------------------------------------------------
+r5969 | jmason | 2003-11-18 07:06:12 +0000 (Tue, 18 Nov 2003) | 2 lines
+
+bug 2705: dropped contributor Dave Lugo
+
+------------------------------------------------------------------------
+r5968 | jmason | 2003-11-18 07:02:38 +0000 (Tue, 18 Nov 2003) | 2 lines
+
+bug 2698: dropped contributor James Henstridge
+
+------------------------------------------------------------------------
+r5966 | jmason | 2003-11-15 04:17:06 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+oops, screwed up the rfc-2822 regexp
+
+------------------------------------------------------------------------
+r5965 | jmason | 2003-11-15 04:13:30 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2710: drop contributor Anirvan Chatterjee
+
+------------------------------------------------------------------------
+r5964 | jmason | 2003-11-15 04:05:28 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2715: dropped a regexp from Mail::Header, since the CLA from Graham Barr seems to have gone AWOL
+
+------------------------------------------------------------------------
+r5960 | jmason | 2003-11-15 02:58:12 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2745: pobox.com Received header format handled
+
+------------------------------------------------------------------------
+r5959 | jmason | 2003-11-15 02:55:30 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2633: spamd giving undef warnings if DNS is not available for spamc client
+
+------------------------------------------------------------------------
+r5956 | jmason | 2003-11-15 02:46:25 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2502: -H switch broken due to lack of srand seeding
+
+------------------------------------------------------------------------
+r5955 | jmason | 2003-11-15 02:44:49 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2591: Received header format for something called 'smtpd' supported; added internal_networks initial implementation
+
+------------------------------------------------------------------------
+r5954 | jmason | 2003-11-15 02:49:32 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2630: spamd_parallel test no longer distributed due to load problems it causes
+
+------------------------------------------------------------------------
+r5953 | jmason | 2003-11-15 02:47:42 +0000 (Sat, 15 Nov 2003) | 2 lines
+
+bug 2504: fixed 'Use of uninitialized value in numeric eq' in spamd
+
+------------------------------------------------------------------------
+r5950 | jmason | 2003-11-13 02:49:46 +0000 (Thu, 13 Nov 2003) | 2 lines
+
+updated MANIFEST
+
+------------------------------------------------------------------------
+r5949 | jmason | 2003-11-13 02:49:10 +0000 (Thu, 13 Nov 2003) | 2 lines
+
+bug 2654: fix pattern to avoid FP on FORGED_MUA_EUDORA
+
+------------------------------------------------------------------------
+r5948 | jmason | 2003-11-12 07:36:19 +0000 (Wed, 12 Nov 2003) | 2 lines
+
+bug 2294: removed FORGED_MUA_THEBAT due to FPs and 2 other better variants being present anyway
+
+------------------------------------------------------------------------
+r5945 | jmason | 2003-11-09 02:18:56 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+oops. a test failure which I thought was due to my user_prefs, was actually a bug in the test code; fixed
+
+------------------------------------------------------------------------
+r5944 | jmason | 2003-11-09 02:14:26 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+oops, typo in previous patch application in spamd request-header parser code
+
+------------------------------------------------------------------------
+r5943 | jmason | 2003-11-09 02:11:52 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+bug 2664: drop contributor Beniamino Galvani
+
+------------------------------------------------------------------------
+r5942 | jmason | 2003-11-09 02:07:03 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+bug 2676: dropped contributor John Levine
+
