2003-06-24 22:33 felicity
* rules/50_scores.cf: updated bayes scores to the 2.54 version, and
applied set 0 and set 1 to 2 and 3 scores with 0.86 multiplier.
2003-06-24 21:34 jmason
* lib/Mail/SpamAssassin/Received.pm: bug 2113: fix to consider 1
public IP addr an indication of the untrusted boundary
2003-06-24 20:23 quinlan
* rules/20_dnsbl_tests.cf: bug 2120: change hil.habeas.com to use
sa-hil.habeas.com
2003-06-24 03:25 quinlan
* lib/Mail/SpamAssassin/Conf.pm, rules/10_misc.cf: fix bug 2119:
always_add_headers 0 broken also change order of firs two
add_headers arguments, improve documentation, and add "on
_HOSTNAME" to end of X-Spam-Checker-Version header
2003-06-23 22:27 felicity
* rules/50_scores.cf: bug 2067: do a quick GA run to get the 2.60
scores in the correct ballpark
2003-06-23 19:33 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 2121: journal writes
could get into a loop if syswrite fails to actually write the data
out (quota limit, disk out of space, etc,) because we were looking
for bytes_written to be < 0 for error, whereas it's actually undef.
2003-06-23 16:21 felicity
* lib/Mail/SpamAssassin/Bayes.pm: bug 2117: undefined warnings if
the message was missing the Message-Id and either the Date or
Received header. add code to deal with that.
2003-06-23 05:34 quinlan
* lib/Mail/SpamAssassin/Dns.pm: bug 2118: use shorter _RBL_ URIs
2003-06-22 08:33 quinlan
* INSTALL, lib/Mail/SpamAssassin/PerMsgStatus.pm, t/html_obfu.t,
t/data/spam/011: bug 1551: improve handling of comment obfuscation
2003-06-22 02:58 jmason
* sa-learn.raw: added note about removing markup with sa-learn
2003-06-22 02:20 jmason
* rules/: 20_head_tests.cf, 30_text_fr.cf, 30_text_sk.cf: bug 2104:
MSGID_SPAMSIGN_ZEROES reinstated over MSGID_OE_SPAM_4ZERO
2003-06-21 06:35 quinlan
* t/data/spam/011: add some tags for legibility and remove
some unused stuff
2003-06-21 03:16 quinlan
* Makefile.PL: fix bug 2110: perl Makefile.PL CONTACT_ADDRESS="blah
blah blah" doesn't work
2003-06-21 02:25 duncf
* debian/: README.Exim3, README.Upgrade, changelog.in, control,
rules, spamassassin.README.Debian, spamassassin.config,
spamassassin.default, spamassassin.examples, spamassassin.init,
examples/gnus: Updated debian packaging for 2.60
2003-06-21 02:16 quinlan
* rules/: 70_broken_rules.cf, 70_cvs_rules_under_test.cf: move
T_RCVD_IN_FABEL to 70_broken.cf for now
2003-06-21 01:55 quinlan
* rules/20_dnsbl_tests.cf: rename RCVD_IN_OSIRU_DUL to
RCVD_IN_OSIRU_DIALUP tweak a few descriptions
2003-06-21 01:55 quinlan
* rules/: 30_text_sk.cf, 50_scores.cf: rename RCVD_IN_OSIRU_DUL to
RCVD_IN_OSIRU_DIALUP
2003-06-21 01:53 quinlan
* rules/: 70_broken_rules.cf, 70_cvs_rules_under_test.cf: move
T_RCVD_IN_DUN_DNSRBL to 70_broken_rules.cf
2003-06-21 01:22 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: promote
__SENDERBASE, but leave it commented out for now
2003-06-20 20:45 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_YAHOO_REDIRECTOR and
T_RCVD_IN_BONDEDSENDER (keep using current BondedSender rule for
now)
2003-06-20 18:54 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 2092: The header
report _REPORT_ has a newline at the end of each line, but the
_process_header function expects the string to not have a newline
at the end as it adds one when putting the header into the message.
Let's trim off trailing newlines from anything going through
_process_header.
2003-06-20 04:25 quinlan
* masses/mass-check: make --rewrite option accept an optional
argument if you don't like /tmp/out
2003-06-20 04:15 quinlan
* masses/mass-check: add --rewrite option change --these_rules to
--rules
2003-06-19 21:50 quinlan
* rules/20_dnsbl_tests.cf: just spaces to tabs
2003-06-19 19:46 quinlan
* lib/Mail/SpamAssassin/Dns.pm: just adding a single defined() test
2003-06-19 19:20 felicity
* README, rules/20_dnsbl_tests.cf, rules/50_scores.cf: bug 1982:
add support for RBL+, disabled by default.
2003-06-19 19:18 felicity
* lib/Mail/SpamAssassin/: Conf.pm, Dns.pm: bug 2086: Make it
possible for people (especially developers and people who will run
mass-check after 2.60 is released) to log real-time DNS responses.
This is particularly important for SenderBase since SenderBase
results change from day-to-day, but it will also help us improve
our assessments (hit rate, false positives, etc.) of the current
set of RBL rules.
2003-06-19 18:42 felicity
* lib/Mail/SpamAssassin/: Conf.pm, PerMsgStatus.pm: bug 2079:
_STARS_ was having problems when the message score was high and
causing folding to have issues. change the header to max at 50
stars, modify how folding works so that the first word in the
header won't be folded (like 50 * chars...)
2003-06-19 08:00 quinlan
* rules/20_html_tests.cf: rename HTML_FONT_INVISIBLE_N to
HTML_FONT_LOW_CONTRAST
2003-06-19 04:06 zelgadis
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Bug 1551: Move stripping
code to *before* the parser invocation, so that spammers can't
trick SA by using "<" and ">".
2003-06-19 03:59 quinlan
* rules/: 70_broken_rules.cf, 70_cvs_rules_under_test.cf: move
T_FAKE_HELO_DELPHI and T_HAS_NUM_AT_DELPHI to broken (too few hits,
we can resurrect later if necessary)
2003-06-19 03:53 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: promote
T_RCVD_IN_EASYNET_DYNABLOCK to RCVD_IN_DYNABLOCK (good results in
my mini-GA run, clear upstream policy, etc.)
2003-06-19 01:44 felicity
* spamd/libspamc.c: bug 2074: libspamc could potentially allow a
NULL character to overwrite another section of memory due to a
missized array not including enough space for the string NULL
terminator.
2003-06-18 21:19 quinlan
* rules/70_cvs_rules_under_test.cf: promote OPM subtests promote 3
of the 4 remaining URI tests broken rules: T_RCVD_IN_DUL_AUPADS,
T_RCVD_IN_INTERSIL, T_RCVD_IN_VOX_SCHPIDER_COM,
T_RCVD_IN_DNSRBL*, T_RCVD_IN_EASYNET_BLACKHOLES,
T_RCVD_IN_FIVETEN* remove HOST_EXCESSIVE_ESCAPES and
PATH_EXCESSIVE_ESCAPES - apparently, these do not work so well on
some corpuses
2003-06-18 21:18 quinlan
* rules/70_broken_rules.cf: broken rules: T_RCVD_IN_DUL_AUPADS,
T_RCVD_IN_INTERSIL, T_RCVD_IN_VOX_SCHPIDER_COM, T_RCVD_IN_DNSRBL*,
T_RCVD_IN_EASYNET_BLACKHOLES, T_RCVD_IN_FIVETEN*
2003-06-18 21:16 quinlan
* rules/20_uri_tests.cf: add some URI tests from 70_cvs
2003-06-18 21:15 quinlan
* rules/20_head_tests.cf: just adding some comments and tweaking
descriptions
2003-06-18 21:14 quinlan
* rules/20_dnsbl_tests.cf: promote OPM subtests
2003-06-18 03:44 duncf
* lib/Mail/SpamAssassin/Conf.pm: Bug 2080: add_header didnt allow
numbers
2003-06-18 00:42 jmason
* rules/20_ratware.cf: bug 2077: incorrect use of MESSAGEID by
FORGED_MUA_THEBAT rule
2003-06-17 21:59 jmason
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: promoted
FAKED_HOTMAIL_DAV and HAS_X_ORIG_HOST
2003-06-17 07:37 jmason
* rules/70_cvs_rules_under_test.cf: are we too late for a couple of
1.0 S/Os? ;)
2003-06-17 04:43 duncf
* rules/30_text_fr.cf: bug 2042: More/better french translations.
The last 3 revisions of this file have been primarily done by
Michel Bouissou
2003-06-17 03:30 quinlan
* lib/Mail/SpamAssassin/: Conf.pm, Dns.pm: bug 1785: improvements
to RBL timeout code
2003-06-16 22:01 duncf
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Bug 2066: cleanup of tags
stuff
2003-06-16 20:23 jmason
* rules/70_cvs_rules_under_test.cf: added 2 testing rules.
probably not suitable for 2.60 but there for 2.70cvs
2003-06-16 19:46 felicity
* sa-learn.raw, lib/Mail/SpamAssassin/ArchiveIterator.pm,
lib/Mail/SpamAssassin/CmdLearn.pm: bug 1646: from the bug messages:
Make sa-learn's basic invocation look like:
sa-learn [options] [file1..fileN]
--mbox becomes an option, stating that the data is in mbox format.
If not specified, the input source is presumed to be seperate
RFC822 messages. --file becomes ignored for backwards
compatability --single becomes ignored for backwards compatability
--directory becomes ignored for backwards compatability
If no files are specified, sa-learn awaits input on stdin. If one
or more files are specified, sa-learn reads them as individual
messages (or mboxes if --mbox is present). If one or more
directories are specified, sa-learn reads them as directories.
2003-06-16 06:38 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_FORGED_MUA_EUDORA_FROM
(I think we have better rules) remove T_FORGED_MUA_OUTLOOK_FROM (I
think we have better rules) promote T_FAKE_HELO_NETSCAPE_COM to
FAKE_HELO_NETSCAPE_COM promote T_BIZ_LINKS_3 to BIZ_TLD and revise
BZ_TLD to be similar replace USERPASS with T_USERPASS remove
T_URI_LINK_AT_SIGN revise *ESCAPE* rules with new version remove
T_BANNED_LINKS4_11 rename remaining *LINKS* rules to more meaninful
names, these are perhaps finally promotable
2003-06-16 06:36 quinlan
* rules/20_fake_helo_tests.cf: promote T_FAKE_HELO_NETSCAPE_COM to
FAKE_HELO_NETSCAPE_COM
2003-06-16 06:35 quinlan
* rules/20_uri_tests.cf: promote T_USERPASS to USERPASS promote
T_BIZ_LINKS_3 to BIZ_TLD and revise BZ_TLD as well promote
T_URI_OFFERS to URI_OFFERS purely cosmetic change to
HTTP_EXCESSIVE_ESCAPES (tested)
2003-06-15 21:06 felicity
* lib/Mail/SpamAssassin/Dns.pm: bug 2062: new version to try to
avoid stupid "scalar context" error (which it's not, but ...)
2003-06-15 17:34 felicity
* lib/Mail/SpamAssassin/Dns.pm: bug 2062: newer versions of DCC may
split their result over multiple headers, so change our code to
properly deal with that.
2003-06-15 00:43 quinlan
* masses/rule-qa/: corpus-hourly, corpus-nightly: switch to
rsync.spamassassin.org
2003-06-15 00:41 quinlan
* masses/rule-qa/corpus-hourly: nice 15 instead of nice 10 change
rsync server add NET.7day output file use most recent CVS revision
for NET instead of most common
2003-06-15 00:40 quinlan
* masses/rule-qa/corpus-nightly: higher persistence-level of retry
code for "cvs update" increase start delay from rand(300) to
rand(600) use -j 8 instead of -j 6 remove update stuff, only upload
when done... other way was too confusing change rsync server to
rsync.kluge.net
2003-06-15 00:38 quinlan
* masses/rule-qa/README.nightly: simplify cron job for new simpler
version of nightly script
2003-06-14 21:42 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/20_html_tests.cf,
rules/30_text_sk.cf: bug 2003: merge HTML_FONT_BIG_B into
HTML_FONT_BIG
2003-06-14 21:35 quinlan
* rules/: 20_html_tests.cf, 30_text_fr.cf, 30_text_sk.cf: bug 2002:
fixed description for HTML_FONT_BIG
2003-06-14 06:46 felicity
* rules/: 20_ratware.cf, 70_broken_rules.cf: bug 1760:
forged_mua_aol fps a bunch. so move to broken for now and work on
fixing it in 2.70.
2003-06-14 01:54 felicity
* rules/20_phrases.cf: this didn't commit for some reason. odd.
replaces US_DOLLARS with NA_DOLLARS
2003-06-14 01:48 felicity
* rules/70_cvs_rules_under_test.cf: replace US_DOLLARS with
NA_DOLLARS
2003-06-14 01:30 jmason
* lib/Mail/SpamAssassin/EvalTests.pm, rules/20_head_tests.cf,
rules/70_cvs_rules_under_test.cf: bug 1839: made FORGED_RCVD_TRAIL
into a meta subrule of CONFIRMED_FORGED; that gives us almost all
the spam hits, kills all the FPs, and it's now better than it was
with the old code. victory ;)
2003-06-14 01:08 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, NoMailAudit.pm: bug 1558: if
there is no message-id in the message we're trying to learn,
generate one based on the sha1() hash of the Date, the first add
(last listed) Received header, and the first N bytes of the body
(1k or 1/2 message length, whichever is smaller).
2003-06-14 00:54 quinlan
* rules/70_cvs_rules_under_test.cf: oops, I already promoted
T_BARGAIN_URL to BARGAIN_URL
2003-06-14 00:51 quinlan
* rules/70_cvs_rules_under_test.cf: promote T_FAKE_HELO_YAHOO to
FAKE_HELO_YAHOO promote T_HTML_LINK_PUSH_HERE to
HTML_LINK_PUSH_HERE some revisions of T_BIZ_LINKS split bizarro
T_STRANGE_CHAR_LINKS rule into the following rules: T_USERPASS,
T_URI_LINK_AT_SIGN, T_HTTP_EXCESSIVE_ESCAPE* remove some more of
the lesser LINKS rules, remaining ones might be okay spin
T_SUSPICIOUS_LINKS_09 into T_OFFERS_URL spin T_SUSPICIOUS_LINKS_04
into T_BARGAIN_URL
2003-06-14 00:46 quinlan
* rules/20_fake_helo_tests.cf: add newline
2003-06-14 00:29 quinlan
* rules/20_meta_tests.cf: move OBFUSCATING_COMMENT to
20_html_tests.cf
2003-06-14 00:28 quinlan
* rules/20_html_tests.cf: promote T_HTML_LINK_PUSH_HERE to
HTML_LINK_PUSH_HERE move OBFUSCATING_COMMENT to here
2003-06-14 00:26 quinlan
* rules/20_uri_tests.cf: add newline
2003-06-14 00:26 quinlan
* rules/20_fake_helo_tests.cf: remove FAKE_HELO_ATTBI promote
T_FAKE_HELO_YAHOO to FAKE_HELO_YAHOO
2003-06-14 00:01 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: when we reenabled learning
to journal, we didn't reenable msgid in journal. it also turns out
that a write attempt to a r/o db doesn't cause an error (the msgid
update to seen would fail, but it wasn't actually generating an
error ...)
2003-06-13 23:34 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Bayes.pm,
SpamAssassin/Conf.pm: bug 2005: added a bayes_auto_learn_journal
option to allow autolearning to go into the journal instead of
always going for the db ...
2003-06-13 23:09 felicity
* spamd/spamd.raw: it looks like File::Path isn't taint safe at
all, so the call to rmtree would fail constantly. since there's
only 2 dirs and 1 file that we use temporarily, just remove them
manually.
2003-06-13 22:53 felicity
* spamd/spamd.raw: bug 1704: turns out that File::Spec->tmpdir()
will set TMPDIR to undef if it's not already defined to anything.
very lame. I added a line to remove the undefined TMPDIR if that
situation occurs at spamd startup.
2003-06-13 22:24 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: added clearer scoping of
signal masks
2003-06-13 22:20 jmason
* rules/: 20_phrases.cf, 30_text_de.cf, 30_text_es.cf,
30_text_fr.cf, 30_text_it.cf, 30_text_sk.cf: attempt to avoid a
description hitting a rule
2003-06-13 22:18 felicity
* lib/Mail/SpamAssassin/Dns.pm: bug 2053: dccifd code didn't know
about new header format ...
2003-06-13 21:38 duncf
* rules/30_text_fr.cf: bug 2042: more french description updates
2003-06-13 21:17 jmason
* USAGE, rules/user_prefs.template: added note for MDaemon users;
removed some obsolete refs to 8-bit rules
2003-06-13 21:03 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 2057: updated to
discussion about where the ignores should go. added sigterm.
found bug in upgradedb code (very minor)
2003-06-13 16:27 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 2057: during critical
bayes work (expire, journal sync, upgrading), ignore sigint and
sighup.
2003-06-13 14:15 felicity
* lib/Mail/SpamAssassin/ArchiveIterator.pm: bug 1959: solves the
possible mass-check "sit-n-spin" problem
2003-06-13 05:28 duncf
* rules/30_text_fr.cf: bug 2042: updated french descriptions
2003-06-13 03:38 felicity
* Makefile.PL, binaries.mk.in, configure, configure.in,
spamd/spamc.c, spamd/spamc.pod: bug 1773: add an ENABLE_SSL option
to Makefile.PL to be able to turn on or off (default) SSL support.
instead of sslspamc, it's now "spamc -S" which is only valid if
ENABLE_SSL is set to on.
2003-06-12 23:38 felicity
* lib/Mail/SpamAssassin.pm: bug 1780: we shouldn't call
create_default_prefs if the dont_copy_prefs option is set since
create_default_prefs will consider that a failure and return an
error code.
2003-06-12 22:22 felicity
* lib/Mail/SpamAssassin/: UnixLocker.pm, Win32Locker.pm: bug 2054:
some unlinks could fail without spitting out a warning due to undef
values.
2003-06-12 07:47 duncf
* tools/desc_length.pl: added script to check rule length and other
misc stuff
2003-06-12 07:45 duncf
* rules/: 20_phrases.cf, 20_uri_tests.cf: a few more desc
shortenings
2003-06-12 07:41 duncf
* rules/: 20_body_tests.cf, 20_dnsbl_tests.cf, 20_html_tests.cf,
20_phrases.cf, 20_porn.cf, 20_uri_tests.cf, 23_bayes.cf,
70_cvs_rules_under_test.cf: Bug 2050: test descriptions too long.
2003-06-12 06:25 felicity
* lib/Mail/SpamAssassin/ArchiveIterator.pm: trying to deal better
with dead children ...
2003-06-12 06:20 duncf
* t/reportheader.t: fix test for change to report_template
2003-06-12 06:12 felicity
* lib/Mail/SpamAssassin/Dns.pm: bug 1704: should, hopefully, take
care of the undefined business.
2003-06-12 05:34 duncf
* lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm, rules/10_misc.cf: Got rid of
terse_report and everything terse_report related. Got rid of
_TERSE_ and moved _SUMMARYT_ to _REPORT_ and _SUMMARRY_ back to
_SUMMARYL_
Basically, users should use the _REPORT_ tag in a header to display
a terse summary like terse_report did.
Fixed folding of headers for headers with a \n in them.
Probably something else. :-)
2003-06-12 05:07 jmason
* spamassassin.raw: doco bug: unclear text
2003-06-12 04:53 jmason
* Makefile.PL, lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm, rules/10_misc.cf: bug 1659:
support address in report revisited. Now uses the 'hostname' idea
as discussed; report includes scanning hostname, and the default
contact address is 'the administrator of that machine' (sic).
2003-06-12 04:21 jmason
* spamd/README.spamd: bug 1851: clean up doco a little
2003-06-12 03:31 felicity
* lib/Mail/SpamAssassin/: EvalTests.pm, Received.pm: bug 2048:
replaced the appropriate IP_IN_RESERVED regexps with the more
standardized ip_in_reserved lookup from Received.pm...
2003-06-12 03:21 felicity
* spamd/spamd.raw: bug 2015: spamd shouldn't call system(), so
since it was a last ditch attempt to remove a directory, I replaced
it with a warn.
2003-06-12 03:00 felicity
* spamassassin.raw: bug 1714: added more descriptive text about the
AWL system and how it's not a replacement for the whitelist_from
config options
2003-06-12 02:51 felicity
* lib/Mail/SpamAssassin/: Conf.pm, Dns.pm, EvalTests.pm: bug 1362:
add dccifd support
2003-06-12 02:47 felicity
* lib/Mail/SpamAssassin/EvalTests.pm: bug 2040:
check_access_database should use the trusted_received stuff.
2003-06-12 01:53 felicity
* spamd/spamd.raw: bug 2043: quoted directly from the bug:
When spamd is run with --ssl, if it receives a connection that does
not successfully negotiate SSL then _all_ the spamd processes quit.
The block starting on line 481 of spamd.raw wants to figure out why
the accept() call did not succeed. In the non-SSL world, !$client
is true only where an OS error happened. However, in the SSL world,
!$client can also happen where the SSL negotiation failed, which
does not imply an OS error and the consequent setting of errno.
The attached patch detects this; if SSL is set and errno is zero,
then we log the fact that there must've been an SSL error, and we
don't allow ourselves to be die()d.
2003-06-12 01:49 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 2046: old versions of
perl (5.005_02 at least) seem to require all three arguments to
syswrite which we weren't doing.
2003-06-11 01:43 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: bug 1839: T_FORGED_RCVD_TRAIL
re-added based on old code so we can compare side by side
2003-06-11 01:29 quinlan
* lib/Mail/SpamAssassin/ArchiveIterator.pm: indentation changes
2003-06-11 01:23 jmason
* lib/Mail/SpamAssassin/EvalTests.pm, rules/20_head_tests.cf,
rules/70_cvs_rules_under_test.cf: removed RCVD_FAKE_HELO_DOTCOM2,
replaced by FAKE_HELO rules; promoted successful no-rDNS and
wrong-IP-net helo rules
2003-06-11 01:10 jmason
* lib/Mail/SpamAssassin/: EvalTests.pm, Received.pm:
IP_IN_RESERVED_MATCH was not matching correctly, fixed
2003-06-10 23:26 quinlan
* rules/: 20_ratware.cf, 70_cvs_rules_under_test.cf: bug 1717:
Incorrect FORGED_MUA_OUTLOOK
2003-06-10 23:23 felicity
* lib/Mail/SpamAssassin/ArchiveIterator.pm, masses/mass-check: fix
mass-check from the updates yesterday and such: - mass-check would
miss the results from all children after it decided to restart
them. (ie: for n children and a setting of x for restart,
mass-check would lose n-1 results every x messages.) oops. -
disable caching of data in the sockets - fix the mass-check
progress date printing
2003-06-10 22:11 quinlan
* rules/70_cvs_rules_under_test.cf: revise HTML_LINK rules
2003-06-10 18:15 felicity
* contrib/mbox-to-check: a script used by run-masses to spit out
the mbox files in the right format for mass-check
2003-06-10 18:11 felicity
* contrib/: run-corpora, run-masses: Script examples for running
mass-checks and doing full corpus runs.
2003-06-10 17:28 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: added more headers checked
for whitelisting, thanks to Nancy McGough for the tip
2003-06-10 06:36 jmason
* rules/70_cvs_rules_under_test.cf: bug 1717: exempt Outlook 2003
from FORGED_MUA_OUTLOOK
2003-06-10 06:29 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Received.pm,
rules/70_cvs_rules_under_test.cf: bug 1990: RCVD_FAKE_HELO_DOTCOM_2
work. added T_NO_RDNS_DOTCOM_HELO to catch hosts HELOing as a big
dot-com, with no reverse DNS
2003-06-10 06:10 jmason
* rules/: 30_text_de.cf, 30_text_fr.cf, 30_text_it.cf: copied
translations for default whitelists
2003-06-10 06:06 jmason
* lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Received.pm, rules/20_head_tests.cf,
rules/50_scores.cf, rules/60_whitelist.cf: bug 846: default
whitelists not so easily forged anymore. also added documentation
on the trusted_networks inference option
2003-06-10 05:43 jmason
* masses/bayes-testing/bayes-static-thresholds: added
static-thresholds script used in recent tests
2003-06-10 03:24 felicity
* lib/Mail/SpamAssassin/ArchiveIterator.pm, masses/mass-check:
trying to make mass-check less likely to blow up. children are
forked pre-message loading, so they should take less memory.
support having children die (via --restart #) after # of messages
have been processed to help deal with any memory leaks and the
like.
2003-06-10 03:14 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: let's try this again. another
fix for FORGED_RCVD_TRAIL et al. Also added T_FORGED_RCVD_IP_HELO
which I noticed during testing
2003-06-10 03:05 felicity
* build/update_stable: modified gpg support
2003-06-10 02:41 jmason
* build/update_stable, lib/Mail/SpamAssassin/Bayes.pm,
lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Message.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm,
lib/Mail/SpamAssassin/Received.pm: added prelim support for
GPG-signing to build/update_stable
2003-06-09 22:43 felicity
* build/README: added GPG signature to the release procedure
2003-06-09 22:20 quinlan
* rules/70_cvs_rules_under_test.cf: T_HTML_PUSH_HERE* variants
2003-06-09 21:02 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_RCVD_IN_DUL_BOFH
(worst of dial-up and dynamic DNSBLs)
2003-06-09 20:57 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: promote
T_DNS_FROM_RFCI_DSN to DNS_FROM_RFCI_DSN
2003-06-09 20:40 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: promote
T_RCVD_IN_OSIRU_OPEN_PROXY to RCVD_IN_OSIRU_OPEN_PROXY just delete
T_RCVD_IN_OSIRU_OPT_IN, it's worthless clean up the OSIRU rules
some
2003-06-09 20:29 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: promote
NJABL sub-tests
2003-06-09 20:26 felicity
* lib/Mail/SpamAssassin/Dns.pm: should run test_log _BEFORE_
got_hit (which calls handle_hit) ...
2003-06-09 20:22 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: promote
T_RCVD_IN_SORBS* to RCVD_IN_SORBS*
2003-06-09 20:21 quinlan
* rules/20_ratware.cf: fix FORGED_MUA_EUDORA
2003-06-09 20:07 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 2035: only handle_hit
should put stuff in the report summary because things like test_log
run before handle_hit, and therefore cause odd-looking results.
2003-06-09 19:49 quinlan
* rules/70_cvs_rules_under_test.cf: promote T_FORGED_MUA_EUDORA to
FORGED_MUA_EUDORA delete T_PUSHHERELINK since T_HTML_PUSH_HERE
looks more viable
2003-06-09 19:49 quinlan
* rules/20_ratware.cf: promote T_FORGED_MUA_EUDORA to
FORGED_MUA_EUDORA
2003-06-09 19:40 quinlan
* rules/: 20_ratware.cf, 70_cvs_rules_under_test.cf: replace
FORGED_MUA_AOL_FROM with T_AOL_FAKE_MAIL2 (better hit rate)
2003-06-09 19:27 quinlan
* rules/70_cvs_rules_under_test.cf: delete some lame MSGID rules
2003-06-09 19:25 quinlan
* rules/: 20_ratware.cf, 70_cvs_rules_under_test.cf: promote
RATWARE_RCVD_BONUS_SPC delete T_INVALID_DATE_HMS and
T_INVALID_DATE_2 (exemptions barely help)
2003-06-09 19:21 felicity
* spamd/utils.h: added new full_read_ssl function to utils.h as
prototype
2003-06-09 18:52 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/20_html_tests.cf,
rules/70_cvs_rules_under_test.cf: promote T_HTML_COMMENT_SHOUT1 to
HTML_COMMENT_SHOUTING, delete others
2003-06-09 08:21 jmason
* rules/70_cvs_rules_under_test.cf: working on DAV rule
2003-06-09 06:16 quinlan
* rules/70_cvs_rules_under_test.cf: AOL rule revisions
2003-06-09 05:58 quinlan
* rules/: 20_ratware.cf, 70_cvs_rules_under_test.cf: promote
RATWARE_RCVD_LC_ESMTP and FORGED_QUALCOMM_TAGS clean-up formatting
and tabs a bit
2003-06-09 05:32 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
look for comments/declarations nested inside of certain tags
2003-06-09 03:25 quinlan
* rules/20_phrases.cf: replace SATISFACTION with T_SATISFACTION
version
2003-06-09 03:24 quinlan
* rules/20_body_tests.cf: add comment
2003-06-09 03:24 quinlan
* rules/70_cvs_rules_under_test.cf: some promotion and deletion
2003-06-09 03:00 quinlan
* rules/: 20_body_tests.cf, 20_html_tests.cf, 20_meta_tests.cf,
20_ratware.cf, 70_cvs_rules_under_test.cf: bug 2034: replace
MIME_HTML_ONLY with __MIME_HTML_ONLY
2003-06-08 09:11 jmason
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: promoted
T_X_PRI_MISMATCH_HI
2003-06-08 09:05 jmason
* rules/70_cvs_rules_under_test.cf: fixed bug in T_LC_ESMTP_RCVD
2003-06-08 08:54 jmason
* rules/70_cvs_rules_under_test.cf: oh dear. Hotmail DAV is a 1.0
spamsign for me. let's see how it does for everyone else...
2003-06-08 01:21 jmason
* lib/Mail/SpamAssassin/Received.pm: some work on forged-HELO stuff
2003-06-08 01:10 jmason
* rules/70_cvs_rules_under_test.cf: remove T_HAS_X_ORIG_EMAIL, try
out a variant
2003-06-08 00:52 jmason
* USAGE, rules/local.cf: updated some doco, and noted
trusted_networks param
2003-06-08 00:45 jmason
* lib/Mail/SpamAssassin/Received.pm: fixed bugs
2003-06-08 00:37 jmason
* lib/Mail/SpamAssassin/Received.pm: update comment
2003-06-08 00:35 jmason
* lib/Mail/SpamAssassin/NetSet.pm: oops, bug if no trusted_networks
were set
2003-06-08 00:34 jmason
* lib/Mail/SpamAssassin/: NetSet.pm, Received.pm: added
Received-header trust inference code
2003-06-07 23:58 jmason
* spamd/spamd.raw: bug 2013: bring SQL checks into line with
documentation
2003-06-07 23:52 jmason
* INSTALL: bug 1996: note about installing DB_File
2003-06-07 08:04 duncf
* rules/70_cvs_rules_under_test.cf: bug 1687 HOME_EMPLOYMENT FPs
2003-06-07 07:34 duncf
* rules/70_cvs_rules_under_test.cf: bug 1580: satisfaction too
broad
2003-06-07 07:18 duncf
* spamd/spamd.raw: Bug 1511: Patch from Ed Henderson to fix
vpopmail stuff.
2003-06-06 22:01 felicity
* lib/Mail/SpamAssassin.pm: bug 2030: if a message doesn't have a
message-id header, then 'spamassassin -r' would learn it as ham
instead of spam. adds a 'scalar()' to the call so that a lack of
message-id will be ok.
2003-06-06 21:20 quinlan
* lib/Mail/SpamAssassin/Dns.pm: bug 2026: DNS log inherited by DCC
check, I think this should fix it
2003-06-06 04:28 jmason
* lib/Mail/SpamAssassin/Dns.pm: enclose URLs from TXT replies in
angle brackets, to ensure correct MUA display
2003-06-05 23:00 duncf
* lib/Mail/SpamAssassin/: Conf.pm, PerMsgStatus.pm: Bug 2019: strip
newlines if headers arent folded; fix documentation errors
2003-06-05 22:26 quinlan
* rules/70_cvs_rules_under_test.cf: remove more LINKS rules (low
hit rate on 2 days of corpus results, I hope everyone's corpus is
up-to-date :-)
2003-06-05 22:22 quinlan
* rules/70_cvs_rules_under_test.cf: remove lowest tier of LINKS
rules (no renumbering)
2003-06-05 22:15 quinlan
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: replace
header 8-bit rules with illegal character rules reorganize
non-English ADV rules too
2003-06-05 22:13 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: replace
check_subject_for_lotsa_8bit_chars and
check_all_headers_for_8bit_chars with check_illegal_chars bug 2014:
Hotmail X-Originating-IP not included in DNSBL checks remove some
obsolete code from DNSBL code
2003-06-04 03:16 duncf
* lib/Mail/SpamAssassin/PerMsgStatus.pm: and fix unintialized value
warnings
2003-06-04 03:05 duncf
* lib/Mail/SpamAssassin/Conf.pm, lib/Mail/SpamAssassin/Dns.pm,
lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm,
lib/Mail/SpamAssassin/Received.pm, rules/10_misc.cf: bug 1473 made
the add_header option which should be used for _all_ configuration
of headers.
_TAGS_ can be used with add_header and the report templates to add
interesting info into the reports
To add a tag, change the documentation in Conf.pm, and either put
the info into $self->{tag_data}->{TAG} or add a anonymous sub ref
to the hash %tags in PerMsgStatus.pm
There is also a bit of backwards compatibility stuff involved here,
although some of it should be permanent (like default headers,
maybe).
I did profile the code, and there doesn't seem to be much of a
change at all.
2003-06-03 22:11 quinlan
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: fix bug
274: replacement rule for FRIEND_AT_PUBLIC fix bug 1734:
replacement rule for FROM_OFFERS
2003-06-03 19:41 felicity
* rules/70_cvs_rules_under_test.cf: no, I mean fix the damn rule.
2003-06-03 19:40 felicity
* rules/70_cvs_rules_under_test.cf: fixed broken T_NA_DOLLARS rule.
2003-06-03 04:51 quinlan
* rules/70_cvs_rules_under_test.cf: renumber the "links" rules
2003-06-03 04:33 quinlan
* rules/70_cvs_rules_under_test.cf: second pass on 8bit rules
2003-06-03 03:58 felicity
* rules/20_head_tests.cf: bug 1980, forgot to fix this dependency.
2003-06-03 03:54 felicity
* rules/: 20_body_tests.cf, 20_compensate.cf, 20_dnsbl_tests.cf,
20_head_tests.cf, 20_html_tests.cf, 20_meta_tests.cf,
20_phrases.cf, 20_ratware.cf, 20_uri_tests.cf,
70_cvs_rules_under_test.cf: bug 1980: make all rules fit convention
I also added T_NA_DOLLARS to try to replace US_DOLLARS. US_DOLLARS
and MILLION_USD overlap a lot, so try to seperate them.
2003-06-03 03:31 jmason
* masses/: CORPUS_SUBMIT, CORPUS_SUBMIT_NIGHTLY: updated
instructions to use rsync.kluge.net
2003-06-02 07:25 jmason
* rules/70_cvs_rules_under_test.cf: couple of test rules
2003-06-02 05:21 felicity
* rules/: 20_head_tests.cf, 20_phrases.cf, 20_uri_tests.cf: fixing
up descriptions and such
2003-06-02 04:06 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: found that the score as
shown in the report wasn't lining up properly if abs(score) >= 10.
modified the layout to fix it.
2003-06-01 21:17 felicity
* lib/Mail/SpamAssassin.pm: bug 1907: if SA is installed into a
non-standard location, and the modules are called by a third party
set of programs that don't pass in PREFIX, the modules will
probably not be able to figure out where the rules and such are.
added a make-time PREFIX default setting to SA.pm such that if
PREFIX isn't passed in, the modules can at least make an educated
guess about where to find the rules and such.
2003-06-01 20:25 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: bug 1586: instead of just
sending in an array with 1 element per header line, support sending
in an array with 1 element per header (and let it include folded
strings, etc.)
2003-06-01 20:04 felicity
* contrib/: spamcheck.py, spamd.py: updated version of spamcheck.py
2003-06-01 18:39 felicity
* lib/Mail/SpamAssassin/ArchiveIterator.pm: bug 1959: try to clean
up mass-check's parallelization function to stop "going out to
lunch".
2003-06-01 16:19 felicity
* rules/70_cvs_rules_under_test.cf: typo in new *links rule... :(
2003-06-01 16:07 felicity
* rules/70_cvs_rules_under_test.cf: split banned and suspicious
links into seperate rules for each phrase
2003-06-01 15:37 felicity
* lib/Mail/SpamAssassin/Conf.pm, rules/10_misc.cf: as discussed on
sa-talk, some documentation was out-of-date.
2003-05-31 08:44 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1833: rbl_finish()
called twice on each message
2003-05-31 07:22 duncf
* rules/70_cvs_rules_under_test.cf: AOL rules...
2003-05-31 03:35 duncf
* masses/extract-message-from-mbox: Added support for simply piping
mass-check results in (-m option). Added support for outputting
headers only (-H option)
2003-05-31 03:07 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: new 8-bit rules!!!
2003-05-30 21:14 quinlan
* rules/: 20_uri_tests.cf, 70_cvs_rules_under_test.cf: replace
PORN_4 with T_PORN_4
2003-05-30 10:23 quinlan
* rules/70_cvs_rules_under_test.cf: bug 274: add T_FRIEND_AT_PUBLIC
2003-05-30 10:09 quinlan
* rules/: 20_body_tests.cf, 20_head_tests.cf, 50_scores.cf: shorten
some rule names
2003-05-30 10:04 quinlan
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: promote
T_MSGID_ADDED_BY_HOTMAIL to MSGID_FROM_MTA_HOTMAIL
2003-05-30 09:43 quinlan
* rules/70_cvs_rules_under_test.cf: T_PORN_4 for bug 1835
2003-05-30 05:51 quinlan
* rules/70_cvs_rules_under_test.cf: add baseline tests for
developing new 8bit and illegal character rules
2003-05-30 01:02 jmason
* rules/70_cvs_rules_under_test.cf: oops, typo
2003-05-30 00:19 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: cleared up more
FORGED_RCVD_TRAIL fps
2003-05-30 00:14 quinlan
* rules/70_cvs_rules_under_test.cf: spin T_FROM_OFFERS*
2003-05-30 00:07 felicity
* sa-learn.raw: modified the expiration docs a bit
2003-05-29 22:04 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1839: massive drop in
accuracy for FORGED_RCVD_TRAIL, fixed I think.
2003-05-29 21:39 jmason
* lib/Mail/SpamAssassin/Bayes.pm,
lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Received.pm, lib/Mail/SpamAssassin/Util.pm,
rules/70_cvs_rules_under_test.cf: received header parser could not
deal with IPv6 or IPv4-mapped-in-IPv6 IP addresses, fixed (related
to bug 1765)
2003-05-29 08:45 quinlan
* rules/20_meta_tests.cf: simplify a slow rule
2003-05-29 08:18 jmason
* lib/Mail/SpamAssassin/Received.pm: trusted-to-untrusted network
transition was totally wrong
2003-05-29 02:23 felicity
* sa-learn.raw, lib/Mail/SpamAssassin.pm,
lib/Mail/SpamAssassin/BayesStore.pm, lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm,
masses/lint-rules-from-freqs, rules/10_misc.cf, spamd/README.spamd:
replace auto_learn* with bayes_auto_learn*, but support the older
version for now. add bayes_auto_expire to specify whether or not
autoexpire ought to run because max_db_size needs to be set for
manual expiry.
2003-05-28 20:54 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1839: major drop in
accuracy for FORGED_RCVD_TRAIL, revert to last working version. I
didn't have much luck getting the new version working with good
accuracy.
2003-05-28 20:08 quinlan
* rules/70_cvs_rules_under_test.cf: remove some rules
2003-05-28 09:58 quinlan
* rules/70_cvs_rules_under_test.cf: FROM_OFFERS additions
2003-05-28 09:05 quinlan
* rules/70_cvs_rules_under_test.cf: T_RCVD_IN_WSFF, we hardly knew
ye
2003-05-28 08:24 felicity
* rules/20_body_tests.cf: added description for the BLANK_LINE
rules.
2003-05-28 08:20 felicity
* MANIFEST: need to add these to the MANIFEST
2003-05-28 07:52 felicity
* build/find_meta_missing_dependencies, rules/20_head_tests.cf,
rules/20_meta_tests.cf, rules/20_ratware.cf,
rules/70_cvs_rules_under_test.cf: solving missing meta dependencies
added script to find missing meta dependencies, bug 1979
2003-05-28 07:44 jmason
* rules/70_cvs_rules_under_test.cf: oops, typo
2003-05-28 07:23 jmason
* rules/70_cvs_rules_under_test.cf: bug 1918: try out an exception
for MSOE-produced 8bit headers
2003-05-28 07:20 felicity
* spamassassin.spec: updated for 2.60
2003-05-28 07:19 felicity
* spamd/: libspamc.c, utils.c: bug 1751: fixed issues with SSL
discrepencies
2003-05-28 07:15 jmason
* lib/Mail/SpamAssassin/NetSet.pm, t/cidrs.t: ok, unit test code
moved out
2003-05-28 06:41 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1517: fix base64
decoding problem
2003-05-28 04:26 duncf
* lib/Mail/SpamAssassin/CmdLearn.pm: Allow sa-learn --folders to
actually work, and support "targets"
2003-05-28 03:49 quinlan
* rules/23_bayes.cf: shorten Bayes descriptions
2003-05-28 03:32 felicity
* masses/: Makefile, lint-rules-from-freqs, runGA: runGA: script I
use to actually run the GA. it automates the procedure, although
you do need to modify the script per scoreset right now.
Makefile: added badfiles target, set athlon to be the default (my
system)
lint-rules-from-freqs: bug 1978, check that all rule names are <=
22 chars, rule descriptions are <= 50 chars, and all rules have a
description.
2003-05-28 02:56 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_URI_CASE_CHANGE and
T_HTTP_ESCAPED_DIRECTORY
2003-05-28 01:35 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm, rules/20_head_tests.cf,
rules/30_text_sk.cf, rules/50_scores.cf: name correction for
MSGID_OUTLOOK_INVALID
2003-05-28 00:14 jmason
* rules/50_scores.cf, t/html_obfu.t, t/data/spam/011,
rules/50_scores.cf: oops, NATURAL_VIAGRA is disabled so make test
would fail under make disttest
2003-05-27 23:42 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Locales.pm, rules/20_head_tests.cf: bug 1441:
cleared up some Chinese-charset FPs
2003-05-27 23:40 quinlan
* rules/70_cvs_rules_under_test.cf: few quick test rules
2003-05-27 23:34 quinlan
* rules/: 20_body_tests.cf, 20_meta_tests.cf: move MIME_HTML* to
20_body_tests.cf to be with rest of MIME tests
2003-05-27 23:11 jmason
* build/cvs2cl-branch, t/cidrs.t: adding some missed files
2003-05-27 23:06 jmason
* MANIFEST, lib/Mail/SpamAssassin/PerMsgStatus.pm, t/html_obfu.t,
t/data/spam/011: bug 1551: HTML comment obfuscation; handle a few
more illegal obfuscatory styles that Outlook supports
2003-05-27 21:12 jmason
* MANIFEST: updated MANIFEST
2003-05-27 21:07 jmason
* MANIFEST, lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/Received.pm, lib/Mail/SpamAssassin/Util.pm,
spamd/spamd.raw, lib/Mail/SpamAssassin/NetSet.pm: bug 1425:
trusted_networks should accept short CIDR specifications. Also
refactored CIDR into a new class, in order to add unit tests, fixed
code to _work_ after adding and running tests ;), and fixed spamd
to share this code for -A option.
2003-05-27 20:10 jmason
* Makefile.PL, build/preprocessor: bug 1977: make install depends
on umask; build/preprocessor now uses PERM_RW var from Makefile to
set modes for files installed
2003-05-27 19:46 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: tweak MIME_MULTIPART_SHORT to
also catch empty text parts (no change on my corpus, but adds hit
for example from bug 1529)
2003-05-27 18:57 quinlan
* rules/20_body_tests.cf, rules/70_cvs_rules_under_test.cf,
lib/Mail/SpamAssassin/EvalTests.pm: bug 1434: bug promote
T_MULTIPART_TRICK_D_00_01 to MIME_MULTIPART_SHORT
2003-05-27 18:35 quinlan
* rules/20_dnsbl_tests.cf: just descriptions and formatting
2003-05-27 08:24 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_RCVD_IN_OSSRC (not a
repeat, but doesn't add much beyond other spam-source RBLs,
especially RCVD_IN_OSIRU_SPAM_SRC)
2003-05-27 07:15 quinlan
* rules/70_cvs_rules_under_test.cf: remove
T_RCVD_IN_RELAYS_KUNDENSERVER (same deal, very high overlap with
other open relay RBLs)
2003-05-27 07:12 quinlan
* rules/70_cvs_rules_under_test.cf: removing
T_RCVD_IN_EASYNET_PROXIES (high overlap with other open proxy RBLs,
but it's getting harder to remove RBLs)
2003-05-27 05:21 quinlan
* rules/: 20_body_tests.cf, 20_head_tests.cf, 30_text_sk.cf,
50_scores.cf, 70_broken_rules.cf: shorten some rule names and
convert "MSG_ID" to "MSGID"
2003-05-27 05:05 jmason
* rules/70_cvs_rules_under_test.cf: couple of other HTML abuse
rules
2003-05-27 05:00 jmason
* rules/70_cvs_rules_under_test.cf: fixed T_OBFUSCATING_BAD_TAG to
avoid some FPs
2003-05-27 04:56 jmason
* rules/70_cvs_rules_under_test.cf: added T_OBFUSCATING_BAD_TAG to
catch my pet-hate spam tags
2003-05-27 04:46 quinlan
* rules/20_head_tests.cf, lib/Mail/SpamAssassin/EvalTests.pm: bug
1919: nuking BAD_HELO_WARNING (very poor results)
2003-05-27 02:52 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm, rules/10_misc.cf:
rearrange report to meet new conventions
2003-05-27 02:38 felicity
* lib/Mail/SpamAssassin/Conf.pm: added documentation about rule /
description name and length conventions
2003-05-27 02:06 duncf
* lib/Mail/SpamAssassin/Conf.pm: Implement ""'s for other
templates.
2003-05-27 01:51 quinlan
* rules/70_cvs_rules_under_test.cf: T_RCVD_IN_ALL has served its
purpose
2003-05-27 01:05 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_RCVD_IN_GIPPER (very
high overlap with SORBS and DSBL)
2003-05-27 01:01 felicity
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: promoted
new versions of FROM_NO_USER and TO_NO_USER hit rate doesn't change
in the nightly run, but does stop an FP as reported in bug 1758
2003-05-27 00:53 felicity
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: promoted
new TO_MALFORMED rule, the spam hit rate is a little lower, but the
fp rate is MUCH lower.
2003-05-27 00:47 felicity
* rules/: 20_phrases.cf, 50_scores.cf, 70_cvs_rules_under_test.cf:
promoting new WE_HONOR_ALL, removing duplicate REMOVE_RESPECT
2003-05-26 22:56 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_RCVD_IN_DNSBLINFO (95%
of hits also in SBL, only hit 0.6501% spam)
2003-05-26 22:18 felicity
* lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm, rules/10_misc.cf: some more
cleanup of the verbose report. added the ability for "report" to
specify the string in quotes if it needs front whitespace.
2003-05-26 21:47 quinlan
* rules/: 20_uri_tests.cf, 70_cvs_rules_under_test.cf: promote
T_IP_LINK_PLUS to IP_LINK_PLUS
2003-05-26 21:45 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_HTML_WITHOUT_HTML*
2003-05-26 21:45 quinlan
* rules/20_html_tests.cf: no real change
2003-05-26 21:36 quinlan
* rules/20_head_tests.cf: promote T_FROM_MIXED_ALPHANUM_08 to
FROM_HAS_MIXED_NUMS3
2003-05-26 21:36 quinlan
* rules/70_cvs_rules_under_test.cf: promote
T_FROM_MIXED_ALPHANUM_08 to FROM_HAS_MIXED_NUMS3 fix T_FROM_OFFERS*
(oops, mistake going from "grep -P" to SA)
2003-05-26 21:26 felicity
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: promote new
GAPPY_SUBJECT rule
2003-05-26 21:24 quinlan
* rules/: 20_dnsbl_tests.cf, 30_text_de.cf, 30_text_es.cf,
30_text_fr.cf, 30_text_it.cf, 30_text_pl.cf, 30_text_sk.cf,
50_scores.cf: add "MAPS" to the rule names of MAPS RBL tests
2003-05-26 21:07 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: left justify logs
normally, right justify if they are too long
2003-05-26 21:01 quinlan
* rules/: 20_meta_tests.cf, 70_cvs_rules_under_test.cf: promote
T_MULTIPART_HTML_ONLY_05 to MIME_HTML_ONLY_MULTI second pass on
T_MULTIPART_TRICK* (figure out which part of range to use)
2003-05-26 20:39 quinlan
* lib/Mail/SpamAssassin/HTML.pm: 80 columns in a few places
2003-05-26 09:16 quinlan
* lib/Mail/SpamAssassin/Dns.pm: tweak to remove duplicate test logs
(due to multi-RBLs that return multiple A records)
2003-05-26 08:51 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: removing
X-Spam-Check-Host, see bug 1473 for alternate solution
2003-05-26 07:15 felicity
* INSTALL: bug 1816, clarify that Sys::Syslog is probably already
installed, and listed a method to check.