+------------------------------------------------------------------------
+r5941 | jmason | 2003-11-09 01:48:20 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+bug 2662: dropped contributor Alexander Kourakos
+
+------------------------------------------------------------------------
+r5940 | jmason | 2003-11-09 01:41:02 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+bug 2693: remove contributor Tony L. Svanstrom
+
+------------------------------------------------------------------------
+r5939 | jmason | 2003-11-09 01:31:53 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+bug 2665: drop contributor Bill O'Hanlon
+
+------------------------------------------------------------------------
+r5938 | jmason | 2003-11-09 01:25:02 +0000 (Sun, 09 Nov 2003) | 2 lines
+
+bug 2560: removed build/cvs2cl-branch from CVS due to licensing
+
+------------------------------------------------------------------------
+r5937 | jmason | 2003-11-08 03:40:35 +0000 (Sat, 08 Nov 2003) | 2 lines
+
+bug 2663: dropped contributor Antonello Nocchi
+
+------------------------------------------------------------------------
+r5936 | jmason | 2003-11-08 03:29:40 +0000 (Sat, 08 Nov 2003) | 2 lines
+
+bug 2673: drop contributor David Greenaway
+
+------------------------------------------------------------------------
+r5935 | jmason | 2003-11-07 03:06:39 +0000 (Fri, 07 Nov 2003) | 2 lines
+
+bug 2580: undefined value in spamd check()
+
+------------------------------------------------------------------------
+r5931 | jmason | 2003-11-05 04:53:13 +0000 (Wed, 05 Nov 2003) | 2 lines
+
+bug 2122: FORGED_MUA rules matching for some mailing lists
+
+------------------------------------------------------------------------
+r5928 | jmason | 2003-10-29 05:39:51 +0000 (Wed, 29 Oct 2003) | 2 lines
+
+try out delphi.com anti-joe-job rule
+
+------------------------------------------------------------------------
+r5927 | felicity | 2003-10-29 05:13:13 +0000 (Wed, 29 Oct 2003) | 3 lines
+
+at the LISA postmaster BoF, the qualcomm postmaster commented that
+no mail comes from eudora.com, but it is faked a lot...
+
+------------------------------------------------------------------------
+r5926 | jmason | 2003-10-28 07:09:34 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+updated MANIFEST
+
+------------------------------------------------------------------------
+r5925 | jmason | 2003-10-28 07:08:30 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+applied patch from bug 2505 to HEAD
+
+------------------------------------------------------------------------
+r5924 | jmason | 2003-10-28 07:02:50 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+added rules to catch 'rndmx' random tags, and newish ratware (using header samples from Steve Champeon)
+
+------------------------------------------------------------------------
+r5923 | jmason | 2003-10-28 07:02:03 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+dns.t fixed; blocked.secnap.net was closed down by owner, so replacement mini-zone set up on spamassassin.org instead.
+
+------------------------------------------------------------------------
+r5922 | jmason | 2003-10-28 06:53:51 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+avoid undef warning in t/strip2.t
+
+------------------------------------------------------------------------
+r5921 | jmason | 2003-10-28 06:47:51 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+removed windows_install.pl, obsolete
+
+------------------------------------------------------------------------
+r5920 | jmason | 2003-10-28 06:45:53 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+avoid an SPF failure
+
+------------------------------------------------------------------------
+r5919 | jmason | 2003-10-28 06:45:31 +0000 (Tue, 28 Oct 2003) | 2 lines
+
+DCC installation docs updated
+
+------------------------------------------------------------------------
+r5916 | climent | 2003-10-23 14:38:42 +0000 (Thu, 23 Oct 2003) | 2 lines
+
+Small change to avoid a critical session.