2003-05-26 07:08 felicity
* lib/Mail/SpamAssassin/HTML.pm: bug 1405: could
set off the HTML_EVENT_UNSAFE rule, even though it's an empty
value. so add a check to see if there's any value before flagging
the rule.
2003-05-26 06:41 felicity
* rules/60_whitelist.cf: added default whitelist for ntbugtraq.
bug 1574
2003-05-26 06:26 felicity
* rules/60_whitelist.cf: additional whitelist for efax.com, bug
1348.
2003-05-26 06:21 felicity
* lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/EvalTests.pm, rules/20_head_tests.cf,
rules/50_scores.cf: bug 883: create a blacklist_to to round out the
black/whitelist groups.
2003-05-26 06:21 duncf
* lib/Mail/SpamAssassin/: Conf.pm, PerMsgStatus.pm: Prettier
prettified report format... right justify excess stuff
2003-05-26 06:13 quinlan
* lib/Mail/SpamAssassin/Dns.pm: fix DNS test logging to work
correctly for repeated hits
2003-05-26 05:42 duncf
* lib/Mail/SpamAssassin/PerMsgStatus.pm, rules/10_misc.cf: Bug
1493: prettify report format (make a table)
2003-05-26 04:25 felicity
* lib/Mail/SpamAssassin/Reporter.pm: if there are no methods
available for reporting, say so with a warning.
2003-05-26 03:33 felicity
* spamassassin.raw, lib/Mail/SpamAssassin/Reporter.pm: bug 1968:
some more verbose messages when reporting fails
2003-05-26 01:39 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: bug 1434: look for
multipart/alternative that isn't
2003-05-26 01:25 quinlan
* rules/20_html_tests.cf: tweak HTML_WEB_BUGS description remove
HTML_IMAGE_WEB_BUGS (no longer needed)
2003-05-26 00:56 felicity
* t/reportheader.t: bug 1965: the text for the default report
changed, so our test for it failed. doh. updated with the new
string.
2003-05-26 00:47 felicity
* rules/10_misc.cf: bug 1797: change unsafe_report to be more clear
about "did not contain plain text".
2003-05-26 00:32 felicity
* spamd/spamd.raw: turns out exists doesn't work there for some
reason. :(
2003-05-26 00:03 felicity
* rules/70_cvs_rules_under_test.cf: bug 1634: missed the Palm
version of Eudora ... I think we should make new rules for
the Windows, Mac, and Palm versions of Eudora. We're getting too
many exceptions to this rule.
2003-05-25 23:58 felicity
* rules/70_cvs_rules_under_test.cf: bug 1634: FPs for Eudora for
Mac OS X. Also found that a removed nice rule was a dependency for
the FORGED rule, so added it back in as such.
2003-05-25 23:42 felicity
* spamd/spamd.raw: bug 1774: if spamd is run with -u, we should
change $ENV{'HOME'} appropriately so that Razor, Pyzor, etc, will
get the correct home directory (not the home directory from the
parent process...)
2003-05-25 21:31 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1603: added a
X-Spam-Check-Host to the message headers to specify which host
marked up the mail.
2003-05-25 21:13 quinlan
* rules/70_cvs_rules_under_test.cf, lib/Mail/SpamAssassin/HTML.pm:
upgrade HTML_WEB_BUGS to new version with much better HTML_MESSAGE
accuracy: 9.341 10.5488 1.7643 0.857 0.63 0.01
HTML_WEB_BUGS:new 18.923 18.2034 23.4373 0.437 0.11 0.20
HTML_WEB_BUGS:old remove HTML_IMAGE_WEB_BUGS (now adds more ham
than spam)
2003-05-25 20:45 felicity
* rules/70_cvs_rules_under_test.cf: possible replacement for
GAPPY_SUBJECT. 7 works best for me, but we'll see how the nightly
results do.
2003-05-25 17:54 felicity
* rules/: 20_head_tests.cf, 20_phrases.cf, 20_ratware.cf,
20_uri_tests.cf, 70_cvs_rules_under_test.cf: promoted some rules,
removed some bad rules ...
2003-05-25 17:38 felicity
* rules/: 20_body_tests.cf, 20_meta_tests.cf,
70_cvs_rules_under_test.cf: promoted some rules ...
2003-05-25 15:14 felicity
* rules/20_ratware.cf: bug 1822: "Outlook Mac" is sometimes
"Outlook for Mac".
2003-05-25 04:05 felicity
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1372: use the pristine
message instead of a possibly munged version for razor, dcc, and
pyzor.
2003-05-25 03:57 felicity
* rules/70_cvs_rules_under_test.cf: possible replacements for
TO_MALFORMED T_TO_MALFORMED2 clears up most of my FPs (which were
at least 3x the valid hits)
2003-05-25 03:44 felicity
* lib/Mail/SpamAssassin/Conf.pm: bug 1848: update the docs to
specify that the body rule type includes the subject header.
2003-05-25 03:20 felicity
* rules/70_cvs_rules_under_test.cf: bug 1874: REMOVE_REQUEST and
WE_HONOR_ALL are very similar. added a new rule which should at
least cover the same messages as both original rules.
2003-05-25 01:43 felicity
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1475: juno.com sends as
untd.com now, so add them into the forged re.
2003-05-25 01:33 felicity
* rules/70_cvs_rules_under_test.cf: bug 1758: FROM_NO_USER was
matching when there was a '( some @ foo )' in the From header.
2003-05-25 00:24 quinlan
* rules/70_broken_rules.cf: just making sure
2003-05-25 00:20 quinlan
* rules/70_cvs_rules_under_test.cf: delete T_RCVD_IN_MONKEYS_UPL
(NEVER PUT THIS RULE BACK INTO SA)
2003-05-24 23:56 jmason
* Makefile.PL, rules/10_misc.cf: bug 1659: Makefile.PL asks for a
support address/URL
2003-05-24 23:55 jmason
* lib/Mail/SpamAssassin/Conf.pm: fixed POD typo in Conf.pm
2003-05-24 22:02 felicity
* lib/Mail/SpamAssassin/Conf.pm: bug 1762: make it more clear that
the whitelist_to settings will try to match the address anywhere as
a recipient in the header...
2003-05-24 21:32 quinlan
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Conf.pm: bug 1623: add
ham == non-spam
2003-05-24 21:31 quinlan
* sa-learn.raw: URL tweak
2003-05-24 20:50 felicity
* rules/70_cvs_rules_under_test.cf: bug 1914: new test version of
see_for_yourself
2003-05-24 07:16 quinlan
* lib/Mail/SpamAssassin/Dns.pm: bug 1733: add initial SenderBase
support add back logging of TXT record DNSBL hits (goes into the
report) and also add logging support for A record DNSBL hits
2003-05-24 07:14 quinlan
* rules/70_cvs_rules_under_test.cf: bug 1733: add initial
SenderBase support also add some dialup blacklists to test
2003-05-24 02:45 quinlan
* rules/70_cvs_rules_under_test.cf: more SORBS tests (are fun)
2003-05-24 01:12 quinlan
* rules/: 20_dnsbl_tests.cf, 30_text_sk.cf, 50_scores.cf,
70_broken_rules.cf, 70_cvs_rules_under_test.cf: revise NJABL rules
2003-05-24 00:10 quinlan
* rules/70_cvs_rules_under_test.cf: remove unneeded rule
2003-05-24 00:00 quinlan
* rules/70_cvs_rules_under_test.cf: bug 1734: replacement rules for
FROM_OFFERS
2003-05-23 23:45 jmason
* USAGE, t/SATest.pm: USAGE: note that Razor requires registration,
and set default patterns for tests to make it easier to write them
;)
2003-05-23 23:35 jmason
* rules/10_misc.cf: updated default report text as we finally
agreed way back on 10 Mar
2003-05-23 23:10 quinlan
* lib/Mail/SpamAssassin/Conf.pm: bug 1953: lower rbl_timeout to 10
seconds
2003-05-23 21:19 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_RCVD_IN_DEADBEEF add
some T_FROM_MIXED_ALPHANUM* variations remove all T_URI_CASE* rules
except for one (probably not good enough)
2003-05-23 19:48 felicity
* spamd/spamd.raw: bug 1952: have spamd write the pid file whether
or not we're daemonized
2003-05-23 19:19 felicity
* spamd/spamd.raw: bug 1952: spamd -r required -d as well. added
documentation.
2003-05-21 19:50 felicity
* rules/: 20_dnsbl_tests.cf, 50_scores.cf, 70_broken_rules.cf:
looks like orbs is down permanently. removing from dnsbl list
2003-05-21 13:44 felicity
* rules/: 20_html_tests.cf, 20_phrases.cf,
70_cvs_rules_under_test.cf: cleaning up the test rules a bit
promoted some, removed some others ...
2003-05-21 10:20 quinlan
* rules/70_cvs_rules_under_test.cf: name fix: RCVD_IN_FIVETEN to
T_RCVD_IN_FIVETEN add URI case change test rules add
T_HTTP_ESCAPED_DIRECTORY
2003-05-20 07:10 jmason
* lib/Mail/SpamAssassin.pm: 2.56 devel cycle started
2003-05-20 07:06 jmason
* Changes: ready for release
2003-05-20 00:59 felicity
* lib/Mail/SpamAssassin/: Conf.pm, PerMsgStatus.pm: bug 1628: added
ability to specify headers that should be copied from the original
message when doing a report_safe encapsulation.
2003-05-19 23:49 quinlan
* lib/Mail/SpamAssassin/HTML.pm,
lib/Mail/SpamAssassin/EvalTests.pm, rules/20_body_tests.cf,
rules/20_html_tests.cf, rules/30_text_fr.cf, rules/30_text_sk.cf,
rules/50_scores.cf, rules/70_broken_rules.cf,
rules/70_cvs_rules_under_test.cf: various rule renaming, revising
(test rules only), and promotion
2003-05-19 23:34 felicity
* Changes, lib/Mail/SpamAssassin.pm: updated changes and devel flag
for 2.55 release
2003-05-19 23:25 felicity
* lib/Mail/SpamAssassin/: Conf.pm, Conf.pm: bug 1929: looks like we
never specified that "score 0" disables a rule from running.
2003-05-19 21:34 felicity
* rules/: 20_head_tests.cf, 30_text_fr.cf, 30_text_it.cf,
30_text_sk.cf, 50_scores.cf, 70_broken_rules.cf: removed
MSGID_CHARS_SPAM
2003-05-19 21:26 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: removed incorrect
statement about safety zone around required_hits (no longer used)
2003-05-19 20:48 felicity
* rules/70_cvs_rules_under_test.cf: bug 1899: new suggestion for
reverse_aging rule.
2003-05-19 20:33 felicity
* lib/Mail/SpamAssassin/: Dns.pm, Received.pm: bug 1858: some
header additions weren't using X-Spam as the prefix, so a
"spamassassin -d" wouldn't remove them. modified those headers to
be X-Spam prefixed.
2003-05-19 19:46 quinlan
* rules/70_cvs_rules_under_test.cf: add T_FROM_MIXED_ALPHANUM_* for
testing
2003-05-19 09:17 quinlan
* rules/70_cvs_rules_under_test.cf: change wirehub.net to
easynet.nl
2003-05-19 08:53 quinlan
* rules/70_cvs_rules_under_test.cf: rename __RCVD_IN_FIVETEN to
RCVD_IN_FIVETEN (let the GA set scores instead of worrying about
it)
2003-05-19 04:41 quinlan
* lib/Mail/SpamAssassin/Conf.pm: bug 1617: add a charset header to
mime reports remove old documentation about "SPAM:" prefix from
"report" documentation
2003-05-19 04:39 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1617: add a charset
header to mime reports
2003-05-19 04:27 felicity
* rules/: 50_scores.cf, 70_cvs_rules_under_test.cf: added some new
suggested rules, fixed razor2_cf_range scores
2003-05-19 01:36 quinlan
* masses/rule-qa/corpus-hourly: use most common CVS revision
instead of current CVS revision for NET class
2003-05-19 01:20 felicity
* spamassassin.raw, lib/Mail/SpamAssassin/Conf.pm,
spamassassin.raw, lib/Mail/SpamAssassin/Conf.pm: bug 1917: docs
still talk about 'spamassassin -c', which was deprecated for
'spamassassin -C' ...
2003-05-19 01:18 felicity
* spamd/: spamd.raw, spamd.raw: bug 1905: in perl 5.005, mkdir
needs a umask parameter
2003-05-19 01:12 felicity
* Makefile.PL: removed more obsolete razor stuff bug 1893: if
razor2 isn't installed, a warning could pop up from the undefined
reference
2003-05-19 01:06 quinlan
* rules/70_cvs_rules_under_test.cf: remove low-ranking rules
2003-05-19 01:03 felicity
* Makefile.PL: bug 1893: a few warnings pop up if razor isn't
installed for testing. simple cleanup.
2003-05-19 01:01 quinlan
* rules/70_cvs_rules_under_test.cf: I hate HTML email
2003-05-19 00:59 quinlan
* rules/20_html_tests.cf: add description for HTML_COMMENT_RATIO
(and shorten prerequisite test name)
2003-05-18 22:27 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
restore HTML area code (whoops) revisions to new web bug tests
2003-05-18 15:48 felicity
* spamassassin.spec: updated spec file for 2.55
2003-05-18 09:40 quinlan
* rules/70_cvs_rules_under_test.cf: add some variations to
T_MIME_HTML_NO_*_TAG*
2003-05-18 09:16 quinlan
* rules/70_cvs_rules_under_test.cf: add some test phrases
2003-05-18 04:39 quinlan
* rules/: 70_broken_rules.cf, 70_cvs_rules_under_test.cf: move
T_RCVD_IN_RELAYS_ORDB_ORG to 70_broken_rules.cf for good... low hit
rate, no sub rules, and low score in almost every mini-GA run
2003-05-18 02:36 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
add some more test variations for web bugs
2003-05-18 02:25 duncf
* debian/rules: Argh. More Debian packaging fixed.
2003-05-18 02:21 duncf
* debian/rules: Argh. VENDORPREFIX annoys me
2003-05-18 01:31 quinlan
* rules/50_scores.cf: remove old firsthop scores
2003-05-18 00:11 duncf
* debian/: rules, rules: More woody build failures :-( This should
work now, I hope. Someone test please.
2003-05-17 23:47 duncf
* debian/: changelog.in, spamassassin.preinst, changelog.in,
spamassassin.preinst: last fix... honestly! handles case where
debconf not installed
2003-05-17 23:19 duncf
* debian/control: Forgot to change debian/control for woody compat.
Fixed
2003-05-17 23:18 duncf
* debian/: control, examples/muttrc, po/POTFILES.in, po/fr.po,
po/pt_BR.po, po/templates.pot: Umm... fixed better packaging ;-)
2003-05-17 23:14 duncf
* debian/: control, logcheck, rules, spamassassin.config,
spamassassin.dirs, spamassassin.postinst, spamassassin.preinst,
spamassassin.templates: Updated Debian packaging for 2.54, better
woody upgrade
2003-05-17 22:59 duncf
* debian/po/fr.po: file fr.po was initially added on branch b2_5_0.
2003-05-17 22:59 duncf
* debian/: control, logcheck, rules, spamassassin.config,
spamassassin.dirs, spamassassin.links, spamassassin.postinst,
spamassassin.postrm, spamassassin.preinst, spamassassin.templates,
examples/muttrc, po/fr.po, po/pt_BR.po, po/templates.pot: Debian
packaging for 2,54. Includes a good upgrade from 2.20-1woody
2003-05-17 22:59 duncf
* debian/examples/muttrc: file muttrc was initially added on branch
b2_5_0.
2003-05-17 21:16 quinlan
* masses/rule-qa/corpus-hourly: allow
CURRENT_CORPORA_WEEKLY_VERSION for network checks produce "all"
results for every class NET class is anything less than 10 days old
with same CVS revision fix some bugs
2003-05-17 21:14 quinlan
* masses/rule-qa/corpus-nightly: add a random sleep to avoid
simultaneous checkout
2003-05-17 18:03 felicity
* rules/: 20_body_tests.cf, 20_html_tests.cf,
70_cvs_rules_under_test.cf: rule cleanup and some promotion
2003-05-17 08:10 quinlan
* rules/: 20_dnsbl_tests.cf, 30_text_de.cf, 30_text_es.cf,
30_text_fr.cf, 30_text_it.cf, 30_text_pl.cf, 30_text_sk.cf,
50_scores.cf, 70_broken_rules.cf, 70_cvs_rules_under_test.cf: first
pass on cleaning up DNS test names (get everything synchronized)
2003-05-17 07:21 quinlan
* lib/Mail/SpamAssassin/Conf.pm: add TODO comment
2003-05-17 06:51 quinlan
* lib/Mail/SpamAssassin/Conf.pm: remove backward compatibility hack
for old RBL tests (won't run now anyway)
2003-05-17 04:40 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
quick second pass on T_HTML_IMAGE_WEB_BUGS*
2003-05-17 03:55 quinlan
* lib/Mail/SpamAssassin/HTML.pm: fix warning
2003-05-17 03:53 quinlan
* rules/: 70_broken_rules.cf, 70_cvs_rules_under_test.cf: add back
T_RCVD_IN_RELAYS_ORDB_ORG for further testing (made a mistake) add
T_RCVD_IN_ALL for testing of Received: header parsing and DNS code
(should get 100% on spam)
2003-05-17 02:28 quinlan
* rules/: 20_dnsbl_tests.cf, 70_broken_rules.cf: move
RCVD_IN_RELAYS_ORDB_ORG to 70_broken_rules.cf: low hit rate, poor
GA results (both current and past), no sub-tests for RBL either
2003-05-17 01:37 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: remove unused eval tests:
check_carriage_returns(), check_signature(), and _check_signature()
2003-05-16 23:54 quinlan
* rules/70_cvs_rules_under_test.cf: pick one
T_MSGID_ADDED_BY_HOTMAIL rule bug 1757: remove T_HELLO_ADDRESS
(horrible results) removed T_ASIAN_LINKS, T_RATWARE_BAD_GECKO,
T_PHRASE_COMBO_LINKS, T_WSDOMAIN, T_MULTIPLEIMAGES removed
T_FAKE_HELO rules with poor results using results from daf, jm,
lan, quinlan, rODbegbie, theo: 0.031 0.0587 0.0154 0.792
0.48 0.01 T_FAKE_HELO_NETSCAPE 0.019 0.0232 0.0161
0.590 0.20 0.01 T_FAKE_HELO_YAHOO_CO_UK
2003-05-16 23:51 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: remove T_RATWARE_MIZSPEELING
(zero hits on all corpuses)
2003-05-15 21:33 felicity
* t/spamd_protocol_10.t: new fix for 1909
2003-05-15 21:17 felicity
* t/spamd_protocol_10.t: bug 1909: oops. while fixing another bug,
I broke the seek call. should be fixed now. :)
2003-05-15 21:04 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Conf.pm: added
bayes_learn_during_report option to disable automatically learning
mails that are being reported. depending on how one wants to train
the bayes system, this may be a good option to disable.
2003-05-15 19:01 quinlan
* rules/70_cvs_rules_under_test.cf, lib/Mail/SpamAssassin/HTML.pm:
add some T_HTML_IMAGE_WEB_BUGS* tests (based on Darren Coleman's
test)
2003-05-15 01:16 jmason
* INSTALL: updated with correct Razor info
2003-05-14 19:14 quinlan
* rules/70_cvs_rules_under_test.cf: changes to YAHOO_REDIR*
2003-05-14 18:57 quinlan
* rules/20_body_tests.cf, rules/70_cvs_rules_under_test.cf,
lib/Mail/SpamAssassin/EvalTests.pm: promote
T_MIME_BASE64_WITHOUT_NAME to MIME_BASE64_NO_NAME
2003-05-14 18:27 felicity
* spamd/redhat-rc-script.sh: bug 1915: didn't list condrestart as
an option under usage
2003-05-14 08:27 quinlan
* masses/rule-qa/README.nightly: add a hint :-)
2003-05-14 08:21 quinlan
* masses/rule-qa/corpus-hourly: add revision checking and fix a few
minor bugs
2003-05-14 08:20 quinlan
* masses/rule-qa/corpus-nightly: fix a typo
2003-05-14 08:08 quinlan
* rules/: 20_body_tests.cf, 30_text_de.cf, 30_text_es.cf,
30_text_fr.cf, 30_text_it.cf, 30_text_sk.cf, 50_scores.cf: rename
BASE64_ENC_TEXT to MIME_BASE64_TEXT
2003-05-14 08:04 quinlan
* rules/20_dnsbl_tests.cf: add comment
2003-05-14 08:04 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: promote T_MIME_BASE64_BLANKS to
MIME_BASE64_BLANKS
2003-05-13 21:10 quinlan
* rules/70_cvs_rules_under_test.cf: reduce versions of
T_MSGID_THREESIXSIX*
2003-05-13 20:57 quinlan
* rules/70_cvs_rules_under_test.cf: promote
T_OBFUSCATING_COMMENT_05B to replace OBFUSCATING_COMMENT and
OBFUSCATING_COMMENT2 remove T_GAPPY_SUBJECT* because none were any
better
2003-05-13 20:54 quinlan
* rules/: 20_body_tests.cf, 20_meta_tests.cf: replace
OBFUSCATING_COMMENT with improved rule remove OBFUSCATING_COMMENT2
2003-05-13 03:53 felicity
* sa-learn.raw, lib/Mail/SpamAssassin/BayesStore.pm: more
documentation changes for bayes sync/expire. optimization in
journal_sync_due.
2003-05-13 02:11 quinlan
* rules/: 20_head_tests.cf, 20_phrases.cf, 20_porn.cf,
20_ratware.cf: 100% overlap removals: ID_X_LIST_UNSUBSCRIBE,
FOR_INSTANT_ACCESS, RATWARE_GROUPMAIL add overlap-related comments
for RATWARE_STORM and X_LIST_UNSUBSCRIBE
2003-05-13 01:43 quinlan
* masses/overlap: ignore extra fields at end of line improve
performance by only counting each pair once (the "-a" option should
only affect output and not add extra work to the inner loop)
2003-05-13 01:37 jmason
* rules/20_head_tests.cf: removed MSGID_SPAMSIGN_ZEROES: totally
overlapping with MSGID_OE_SPAM_4ZERO
2003-05-13 00:31 quinlan
* masses/mass-check: add CVS revision to log comment
2003-05-13 00:11 quinlan
* rules/70_cvs_rules_under_test.cf: add $Revision tag
2003-05-12 21:55 felicity
* sa-learn.raw, lib/Mail/SpamAssassin/BayesStore.pm,
t/spamd_protocol_10.t: some doc reworking fixed some errors spewed
out by spamd_procotol_10
2003-05-12 21:35 jmason
* lib/Mail/SpamAssassin/Bayes.pm: latest bayes tweaks: best results
checked in. Robinson S constant reduced back down to 0.1; previous
good results were actually artifact of bad testing methodology;
this reduces FPs/FNs. 'chewing' of URIs and addresses in headers
and body, where hostnames are broken into domain components;
increases accuracy in general.
2003-05-12 20:23 jmason
* lib/Mail/SpamAssassin.pm: 2.55 devel cycle started, not that
we'll need it of course ;)
2003-05-12 20:20 jmason
* build/2.54_change_summary: 2.54 RELEASED
2003-05-12 20:15 jmason
* build/tag_weekly_mc: weekly mass-check fixed
2003-05-12 19:58 quinlan
* masses/rule-qa/README.nightly: few additions
2003-05-12 19:52 quinlan
* masses/rule-qa/: README.nightly, corpus-hourly, corpus-nightly,
corpus-tagtime: nightly mass-check scripts and README
2003-05-12 08:30 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_RCVD_IN_OSSOCKS (99%
of hits were in T_X_OSIRU_OPEN_PROXY)
2003-05-12 08:27 quinlan
* rules/20_dnsbl_tests.cf: delete RCVD_IN_UNCONFIRMED_DSBL (97.8%
of the spam hits are in RCVD_IN_DSBL and it has 5.4% FPs vs. 0.4%
FPs for RCVD_IN_DSBL)
2003-05-12 08:22 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_RCVD_IN_OSRELAY since
it duplicates X_OSIRU_OPEN_RELAY
2003-05-12 07:25 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: improve speed/accuracy of
T_FROM_IN_RFCI_DSN
2003-05-12 06:55 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: minor tweak to avoid
fetching twice
2003-05-12 05:22 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_FROM_IN_PIGS and
T_FROM_IN_DEADBEEF (poor results) tweak T_WS_PHRASE
2003-05-12 03:39 felicity
* rules/70_cvs_rules_under_test.cf: needed to add tflags
2003-05-12 02:21 quinlan
* rules/20_head_tests.cf: bug 1582: consistent false-positives for
SquirrelMail (die die die!)
2003-05-12 01:27 quinlan
* rules/70_cvs_rules_under_test.cf: more T_OBFUSCATING_COMMENT* -
tired of this yet?
2003-05-12 01:05 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm, rules/20_dnsbl_tests.cf,
rules/50_scores.cf: fix bug 1871: hil nil ril real
2003-05-12 00:12 quinlan
* rules/70_cvs_rules_under_test.cf: add some T_MSGID_THREESIXSIX
variants (missing about 27% of hits before)
2003-05-11 23:28 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_X_HEADER_PITCHID
2003-05-11 22:27 felicity
* rules/70_cvs_rules_under_test.cf: try to simplify/get better
results with razor2 range tests
2003-05-11 21:50 quinlan
* rules/70_cvs_rules_under_test.cf: various T_YAHOO_REDIR* rule
ideas
2003-05-11 21:32 quinlan
* rules/: 70_cvs_rules_under_test.cf, 20_ratware.cf: promote
T_FORGED_THEBAT_HTML to FORGED_THEBAT_HTML
2003-05-11 21:06 quinlan
* rules/70_cvs_rules_under_test.cf: new T_OBFUSCATING_COMMENT*
tests new T_GAPPY_SUBJECT* tests
2003-05-11 20:40 msquadrat
* rules/70_cvs_rules_under_test.cf: T_GAPPY_PORN is broken -- some
alternatives
2003-05-11 18:02 felicity
* lib/Mail/SpamAssassin/BayesStore.pm, masses/mass-check:
mass-check needs to make sure that bayes will expire more often
then once every 12 hours since it's going to churn through stuff
fairly quickly. added a flag to M::SA (passed in by mass-check) to
disable that check.
2003-05-11 17:37 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: more logic changes to the
expire code
2003-05-11 07:03 felicity
* lib/Mail/SpamAssassin/Bayes.pm: stop dividing by 0 johnson!
2003-05-11 06:59 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: few more bug fixes
2003-05-11 06:35 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: added a bunch of debugging
statements to the expire area refixed 1888 added in some other
checks, useful for force expire where an expire won't occur.
2003-05-11 05:22 jmason
* build/2.54_change_summary: file 2.54_change_summary was initially
added on branch b2_5_0.
2003-05-11 05:22 jmason
* build/2.54_change_summary: change summary for 2.54 release
2003-05-11 05:21 felicity
* sa-learn.raw, lib/Mail/SpamAssassin/Bayes.pm,
lib/Mail/SpamAssassin/CmdLearn.pm, tools/check_bayes_db: replaced
check_bayes_db since it only understand db format 0. it now just
calls sa-learn with the --dump option, passing anything else along
with it.
added the --dbpath and --regexp options from check_bayes_db to
sa-learn/CmdLearn. They only have an effect via --dump.
2003-05-11 05:19 jmason
* build/README, lib/Mail/SpamAssassin/Bayes.pm, build/README:
updated build README
2003-05-11 05:17 jmason
* Changes, build/README, build/cvs2cl-branch: added cvs2cl-branch,
fixed Changes file
2003-05-11 05:17 jmason
* build/cvs2cl-branch: file cvs2cl-branch was initially added on
branch b2_5_0.
2003-05-11 04:38 felicity
* Changes: Updated for 2.54rc1
2003-05-11 02:05 duncf
* lib/Mail/SpamAssassin.pm: Ready to release?
2003-05-11 01:34 felicity
* rules/: 20_head_tests.cf, 30_text_de.cf, 30_text_fr.cf,
30_text_it.cf, 50_scores.cf, 70_broken_rules.cf: bug 1891:
RCVD_IN_VISI is no longer available
2003-05-11 01:32 felicity
* rules/70_cvs_rules_under_test.cf: added rules for testing
2003-05-11 01:29 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1888: the expiry system
tries to figure out if the new expire is largely different from the
last expire. if they are, there was probably a surge at some point
and it would make a bad estimation for this expire, so it'll
recompute atime delta. but 0 is a valid last expire count (new
db), so we got a div by zero error. :(
changed expiry_due to check for >100000 tokens. if there were
exactly 100k tokens, and we did the expire, it would get confused
because to leave 100k tokens in the db (safety), we would want to
remove 0 tokens.
if we couldn't find a good exponent to expire from in the first
pass, we'd return without removing the expiry_running flag. it
would eventually time out, but if we know we're not expiring
anymore, we should remove it.
had a stupid bug where expiry would expire to 100000 if it wanted
to keep more than 100k tokens. teaches you to use the correct <>
symbol. ;) (it now says keep 100k if the "to expire to" point is <
100k, which is the correct logic...)
added a "need to do first pass" check if the last expire count is
0. whenever that happens, last atime delta should also be 0, and
that's already checked, but ...
2003-05-10 23:40 jmason
* rules/70_broken_rules.cf: just removed VISI test, it's dead as a
dodo these days
2003-05-10 04:51 felicity
* rules/70_cvs_rules_under_test.cf: removed unused header count
tests that didn't do so well
2003-05-10 04:48 felicity
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: promoting
rule
2003-05-10 04:10 felicity
* rules/70_cvs_rules_under_test.cf: added a bunch of suggested
rules
2003-05-10 03:50 felicity
* lib/Mail/SpamAssassin/Conf.pm: forgot to set the default value.
haha.
2003-05-10 01:46 felicity
* lib/Mail/SpamAssassin/CmdLearn.pm: bug 1887: sa-learn didn't
accept parameters to -D
2003-05-10 01:28 felicity
* spamd/: spamd.raw, spamd.raw: bug 1818: spamd fails to create
~/.spamassassin when running setuid with sql
2003-05-09 23:36 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Bayes.pm,
SpamAssassin/BayesStore.pm, SpamAssassin/CmdLearn.pm,
SpamAssassin/Conf.pm: bug 1719: enable learning to journal. jm did
this, but it was waiting for the 1881 fix.
bug 1881: we will now learn to the journal instead of requiring the
db to be r/w. however, msgids (the seen db) weren't updated by
journal learning. added code to support learning and forgetting
(but don't do that, the code is actually disabled to forget to
journal).
bug 1821: we decided on a new bayes format again. this time it's
token atime is based on message received time. this brought up a
whole bunch of issues, but mostly required new pack/unpack,
upgrade_db, expire, and a bunch of other changes. while I was
milling about I fixed a few other random bugs. I don't quite
remember what they are right now, so just assume everything's
peachy.
2003-05-09 09:45 quinlan
* rules/: 20_ratware.cf, 70_cvs_rules_under_test.cf: rename a few
forgery rules to be more consistently named promote
T_HTML_FAKED_OUTLOOK to FORGED_OUTLOOK_TAGS fix
T_FORGED_THEBAT_HTML
2003-05-09 09:15 quinlan
* rules/: 20_head_tests.cf, 20_phrases.cf, 70_broken_rules.cf,
70_cvs_rules_under_test.cf: promote DATE_SPAMWARE_Y2K - this rule
rocks promote T_ORDER_NOW to ORDER_NOW - barely perceptible
improvement move USER_AGENT rules to broken - may be gold in there
for forgeries
2003-05-09 04:57 jmason
* spamd/: libspamc.c, libspamc.c: bug 1879: libspamc was closing
fds it didn't open
2003-05-09 01:29 quinlan
* rules/70_cvs_rules_under_test.cf: more T_GAPPY_SUBJECT_* rules
2003-05-09 01:25 quinlan
* rules/70_cvs_rules_under_test.cf: T_OBFUSCATING_COMMENT* - will
it ever end?
2003-05-09 00:54 jmason
* spamd/README.spamd: bug 1799: doco updated to note protocol
change compared to 2.4x, 2.5x
2003-05-09 00:48 quinlan
* lib/Mail/SpamAssassin/HTML.pm: remove
$self->{html}{comment_no_alphanum}
2003-05-09 00:28 jmason
* spamd/: README.spamd, spamc.pod: bug 1799: doco updated to note
protocol change compared to 2.4x, 2.6x
2003-05-08 22:47 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1869: came up with a
kluge for the "tokens have 16-bit atime" problem. if the expire
code detects that scancount is > 65535, reset all the counters to
an atime of 0. it means one expire run doesn't actually expire
anything, but it should keep the useful tokens around instead of
the semi-random expire that was occuring in earlier versions.
2003-05-08 22:42 jmason
* spamd/libspamc.c: bug 1762: sscanf should not use %f due to
locale issues
2003-05-08 21:49 felicity
* rules/70_cvs_rules_under_test.cf: trim out some of the not so
good rules
2003-05-08 20:03 jmason
* spamd/: libspamc.c, spamc.c: bug 1764: found a bug in
locale-independent float parsing code, fixed. added unit tests for
same
2003-05-08 07:41 jmason
* spamd/libspamc.c: bug 1799: REPORT and REPORT_IFSPAM was
inconsistent between 2.4x and 2.5x; now matching 2.4x again
2003-05-08 04:41 jmason
* spamd/: spamd.raw, spamd.raw: bug 1802: REPORT_IFSPAM is
unhandled if message is ham
2003-05-08 00:50 felicity
* rules/: STATISTICS-set1.txt, STATISTICS-set2.txt,
STATISTICS-set3.txt, STATISTICS.txt: bug 1589: generated
STATISTICS* files for the new GA run
2003-05-07 23:23 felicity
* lib/Mail/SpamAssassin/: PerMsgStatus.pm, PerMsgStatus.pm: bug
1868: X-Spam-Checker-Version may or may not have parens depending
on if the message was considered spam or not. Just add in the
parens...
2003-05-07 22:16 jmason
* spamd/libspamc.c: bug 1764: %f in sscanf breaks version check due
to locale LC_NUMERIC setting
2003-05-07 21:06 jmason
* spamd/spamd.raw, lib/Mail/SpamAssassin/Bayes.pm, spamd/spamd.raw:
bug 1725: spamd taint mode problem with ARGV fixed
2003-05-07 13:38 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1782: a potential but in
stdio could cause journal data to be corrupted by multiple writes
to the same area of the journal. switch to using syswrite()
instead of just print (stdio) to help address this issue.
2003-05-07 07:26 jmason
* lib/Mail/SpamAssassin/Bayes.pm: bug 1860: Bayes was not using
hapaxes
2003-05-07 00:51 duncf
* rules/20_body_tests.cf: Fix to check ROT13 encoding of e-mail
address that don't end in .com .net .org or .ca. (Seeing how we're
internation and all... obviously)
2003-05-06 14:14 felicity
* rules/: 20_meta_tests.cf, 50_scores.cf: bug 1589: spammers have
been targeting our nice rules to get themselves negative overall
scores. ran the GA again to lower the nice rule scores. this
should help deal with the forgeries. also added a "TOO_MANY_MUA"
rule that will catch when multiple USER_AGENT rules hit.
2003-05-06 13:33 felicity
* masses/score-ranges-from-freqs: don't use a system call when we
can just run mkdir allow learn rules to get larger scores some
cleanup
2003-05-06 08:32 quinlan
* rules/: 20_phrases.cf, 70_cvs_rules_under_test.cf: promote
revised PENIS_ENLARGE, PENIS_ENLARGE2, AS_SEEN_ON, SEE_FOR_YOURSELF
promote new HTML_IMAGE_WEB_BUGS delete some lesser rules revisions
of T_GAPPY_SUBJECT (try to find something better, it was a good
idea)
2003-05-06 08:27 quinlan
* rules/20_html_tests.cf: finish image_web_bugs rule
2003-05-06 08:23 quinlan
* lib/Mail/SpamAssassin/HTML.pm: finish image_web_bugs rule
2003-05-06 05:36 quinlan
* rules/70_cvs_rules_under_test.cf: nearing closure on
T_OBFUSCATING_COMMENT*
2003-05-05 22:36 jmason
* lib/Mail/SpamAssassin/Bayes.pm: additional fix for bayes
2003-05-05 21:58 jmason
* sa-learn.raw, lib/Mail/SpamAssassin/Bayes.pm: oh... my... ghod.
we've been ignoring hapaxes since the start of December :(
2003-05-05 21:45 jmason
* rules/70_cvs_rules_under_test.cf: some new ratware testing rules
2003-05-04 01:35 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: finish off
bug 1853: fix dialup tests to be -notfirsthop
2003-05-03 09:44 quinlan
* lib/Mail/SpamAssassin/Dns.pm, lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: bug 1853: allow primary RBL tests
to match only on certain responses
2003-05-02 22:51 quinlan
* lib/Mail/SpamAssassin/Conf.pm, rules/20_dnsbl_tests.cf,
rules/70_cvs_rules_under_test.cf: remove rbl_res_evals
2003-05-02 22:41 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
replace HTML_FONT_COLOR_NAME with T_HTML_FONT_COLOR_NAME
2003-05-02 12:06 quinlan
* rules/: 20_dnsbl_tests.cf, 70_broken_rules.cf: move
RCVD_IN_MULTIHOP_DSBL to broken rules
2003-05-02 10:26 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: OPM
supports TXT queries, but A queries are needed for sub-tests
2003-05-02 07:31 jmason
* spamd/spamd.raw: match 2.4x REPORT format
2003-05-02 07:25 jmason
* spamd/spamd.raw: bug 1799: REPORT should use local report
template, not default
2003-05-02 07:19 jmason
* sa-learn.raw, spamassassin.raw: bug 1737: Bayes doco improved
2003-05-02 02:25 jmason
* MANIFEST, lib/Mail/SpamAssassin/Bayes.pm, spamd/libspamc.c,
spamd/libspamc.h, spamd/spamc.c, spamd/spamc.pod, spamd/spamd.raw,
spamd/utils.c, spamd/utils.h, t/SATest.pm, t/spamd_unix.t: UNIX
domain socket support added to spamd/spamc, thanks to steve at
unixwiz dot net
2003-05-01 21:53 jmason
* build/tag_nightly_mc: oops, logic bug in tagging script
2003-04-30 22:50 quinlan
* rules/70_cvs_rules_under_test.cf: more clean up of
T_OBFUSCATING_COMMENT*
2003-04-30 21:33 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: add alternate
hostname_to_domain2() function add check_rbl_from() for checking
sender domains in domain-based RBLs
2003-04-30 21:31 quinlan
* rules/70_cvs_rules_under_test.cf: bug 1017: domain-based
blacklists (initially a few using sender addresses)
2003-04-30 07:30 quinlan
* rules/70_cvs_rules_under_test.cf: remove worst of
T_OBFUSCATING_COMMENT* variations
2003-04-30 07:16 quinlan
* rules/70_cvs_rules_under_test.cf: T_OBFUSCATING_COMMENT* - run
them all and let mass-check sort theem out
2003-04-29 21:21 quinlan
* lib/Mail/SpamAssassin/Dns.pm: switch part of dnscache to be an
array instead of a hash update launch time for each DNS query (so
last one gets full timeout) some clean up of dnscache
2003-04-29 00:59 jmason
* rules/20_uri_tests.cf: oops, avoid var-length lookbehind
2003-04-29 00:56 jmason
* spamd/spamd.raw: try spamd change to avoid RH9 warning
2003-04-29 00:53 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: seems the misspelling ratware
can use 3-component hostnames as well
2003-04-29 00:36 jmason
* lib/Mail/SpamAssassin/Conf.pm: enabled Chi^2-combining (aka
Fisher combining in bogofilterland)
2003-04-29 00:35 jmason
* lib/Mail/SpamAssassin/Bayes.pm,
lib/Mail/SpamAssassin/EvalTests.pm, rules/20_uri_tests.cf,
tools/split_corpora: PORN_4: avoided FP on basex.com
2003-04-28 23:44 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: fix bug 1827: warnings in
EvalTests::check_for_rdns_helo_mismatch
2003-04-28 07:55 quinlan
* rules/70_cvs_rules_under_test.cf: remove duplicate
__XM_BAD_DATE_MUAS consolidate T_AS_SEEN_ON
2003-04-28 07:47 duncf
* tools/sa-stats.pl: Initial revisions of sa-stats.pl adapted from
suggestions in bug 1794
2003-04-28 07:14 quinlan
* lib/Mail/SpamAssassin/Dns.pm: reduce number of times "require
Razor2::Client::Agent;" is evaluated remove unnecessary
prerequisite checks in razor2_lookup() fix some occasional warnings
in DNS, Razor, and DCC code
2003-04-28 07:06 quinlan
* rules/: 70_broken_rules.cf, 70_cvs_rules_under_test.cf: move
T_RCVD_IN_MONKEYS_FORMMAIL, T_RCVD_IN_DUINV, T_RCVD_IN_PDL to
broken rules
2003-04-28 02:38 jmason
* lib/Mail/SpamAssassin/: Bayes.pm, PerMsgStatus.pm: still getting
leftover locks after sa-learn --ham/spam. now fixed
2003-04-28 02:26 jmason
* tools/split_corpora: split_corpora was broken by ArchiveIterator
changes
2003-04-27 04:14 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: didn't need extra $db_ver
variable, just use $self->{db_version}
2003-04-27 01:39 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: thought of a bug where if
we're upgrading, and don't set $self->{db_version}, some functions
(like tok_unpack) will do the wrong thing. so let's change it
while we're doing the upgrade and things will work as expected. :)
2003-04-26 09:41 quinlan
* lib/Mail/SpamAssassin/Dns.pm, lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm, rules/20_dnsbl_tests.cf,
rules/70_cvs_rules_under_test.cf: bug 1824: rewrite of DNS test
engine
2003-04-26 07:16 jmason
* MANIFEST, build/tag_nightly_mc, build/tag_weekly_mc: tagging;
added weekly version and sanity check to ensure we're running at
0859 UTC
2003-04-24 12:17 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: fixed bug where upgrading
the DB, and having the atime reset to last expire time would cause
oldest token to be incorrectly set.
2003-04-24 06:10 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1776: decided on atime
== msgcount, so re-enabled bayes r/w mode expire wasn't setting the
bayes db version, oops. fixed. a few other small cleanups
2003-04-24 05:48 felicity
* t/spamd_utf8.t: bug 1708: UTF-8 locale was specified incorrectly
(glibc version)
2003-04-23 15:41 felicity
* lib/Mail/SpamAssassin/Bayes.pm: Oops, cut/paste from another
section and forgot to set the bayes version. :|
2003-04-23 07:31 jmason
* MANIFEST, rules/20_anti_ratware.cf, rules/20_fake_helo_tests.cf,
rules/20_head_tests.cf, rules/20_ratware.cf, rules/20_uri_tests.cf,
rules/70_cvs_rules_under_test.cf: the bi-weekly trawl through
T_land. promoted a few good ones, dropped one or two bad ones
2003-04-23 07:11 jmason
* lib/Mail/SpamAssassin/PerMsgStatus.pm: MIME decoder was
occasionally missing related parts
2003-04-23 03:36 felicity
* lib/Mail/SpamAssassin/Bayes.pm: missed a magic token prefix
change during tokenization ... oops. :|
2003-04-23 03:32 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm: bugs 1523, 1666,
1775, and 1776: set a new bayes db format for 2.6x, version 1 (2.5x
had version 0). atimes are now unsigned ints (32-bit) instead of a
short (16-bit) so we won't get overflow/expire issues. code is now
included to automatically upgrade to v1 when the db is opened for
writing. v0 is still supported for readonly mode. bayes tokens
are now packed in little-endian (VAX) format. this means that the
same DB will work on different endian platforms.
at the moment, r/w access to the bayes db has been disabled pending
final resolution of bug 1776 and the "should atime be message count
or timestamp" question. right now, the code assumes message count.
2003-04-22 13:14 felicity
* USAGE: bug 1739: USAGE file says that a test message will have
the subject rewritten, but rewrite_subject is off by default in
2.5x. change the docs to reflect this.
2003-04-22 13:03 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1706 updates
2003-04-22 13:00 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1706: fixed a small
issue where the code was expecting a shift but got a pop, so we
need to look at the end of the array, not the beginning. :(
2003-04-22 12:54 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1706: when bayes was
expiring tokens, it would check to see if there were enough tokens
left at the end and reprieve tokens if necessary. however, the
reprieval was happening in a random order, so the oldest tokens
would likely be left in the DB. the code will now sort the tokens
by atime and reprieval the most recently used tokens first.
2003-04-22 08:31 jmason
* sa-learn.raw, lib/Mail/SpamAssassin.pm,
lib/Mail/SpamAssassin/Bayes.pm,
lib/Mail/SpamAssassin/BayesStore.pm,
lib/Mail/SpamAssassin/CmdLearn.pm: re-added journalled learning2
2003-04-22 08:06 jmason
* lib/Mail/SpamAssassin/Dns.pm: bug 1784: IP_IN_RESERVED_RANGE out
of date2
2003-04-22 08:03 jmason
* spamd/spamd.raw: bug 1799: REPORT verb ignores local config
2003-04-22 08:01 jmason
* USAGE: bug 1739: USAGE doco incorrect
2003-04-22 07:59 jmason
* MANIFEST: bug 1805: file was missing
2003-04-22 07:56 jmason
* t/spamd_utf8.t: bug 1708: UTF-8 locale was used incorrectly2
2003-04-22 07:53 jmason
* lib/Mail/SpamAssassin/Dns.pm, spamd/spamd.raw: couple of minor
tweaks to Net::DNS use2
2003-04-22 07:05 jmason
* build/tag_nightly_mc: use rtag -a instead of tag
2003-04-22 06:21 jmason
* rules/70_cvs_rules_under_test.cf: previous rule already caught by
_0 variant
2003-04-22 02:15 quinlan
* lib/Mail/SpamAssassin/HTML.pm: add back T_HTML_IMAGE_WEB_BUGS_1
2003-04-22 01:26 jmason
* rules/70_cvs_rules_under_test.cf: [no log message]
2003-04-21 09:58 quinlan
* rules/70_cvs_rules_under_test.cf: add case-insensitive version of
USERPASS
2003-04-21 05:09 duncf
* debian/: README.CVS, control, spamassassin.templates,
po/POTFILES.in, po/pt_BR.po, po/templates.pot: Get the Debian stuff
to work on Debian woody.
2003-04-20 10:14 quinlan
* rules/70_cvs_rules_under_test.cf: add some more exploratory tests
for web bug images (based on area) add a few additional DNSBLs for
testing
2003-04-20 10:12 quinlan
* lib/Mail/SpamAssassin/HTML.pm: add some more exploratory tests
for web bug images (based on area)
2003-04-19 20:35 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1800: messages with
large positive scores (if they hit GTUBE at 1000 points for
instance), can have a X-Spam-Level header with more than the RFC
allowed number of chars. so we now limit X-Spam-Level to show a
max score of 100. that makes it obvious the message is really spam
while staying RFC compliant.
2003-04-19 00:04 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1800: limit
X-Spam-Level to 100 chars. Keeps high scores from becoming RFC
incompliant (997 char limit), and frankly at 100 you get the point
that it's spam.
2003-04-18 22:54 jmason
* MANIFEST, spamd/spamd.raw: spamd: support kill -1 to restart and
reread config
2003-04-18 22:08 jmason
* build/preprocessor: build/preprocessor: REMOVEFORINST was
rewritten to comment lines. nope, REMOVE them ;)
2003-04-18 17:17 jmason
* rules/70_cvs_rules_under_test.cf: added a bz URI test
2003-04-18 05:43 jmason
* rules/70_cvs_rules_under_test.cf: added rd.yahoo rule
2003-04-15 21:02 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1782: bayes journal
writes may be split up by stdio, causing atime updates to be missed
2003-04-15 18:21 jmason
* sa-learn.raw: FAQ on learning
2003-04-15 06:29 quinlan
* lib/Mail/SpamAssassin/: Dns.pm, EvalTests.pm: bug 1778: remove
check_two_rbl_results() from EvalTests.pm bug 1781: TXT to
127.0.0.2 hack causes problems (changes dnscache to be
type-specific)
2003-04-15 03:59 quinlan
* masses/freqdiff: add -b option
2003-04-15 03:33 quinlan
* rules/70_cvs_rules_under_test.cf: add tests for detecting
ultra-small web bug images remove Reynolds RBLs since they require
subscriptions for commercial use and non-small sites
2003-04-15 03:31 quinlan
* lib/Mail/SpamAssassin/HTML.pm: add test code for detecting
ultra-small web bug images
2003-04-14 17:15 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: remove T_NO_DNS_FOR_RECIPS_*
(only tested well on older corpus email)
2003-04-14 12:08 quinlan
* rules/70_cvs_rules_under_test.cf: add missing net tflags
2003-04-14 12:02 quinlan
* rules/70_cvs_rules_under_test.cf: minor typo
2003-04-14 12:01 quinlan
* rules/70_cvs_rules_under_test.cf: add some fiveten tests back
since I didn't properly test them before (they're free anyway since
it's a multi-rbl)
2003-04-14 11:37 quinlan
* rules/: 20_dnsbl_tests.cf, 70_cvs_rules_under_test.cf: fix a
DNSBL bug where the post-tests were not running after harvesting
because we weren't using rblreseval as necessary with the current
RBL code
2003-04-14 10:03 quinlan
* rules/70_cvs_rules_under_test.cf: add an A record version of
T_RCVD_IN_MONKEYS_UPL2 for testing fix a typo in the SORBS
definition
2003-04-14 08:00 quinlan
* rules/70_cvs_rules_under_test.cf: add some RBL results comments
add the best of the RBLs I've been testing (some are definitely
better than some of our current ones), note that I do not intend to
significantly increase the number of RBLs that we're using!