+
+------------------------------------------------------------------------
+r5915 | climent | 2003-10-23 12:48:55 +0000 (Thu, 23 Oct 2003) | 2 lines
+
+FRIEND AT PUBLIC small change
+
+------------------------------------------------------------------------
+r5914 | quinlan | 2003-10-20 01:37:19 +0000 (Mon, 20 Oct 2003) | 2 lines
+
+minor documentation clarification
+
+------------------------------------------------------------------------
+r5913 | msquadrat | 2003-10-15 21:16:11 +0000 (Wed, 15 Oct 2003) | 2 lines
+
+removed stale qmail entries
+
+------------------------------------------------------------------------
+r5912 | msquadrat | 2003-10-15 16:01:40 +0000 (Wed, 15 Oct 2003) | 2 lines
+
+bug 2596: missing space kept qmail/qmail-spamc and t/do_net from being removed on 'make clean'
+
+------------------------------------------------------------------------
+r5911 | quinlan | 2003-10-15 08:08:05 +0000 (Wed, 15 Oct 2003) | 2 lines
+
+remove debugging statement
+
+------------------------------------------------------------------------
+r5910 | quinlan | 2003-10-15 07:53:12 +0000 (Wed, 15 Oct 2003) | 3 lines
+
+improved title testing (test each title separately instead of together
+or as a whole)
+
+------------------------------------------------------------------------
+r5909 | quinlan | 2003-10-15 07:51:49 +0000 (Wed, 15 Oct 2003) | 4 lines
+
+small simplification of _get_date_header_time()
+add html_title()
+trailing line comment after { is just wrong
+
+------------------------------------------------------------------------
+r5908 | quinlan | 2003-10-15 04:58:10 +0000 (Wed, 15 Oct 2003) | 4 lines
+
+add some HTML <title> test rules
+bug 2589: problems detecting html titles (handling of titles is now same as
+IE and Mozilla, first title is used regardless of position in document)
+
+------------------------------------------------------------------------
+r5907 | quinlan | 2003-10-14 05:17:23 +0000 (Tue, 14 Oct 2003) | 2 lines
+
+add some HTML entity rules
+
+------------------------------------------------------------------------
+r5906 | quinlan | 2003-10-14 03:42:45 +0000 (Tue, 14 Oct 2003) | 3 lines
+
+replace old fallback base64 decoder with new one
+fix minor bug in base64 decoder
+
+------------------------------------------------------------------------
+r5905 | quinlan | 2003-10-13 21:51:21 +0000 (Mon, 13 Oct 2003) | 2 lines
+
+speed up spamd testing on fast machines, wait longer before failing too
+
+------------------------------------------------------------------------
+r5904 | msquadrat | 2003-10-11 13:49:24 +0000 (Sat, 11 Oct 2003) | 2 lines
+
+bug 2557: enhanced qmail-spamc can now feed command line options to spamc (thanks to John Peacock)
+
+------------------------------------------------------------------------
+r5903 | msquadrat | 2003-10-10 19:57:49 +0000 (Fri, 10 Oct 2003) | 2 lines
+
+The previous commit still used $_ instead of $line in the first check.
+
+------------------------------------------------------------------------
+r5902 | msquadrat | 2003-10-10 19:49:39 +0000 (Fri, 10 Oct 2003) | 2 lines
+
+bug 2580: calling subroutines can make $_ undef
+
+------------------------------------------------------------------------
+r5901 | quinlan | 2003-10-07 08:11:25 +0000 (Tue, 07 Oct 2003) | 2 lines
+
+removing obsolete code
+
+------------------------------------------------------------------------
+r5900 | quinlan | 2003-10-07 04:15:36 +0000 (Tue, 07 Oct 2003) | 2 lines
+
+note this was idea
+
+------------------------------------------------------------------------
+r5899 | quinlan | 2003-10-06 01:08:47 +0000 (Mon, 06 Oct 2003) | 2 lines
+
+re-enable Habeas
+
+------------------------------------------------------------------------
+r5898 | quinlan | 2003-10-06 00:31:02 +0000 (Mon, 06 Oct 2003) | 2 lines
+
+add HTML renderer
+
+------------------------------------------------------------------------
+r5897 | quinlan | 2003-10-06 00:19:35 +0000 (Mon, 06 Oct 2003) | 2 lines
+
+modularize the HTML rendering code
+
+------------------------------------------------------------------------
+r5896 | quinlan | 2003-10-04 03:08:09 +0000 (Sat, 04 Oct 2003) | 2 lines
+
+fix use_auto_whitelist
+
+------------------------------------------------------------------------
+r5895 | quinlan | 2003-10-04 02:55:13 +0000 (Sat, 04 Oct 2003) | 2 lines
+
+add T_MSGID_DOLLARS_*
+
+------------------------------------------------------------------------
+r5894 | quinlan | 2003-10-04 02:44:45 +0000 (Sat, 04 Oct 2003) | 2 lines
+
+bug 2538: address problems with Outlook forgery rules
+
+------------------------------------------------------------------------
+r5892 | jmason | 2003-10-03 20:37:38 +0000 (Fri, 03 Oct 2003) | 2 lines
+
+re-added relays.visi.com, it's back
+
+------------------------------------------------------------------------
+r5891 | quinlan | 2003-10-03 00:39:12 +0000 (Fri, 03 Oct 2003) | 2 lines
+
+disable Habeas tests by default
+
+------------------------------------------------------------------------
+r5890 | quinlan | 2003-10-02 23:01:01 +0000 (Thu, 02 Oct 2003) | 2 lines
+
+add $re_start for HTML rendering heuristic
+
+------------------------------------------------------------------------
+r5889 | quinlan | 2003-10-02 22:59:00 +0000 (Thu, 02 Oct 2003) | 2 lines
+
+add text rendering place-holder
+
+------------------------------------------------------------------------
+r5888 | felicity | 2003-10-02 19:29:36 +0000 (Thu, 02 Oct 2003) | 2 lines
+
+Clean up the render area a little bit, add some docs.