2003-04-14 04:08 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1775: initial work to
deal with bayes db upgrade
2003-04-13 07:19 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: change
check_rbl_results_for() so it can also be used for bitmask tests
like opm.blitzed.org or regular expression tests (might come in
handy)
2003-04-13 04:38 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1706: slight
improvement, don't bother sorting and the like unless we actually
need to reprieve
2003-04-13 04:28 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1706: sort the deleted
tokens for reprieval so we get the newer ones first. also,
reprieve all tokens from the same atime at once.
2003-04-13 03:25 quinlan
* rules/70_cvs_rules_under_test.cf: add some juno.com anti-forgery
rules
2003-04-12 23:50 jmason
* rules/70_cvs_rules_under_test.cf: fixed T_FAKE_HELO_JUNO, should
be able to replace other juno-forgery rules
2003-04-12 08:45 quinlan
* rules/70_cvs_rules_under_test.cf: bug 1769, possible replacement
rules
2003-04-12 04:44 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/20_html_tests.cf,
rules/70_cvs_rules_under_test.cf: promote HTML_FONT_NEAR_INVISIBLE
2003-04-11 08:33 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
narrow down number of invisible text rules
2003-04-11 03:26 jmason
* rules/70_cvs_rules_under_test.cf: couple more T_FAKE_HELO rules
2003-04-10 23:09 jmason
* rules/: 20_head_tests.cf, 20_phrases.cf, 20_uri_tests.cf,
70_cvs_rules_under_test.cf: promoted some 1.0-S/O rules
2003-04-10 23:00 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: fixed bugs with T_FAKE_HELO rules
2003-04-10 22:43 jmason
* lib/Mail/SpamAssassin/: EvalTests.pm, Received.pm: deal with
Apple's Mail.app brokenness with HELO; recognize attbi.com REceived
hdrs
2003-04-10 16:27 felicity
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1524: EvalTests looking
for/storing IPs would sometimes catch extra chars that occur around
the IP address (such as [ in Received headers). Use a new regexp
to only match IP addresses.
2003-04-10 07:37 quinlan
* lib/Mail/SpamAssassin/Dns.pm: small indentation change
2003-04-09 23:24 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: updated T_FAKE_HELO tests to
avoid hotmail.com matching mail.com
2003-04-09 23:20 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: updated T_FAKE_HELO tests to
double-check possibly-forged rcvd lines
2003-04-09 22:57 jmason
* MANIFEST, rules/30_text_sk.cf: added Slovak translation from
Peter Mann (peter DOT mann AT tuke DOT sk)
2003-04-09 20:09 jmason
* spamd/spamc.pod: mention -x in -f section
2003-04-09 11:02 quinlan
* lib/Mail/SpamAssassin/HTML.pm: improvements to invisible text
tests - correctly track fgcolor - only trigger if invisible
text is actually printed (thinking ahead to possible Bayes
tweaks) - don't retest tested color combinations - weight R, G,
B based on perceived luminosity (Rec. 709 values) - scale both
difference formulas to 0..765 for testing phase - separate
near-invisibility from total-invisibility (near invisibility
seems to have a better S/O since it is more likely to be
intentional)
2003-04-08 20:03 jmason
* lib/Mail/SpamAssassin/Received.pm: duh, typo
2003-04-08 08:25 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Received.pm, rules/20_dnsbl_tests.cf,
rules/70_cvs_rules_under_test.cf: T_FAKE_HELO rule-set added
2003-04-08 06:11 jmason
* lib/Mail/SpamAssassin/Util.pm: netmask support not supporting
192.168/16
2003-04-08 02:59 jmason
* lib/Mail/SpamAssassin/: Dns.pm, Received.pm: removed markup with
semi-trusted hdr; removed PTR lookup for reserved IPs
2003-04-08 01:17 jmason
* MANIFEST, lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/Dns.pm, lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Received.pm, rules/20_dnsbl_tests.cf,
rules/70_cvs_rules_under_test.cf: added 20_dnsbl_tests; implemented
notfirsthop and lastuntrusted set modifiers; fixed some warnings;
updated reserved-IP list; document trusted/untrusted relay
differentiation
2003-04-08 00:14 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1455: fix IP_ADDRESS
regexp
2003-04-07 23:33 jmason
* lib/Mail/SpamAssassin/: Dns.pm, EvalTests.pm: removed
one-match-per-IP code, which seems like a bizarre idea; and fixed
an undef warning
2003-04-07 23:27 jmason
* lib/Mail/SpamAssassin/Conf.pm, lib/Mail/SpamAssassin/Dns.pm,
lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm, rules/20_head_tests.cf,
rules/70_broken_rules.cf, rules/70_cvs_rules_under_test.cf: massive
DNS functionality improvement. Some DNSBLs were being missed, due
to all of them sharing 'relay' for a zone id, fixed; DNSBL lookups
now use TXT records on the zones that support it, and log that in
the report; DNSBL lookup code simplified; removed obsolete
dialup_codes functionality; added lookups for formmail and
proxies.relays.monkeys.org.
2003-04-07 08:55 quinlan
* lib/Mail/SpamAssassin/: HTML.pm, PerMsgStatus.pm: major
improvements to HTML bgcolor code add a few more named colors to
color hash also replace HTML_WITH_BGCOLOR with T_HTML_WITH_BGCOLOR
2003-04-07 08:48 quinlan
* rules/70_cvs_rules_under_test.cf: add T_HTML_FONT_COLOR_NAME as
possible replacement rule
2003-04-07 03:08 duncf
* t/: mailing_lists.t, spamd_report.t, data/nice/007,
data/nice/ezmlm_message.txt, data/nice/mailman_reminder.txt: Bug
1721: make test fails. removed mailing list test, and data, since
we have no mailing list rules any more. Added t/data/nice/007 which
contains habeas mark (one of few ham signs)
2003-04-07 02:29 felicity
* spamd/: utils.c, utils.c: bug 1755: a c++ comment slipped in, and
some compilers don't know what to do with it.
2003-04-06 20:58 felicity
* lib/Mail/SpamAssassin/Util.pm: changed untaint_file_path to be
the same as the 2.54 version
2003-04-06 20:56 felicity
* lib/Mail/SpamAssassin/Util.pm: bug 1741: allow ~ and space in
file path (good for Windows), return blank if blank passed in,
fixed return only if RE matches
2003-04-06 20:52 felicity
* spamassassin.spec: bug 1754: typo in the new spec file.
2003-04-06 02:27 jmason
* masses/corpora/mass-find-nonspam: added --single flag
2003-04-06 00:30 jmason
* lib/Mail/SpamAssassin/Received.pm: slight reordering of logic in
Received.pm; now attempts to infer _some_ data at least before
giving up
2003-04-06 00:06 jmason
* lib/Mail/SpamAssassin/Conf.pm: added little bit of doco
2003-04-06 00:03 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1740: remove dups from
check_rbl call
2003-04-05 23:58 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1455: better IP_ADDRESS
regexp
2003-04-05 23:53 jmason
* lib/Mail/SpamAssassin/: Conf.pm, Received.pm, Util.pm: Received
parser: added 'trusted_networks' config param so users can specify
which relays are trusted, improves DNSBL accuracy and speed. Still
need to come up with code to figure this out algorithmically, which
I think Dan plans to do
2003-04-05 23:45 felicity
* rules/70_cvs_rules_under_test.cf: bug 1711: added a new
T_NIGERIAN_SUBJECT2 test rule to possibly replace the original one
2003-04-05 23:24 felicity
* lib/Mail/SpamAssassin/Received.pm: postfix header parsing needed
an extra if statement
2003-04-05 22:27 jmason
* lib/Mail/SpamAssassin/Received.pm: test failure with new Received
code, worked around
2003-04-05 22:08 jmason
* lib/Mail/SpamAssassin/: EvalTests.pm, Received.pm: fixed some
eval tests to use new Received-header parser
2003-04-05 21:15 jmason
* MANIFEST, lib/Mail/SpamAssassin/PerMsgStatus.pm,
lib/Mail/SpamAssassin/Received.pm: new Received header parser added
2003-04-05 20:58 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm: bug 1732: we
were looking for caller_will_untie in finish() which was a very bad
idea (it is the function that the caller is supposed to call...),
but sync_journal and expire_old_tokens were untieing whenever it
felt like it.
2003-04-05 20:36 felicity
* lib/Mail/SpamAssassin/CmdLearn.pm: bug 1691: sa-learn was sending
non-error output to stderr instead of stdout (specifically the
"Learned # messages" string)
2003-04-05 19:53 jmason
* rules/20_ratware.cf: removed a duplicate BAT_MSGID rule
2003-04-05 04:23 duncf
* lib/Mail/SpamAssassin/Conf.pm: minor cleanup... hanging around in
my tree... should have been committed a while ago
2003-04-05 02:05 felicity
* spamassassin.spec, spamd/redhat-rc-script.sh: Standardized the
RPM spec file and spamd rc script to be more RedHat-ish. Merged in
some changes from the RH SA RPM.
2003-04-05 01:11 jmason
* lib/Mail/SpamAssassin/Util.pm, rules/70_cvs_rules_under_test.cf:
added T_WALAA, still chuckling
2003-04-04 23:22 jmason
* spamassassin.raw, sa-learn.raw: bug 1737: bayes doco is weak,
improved
2003-04-04 20:28 felicity
* sa-learn.raw, lib/Mail/SpamAssassin.pm,
lib/Mail/SpamAssassin/Bayes.pm,
lib/Mail/SpamAssassin/BayesStore.pm,
lib/Mail/SpamAssassin/CmdLearn.pm: bug 1552: merged the
capabilities of check_bayes_db into sa-learn via the --dump
parameter. also forward ported BayesStore::nspam_nham_get() from
2.5 which was missed. doh.
2003-04-04 19:22 felicity
* lib/Mail/SpamAssassin/Bayes.pm: bug 1723: X-Perlmx-Spam is the
PerlMX equiv of X-Spam-Status, so lets ignore it.
2003-04-04 15:17 felicity
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: rfc violations aren't necessarily
great spam indicators. all of the header_count_bad rules suck,
except for the ctype one which seems to do a pretty decent job,
even though it's not an rfc violation. to deal with fps, I added
code to treat multiple headers with the exact same content as a
single header.
2003-04-04 14:58 felicity
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
the blank comment rule didn't perform well, so it's gone. the
alphanum comment rule is ok, but it looks like some versions of
Mozilla add in a spurious empty comment which triggers the rule, so
add a meta and see what happens. :)
2003-04-04 14:35 felicity
* spamassassin.spec: update spec file to 2.54 version
2003-04-04 03:45 jmason
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: misspelling ratware detection
eval test added
2003-04-04 02:48 duncf
* debian/: changelog.in, control, logcheck, spamassassin.default,
spamassassin.templates, watch, po/POTFILES.in, po/pt_BR.po,
po/templates.pot: Updated debian packaging for CVS builds; adds
po-debconf support
2003-04-04 02:48 felicity
* spamassassin.spec: updated spec file since 2.51 prepared for 2.54
release
2003-04-04 02:34 jmason
* build/preprocessor.pm: backport: preprocessor being indexed by
CPAN, avoid for next version
2003-04-04 02:33 jmason
* lib/Mail/SpamAssassin.pm: Starting 2.54
2003-04-04 02:22 duncf
* debian/: changelog.in, control, logcheck, spamassassin.default,
spamassassin.templates, watch, po/POTFILES.in, po/pt_BR.po,
po/templates.pot: Update packaging for Debian; notably, adds
po-debconf support
2003-04-03 20:03 jmason
* MANIFEST, t/SATest.pm, t/spamd_protocol_10.t: spamd protocol 1.0
test added
2003-04-03 04:32 quinlan
* rules/70_cvs_rules_under_test.cf,
lib/Mail/SpamAssassin/EvalTests.pm: remove
T_MIME_BASE64_SHORT_LINES since it doesn't work that well
2003-04-02 20:20 jmason
* lib/Mail/SpamAssassin/EvalTests.pm, spamd/spamd.raw: iPlanet
rewrites msgids
2003-04-02 18:40 climent
* rules/30_text_es.cf: 8 bits typo
2003-04-02 09:24 quinlan
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
add T_HTML_WITH_BGCOLOR which only tests background color change
HTML_FONT_INVISIBLE to only look at current bgcolor add
T_HTML_FONT_INVISIBLE* tests which allow varying degrees of color
differences to be considered the same
2003-04-02 04:31 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1649: remove the
required_hits safety thresholds.
2003-04-02 02:03 jmason
* rules/: 20_ratware.cf, 30_text_fr.cf, 70_cvs_rules_under_test.cf:
promoted some 1.0-S/O testing rules. deleted a couple of
non-performers
2003-04-02 00:36 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1254: avoid SUBJ_ALL_CAPS
on KOI8-R-encoded mail
2003-04-01 03:06 jmason
* build/README, build/update_stable,
lib/Mail/SpamAssassin/CmdLearn.pm: output Learned message to stdout
instead of stderr
2003-04-01 02:42 jmason
* Changes, build/README, lib/Mail/SpamAssassin/BayesStore.pm: ready
for release
2003-04-01 01:43 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1689: exists method does
not exist in NDBM_File for some reason. worked around
2003-04-01 01:08 jmason
* lib/Mail/SpamAssassin/Conf.pm: moved dialup_codes into more
privileged area
2003-04-01 00:39 jmason
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm: bug 1681: avoid
NDBM_File error by using defined instead of exists
2003-03-30 14:16 msquadrat
* lib/Mail/SpamAssassin/Conf.pm: fix by Aaron Sherman
: don't word wrap the DUL examples
2003-03-30 05:11 duncf
* lib/Mail/SpamAssassin/Conf.pm: Fixing allow_user_rules stuff
2003-03-30 04:26 jmason
* lib/Mail/SpamAssassin.pm: ready to release
2003-03-30 04:25 quinlan
* rules/: 20_anti_ratware.cf, 20_compensate.cf, 20_head_tests.cf,
20_html_tests.cf, 30_text_de.cf, 30_text_fr.cf, 30_text_it.cf,
50_scores.cf, 70_broken_rules.cf, regression_tests.cf: bug 1680:
phase out all nice rules that are abusable
2003-03-30 03:42 quinlan
* rules/70_cvs_rules_under_test.cf: remove T_GOOD_LIST_HEADERS,
just another nice rule we don't need...
2003-03-30 02:52 felicity
* MANIFEST, Makefile.PL, t/SATest.pm, t/rule_tests.t,
t/zz_cleanup.t: bug 1703: Makefile.PL would set a razor1 test if
razor1 wasn't installed. rule_tests.t was still trying to use
bytes.pm. spamd tests were using user's preferences/bayes/etc.
2003-03-30 02:42 felicity
* t/zz_cleanup.t: file zz_cleanup.t was initially added on branch
b2_5_0.
2003-03-30 02:42 felicity
* MANIFEST, Makefile.PL, t/SATest.pm, t/rule_tests.t,
t/zz_cleanup.t: bug 1703: Makefile.PL would set a razor1 test if
razor1 wasn't installed. rule_tests.t was still trying to use
bytes.pm. spamd tests were using user's preferences/bayes/etc.
2003-03-30 02:28 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1707: When syncing the
journal, the code took advantage of a UNIX-ism where you can open a
file, then rename it, and still access the file data.
Unfortunately, Windows doesn't let you do that. So we now rename
before we open, and add a "can we read the file" check beforehand
to avoid messy cleanup.
2003-03-30 02:10 felicity
* lib/Mail/SpamAssassin/Bayes.pm: bug 1699: a slight change to the
last patch, during tokenization, strip the extra chars before
skipping magic tokens.
2003-03-30 00:58 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: removed a redundant stat
2003-03-29 21:24 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1707: windows can't
handle the open then rename thing like UNIX does, so journal sync
is broken in 2.52 on windows. :(
2003-03-29 16:45 felicity
* lib/Mail/SpamAssassin/Bayes.pm: We check to see if it's a magic
token, then modify the token to strip off some non-alphanum chars.
This means that "**NSPAM" would still get through.
2003-03-29 16:38 felicity
* lib/Mail/SpamAssassin/Bayes.pm: moved # of tokens check up a
little, skips unneeded probability check
2003-03-29 01:33 jmason
* rules/20_phrases.cf: HGH has FP in common Japanese phrase; worked
around
2003-03-29 00:19 jmason
* rules/: 20_head_tests.cf, 20_phrases.cf: oops, someone didn't run
make test ;)
2003-03-28 23:53 jmason
* t/SATest.pm: bug 1703: fixed tests to use a copy of the rules
dir, instead of a -p user-prefs file -- which spamd does not
support
2003-03-28 23:50 jmason
* spamd/: libspamc.c, libspamc.h, spamc.pod: libspamc: added a
'private' member to struct message, to solve bug 1595 (-B was
colliding with -c in terms of usage of struct message fields)
2003-03-28 23:48 jmason
* lib/Mail/SpamAssassin/: BayesStore.pm, Conf.pm,
DBBasedAddrList.pm, Locker.pm, UnixLocker.pm, Win32Locker.pm:
updated lock-timeout code to match 2.53
2003-03-28 21:28 jmason
* MANIFEST, spamd/libspamc.c, spamd/libspamc.h, t/spamc_B.t,
t/data/spam/bsmtp: bug 1595: BSMTP mode (spamc -B) fixed
2003-03-28 20:26 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm: bug 1699:
slightly better regex. ignore actual magic token strings, not just
magic token prefixes. this leaves more tokens able to be used.
2003-03-28 20:23 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm: slightly better
version for bug 1699. keeps **No as valid token, but ignores magic
token **NO.
2003-03-28 20:14 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm, Bayes.pm,
BayesStore.pm: Bug 1699: The code wasn't protecting the Bayes DB
"magic" tokens (tracks num ham/spam, etc.) from being overwritten
if they occured in a message. This allows mails to corrupt the DB
either maliciously or if they discuss the tokens.
2003-03-28 09:24 quinlan
* rules/: 20_ratware.cf, 70_cvs_rules_under_test.cf: promote
RATWARE_STORM_URI remove some test rules that don't work very well
2003-03-28 08:22 quinlan
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: bug 1533:
new Received: header forgery tests using HELO
2003-03-28 08:11 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm, rules/20_head_tests.cf,
rules/70_cvs_rules_under_test.cf: bug 1581: new MSG_ID_ADDED_BY_MTA
tests are done
2003-03-28 07:24 quinlan
* rules/: 70_cvs_rules_under_test.cf, 20_phrases.cf: promote
REMOVE_REMOVAL_*
2003-03-28 06:41 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin/BayesStore.pm,
SpamAssassin/CmdLearn.pm, SpamAssassin/DBBasedAddrList.pm,
SpamAssassin/Locker.pm, SpamAssassin/UnixLocker.pm,
SpamAssassin/Win32Locker.pm: bug 1690: now has different lock
timeouts for different bayes operations
2003-03-28 03:25 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, Bayes.pm: bug 1696: you could
continuously "forget" tokens from messages you've learned, and even
those you hadn't learned.
2003-03-28 03:18 felicity
* spamassassin.raw, t/reportheader.t, spamassassin.raw,
t/reportheader.t: bug 1695: report_header was removed in 2.5x but
still was in the POD docs and a test config.
2003-03-28 03:16 felicity
* sa-learn.raw, sa-learn.raw: bug 1502: sa-learn wasn't paying
attention to the compile-time paths for rules and the like. The
problem was that the variables were declared my() which aren't
accessible by other packages.
2003-03-28 01:27 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin/BayesStore.pm,
SpamAssassin/CmdLearn.pm, SpamAssassin/Conf.pm,
SpamAssassin/DBBasedAddrList.pm, SpamAssassin/Locker.pm,
SpamAssassin/UnixLocker.pm, SpamAssassin/Win32Locker.pm: backing
out lock_timeout patch for redo
2003-03-28 01:01 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin/BayesStore.pm,
SpamAssassin/CmdLearn.pm, SpamAssassin/Conf.pm,
SpamAssassin/DBBasedAddrList.pm, SpamAssassin/Locker.pm,
SpamAssassin/UnixLocker.pm, SpamAssassin/Win32Locker.pm: bug 1690:
now has different lock timeouts for different bayes operations
2003-03-28 00:49 jmason
* lib/Mail/SpamAssassin/: BayesStore.pm, Conf.pm,
DBBasedAddrList.pm: updated fix for multiple lock timeouts
2003-03-28 00:26 jmason
* spamassassin.raw, spamassassin.raw: bug 1608: spamassassin.1
lists incorrect paths for config files
2003-03-28 00:15 jmason
* spamd/: libspamc.c, spamc.c, spamc.pod: bug 1697: spamc fails to
dump message when running out of memory
2003-03-28 00:00 jmason
* rules/70_cvs_rules_under_test.cf: tweaked the X_ORIG_IP_X2 rule
2003-03-27 21:28 jmason
* spamd/: libspamc.c, spamc.c, spamc.pod: documented spamc -x arg,
fixed bug where mem shortage resulted in lost message
2003-03-27 21:18 jmason
* spamd/spamd.raw: bug 1536: kernel reporting SIGWAIT problem on
recent Red Hat kernels
2003-03-27 21:13 jmason
* lib/Mail/SpamAssassin/DBBasedAddrList.pm: bug 1661: user_prefs
was world-writable after AWL tie, due to failure to set umask
correctly
2003-03-27 21:12 jmason
* lib/Mail/SpamAssassin/: EvalTests.pm, EvalTests.pm: bug 1401:
check_rbl was checking incorrect number of Received hdrs
2003-03-27 20:58 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin/BayesStore.pm,
SpamAssassin/CmdLearn.pm, SpamAssassin/Conf.pm,
SpamAssassin/DBBasedAddrList.pm, SpamAssassin/Locker.pm,
SpamAssassin/UnixLocker.pm, SpamAssassin/Win32Locker.pm: bug 1690:
now has different lock timeouts for different bayes operations
2003-03-27 20:35 felicity
* lib/Mail/SpamAssassin/: Dns.pm, Reporter.pm, Dns.pm, Reporter.pm:
bug 1632: Razor modifies %ENV so we tried to use a local() to make
a temp version, which works, but kills %ENV for the Razor code as
well. So now we'll let the helper_mode functions handle backing
up/restoring %ENV as appropriate.
2003-03-27 20:11 jmason
* rules/70_cvs_rules_under_test.cf: new rule for test
2003-03-27 17:54 felicity
* spamd/: spamd.raw, spamd.raw: bug 1514: spamd was trying to use
"../lib" which is ok most of the time, but during a "make test" on
perl 5.005 it errored out with the "bytes.pm" error. so we now
check for a "../blib/lib" first and use it if it exists.
otherwise, we continue with "../lib".
2003-03-27 17:48 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin.pm: bug 1685: we should
clean PATH when we start if we're in taint mode. This fixes a
Sys::Hostname problem in 5.005, but it's a good idea anyway since
we won't be able to use PATH unless it's cleaned in taint mode
anyway.
2003-03-27 08:56 quinlan
* rules/70_cvs_rules_under_test.cf: narrow down to T_REMOV* some
possible replacements for MSG_ID_ADDED_BY_MTA_2
2003-03-27 02:58 jmason
* rules/70_cvs_rules_under_test.cf: couple of rules for test
2003-03-27 02:49 jmason
* rules/70_cvs_rules_under_test.cf, spamd/spamd.raw: let's try a
forged-Eudora variant by From address
2003-03-27 01:53 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin.pm: bug 1630:
add-all-addrs-to-blacklist was a no-op, due to > vs >= bug
2003-03-27 01:04 jmason
* lib/Mail/SpamAssassin/DBBasedAddrList.pm,
rules/20_anti_ratware.cf, rules/20_body_tests.cf,
rules/20_head_tests.cf, rules/20_phrases.cf, rules/20_ratware.cf,
rules/70_cvs_rules_under_test.cf: fixed DBBasedAddrList umask issue
in HEAD; moved some 1.0-S/O T_ rules into the live set
2003-03-26 21:10 jmason
* MANIFEST, spamd/libspamc.c, spamd/libspamc.h, t/spamc_B.t,
t/data/spam/bsmtp: bug 1595: BSMTP mode broken, fixed
2003-03-26 20:18 jmason
* sa-learn.raw, lib/Mail/SpamAssassin/CmdLearn.pm, sa-learn.raw,
lib/Mail/SpamAssassin/CmdLearn.pm: bug 1572: sa-learn doco was
still talking about auto-whitelists, although this was never
implemented
2003-03-26 20:07 jmason
* spamd/: README.spamd, README.spamd: bug 1674: README.spamd didn't
know that auto-learning had been implemented
2003-03-26 20:06 jmason
* MANIFEST.SKIP, MANIFEST.SKIP: updated MANIFEST.SKIP likewise
2003-03-26 20:05 jmason
* spamd/: .cvsignore, .cvsignore: updated cvsignore to ignore
sslspamc
2003-03-26 20:04 jmason
* spamd/spamd.raw: bug 1684: syslog-socket had a typo, fixed
2003-03-26 20:01 jmason
* spamd/libspamc.c, spamd/utils.c, spamd/utils.h, .cvsignore,
.cvsignore, spamd/libspamc.c, spamd/utils.c, spamd/utils.h: bug
1637: compilation of spamd/sslspamc broken, fixed
2003-03-26 19:27 jmason
* .cvsignore, config.h.in, configure, configure.in: bug 1542:
configure doesn't detect SSL_CTX_free on NetBSD, fixed
2003-03-26 19:13 jmason
* config.h.in, configure, configure.in, .cvsignore, .cvsignore: bug
1542: configure doesn't detect SSL_CTX_free on NetBSD, fixed
2003-03-26 18:28 msquadrat
* configure, Makefile.PL, .cvsignore: bug 1542: configure wasn't
updated, so the fix for this bug didn't work per default
2003-03-26 18:20 msquadrat
* configure, Makefile.PL, .cvsignore: bug 1542: configure wasn't
updated, so the fix for this bug didn't work per default
2003-03-26 07:26 quinlan
* rules/70_cvs_rules_under_test.cf: add a bunch of T_REMOV*_REMOV*
tests for further testing
2003-03-26 02:55 jmason
* rules/70_cvs_rules_under_test.cf: one more forgery test, for IMS
this time
2003-03-26 01:55 jmason
* rules/70_cvs_rules_under_test.cf: renamed some meta subrules
2003-03-26 01:17 jmason
* rules/70_cvs_rules_under_test.cf: added some anti-Bat-forgery
rules based on excellent advice from the Bat developers
2003-03-26 00:48 jmason
* build/README: added note about md5sums
2003-03-25 17:29 msquadrat
* spamd/spamd.raw: bug: 1684
: typo broke fix for --syslog-socket (doh\!)
2003-03-25 01:22 jmason
* lib/Mail/SpamAssassin.pm: 2.53 devel cycle started
2003-03-25 01:20 jmason
* Changes: 2.52 RELEASED
2003-03-25 01:11 jmason
* INSTALL: DB_File recommended in INSTALL
2003-03-25 01:11 jmason
* INSTALL: last-minute doco change: recommend using DB_File
2003-03-24 22:39 msquadrat
* lib/Mail/SpamAssassin.pm: ready for release
2003-03-24 22:21 msquadrat
* spamassassin.raw, spamassassin.raw: bug 1260: corrected
description of --nocreate-prefs option
2003-03-24 22:17 msquadrat
* spamd/: netbsd-rc-script.sh, netbsd-rc-script.sh: bug 1656:
updated init script for NetBSD
2003-03-24 21:38 felicity
* sa-learn.raw, lib/Mail/SpamAssassin/CmdLearn.pm: bug 1572: -a was
never implemented in sa-learn. removing references to it.
2003-03-24 21:28 msquadrat
* configure.in, configure.in: bug 1542: configure didn't detect
SSL_CTX_free on NetBSD
2003-03-24 21:15 msquadrat
* spamd/spamd.raw: bug 1601: option --syslog-socket wasn't
implemented
2003-03-24 21:11 msquadrat
* spamd/spamd.raw: bug 1601: option --syslog-socket wasn't
implemented
2003-03-24 08:58 quinlan
* rules/70_cvs_rules_under_test.cf: another
T_MSGID_ADDED_BY_HOTMAIL variation
2003-03-24 08:46 quinlan
* rules/70_cvs_rules_under_test.cf: add some
T_MSGID_ADDED_BY_HOTMAIL variants
2003-03-24 07:43 quinlan
* rules/70_cvs_rules_under_test.cf: merge T_FAKED_THEBAT2 and
T_FORGED_MUA_THEBAT into T_RATWARE_NOT_THEBAT add a bunch of MSGID
format rules
2003-03-24 04:26 jmason
* rules/70_cvs_rules_under_test.cf: some new User-Agent rules
against forgery
2003-03-24 03:22 jmason
* rules/20_body_tests.cf: note about Send-Safe ratware
2003-03-24 01:14 quinlan
* lib/Mail/SpamAssassin/TextCat.pm: bug 692: some major performance
improvements from Sidney Markowitz plus a few additional tweaks
2003-03-24 01:10 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm: bug 1664: the
only entries in the bayes journal are token access time updates
(from scans). we can exploit the fact that most of the tokens
appear multiple times by caching in memory the updates per token,
then just writing to db the end result times for all tokens.
also, scan() tries to opportunistically do a journal sync and
expire which causes lots of contention if there is a lot of
messages for a single db (this is usually seen with site-wide dbs).
so we flag when we are doing the sync and expire and let all other
pids not be so opportunistic. this should greatly lower
contention. we also let the db tie r/w happen before we try to
read from the journal, and also exploit the fact we can open the
journal before renaming it so the most common error occurs first,
and then a journal read failure means we don't need to rename the
temp file back.
2003-03-24 00:52 felicity
* lib/Mail/SpamAssassin/Bayes.pm: bug 1672: a typo in a backported
patch for 2.51 caused bayes to sometimes not unlock the db.
2003-03-23 23:04 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm: Merging in patch
796 -- lowering opportunistic expiry contention
2003-03-23 22:51 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: Merging with patch 793
2003-03-23 22:48 jmason
* rules/70_cvs_rules_under_test.cf: T_FORGED_MUA_THEBAT,
replacement for FORGED_MUA_THEBAT
2003-03-23 11:40 quinlan
* rules/70_cvs_rules_under_test.cf: fix T_FAKED_THEBAT2 add
T_JAPANESE_UCE_SUBJECT* add T_RUSSIAN_UCE_SUBJECT
2003-03-23 09:50 quinlan
* rules/70_cvs_rules_under_test.cf: promote
T_MSGID_SPAMSIGN_99X9XX99, T_RCVD_6_CAPS_ESMTP_ID,
T_BASE64_ENC_TEXT, T_MIME_BASE64_LATIN, T_PORN_15, T_OFFERS_ETC,
T_HOT_NASTY, T_CELEBRITY_PORN, T_VACATION_SCAM, T_HAIR_LOSS
replace T_INVALID_DATE_HMS with T_INVALID_DATE_HMS2 tweak
T_REMOVE_REMOVAL*, T_ID*, add T_RATWARE_HASH_DASH, T_FAKED_THEBAT2,
T_MIME_HTML_NO_MAJOR_TAG a few other minor changes
2003-03-23 09:44 quinlan
* rules/: 20_phrases.cf, 20_porn.cf: replace rules with faster
versions
2003-03-23 09:43 quinlan
* rules/20_head_tests.cf: add MSGID_SPAMSIGN_99X9XX99 and
RCVD_6_CAPS_ESMTP_ID
2003-03-23 09:43 quinlan
* rules/20_body_tests.cf: add MIME_BASE64_LATIN
2003-03-23 09:41 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: replace BASE64_ENC_TEXT with
T_BASE64_ENC_TEXT promote T_MIME_BASE64_LATIN to MIME_BASE64_LATIN
2003-03-22 19:13 jmason
* rules/70_cvs_rules_under_test.cf: new forged-Yahoo test added
2003-03-22 15:10 felicity
* rules/: 20_head_tests.cf, 50_scores.cf,
70_cvs_rules_under_test.cf: Promoted accessdb. still disabled by
default.
2003-03-21 01:20 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: another update for bug 1629
2003-03-20 21:26 msquadrat
* build/preprocessor: add support for -Dfilename for later use (the
command line is getting too long)
2003-03-20 20:51 jmason
* build/README: build process updated; new README
2003-03-20 20:49 jmason
* Changes, build/update_devel,
lib/Mail/SpamAssassin/AuditMessage.pm,
lib/Mail/SpamAssassin/EncappedMessage.pm: build process update
2003-03-20 20:38 msquadrat
* build/preprocessor: make the preprocessor use strict
2003-03-20 20:31 jmason
* lib/Mail/SpamAssassin.pm: 2.51 devel cycle started
2003-03-20 20:27 jmason
* Changes, build/update_devel, lib/Mail/SpamAssassin.pm: 2.51
RELEASED
2003-03-20 20:07 jmason
* Changes, lib/Mail/SpamAssassin.pm: 2.5.1 RELEASED
2003-03-20 19:38 jmason
* lib/Mail/SpamAssassin.pm: ready to release
2003-03-20 18:55 felicity
* rules/70_cvs_rules_under_test.cf: added rule to catch pxlg.com
mails, works nicely for me. :)
2003-03-20 06:53 jmason
* sa-learn.raw, lib/Mail/SpamAssassin/CmdLearn.pm: bug 1638:
--force-rebuild doco is unclear. fixed
2003-03-20 06:51 quinlan
* lib/Mail/SpamAssassin/: Conf.pm, Conf.pm: trivial documentation
correction
2003-03-19 22:26 msquadrat
* Makefile.PL: bug 1652: Perl 5.6.0 and 5.6.1 ship different
versions of ExtUtils::MakeMaker; both carry the version 5.45 though
2003-03-19 21:32 felicity
* Makefile.PL: new fix for bug 1652
2003-03-19 19:29 felicity
* Makefile.PL: bug 1652: the suggested patch doesn't work. perl
5.6.0 and perl 5.6.1 both have EU:MM 5.45, but the one in 5.6.1 has
PM_FILTER. so reverting the initial patch since it doesn't take
this into account.
2003-03-19 14:23 felicity
* lib/Mail/SpamAssassin/Bayes.pm: for some reason,
sanity_check_is_untied was defined twice
2003-03-19 05:27 felicity
* lib/Mail/SpamAssassin/Bayes.pm: bug 1655: we tie to check # of
Bayes tokens, but we don't untie if # ham/spam is too low. fixed.
2003-03-19 05:15 felicity
* lib/Mail/SpamAssassin/Bayes.pm: bug 1655: if we tie to check # of
Bayes tokens, we don't untie if # ham/spam is too low
2003-03-19 04:52 jmason
* MANIFEST, MANIFEST: bug 1541: removed trim_bayes_db from distro
2003-03-19 02:38 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1636: incorrect errors
from bayes if an error occurred during big scan-counter increment
2003-03-19 02:30 jmason
* spamd/README.spamd: README.spamd is well out of date on the spamd
protocol. Partial fix, more to follow (still missing recent method
additions)
2003-03-19 02:30 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1636: incorrect errors
from bayes if an error occurred during big scan-counter increment
2003-03-19 01:56 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1626: if a token's atime
in the bayes db is > current scan count, there was an expiry error,
so we'll reset the atime to the last expire time.
2003-03-18 23:13 quinlan
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Conf.pm,
SpamAssassin/PerMsgStatus.pm: bug 1576: need safer way to render
attachments for broken clients
2003-03-18 22:52 felicity
* t/: razor.t, razor2.t: bug 1619: razor tests were failing because
the rule description changed from what was being tested
2003-03-18 22:21 felicity
* lib/Mail/SpamAssassin/Bayes.pm: reminents of patch for bug 1566.
added some sanity-checks for bayes ...
2003-03-18 22:11 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Bayes.pm,
SpamAssassin/BayesStore.pm, SpamAssassin/PerMsgStatus.pm: bug 1566:
if spamd is run with -x, spamd only ties the Bayes DBs once, so SA
doesn't detect when nspam/nham have gone over the min # required
(200) and enable automatically.
2003-03-18 22:04 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1579: there were several
undef warnings in the Bayes code. cleaned that out.
2003-03-18 21:58 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm, PerMsgStatus.pm:
bug 1556: PerMsgStatus wasn't cleaning up bayes DB ties after it
was done scanning which caused problems with MailScanner.
2003-03-18 21:55 felicity
* lib/Mail/SpamAssassin/HTML.pm: bug 1653: HTML tests had undef
warnings sometimes. do better parsing via RE.
2003-03-18 03:29 quinlan
* t/razor2.t: rest of fix for bug 1619: make test: razor2 test
fails
2003-03-18 03:27 felicity
* spamd/spamd.raw: bug 1640: spamd was sending the Content-Length
header instead of the Spam: header for protocol 1.2 clients. it
should have been sending just the Spam header, not the
Content-Length one.
2003-03-18 03:24 quinlan
* lib/Mail/SpamAssassin/HTML.pm: bug 1653 for HEAD: patch to avoid
'undef' warning in HTML.pm
2003-03-18 03:21 felicity
* lib/Mail/SpamAssassin/Conf.pm: bug 1648: unwhitelist_from_rcvd
was non-functional due to typo. also updated some docs.
2003-03-18 03:17 felicity
* Makefile.PL: bug 1652, we looked for version < 5.45, but feature
was added > 5.45, so == 5.45 had problems
2003-03-18 03:14 felicity
* Makefile.PL: bug 1652: we checked for version < 5.45, but the
feature was added > 5.45, so we get confused when == 5.45.
2003-03-17 23:56 felicity
* rules/: 10_misc.cf, 20_anti_ratware.cf, 20_body_tests.cf,
20_compensate.cf, 20_head_tests.cf, 20_html_tests.cf,
20_meta_tests.cf, 20_phrases.cf, 20_porn.cf, 20_ratware.cf,
20_uri_tests.cf, 23_bayes.cf, 25_body_tests_es.cf,
25_body_tests_pl.cf, 25_head_tests_es.cf, 25_head_tests_pl.cf,
30_text_de.cf, 30_text_es.cf, 30_text_fr.cf, 30_text_it.cf,
30_text_pl.cf, 50_scores.cf, 60_whitelist.cf, 70_broken_rules.cf,
70_cvs_rules_under_test.cf, regression_tests.cf: bug 1607: add
license information to the rule files
2003-03-17 23:41 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: just moving the conf
setting into the initialization section right above it.
2003-03-17 20:34 jmason
* lib/Mail/SpamAssassin/HTML.pm: updated 'undef warning in HTML.pm'
patch
2003-03-17 20:28 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Bayes.pm,
SpamAssassin/BayesStore.pm, SpamAssassin/EvalTests.pm,
SpamAssassin/PerMsgStatus.pm: added sanity checks for bayes db
untieing
2003-03-17 19:55 jmason
* lib/Mail/SpamAssassin/Conf.pm, t/razor2.t: bug 1648:
unwhitelist_from_rcvd was not working
2003-03-17 19:51 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: deletion of tokens was
giving 'undefined value as ARRAY reference' warnings
2003-03-17 19:07 felicity
* lib/Mail/SpamAssassin/Bayes.pm: bug 1651: if for some reason a
Bayes scan() doesn't occur (too few tokens, tokens without enough
probability, etc,) but the DBs were already tied, we didn't untie
them.
2003-03-17 18:33 quinlan
* masses/mass-check: commented-out debugging code for rewritten
message output
2003-03-17 18:22 quinlan
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Conf.pm,
SpamAssassin/PerMsgStatus.pm: bug 1576: need safer way to render
attachments for broken clients
2003-03-17 15:24 felicity
* sa-learn.raw, spamassassin.raw, spamd/spamc.pod: bug 1605: SA
documentation ought to specify what "ham" is for people who don't
necessarily know it means non-spam.
2003-03-17 15:19 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Bayes.pm,
SpamAssassin/EvalTests.pm, SpamAssassin/PerMsgStatus.pm: fix for
bug 1591: Bayes is_available() didn't pay attention to use_bayes,
so it would still attempt a tie and possibly create DBs even if it
wasn't going to use the bayes stuff. renamed the function to
is_scan_available() since other modules call it to see if a scan
would be done (determining scoreset, etc.) cleaned up scan() a
little (let is_scan_available check for # of tokens)
2003-03-17 15:12 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: fix for bug 1610: after a
bayes expire, the current message scan count would return to 0.
this is because the magic token in the db (scancount_base) wasn't
being copied over. oops. also added checks for return values to
make sure we don't delete the journal unless we really should.
modified expiry_due to do an expire if the last expire time is >
current message scan. this should never happen in actual usage and
only appears if this bug has occured.
2003-03-17 07:25 quinlan
* debian/: README.Upgrade, changelog.in, copyright,
spamassassin.README.Debian, spamassassin.docs,
spamassassin.examples: bug 1643: backport Debian 2.50 changes to
2.50 branch
2003-03-17 04:35 quinlan
* lm/: LICENSE, README, af.lm, am.utf-8.lm, ar.iso-8859-6.lm,
ar.windows-1256.lm, be.windows-1251.lm, bg.iso-8859-5.lm, bs.lm,
build.pl, ca.lm, cs.iso-8859-2.lm, cy.lm, da.lm, de.lm,
el.iso-8859-7.lm, en.lm, eo.lm, es.lm, et.lm, eu.lm, fa.lm, fi.lm,
fr.lm, fy.lm, ga.lm, gd.lm, he.iso-8859-8.lm, hi.lm,
hr.us-ascii.lm, hu.lm, hy.lm, id.lm, is.lm, it.lm, ja.euc-jp.lm,
ja.iso-2022-jp.ln, ja.shift-jis.lm, ka.lm, ko.lm, la.lm, lt.lm,
lv.lm, mr.lm, ms.lm, ne.lm, nl.lm, no.lm, pl.lm, pt.lm, qu.lm,
rm.lm, ro.lm, ru.iso-8859-5.lm, ru.koi8-r.lm, ru.windows-1251.lm,
sa.lm, sco.lm, sk.us-ascii.lm, sk.windows-1250.lm,
sl.iso-8859-2.lm, sl.us-ascii.lm, sq.lm, sr.us-ascii.lm, sv.lm,
sw.lm, ta.lm, th.lm, tl.lm, tr.iso-8859-9.ln, tr.unknown.lm,
uk.koi8-r.lm, vi.lm, yi.utf-8.lm, zh.big5.lm, zh.gb2312.lm,
inactive/README, inactive/br.lm, inactive/gv.lm, inactive/haw.lm,
inactive/middle-frisian.lm, inactive/mingo.lm: add language model
source code
2003-03-16 20:09 felicity
* lib/Mail/SpamAssassin/Bayes.pm: don't try untieing if we're not
already tied. ;)
2003-03-16 19:32 felicity
* lib/Mail/SpamAssassin/Bayes.pm: We only untie if the scan()
completed successfully, not if there's an error (no tokens, etc.)
2003-03-16 04:04 quinlan
* rules/70_cvs_rules_under_test.cf: add some URL rules for some
low-scoring spam in my corpus
2003-03-15 21:40 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: AnyDBM_File use in EvalTests
removed; DB_File used directly instead
2003-03-15 18:29 msquadrat
* lib/Mail/SpamAssassin/Dns.pm: bug 1641: don't load razor if it is
not available
2003-03-15 18:24 msquadrat
* t/: SATest.pm, db_awl_path.t, reportheader.t, whitelist_to.t: bug
1639: permission issue in testbench
2003-03-15 18:19 msquadrat
* lib/Mail/SpamAssassin/UnixLocker.pm: bug 1635: warn when removing
a stale lock file fails
2003-03-15 18:16 msquadrat
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1625: problem with
sync_journal() when journal is out-of-date
2003-03-15 16:41 msquadrat
* lib/Mail/SpamAssassin/CmdLearn.pm: bug 1545: possible warning
about undefined veriable in sa-learn
2003-03-15 16:33 msquadrat
* sa-learn.raw, lib/Mail/SpamAssassin/CmdLearn.pm: bug 1578: fix
problem with --single in sa-learn
2003-03-15 16:23 msquadrat
* rules/30_text_de.cf: bug 1568: small type in German translation
2003-03-15 16:19 msquadrat
* sa-learn.raw, lib/Mail/SpamAssassin/BayesStore.pm: bug 1545:
extended docu for sa-learn
2003-03-15 16:10 msquadrat
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm, UnixLocker.pm,
Win32Locker.pm: bug 1545: locking problems in combination with
GDBM_File
2003-03-15 16:03 msquadrat
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1513: permission issues
with bayes_msgcount
2003-03-15 15:56 msquadrat
* lib/Mail/SpamAssassin/Conf.pm: bug 1528: wrong clear statement in
unsafe_report docu
2003-03-15 15:50 msquadrat
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1570: tab instead of
space in X-Spam-Status header
2003-03-14 19:39 jmason
* sa-learn.raw, lib/Mail/SpamAssassin/CmdLearn.pm,
lib/Mail/SpamAssassin/Dns.pm, lib/Mail/SpamAssassin/EvalTests.pm,
spamd/spamd.raw, t/SATest.pm, t/db_awl_path.t, t/reportheader.t,
t/whitelist_to.t: spamd backwards compat broken for SYMBOLS method,
fixed; razor modules were loaded for -L; AnyDBM_File preferences
not stated in EvalTests; test scripts using HOME/.spamassassin dbs;
minor sa-learn doco patch
2003-03-14 18:49 quinlan
* lib/Mail/SpamAssassin/Conf.pm: bug 1554: documentation bug for
ok_locales
2003-03-14 18:48 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: fix bug 1546: use our
local_tz() function instead of non-standard strftime %z extension
to get local timezone
2003-03-14 13:24 felicity
* lib/Mail/SpamAssassin/UnixLocker.pm: should warn if we can't
unlink
2003-03-14 03:59 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: fix for bug 1611
2003-03-14 03:34 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: fix for bug 1544
2003-03-14 03:23 felicity
* spamd/spamd.raw: fix for bug 1604
2003-03-14 03:10 felicity
* lib/Mail/SpamAssassin/EncappedMessage.pm,
masses/rewrite-cf-with-new-scores, masses/score-ranges-from-freqs:
bug 1558
2003-03-14 02:59 felicity
* spamassassin.spec: bug 1593, and updated spec file to 2.51
version
2003-03-14 02:50 felicity
* lib/Mail/SpamAssassin/Util.pm: fix for bug 1559
2003-03-13 22:26 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: add T_MIME_BASE64_LATIN3
2003-03-13 18:04 felicity
* rules/: 20_head_tests.cf, 50_scores.cf: Added the MAPS
nonconfirmed mail list bl. default score of 0 as usual.
2003-03-13 03:20 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1626. if via a new
expire run, a token has an atime > current scan count, a previous
db error is likely to have been the cause. So let's reset the
atime to the last expire time before processing.
2003-03-13 03:04 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: bug 1629: all we really care
about are token updates. since the likelihood is that we'll be
updating the same tokens multiple times, we can save I/O by only
updating the token with the final value.
2003-03-12 16:54 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: fix for bug 1625. we only
touch the token when syncing the journal, and the journal only has
updates from scan() whereas tokens are automatically updated from
learn(), so we will likely end up lowering the atime of the token
when we sync. doh. :(
2003-03-11 18:36 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: tweak UNDESIRED_LANGUAGE_BODY to
reduce false positives (sco looks very similar to en) add some
additional charsets to T_MIME_BASE64_LATIN add T_MIME_BASE64_LATIN2
variation to test
2003-03-11 05:05 quinlan
* rules/70_cvs_rules_under_test.cf: some good unique ID tests a
better stormpost ratware rule (works for more recent versions, I
think)
2003-03-10 06:34 quinlan
* rules/70_cvs_rules_under_test.cf: some generic rules
2003-03-10 02:49 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: add T_MIME_QP control rule add
T_MIME_BASE64_LATIN rule (should be better than
T_MIME_BASE64_ISO_8859) add T_MIME_QP_INLINE_NO_CHARSET (another
product of permutation testing) remove T_MIME_BASE64_NO_BLANK (no
hits for current version, other versions have too many false
positives)
2003-03-10 01:07 felicity
* lib/Mail/SpamAssassin/: Conf.pm, EvalTests.pm: Conf cleanup for
whitelist -- the eval test does the lc, so we should do the lc when
we add otherwise the two may not be equal. Eval cleanup for
accessdb lookup.