+
+------------------------------------------------------------------------
+r5887 | quinlan | 2003-10-02 02:30:33 +0000 (Thu, 02 Oct 2003) | 4 lines
+
+don't open results files until we get to the point of writing results
+to avoid overwriting spam.log and ham.log accidentally (when trying to
+use an option to print to stdout, but failing to find a message)
+
+------------------------------------------------------------------------
+r5886 | felicity | 2003-10-01 05:50:51 +0000 (Wed, 01 Oct 2003) | 2 lines
+
+work on the docs a little bit
+
+------------------------------------------------------------------------
+r5885 | felicity | 2003-10-01 05:31:56 +0000 (Wed, 01 Oct 2003) | 5 lines
+
+If a multipart message has no boundary defined, or the boundary is defined
+but not found in the part, treat the whole part as the data to use.
+Otherwise, if a boundary is defined and it's found, just use the data
+between the boundaries.
+
+------------------------------------------------------------------------
+r5884 | felicity | 2003-10-01 04:39:48 +0000 (Wed, 01 Oct 2003) | 2 lines
+
+Make everything but parse() a "private" function.
+
+------------------------------------------------------------------------
+r5883 | felicity | 2003-10-01 04:36:21 +0000 (Wed, 01 Oct 2003) | 3 lines
+
+If there is no encoding, there will be no decoding. So raw == decoded,
+just point to the same data to save memory.
+
+------------------------------------------------------------------------
+r5882 | felicity | 2003-10-01 03:18:18 +0000 (Wed, 01 Oct 2003) | 2 lines
+
+parsed -> rendered
+
+------------------------------------------------------------------------
+r5881 | felicity | 2003-10-01 01:49:42 +0000 (Wed, 01 Oct 2003) | 2 lines
+
+more mime headers being kept, default text/plain charset per the RFC
+
+------------------------------------------------------------------------
+r5880 | felicity | 2003-10-01 01:39:19 +0000 (Wed, 01 Oct 2003) | 2 lines
+
+Fixed decoding issue of main body if not multipart
+
+------------------------------------------------------------------------
+r5879 | msquadrat | 2003-09-30 23:11:50 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+Added a short comment about the optional module Time::HiRes and what its used for.
+
+------------------------------------------------------------------------
+r5878 | felicity | 2003-09-30 22:48:50 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+The MIME parser would miss the last header in a message. Fixed. :)
+
+------------------------------------------------------------------------
+r5877 | felicity | 2003-09-30 18:57:24 +0000 (Tue, 30 Sep 2003) | 3 lines
+
+bug 2140: eudora sometimes has a different message-id than we expected
+before, so account for those differences.
+
+------------------------------------------------------------------------
+r5875 | quinlan | 2003-09-30 18:32:31 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+whoops, regexps can't work backwards
+
+------------------------------------------------------------------------
+r5874 | felicity | 2003-09-30 16:12:28 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+sometimes the parser can try to run through a null body part, so ignore those.