2003-03-10 00:17 felicity
* masses/score-ranges-from-freqs: don't let negative rules get so
large. limit to -4.5.
2003-03-09 19:30 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: add T_MIME_BASE64 control test
(for baseline comparison with BASE64 tests since BASE64 is more
common in spam)
2003-03-09 16:35 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: patch to
backwards-compatibly fix bug 1610 databases (all of them). Current
scan count should never be less than last expire count unless the
bug from 1610 hits.
2003-03-09 16:25 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: more patching for bug 1610
2003-03-08 17:57 felicity
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: Code to check an access database
(sendmail and postfix at least) Since it's userconf and requires
that you have a database to look at, I force the score to 0.
2003-03-08 16:18 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: Patch for bug 1610
2003-03-08 09:02 quinlan
* rules/: 20_head_tests.cf, 70_cvs_rules_under_test.cf: bug 1587:
update ADVERT_CODE2 to not hit when ADVERT_CODE does
2003-03-08 08:52 quinlan
* rules/70_cvs_rules_under_test.cf: add additional T_MIME_BASE64_*
tests add comment for repeated To/Cc/Bcc header tests
2003-03-08 08:51 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: add additional
T_MIME_BASE64_* tests
2003-03-07 23:16 felicity
* spamassassin.spec: Added more files to be installed as docs
2003-03-07 17:34 quinlan
* rules/: 10_misc.cf, 20_anti_ratware.cf, 20_body_tests.cf,
20_compensate.cf, 20_head_tests.cf, 20_html_tests.cf,
20_meta_tests.cf, 20_phrases.cf, 20_porn.cf, 20_ratware.cf,
20_uri_tests.cf, 23_bayes.cf, 25_body_tests_es.cf,
25_body_tests_pl.cf, 25_head_tests_es.cf, 25_head_tests_pl.cf,
30_text_de.cf, 30_text_es.cf, 30_text_fr.cf, 30_text_it.cf,
30_text_pl.cf, 50_scores.cf, 60_whitelist.cf, 70_broken_rules.cf,
70_cvs_rules_under_test.cf, regression_tests.cf: include license
text at top of file (in addition to reference to "License")
2003-03-07 12:51 climent
* rules/25_body_tests_es.cf: New tweaks
2003-03-07 06:03 quinlan
* masses/mass-check: bug 1596: make progress bar optional in
mass-check
2003-03-07 06:01 quinlan
* rules/70_cvs_rules_under_test.cf: bug 1261: add T_BASE64_ENC_TEXT
new T_MIME_BASE64_* tests possible replacements for ADVERT_CODE and
ADVERT_CODE2
2003-03-07 06:00 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: tweaks to
T_MSG_ID_ADDED_BY_MTA_* new T_MIME_BASE64_* tests always test MIME
headers in message header (big improvement for
MIME_HTML_NO_CHARSET)
2003-03-06 23:01 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: standardize the "incorrect
data, not just a number" check with the others. :)
2003-03-06 22:34 felicity
* lib/Mail/SpamAssassin/: BayesStore.pm, CmdLearn.pm: Fix for bug
1611, we ought to update the token count. ;) I also noticed that
sa-learn wasn't taking an argument for -D so you couldn't get
detailed debug info. :(
2003-03-06 02:37 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: we should return if we can't
open instead of just warning and continuing
2003-03-05 21:43 felicity
* rules/70_cvs_rules_under_test.cf: 2 new rules. HTML mail should
have a HTML and BODY tag.
2003-03-05 18:48 felicity
* spamd/spamd.raw: fix for bug 1604. do a setgid() when we do the
setuid().
2003-03-05 18:22 felicity
* sa-learn.raw, spamassassin.raw, spamd/spamc.pod: some small doc
changes for bug 1605. we should use "ham" everywhere, and we
should specify in each doc file that "ham means non-spam".
2003-03-04 19:11 felicity
* rules/70_cvs_rules_under_test.cf: added a check for
UNUSABLE_MSGID in the T_FORGED_MUA rules
2003-03-04 16:11 felicity
* spamd/spamd.raw: Cleaned up the documentation about --username a
little to make it (hopefully) very obvious that setting
"--username" disables the setuid() functionality in the code, even
if the specified username is "root".
2003-03-04 09:15 quinlan
* lib/Mail/SpamAssassin/Util.pm: use a hash for month mappings to
avoid repeating work
2003-03-04 08:37 quinlan
* rules/70_cvs_rules_under_test.cf: test some faster revisions for
six slow rules add some rawbody T_REMOVE_REMOVAL_* variants bug
821: test for "good" RFC2369 headers (final attempt at this idea,
we hope)
2003-03-04 04:05 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, PerMsgStatus.pm: Updated a few
debug statements
2003-03-03 22:21 felicity
* spamassassin.spec: fix for bug 1593 apparently sometimes
MakeMaker and RPM have a disagreement about where they think the
man pages should be, and our previous kluge for INSTALLMAN1DIR
wasn't actually doing anything since "make install" uses
INSTALLSITEMAN1DIR. So this patch forces all 3 different man paths
to be the same as RPM is expecting.
2003-03-03 19:58 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Bayes.pm,
SpamAssassin/EvalTests.pm, SpamAssassin/PerMsgStatus.pm: Fix for
bug 1591. renamed Bayes::is_available to Bayes::is_scan_available
since it's specifically geared towards scan() (can I tie the DB,
and are there enough tokens?) cleaned up scan() a little to just
call is_scan_available (checked to tie the DB, enough tokens, etc.)
2003-03-03 19:07 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Don't run eval rules from
the wrong scoreset. ie: don't run tflag learn rules unless
scoreset&2 is set, and don't run tflag net rules unless scoreset&1
is set. This way we don't need each eval code to check for
local_tests_only and the like.
2003-03-03 18:47 jmason
* rules/70_cvs_rules_under_test.cf: some more T_ rules for a new
spamware tool
2003-03-03 15:51 quinlan
* masses/mass-check: add comment
2003-03-03 14:20 felicity
* rules/70_cvs_rules_under_test.cf: Different versions of
T_HTML_COMMENT_RATIO rules.
2003-03-03 00:43 felicity
* lib/Mail/SpamAssassin/HTML.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm,
rules/70_cvs_rules_under_test.cf: per bug 1373, added check for
excessive comments in HTML only messages
2003-03-02 23:59 felicity
* rules/: 20_uri_tests.cf, 30_text_fr.cf, 30_text_it.cf,
50_scores.cf, 70_cvs_rules_under_test.cf: per bug 1424, removed
HTTP_USERNAME_USED, replaced USERPASS with new version
2003-03-02 23:44 jmason
* sa-learn.raw, lib/Mail/SpamAssassin/BayesStore.pm,
lib/Mail/SpamAssassin/CmdLearn.pm: fixes to bayes code; better
error messages; more doco; doco bug fixed in links to third-party
sites
2003-03-02 23:06 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: bayes tie/lock failure
messages now make more sense
2003-03-02 22:50 jmason
* rules/70_cvs_rules_under_test.cf: some tweaks to T_ rules based
on my FPs
2003-03-02 16:49 felicity
* rules/70_cvs_rules_under_test.cf: meta subrule got renamed test
rules should start with a T_
2003-03-02 13:20 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: add code for
T_NO_DNS_FOR_RECIPS_* to detect cases where one or more addresses
in recipient list do not have DNS records (meaning they are likely
to be old addresses in spammer databases) also add some code to
better support qmail for T_MSG_ID_ADDED_BY_MTA_*
2003-03-02 13:17 quinlan
* rules/70_cvs_rules_under_test.cf: add T_NO_DNS_FOR_RECIPS_* add
T_REMOVE_REMOVALS_*
2003-03-02 08:02 zelgadis
* lib/Mail/SpamAssassin.pm, masses/mass-check: Bug #1583:
mass-check: specify which rules to run on command line
2003-03-01 20:09 jmason
* lib/Mail/SpamAssassin/HTML.pm: some warning cleanups, very minor
2003-03-01 15:33 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: fixes some undef warnings in
Bayes code
2003-03-01 14:33 jmason
* sa-learn.raw, lib/Mail/SpamAssassin/CmdLearn.pm: bug 1578: 2
buglets from FreeBSD ports maintainer fixed
2003-03-01 06:17 felicity
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: added code to check for the ratio
of blank lines to overall lines in the message.
2003-03-01 06:13 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: the missing eval code
2003-03-01 03:55 quinlan
* rules/70_cvs_rules_under_test.cf: change a few rule names
2003-02-28 22:22 felicity
* lib/Mail/SpamAssassin/EvalTests.pm,
rules/70_cvs_rules_under_test.cf: test rules to look for RFC2822 "#
of header" violations, also look for multiple Content-Type and
Content-Transfer-Encoding headers which shouldn't exist in ham.
2003-02-28 20:33 felicity
* lib/Mail/SpamAssassin/HTML.pm, rules/70_cvs_rules_under_test.cf:
Added two new comment rules. Work well for me. :)
2003-02-28 19:01 jmason
* build/preprocessor.pm, lib/Mail/SpamAssassin/Bayes.pm,
lib/Mail/SpamAssassin/BayesStore.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1556: issues with Bayes
and MailScanner -- sa was killing MailScanner due to lack of bayes
untieing, it seems.
2003-02-28 18:34 jmason
* lib/Mail/SpamAssassin/: Bayes.pm, BayesStore.pm, UnixLocker.pm,
Win32Locker.pm: bug 1545: bayes dbs were corrupted if a lock
attempt failed, when using GDBM_File at least
2003-02-28 18:18 felicity
* rules/70_broken_rules.cf: bug 1573. it could hurt autolearning
for scores which got overwritten here.
2003-02-28 03:44 felicity
* rules/70_cvs_rules_under_test.cf: more anti-ratware forging stuff
2003-02-28 02:00 duncf
* rules/70_cvs_rules_under_test.cf: Added check for rot13 encoding
of email address in body. stupid spam mailers
2003-02-27 23:27 msquadrat
* lib/Mail/SpamAssassin/PerMsgStatus.pm: fix bug 1570: tab instead
of space in X-Spam-Status after autolearn (plus some code cleanup)
2003-02-27 21:41 duncf
* debian/: README.Upgrade, changelog.in, copyright,
spamassassin.README.Debian, spamassassin.docs,
spamassassin.examples, spamassassin.links: Update debian packaging
to correspond with the 2.50 release. (Also allows my CVS builds to
work again)
2003-02-27 21:14 felicity
* lib/Mail/SpamAssassin/Util.pm: fix for bug 1559. stupid array vs
scalar issue.
2003-02-27 20:19 felicity
* rules/70_cvs_rules_under_test.cf: more anti-forging work
2003-02-27 20:05 felicity
* lib/Mail/SpamAssassin/: Bayes.pm, Conf.pm: Made minimum ham/spam
corpus size for bayes a configurable item. As a side tangent, I
added 'conf' to Bayes so the individual subs don't need to look for
$self->{main}->{conf} ...
2003-02-27 14:49 felicity
* rules/70_cvs_rules_under_test.cf: more anti-forging work
2003-02-27 11:11 msquadrat
* rules/30_text_de.cf: 30_text_de.cf
2003-02-27 09:36 quinlan
* rules/70_cvs_rules_under_test.cf: add __DATE_TIME_INVALID and
T_INVALID_DATE_HMS_2 clean up HTML mail without stuff
2003-02-27 09:31 quinlan
* rules/: 20_html_tests.cf, 50_scores.cf: remove
HTML_FONT_COLOR_NOHASH
2003-02-27 09:30 quinlan
* lib/Mail/SpamAssassin/HTML.pm: delete HTML_FONT_COLOR_NOHASH fix
code
2003-02-27 09:29 quinlan
* masses/hit-frequencies: allow -a to show subrules
2003-02-27 05:37 duncf
* lib/Mail/SpamAssassin/Conf.pm: Ummm.... sorry, I guess I just
fixed this for Debian (This has been in my Debian source tree for a
while, and I guess I just never committed it to CVS)
2003-02-27 04:19 felicity
* rules/70_cvs_rules_under_test.cf: trying to strengthen the
anti-ratware rules more work to come hopefully. :)
2003-02-27 00:40 quinlan
* rules/70_cvs_rules_under_test.cf: add some more tests
2003-02-26 17:35 felicity
* rules/70_cvs_rules_under_test.cf: Look for just header ctype
instead of MIME_HTML_ONLY (is valid if there's a single HTML
attachment)
2003-02-26 05:23 zelgadis
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Bug #1551: Standards
violating comment obfuscation
2003-02-26 03:45 felicity
* rules/70_cvs_rules_under_test.cf: possible replacement for
USER_AGENT_APPLEMAIL, bug 1561
2003-02-26 00:42 zelgadis
* rules/70_cvs_rules_under_test.cf: Misnamed
T_HTTP_EXCESSIVE_ENTITIES
2003-02-25 22:22 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: replace_header() will now
put the new header in the same location with the same header case
as the original. :)
2003-02-25 21:47 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: ok, now this better be the
last time...
2003-02-25 21:02 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1424: slight
improvement to scheme-less URI regular expression bug 1546: fix
timezone in Win32 Received: header when report_safe is on
2003-02-25 20:04 felicity
* rules/70_cvs_rules_under_test.cf: no
check_for_mime_multipart_trick() in EvalTests. commenting out for
now.
2003-02-25 20:02 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: Stupid debugging comments.
2003-02-25 19:58 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: hopefully the last patch
for bug 1544.
2003-02-25 19:39 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: More efficient version
2003-02-25 19:32 felicity
* masses/: rewrite-cf-with-new-scores, score-ranges-from-freqs:
added a comment about the spaces
2003-02-25 19:28 felicity
* lib/Mail/SpamAssassin/EncappedMessage.pm,
masses/rewrite-cf-with-new-scores, masses/score-ranges-from-freqs:
Kluge for bug 1558. It's really bugginess in RPM 4.1, but at the
moment it's easier to work around their issue then to explain to
all the users why it's RPM's fault, so that's why the SA RPM
doesn't build right.
2003-02-25 15:36 quinlan
* rules/70_cvs_rules_under_test.cf: add a few more
T_HTML_NO_DOCTYPE variants
2003-02-25 14:51 msquadrat
* Makefile.PL: simplify razor-check code (bug 1482)
2003-02-25 14:50 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: added an additional check
for headers_only
2003-02-25 14:10 quinlan
* lib/Mail/SpamAssassin/Conf.pm: fix ok_languages and ok_locales
examples
2003-02-25 09:27 zelgadis
* rules/70_cvs_rules_under_test.cf: Bug #1553: Character entity URL
obfuscation
2003-02-25 08:59 zelgadis
* rules/70_cvs_rules_under_test.cf: Bug Bug #1551: Standards
violating comment obfuscation
2003-02-25 04:52 quinlan
* rules/70_cvs_rules_under_test.cf: add variants to FAKED rule and
rename them all to be T_FAKED_*
2003-02-24 22:35 felicity
* lib/Mail/SpamAssassin/Conf.pm: small doc fix, per bug 1528
2003-02-24 22:32 felicity
* rules/70_cvs_rules_under_test.cf: added new test rules, cleaning
up the bug list some
2003-02-24 21:43 felicity
* INSTALL, MANIFEST, Makefile.PL, lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/Dns.pm, lib/Mail/SpamAssassin/EvalTests.pm,
lib/Mail/SpamAssassin/Reporter.pm, rules/20_body_tests.cf,
rules/30_text_de.cf, rules/30_text_es.cf, rules/30_text_fr.cf,
rules/30_text_it.cf, rules/30_text_pl.cf, rules/50_scores.cf,
t/.cvsignore, t/razor.t: Per bug 1482, removing Razor1 support from
the code. I think I got everything, but I wouldn't be surprised if
it's still in here somewhere.
2003-02-24 16:09 msquadrat
* lib/Mail/SpamAssassin/Conf.pm: backport fix for bug 1537: better
docu of dns_available
2003-02-24 08:40 msquadrat
* rules/70_cvs_rules_under_test.cf: Fixed __INVALID_DATE
2003-02-24 03:18 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: fix for bug 1544. add
option to _get_header_list to retrieve header names that match case
insensitively instead of just returning the values from all of
those same headers. instead of replace_header seeing if there are
versions that need to be removed, just call delete_header and it'll
figure it out. delete_header now removes all case insensitive
matches (via _get_header_list) of a header instead of the case
sensitive match. according to rfc2822, header names are case
insensitive so we should deal with that appropriately.
2003-02-24 00:08 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: When looking at Bayes in
scoreset, look for bit 2 instead of lt/gt/%/etc.
2003-02-23 23:46 felicity
* rules/70_cvs_rules_under_test.cf: one test should have been a
meta, not a header. __INVALID_DATE is missing a paren in the RE.
it's so complex, I didn't want to try figuring out where to put the
')'... disabled for now.
2003-02-23 20:47 felicity
* rules/70_cvs_rules_under_test.cf: enabled all new testing rules
2003-02-23 20:46 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Bayes.pm,
SpamAssassin/BayesStore.pm: during bayes expiry, we should call
sync() instead of the individual BayesStore functions. added new
"bayes" debug type. added some bayes debug statements around
expiry.
2003-02-23 19:58 felicity
* lib/Mail/SpamAssassin/Conf.pm: Added documentation per bug 1537
2003-02-23 14:03 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: fix typo
2003-02-23 12:17 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: add code needed for these
bugs: bug 1376: HTML mails without bug 1533: new
Received: header forgery tests using HELO bug 1518: fallback to
Bcc: for SORTED and SUSP RECIPS
2003-02-23 12:16 quinlan
* rules/70_cvs_rules_under_test.cf: bug 1376: HTML mails without
bug 1533: new Received: header forgery tests using HELO
2003-02-23 12:15 quinlan
* lib/Mail/SpamAssassin/: HTML.pm, PerMsgStatus.pm: parse HTML
declarations for bug 1376: HTML mails without
2003-02-23 10:27 climent
* rules/25_body_tests_es.cf: New "regalo" entry
2003-02-22 21:45 felicity
* lib/Mail/SpamAssassin.pm: head is now 2.60 development
2003-02-21 23:47 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: fix for bug 1513, do umask
instead of chmod
2003-02-20 19:20 jmason
* build/README: updated the (forgotten) checklist
2003-02-20 14:07 jmason
* Changes, lib/Mail/SpamAssassin.pm: 2.50 RELEASED
2003-02-20 13:12 jmason
* lib/Mail/SpamAssassin.pm: getting ready for 2.50 release
2003-02-20 13:10 jmason
* rules/20_uri_tests.cf: Dan's fix for bug 1501: WEIRD_PORT
misfiring on port 443 for https, which is the https port
2003-02-20 13:05 jmason
* BUGS, t/razor.t: updated with 2 new known bugs
2003-02-20 13:01 jmason
* BUGS, MANIFEST, MANIFEST.SKIP, MailAudit.patch,
masses/RUNME.TWOSTEP, masses/craig-evolve.c.ALTIVEC,
rules/40_spam_phrases.cf, t/reportheader.t, t/spamd_report.t,
t/data/nice/001, wordfreqs/.cvsignore, wordfreqs/README,
wordfreqs/RUNME, wordfreqs/RUNME.DOGMA, wordfreqs/RUNME.SONIC,
wordfreqs/join-and-settle-phrases, wordfreqs/mass-phrase-freq,
wordfreqs/mass-phrase-freq-for-jm, wordfreqs/settle-phrases:
removed more obsolete code from CVS and MANIFEST
2003-02-20 12:23 jmason
* sql/: README, spamassassin.sql: changed suggested username length
limit to 100 chars in SQL setup, so that virtual domains can work
easily
2003-02-20 11:22 quinlan
* MANIFEST: sort
2003-02-20 11:14 quinlan
* MANIFEST: adding stuff to MANIFEST
2003-02-20 10:55 msquadrat
* .cvsignore, masses/.cvsignore, qmail/.cvsignore, t/.cvsignore:
Updating cvsignore
2003-02-20 05:04 felicity
* .cvsignore: we merged the sa-learn* scripts...
2003-02-20 04:50 quinlan
* MANIFEST: remove duplicated triplets.txt file
2003-02-19 21:48 msquadrat
* Makefile.PL: remove the spamassassin.conf manpage again: it's a
nasty hack violating the FHS
2003-02-19 19:27 jmason
* rules/: 20_phrases.cf, 20_porn.cf, 25_body_tests_es.cf,
25_body_tests_pl.cf, 30_text_de.cf, 30_text_es.cf, 30_text_fr.cf,
30_text_it.cf, 30_text_pl.cf: old removed rule descriptions removed
from translation files
2003-02-19 19:01 jmason
* rules/20_uri_tests.cf: bug 1497: PORN_4 still fping, hopeuflly
fixed a little
2003-02-19 18:48 jmason
* INSTALL: noting that RAzor tests do not work on windows
2003-02-19 18:22 jmason
* rules/: STATISTICS-set1.txt, STATISTICS-set2.txt,
STATISTICS-set3.txt, STATISTICS.txt: removed command lines ;)
2003-02-19 18:05 felicity
* masses/mk-baseline-results: Updated slightly to deal with
scoresets better
2003-02-19 18:03 jmason
* README: patch from John Peacock to README
2003-02-19 17:58 felicity
* rules/: STATISTICS-set1.txt, STATISTICS-set2.txt,
STATISTICS-set3.txt, STATISTICS.txt: Fixed the stats files. Oops.
2003-02-19 15:17 felicity
* sa-learn.raw: fixed typo in docs
2003-02-19 13:06 climent
* rules/25_body_tests_es.cf: New tweaks, additions and corrections.
2003-02-19 06:22 felicity
* rules/: STATISTICS-set1.txt, STATISTICS-set2.txt,
STATISTICS-set3.txt, STATISTICS.txt: Statistics files for the 2.50
GA'ed scores, all score sets. :)
2003-02-19 01:14 msquadrat
* Makefile.PL: implement bug 1496: a (kludgy) spamassassin.conf man
page
2003-02-19 00:49 msquadrat
* Makefile.PL, t/razor.t: Make RUN_RAZOR[12]_TESTS less strict and
put it all into a nice clean block
2003-02-18 22:36 msquadrat
* Makefile.PL, build/preprocessor.pm: Bugger about
ExtUtils::MakeMaker < 5.45 at the right place
2003-02-18 22:33 quinlan
* MANIFEST: add t/rule_tests.t
2003-02-18 22:19 jmason
* rules/70_cvs_rules_under_test.cf: some 419 tests for later
2003-02-18 22:02 jmason
* rules/70_cvs_rules_under_test.cf: new nigerian scam phrase test
2003-02-18 21:35 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: We should only add
autolearn= if we actual did learning. If the message was already
seen, etc, let's not add the token.
2003-02-18 21:19 msquadrat
* build/preprocessor.pm: fix a problem with missing inheritance
from ExtUtils::Install
2003-02-18 20:16 jmason
* USAGE: added site-wide user-notification request
2003-02-18 17:54 felicity
* lib/Mail/SpamAssassin/CmdLearn.pm: Now that rebuild is an option,
not a parameter to the function, it's not in $opt{} anymore.
2003-02-18 17:47 felicity
* lib/Mail/SpamAssassin/BayesStore.pm: All of the bayes_* files
ought to have the same file permissions, so we need to do a chmod
after the open for msgcount and journal.
2003-02-18 15:55 jmason
* Makefile.PL: oops, broke Makefile.PL there
2003-02-18 15:39 jmason
* Makefile.PL: Mike Bell: suggested fix for win32 errors when
installing local.cf
2003-02-18 15:37 jmason
* lib/Mail/SpamAssassin/: PerMsgStatus.pm, Util.pm: another way to
do the getpwuid wrapper trick; should be safe on win32 for sure
2003-02-18 12:02 jmason
* masses/mk-baseline-results: allow caller to specify output using
redirection, so we can generate stats for each scoreset
2003-02-17 21:16 felicity
* sa-learn.raw: Changed POD docs a little, s/nonspam/ham/, added
--no-rebuild option to sample command.
2003-02-17 18:18 felicity
* INSTALL: Added note about requiring Sys::Syslog if you're using
spamd
2003-02-17 18:14 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: Fix for bug 1487. With the
Date header at the end of the message, the RE was grabbing the
blank lines as "folded"...
2003-02-17 10:54 msquadrat
* lib/Mail/SpamAssassin/ArchiveIterator.pm: fix bug 1491: glob only
leading tilde; escape white spaces
2003-02-17 10:20 felicity
* lib/Mail/SpamAssassin/Reporter.pm: I think this fixes bug 1492.