+
+------------------------------------------------------------------------
+r5873 | jmason | 2003-09-30 06:05:02 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+avoid duplicating headers in mass-check-results-to-mbox output
+
+------------------------------------------------------------------------
+r5872 | jmason | 2003-09-30 05:20:20 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+try using SPF in a DRIP style
+
+------------------------------------------------------------------------
+r5871 | jmason | 2003-09-30 05:19:19 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+try using SPF in a DRIP style
+
+------------------------------------------------------------------------
+r5870 | jmason | 2003-09-30 05:00:54 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+another Received-line format that was being missed
+
+------------------------------------------------------------------------
+r5869 | jmason | 2003-09-30 04:52:55 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+fixed 'use 5.6.1' to be better for backwards compat
+
+------------------------------------------------------------------------
+r5868 | felicity | 2003-09-30 03:34:11 +0000 (Tue, 30 Sep 2003) | 2 lines
+
+initial step to get MIME parsing in place
+
+------------------------------------------------------------------------
+r5866 | msquadrat | 2003-09-29 21:58:46 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+Increase the required version of Perl.
+
+------------------------------------------------------------------------
+r5865 | msquadrat | 2003-09-29 21:45:38 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+bug 2492 revisited: somehow a part of the MICROSOFT_EXECUTABLE got lost
+
+------------------------------------------------------------------------
+r5863 | msquadrat | 2003-09-29 21:34:31 +0000 (Mon, 29 Sep 2003) | 3 lines
+
+* Removed compatibility code for old (before 5.45) versions of ExtUtils::MakeMaker
+* Updated the docu at several places
+
+------------------------------------------------------------------------
+r5862 | felicity | 2003-09-29 20:59:28 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+fixed comment typo in spec file
+
+------------------------------------------------------------------------
+r5860 | felicity | 2003-09-29 20:10:38 +0000 (Mon, 29 Sep 2003) | 9 lines
+
+more username in subject tests:
+
+ 0.503 0.9122 0.0000 1.000 1.00 0.01 T_USERNAME_IN_SUBJECT3
+ 0.453 0.8213 0.0000 1.000 1.00 0.01 T_USERNAME_IN_SUBJECT2
+ 0.051 0.0928 0.0000 1.000 1.00 2.90 USERNAME_IN_SUBJECT
+ 2.679 4.8392 0.0243 0.995 0.99 0.01 T_USERNAME_IN_SUBJECT1
+
+:)
+
+------------------------------------------------------------------------
+r5859 | felicity | 2003-09-29 19:42:00 +0000 (Mon, 29 Sep 2003) | 3 lines
+
+bug 1889: rbl_check() wasn't able to parse a certain Received header if
+an ident string was included
+
+------------------------------------------------------------------------
+r5857 | jmason | 2003-09-29 18:13:08 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+Received: use of MX data to compute trust turned off again, it's too slow. Added timeout for ALL_RELAYS_NEAR_MXES test
+
+------------------------------------------------------------------------
+r5856 | jmason | 2003-09-29 05:58:50 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+avoid a warning
+
+------------------------------------------------------------------------
+r5855 | jmason | 2003-09-29 05:43:48 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+avoid new 'undefined value' warning
+
+------------------------------------------------------------------------
+r5854 | jmason | 2003-09-29 05:34:35 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+bug 2142: code to infer MAIL FROM: address from message added, using Return-Path, X-Envelope-From and a few other common ones, and producing a synthetic fake header, EnvelopeFrom. bug 2143: SPF support added, first draft, using Mail::SPF::Query for now. bug 2426: spamcop-ish trusted_networks determination scheme, using MX records added; seems to work well in most setups although may extend trust further than necessary in the open-relay case. Added nice test to detect messages that travelled entirely via domains and IPs where the domain's MX and the IP correlate.
+
+------------------------------------------------------------------------
+r5853 | felicity | 2003-09-29 04:37:05 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+for quoted-printable, go call M::SA::Util ...
+
+------------------------------------------------------------------------
+r5852 | quinlan | 2003-09-29 04:33:41 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+add qp_decode()
+
+------------------------------------------------------------------------
+r5851 | felicity | 2003-09-29 04:21:21 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+we should test out some other "username in subject" style rules ...
+
+------------------------------------------------------------------------
+r5850 | felicity | 2003-09-29 04:03:35 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+more mime work
+
+------------------------------------------------------------------------
+r5849 | quinlan | 2003-09-29 04:02:29 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+add HAS_MIME_BASE64
+
+------------------------------------------------------------------------
+r5848 | quinlan | 2003-09-29 03:57:22 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+add base64_decode()
+
+------------------------------------------------------------------------
+r5847 | felicity | 2003-09-29 03:53:18 +0000 (Mon, 29 Sep 2003) | 3 lines
+
+after a discussion with quinlan, changes to MIME parser to leave "raw"
+parts raw, and not slightly cooked.
+
+------------------------------------------------------------------------
+r5846 | felicity | 2003-09-29 03:05:05 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+fixed a few issues ...
+
+------------------------------------------------------------------------
+r5845 | felicity | 2003-09-29 02:59:36 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+Deal with some bad base64 encoding details, etc.
+
+------------------------------------------------------------------------
+r5843 | felicity | 2003-09-29 01:10:32 +0000 (Mon, 29 Sep 2003) | 2 lines
+
+added documentation about add_header name convention
+
+------------------------------------------------------------------------
+r5842 | felicity | 2003-09-28 16:38:44 +0000 (Sun, 28 Sep 2003) | 7 lines
+
+bug 2494:
+1) deal with bz2 builds... ;)
+2) allow release to be overriden via commandline
+3) allow CFLAGS to be modified in the usual method (RPM_OPT_FLAGS)
+4) add more documentation to be installed
+5) update to 2.61
+
+------------------------------------------------------------------------
+r5841 | felicity | 2003-09-28 16:24:41 +0000 (Sun, 28 Sep 2003) | 2 lines
+
+re-remove the 2.60 announcement from MANIFEST
+
+------------------------------------------------------------------------
+r5840 | felicity | 2003-09-28 16:24:13 +0000 (Sun, 28 Sep 2003) | 2 lines
+
+updated docs per bug 2501 (bayes changes in README and INSTALL)
+
+------------------------------------------------------------------------
+r5836 | felicity | 2003-09-27 01:46:55 +0000 (Sat, 27 Sep 2003) | 3 lines
+
+bug 2510: we should remove world-writable directories from PATH since
+it can cause issues with taint mode
+
+------------------------------------------------------------------------
+r5834 | felicity | 2003-09-27 01:26:29 +0000 (Sat, 27 Sep 2003) | 2 lines
+
+bug 2511: clean the path before running Cwd
+
+------------------------------------------------------------------------
+r5833 | jmason | 2003-09-26 19:25:26 +0000 (Fri, 26 Sep 2003) | 2 lines
+
+fix doco due to broken link
+
+------------------------------------------------------------------------
+r5832 | jmason | 2003-09-26 19:16:14 +0000 (Fri, 26 Sep 2003) | 2 lines
+
+changed version for bleeding-edge website doc updater to build
+
+------------------------------------------------------------------------
+r5831 | climent | 2003-09-25 10:44:49 +0000 (Thu, 25 Sep 2003) | 2 lines
+
+Postfix README added
+
+------------------------------------------------------------------------
+r5830 | felicity | 2003-09-25 03:38:04 +0000 (Thu, 25 Sep 2003) | 2 lines
+
+more mime work
+
+------------------------------------------------------------------------
+r5829 | quinlan | 2003-09-25 03:11:24 +0000 (Thu, 25 Sep 2003) | 2 lines
+
+bug 2459: more work on reorganizing auto-whitelist implementation
+
+------------------------------------------------------------------------
+r5828 | quinlan | 2003-09-25 03:11:01 +0000 (Thu, 25 Sep 2003) | 3 lines
+
+bug 2459: more work on reorganizing auto-whitelist implementation
+Bug 2492: MICROSOFT_EXECUTABLE slightly broken
+
+------------------------------------------------------------------------
+r5827 | jmason | 2003-09-24 22:47:43 +0000 (Wed, 24 Sep 2003) | 2 lines
+
+fixed typo
+
+------------------------------------------------------------------------
+r5826 | felicity | 2003-09-24 22:06:03 +0000 (Wed, 24 Sep 2003) | 2 lines
+
+bug 2501: put the release announcement mail into the distro.
+
+------------------------------------------------------------------------
+r5825 | felicity | 2003-09-24 22:01:12 +0000 (Wed, 24 Sep 2003) | 2 lines
+
+convert CRLF->LF in decoded MIME parts
+
+------------------------------------------------------------------------
+r5824 | felicity | 2003-09-24 21:33:33 +0000 (Wed, 24 Sep 2003) | 2 lines
+
+a bit more code cleanup
+
+------------------------------------------------------------------------
+r5823 | felicity | 2003-09-24 19:30:32 +0000 (Wed, 24 Sep 2003) | 2 lines
+
+a few updates for MIME parsing
+
+------------------------------------------------------------------------
+r5822 | quinlan | 2003-09-24 08:37:25 +0000 (Wed, 24 Sep 2003) | 2 lines
+
+minor change
+
+------------------------------------------------------------------------
+r5821 | felicity | 2003-09-24 06:18:38 +0000 (Wed, 24 Sep 2003) | 4 lines
+
+Modules for new parsing code. Mostly from Matt Sergeant's work in the SA3
+tree. I trimmed it down a lot, and got it to deal with variable/reference
+passing instead of temporary files.
+
+------------------------------------------------------------------------
+r5820 | jmason | 2003-09-24 03:11:33 +0000 (Wed, 24 Sep 2003) | 2 lines
+
+T_ rules enabled; a double-dash spam-sign
+
+------------------------------------------------------------------------
+r5819 | quinlan | 2003-09-23 23:58:03 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+use GTUBE for test email
+
+------------------------------------------------------------------------
+r5818 | quinlan | 2003-09-23 23:53:39 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+add CBL as test blacklist
+
+------------------------------------------------------------------------
+r5817 | quinlan | 2003-09-23 23:52:43 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+uncomment test rules!
+
+------------------------------------------------------------------------
+r5816 | quinlan | 2003-09-23 21:39:13 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+bug 2482: add DNSBL regression tests
+
+------------------------------------------------------------------------
+r5815 | quinlan | 2003-09-23 21:37:03 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+bug 2459: reorganize auto-whitelist implementation
+
+------------------------------------------------------------------------
+r5813 | felicity | 2003-09-23 20:34:39 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+fixed small doc thingy about razor patch
+
+------------------------------------------------------------------------
+r5812 | duncf | 2003-09-23 20:30:26 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+Bayes should ignore Gnus annotation, patch from Juergen Kreileder <jk...@blackdown.de>
+
+------------------------------------------------------------------------
+r5811 | felicity | 2003-09-23 18:18:49 +0000 (Tue, 23 Sep 2003) | 3 lines
+
+Allow "--define 'release 1_rh6'" and "--define 'srcext .bz2'" overrides
+from the rpmbuild commandline.
+
+------------------------------------------------------------------------
+r5810 | felicity | 2003-09-23 14:45:50 +0000 (Tue, 23 Sep 2003) | 2 lines
+
+2.70 devel cycle started
+
+------------------------------------------------------------------------
Modified: spamassassin/trunk/lib/Mail/SpamAssassin.pm
==============================================================================
--- spamassassin/trunk/lib/Mail/SpamAssassin.pm (original)
+++ spamassassin/trunk/lib/Mail/SpamAssassin.pm Fri Aug 27 18:02:45 2004
@@ -99,7 +99,7 @@
};
$VERSION = "3.000000"; # update after release (same format as perl $])
-$IS_DEVEL_BUILD = 1; # change for release versions
+$IS_DEVEL_BUILD = 0; # change for release versions
@ISA = qw();
@@ -108,7 +108,7 @@
# If you hacked up your SA, you should add a version_tag to you .cf files.
# This variable should not be modified directly.
-@EXTRA_VERSION = qw(rc1);
+@EXTRA_VERSION = qw(rc2);
if (defined $IS_DEVEL_BUILD && $IS_DEVEL_BUILD) {
push(@EXTRA_VERSION, ( 'r' . qw{$LastChangedRevision$ updated by SVN}[1] ));
}