2003-02-17 08:15 felicity
* lib/Mail/SpamAssassin/Util.pm: $AM_TAINTED is always defined, we
care what the value is
2003-02-17 00:42 felicity
* sa-forget.raw, sa-learn-nonspam.raw, sa-learn-rebuild.raw,
sa-learn-spam.raw: All of these got merged into sa-learn.raw. :)
2003-02-17 00:23 felicity
* rules/: 20_html_tests.cf, 50_scores.cf: bug 1429:
HTML_IMAGE_AREA_09 was autodisabled by GA, and _10 gets very good
results, so rename _10 and lower the range to cover 09 as well.
2003-02-16 23:24 felicity
* MANIFEST: updated for the merged sa-learn script
2003-02-16 23:03 felicity
* rules/70_broken_rules.cf: added regression tests for rules that
were disabled
2003-02-16 23:02 felicity
* rules/regression_tests.cf: removed tests for rules that no longer
exist
2003-02-16 22:33 felicity
* t/rule_tests.t: Since we can test subrules now, we can't count on
hits but have to look at what rules got hit.
2003-02-16 22:11 msquadrat
* rules/20_ratware.cf: fix bug 1488: __OUTLOOK_MSGID too
restrictive
2003-02-16 22:10 felicity
* sa-learn.raw: This script is the now single tool to modify the
learning systems in SpamAssassin. Used to be sa-learn-nonspam, the
documentation was updated to cover all of the new functionality.
2003-02-16 22:07 felicity
* Makefile.PL, README: added support for merge of sa-learn-* and
sa-forget
2003-02-16 22:01 felicity
* lib/Mail/SpamAssassin/CmdLearn.pm: added support for merge of
sa-learn-* and sa-forget
2003-02-16 16:32 jmason
* lib/Mail/SpamAssassin/BayesStore.pm, rules/20_meta_tests.cf,
rules/regression_tests.cf: fixed some wierd warnings;
OBFUSCATING_COMMENT regression test still failing, needs to test
the __ version of the rule instead of the whole meta rule.
2003-02-16 07:39 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Headers that should be
added (X-Pyzor, etc.) weren't being added to the report_safe
header.
2003-02-16 07:31 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: If we call
delete_fulltext_tmpfile() we ought to forget the previous filename
that we used.
2003-02-16 05:46 felicity
* spamd/libspamc.c: debug statement outside of an ifdef. oops.
2003-02-16 03:58 felicity
* rules/: 20_html_tests.cf, 20_meta_tests.cf: fix for bug 1423.
OBFUSCATING_COMMENT looks for larger comments and is now a meta
with MIME_HTML_ONLY
2003-02-16 02:19 felicity
* rules/: 50_scores.cf, 70_broken_rules.cf: re-enabled 3 rules
which are needed for meta rules. left USER_AGENT_* rules disabled
since USER_AGENT is the meta rule which has the dependencies, and
it's disabled too.
2003-02-16 01:58 felicity
* rules/: 50_scores.cf, 70_broken_rules.cf: Lower any abs(score) <
0.01 to 0. Added entries to broken_rules to set those rules to
score 0.001 so that mass-check will still use them, but not the
released version.
2003-02-15 23:24 felicity
* rules/20_head_tests.cf: Changed description for NO_DNS_FOR_FROM
2003-02-15 23:20 felicity
* lib/Mail/SpamAssassin/EvalTests.pm, rules/20_head_tests.cf,
rules/30_text_de.cf, rules/30_text_es.cf, rules/30_text_fr.cf,
rules/30_text_it.cf, rules/30_text_pl.cf, rules/50_scores.cf: bug
1381, renamed NO_MX_FOR_FROM to NO_DNS_FOR_FROM since the code
checks MX and A records...
2003-02-15 15:02 msquadrat
* t/spamc.t: Port 9 (discard) is the correct one to use in this
test
2003-02-15 14:28 felicity
* rules/50_scores.cf: GA set 3 scores
2003-02-15 11:07 climent
* rules/25_body_tests_es.cf: Tweaks and new entries in old rules
2003-02-15 02:23 felicity
* rules/50_scores.cf: small mistakes in a few net test scores for
set 3 causing ga issues ...
2003-02-15 02:06 felicity
* rules/50_scores.cf: Set 2 scores put in
2003-02-14 22:20 felicity
* lib/Mail/SpamAssassin/: Dns.pm, Reporter.pm: Pyzor and DCC
options weren't being set correctly due to non-match of the RE to
untaint them.
2003-02-14 21:57 felicity
* lib/Mail/SpamAssassin/Util.pm: Fixed debug output for "Are we
tainted?"
2003-02-14 21:57 jmason
* tools/check_bayes_db: oops
2003-02-14 21:50 jmason
* lib/Mail/SpamAssassin/BayesStore.pm,
lib/Mail/SpamAssassin/Conf.pm, lib/Mail/SpamAssassin/HTML.pm,
tools/check_bayes_db: bug 1460: bayes_expiry_use_scan_count is now
the default, using a method that should not be much slower than
using the time. This now means that expiry can take place during a
mass-check, which is very important. Implementation uses a
separate file, bayes_msgcount, to track scan-count. This will not
require rescanning the bayes dbs.
2003-02-14 21:10 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Dns.pm: Fixed up the
dcc/pyzor debugging statements
2003-02-14 21:06 felicity
* lib/Mail/SpamAssassin/Dns.pm: Added a little more debugging info
around pyzor and dcc
2003-02-14 20:59 felicity
* README: Added some documentation about default file locations
2003-02-14 20:22 jmason
* spamassassin.raw, lib/Mail/SpamAssassin/Conf.pm: oops, using
'blib' meant that rules were not being found when run from build
dir. fixed
2003-02-14 19:48 felicity
* lib/Mail/SpamAssassin.pm: need to untaint user prefs file before
trying to write to it
2003-02-14 19:05 jmason
* t/spamc_c.t: oops, too many tests
2003-02-14 19:04 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin/Dns.pm,
SpamAssassin/Util.pm: bug 1479: getpwuid on Win32 does really wierd
things
2003-02-14 18:58 jmason
* lib/Mail/SpamAssassin/: Conf.pm, PerMsgStatus.pm: bug 1462: tag
rule failures with rule names where possible
2003-02-14 18:52 felicity
* lib/Mail/SpamAssassin/Util.pm: Added more debug output
2003-02-14 18:43 jmason
* lib/Mail/SpamAssassin/AutoWhitelist.pm,
lib/Mail/SpamAssassin/ConfSourceSQL.pm, spamd/spamd.raw: bug 1457:
added skeleton POD doc to private classes
2003-02-14 18:38 jmason
* MANIFEST, spamd/libspamc.c, t/spamc_c.t: bug 1481: spamc -c
broken, fixed by Eugene Miretsky (eugene .AT. invision .DOT. net);
added test too
2003-02-14 11:59 climent
* rules/25_body_tests_es.cf: Small addition
2003-02-13 16:16 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Added a section to
X-Spam-Status specifying that autolearning has taken place, and
whether it was ham or spam.
2003-02-13 13:21 jmason
* spamassassin.raw, lib/Mail/SpamAssassin.pm,
lib/Mail/SpamAssassin/Reporter.pm: did I forget to commit the
revoke patch? if so, here it is.
2003-02-12 23:05 jmason
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1414: unicode chars
still creeping in via HTML::Parser
2003-02-12 22:57 jmason
* spamd/libspamc.c, spamd/libspamc.h, spamd/spamc.c,
spamd/spamd.raw, t/SATest.pm: bug 1452: libspamc: message is always
reported as EX_TOOBIG in m.is_spam. This requires spamc protocol
version 1.3, which can now support multiple headers in reply from
spamd.
2003-02-12 21:48 felicity
* spamassassin.raw, t/SATest.pm: fix for bug 1464. makes
spamassassin look for a "blib/lib", and if it exists, use it
exclusively.
2003-02-12 21:11 felicity
* rules/20_head_tests.cf: missing end periods on 3 tests.
2003-02-12 18:10 jmason
* build/: update_devel, update_stable: don't add development doco
to website every night
2003-02-12 17:41 felicity
* rules/50_scores.cf: Added some comments to the scores file.
Organized the network/etc area a little bit more.
2003-02-12 15:57 felicity
* lib/Mail/SpamAssassin/Dns.pm: chomp the response to dcc and
pyzor. don't need the newline and makes the debug statements look
better. :)
2003-02-11 19:44 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Modified the debug
statements around auto-learning. Show the thresholds, etc.
2003-02-11 16:59 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Fix for bug 1470, modified
version for bug 1417. Basically HTML parsing was broken due to
forcing byte mode during the parse. pack was called with bad
parameters which nullified the text. doh.
2003-02-11 16:29 felicity
* rules/50_scores.cf: 1 rule got disabled in set1 that's a meta
dependency. re-enabled it...
2003-02-11 16:26 felicity
* rules/50_scores.cf: A few more minor score cleanups
2003-02-11 16:16 felicity
* rules/50_scores.cf: old version was fscked up (only 607
scores!?!) this one has all 844.
2003-02-11 13:13 jmason
* t/SATest.pm: bug 1464: use 'perl -I../blib' to run test commands
2003-02-11 13:09 jmason
* rules/: 70_cvs_rules_under_test.cf, 70_cvs_rules_under_test.cf:
checked in possible replacement T_PENIS_ENLARGE rules for test
after release
2003-02-11 13:07 jmason
* rules/10_misc.cf, rules/20_anti_ratware.cf,
rules/20_body_tests.cf, rules/20_compensate.cf,
rules/20_head_tests.cf, rules/20_html_tests.cf,
rules/20_meta_tests.cf, rules/20_phrases.cf, rules/20_porn.cf,
rules/20_ratware.cf, rules/20_uri_tests.cf, rules/23_bayes.cf,
rules/25_body_tests_es.cf, rules/25_body_tests_pl.cf,
rules/25_head_tests_es.cf, rules/25_head_tests_pl.cf,
rules/30_text_de.cf, rules/30_text_es.cf, rules/30_text_fr.cf,
rules/30_text_it.cf, rules/30_text_pl.cf, rules/50_scores.cf,
rules/60_whitelist.cf, rules/70_broken_rules.cf,
rules/70_cvs_rules_under_test.cf, rules/regression_tests.cf,
t/data/spam/009: added more spammy phrases to utf8 source message.
But all those rules have disappeared too. ;) something's up with
the scores file for sure...
2003-02-11 10:59 jmason
* sa-learn-nonspam.raw: oops, documentation was out of date on
auto_learn
2003-02-10 23:17 msquadrat
* Makefile.PL: Don't require ExtUtils::MakeMaker >= 5.45; we have
compatibility code in build/preprocessor.pm
2003-02-10 23:01 msquadrat
* lib/Mail/SpamAssassin/Conf.pm: Backported fix for Bug 1120:
dcc_path docu mixup (patch 1.115 -> 1.116)
2003-02-10 21:50 msquadrat
* t/razor2.t: Backport fix for Bug 1117: Razor tests ignore
enabled/disabled flag
2003-02-10 21:49 msquadrat
* t/razor.t: Backport fix for Bug 1117: Razor tests ignore
enabled/disabled flag (patch 1.5 -> 1.6)
2003-02-10 21:26 msquadrat
* lib/Mail/SpamAssassin/Reporter.pm: Backport workaround for Razor2
disabling the timeout (patch 1.39 -> 1.40)
2003-02-10 18:04 jmason
* lib/Mail/SpamAssassin/PerMsgStatus.pm: test_logs were persisting
from one line to the next. fixed
2003-02-10 11:20 jmason
* lib/Mail/SpamAssassin/Conf.pm, rules/70_cvs_rules_under_test.cf,
t/data/nice/cjk/gb2312.2: documented always-add-report 'this mail
is spam' message when the mail is not spam. Also added another
good gb2312 mail to t/data/nice/cjk
2003-02-10 00:18 felicity
* lib/Mail/SpamAssassin/Dns.pm: If we get a response, we're not
skipping it, are we?
2003-02-10 00:09 felicity
* lib/Mail/SpamAssassin/Dns.pm: Another stupid RAZOR2_CF_* bug ...
If the first part didn't get a response (like when it's blank and
Razor says "skipme=1"), the other parts aren't checked for a
response. Fixed now. Doh. :)
2003-02-09 15:03 jmason
* lib/Mail/SpamAssassin/Conf.pm: bayes_use_chi2_combining POD fixed
to reflect default of 0
2003-02-09 14:46 jmason
* lib/Mail/SpamAssassin/Reporter.pm: tidying up reporter undef
value fix to match Dns.pm
2003-02-09 01:29 felicity
* lib/Mail/SpamAssassin/Reporter.pm: possible undefined warning?
dunno, it was added to dns ...
2003-02-09 01:17 felicity
* lib/Mail/: SpamAssassin.pm, SpamAssassin/CmdLearn.pm: fixes for
1456. makes bayes_path work properly. basically the scoreset was
chosen before the configs were read, so if the bayes stuff doesn't
live in the user's directory it'll only choose sets 0 or 1, then
bayes never runs.
2003-02-08 16:38 jmason
* lib/Mail/SpamAssassin/Dns.pm: poss fix for that undefined warning
in Dns.pm line 693
2003-02-08 16:21 jmason
* lib/Mail/SpamAssassin/PerMsgStatus.pm: always de-unicode strings
going into HTML::Parser, since recent versions do their own Unicode
stuff
2003-02-08 15:44 jmason
* USAGE: added UPPERCASE rules to set disabled for CJK users.
added more doco on CJK rules and spamtraps
2003-02-08 15:31 jmason
* rules/user_prefs.template: added UPPERCASE rules to set disabled
for CJK users
2003-02-08 15:30 jmason
* t/data/nice/cjk/gb2312.1: the easy way around the 8-bit-FP
problem, for now, without redoing GA or mass-check runs; tell the
users to disable the FPing rules. Also a collection of known-ham
in CJK charsets
2003-02-08 15:26 jmason
* INSTALL, MANIFEST.SKIP, rules/user_prefs.template,
t/data/nice/cjk/big5-base64.1, t/data/nice/cjk/gb2312.1,
t/data/nice/cjk/iso-2022-jp.1, t/data/nice/cjk/iso-2022-jp.2: the
easy way around the 8-bit-FP problem, for now, without redoing GA
or mass-check runs; tell the users to disable the FPing rules
2003-02-07 20:25 jmason
* masses/: CORPUS_SUBMIT, CORPUS_SUBMIT_NIGHTLY: removed -u (update
only, don't overwrite newer files) from rsync flags in
corpus-submit doco, since it's just caused me some confusion at
least ;)
2003-02-07 19:20 felicity
* lib/Mail/SpamAssassin/Reporter.pm: Need to clean up after the tmp
file ...
2003-02-07 19:15 felicity
* lib/Mail/SpamAssassin/: Dns.pm, Reporter.pm: cleaned up a few
things in dns. cleaned up a _lot_ of stuff in reporter. while
working on a fix for bug 1447 I found that it was just easier to
clean things up and fix it at the same time. reporter now calls
some code directly in dns to avoid the "one gets updated without
the other one" problem we keep having... Report and check are
still seperate, but the *_available code is now the same. Reporter
also does helper_mode now (which fixed the tainted path issue).
I've tested against razor and dcc and they seem to work fine. I
don't have pyzor installed, but since dcc works I don't see why
pyzor wouldn't ...
2003-02-06 21:35 jmason
* rules/70_cvs_rules_under_test.cf: some more new rules for when
CVS reopens
2003-02-06 18:44 felicity
* masses/mass-check: nullified --bayes option (still accepted, just
doesn't do anything)
2003-02-06 16:43 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Doh. Like an idiot I put
a good fix in, tested it, then changed the code and submitted
(decided to do $set&2 instead of $set<2) and didn't pay attention.
So now I reverted to the code that did work, tested it, and am now
committing it. Somebody please slap me around for doing this ...
2003-02-06 16:25 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Fixes bug 1451. If Bayes
isn't enabled, PerMsgStatus::check() will now look to see if we can
enable it (ie: change from scoreset {0,1} => {2,3}) and do it if
possible.
2003-02-05 19:21 jmason
* rules/70_cvs_rules_under_test.cf: some rules noted for
uncommenting after release
2003-02-04 20:23 climent
* rules/25_body_tests_es.cf: vim tweak
2003-02-04 11:43 jmason
* rules/60_whitelist.cf: bug 1348: some default whitelists
2003-02-04 10:43 msquadrat
* lib/Mail/SpamAssassin/Dns.pm: Backport workaround for Razor2
disabling the timeout (patch 1.67 -> 1.68)
2003-02-04 10:21 msquadrat
* spamd/libspamc.c: Fixed Bug 1442: Segfault when connection to
spamd fails
2003-02-04 06:23 felicity
* masses/craig-evolve.c: craig-evolve code used to generate sets 0
and 1 for 2.50 applies allen's patch to give score limits
2003-02-04 05:42 felicity
* rules/50_scores.cf: USER_AGENT_MOZILLA_XM was autodisabled by the
GA but is a dependency of USER_AGENT meta rule re-enabled with
-0.001 score
2003-02-04 03:05 felicity
* rules/50_scores.cf: GA generated scores for sets 0 and 1
2003-02-04 02:09 felicity
* rules/20_phrases.cf: Oops, some of my tflags made it in here
somehow ... :(
2003-02-04 02:05 felicity
* rules/: 20_head_tests.cf, 20_phrases.cf, 20_ratware.cf: Updated
to remove rules that went in broken rules
2003-02-03 19:50 felicity
* lib/Mail/SpamAssassin/NoMailAudit.pm: fixed bug 1439: blank
header would cause rewrite_as_spam to see next header in message...
2003-02-03 16:59 duncf
* debian/: changelog.in, control, rules, spamassassin.config: Bump
version number. Misc fixes
2003-02-03 14:22 felicity
* rules/50_scores.cf: New set0 scores w/out BODY_8BITS
2003-02-03 14:15 jmason
* Makefile.PL: fixed 'spamc not being rebuilt after changes'
dependency problem for DuncF
2003-02-03 12:54 jmason
* lib/Mail/SpamAssassin/PerMsgStatus.pm: always add
X-Spam-Checker-Version header for all mails (bug 1342)
2003-02-03 12:42 jmason
* lib/Mail/SpamAssassin.pm: bug 1392: languages_filename cannot be
set if using config_text. now fixed
2003-02-03 12:37 jmason
* lib/Mail/SpamAssassin.pm: added main->{username} attribute so SA
code can usefully figure out the current username
2003-02-03 07:23 quinlan
* lib/Mail/SpamAssassin/: UnixLocker.pm, Util.pm, Win32Locker.pm:
replace Win32 mkdir-based locking with sysopen-based locking add
some debugging, comments, and warnings to locking code
2003-02-03 07:22 quinlan
* rules/20_body_tests.cf: fix bug 1428: BODY_8BITS: will hit lots
of Asian ham
2003-02-03 07:22 quinlan
* lib/Mail/SpamAssassin/PerMsgStatus.pm: fix bug 1371: some tests
don't work on CRLF line ends
2003-02-03 07:21 quinlan
* lib/Mail/SpamAssassin/EvalTests.pm: fix bug 1371: some tests
don't work on CRLF line ends fix bug 1428: BODY_8BITS: will hit
lots of Asian ham
2003-02-03 07:15 quinlan
* lib/Mail/SpamAssassin/Dns.pm: make a dbg() statement a bit
clearer
2003-02-03 04:35 felicity
* spamassassin.spec, spamd/redhat-rc-script.sh: add support for
"condrestart" for redhat installs
2003-02-02 19:29 jmason
* masses/corpora/mass-find-nonspam: bug 1432: mass-find-nonspam was
broken. fixed
2003-02-02 19:21 jmason
* sa-forget.raw, sa-learn-nonspam.raw, sa-learn-spam.raw,
lib/Mail/SpamAssassin/CmdLearn.pm: bug 1433: added --local switch
to learner UIs, for future use
2003-02-02 19:07 felicity
* masses/score-ranges-from-freqs, rules/50_scores.cf: Set 0 scores
from GA run My version of score-ranges for said scores
2003-02-02 15:30 jmason
* spamd/libspamc.c: removed connect() debugging from libspamc, bug
seems fixed now
2003-02-02 00:09 msquadrat
* lib/Mail/SpamAssassin/Conf.pm: A more sophisticated language
handling
2003-02-01 20:01 felicity
* masses/logs-to-c: fixed bug where range_lo == range_hi means
score==1 ... doh!
2003-02-01 17:06 jmason
* sa-learn-nonspam.raw, lib/Mail/SpamAssassin/ArchiveIterator.pm,
masses/mass-check: bug 1391: globbing in mass-check (and
sa-learn-whatever)
2003-02-01 16:57 jmason
* spamd/spamd.raw: bug 773: more flexible --virtual-config switch,
--virtual-config-dir. bit more extension
2003-02-01 16:52 jmason
* MANIFEST, spamd/spamd.raw: bug 773: more flexible
--virtual-config switch, --virtual-config-dir
2003-02-01 16:21 jmason
* spamd/libspamc.c, spamd/libspamc.h, spamd/spamc.c,
spamd/spamc.pod, spamd/spamd.raw, t/SATest.pm, t/spamd_report.t,
t/spamd_report_ifspam.t, t/spamd_symbols.t: added support for
REPORT, REPORT_IFSPAM, and SYMBOLS methods in spamc: bug 1393, bug
191. Thanks to Eugene Miretsky (eugene at invision dot net) for
the patch.
2003-02-01 15:45 felicity
* lib/Mail/SpamAssassin/Conf.pm: Updated docs for auto-training a
little bit
2003-01-31 22:57 jmason
* rules/70_cvs_rules_under_test.cf: added commented
T_MSGID_SPAMSIGN_99X9XX99 for later use
2003-01-31 22:40 jmason
* lib/Mail/SpamAssassin/BayesStore.pm,
lib/Mail/SpamAssassin/Conf.pm, spamd/spamd.raw: verified that
per-user bayes dbs work with spamd, fixed some bugs and some doco
bugs as a result.
2003-01-31 22:11 jmason
* lib/Mail/SpamAssassin/Conf.pm: updated 'meta' doco
2003-01-31 22:07 jmason
* lib/Mail/SpamAssassin/Conf.pm: update bayes_path doco
2003-01-31 15:35 jmason
* lib/Mail/SpamAssassin.pm: 2.44 RELEASED. starting
hopefully-not-required 2.45 cycle
2003-01-31 15:14 jmason
* Changes, build/update_devel, build/update_stable: preparing build
scripts for 2.44 release
2003-01-30 17:45 felicity
* lib/Mail/SpamAssassin.pm: Now we can either comment out or set to
0 for release. Hopefully stop the confusion that keeps happening.
:)
2003-01-30 17:40 felicity
* masses/logs-to-c: [no log message]
2003-01-30 16:15 msquadrat
* lib/Mail/SpamAssassin.pm: A more obvious way to disable the DEVEL
flag ;-)
2003-01-30 15:41 felicity
* rules/50_scores.cf: Reset rules to pre-2.50 GA run.
2003-01-30 15:36 felicity
* rules/24_auto_meta.cf: Per discussion with other developers,
auto_meta rules will wait for 2.60.
2003-01-30 13:09 msquadrat
* lib/Mail/SpamAssassin.pm: Prepare for release
2003-01-30 13:02 msquadrat
* spamd/spamd.raw: Backport workaround from Bug 526: Failed sanity
check because of clobbered STDOUT (patch 1.147 -> 1.148)
2003-01-30 12:55 msquadrat
* spamd/libspamc.c: Backport fix for Bug 1306: Possible buffer
overflow in libspamc when running in BSMTP mode (patch 1.15 ->
1.18)
2003-01-30 12:29 msquadrat
* spamd/libspamc.c: Backport fix for Debian Bug 160206:
Insufficient buffer in libspamc (patch 1.8 -> 1.9)
2003-01-30 12:23 msquadrat
* lib/Mail/SpamAssassin.pm: Backport fix for warinings in sed_path
(patch 1.141 -> 1.142)
2003-01-30 12:20 msquadrat
* lib/Mail/SpamAssassin/NoMailAudit.pm: Backport fix for Bug 1127:
Existing lowercase x-spam-status header kills SpamAssassin (patch
1.40 -> 1.41)
2003-01-29 21:08 felicity
* rules/50_scores.cf: Made sure that rules used by meta rules are
not disabled
2003-01-29 20:08 felicity
* rules/: 24_auto_meta.cf, 50_scores.cf: Updated score sets 0 and 1
per Allen's results. Added 24_auto_meta for the auto generated
meta rules.
2003-01-29 17:56 jmason
* lib/Mail/SpamAssassin/: BayesStore.pm, CmdLearn.pm: bug 1326:
sa-learn-rebuild now shows dots
2003-01-29 12:12 jmason
* lib/Mail/SpamAssassin/EvalTests.pm: bug 1427: SUBJ_FULL_OF_8BITS
should use raw, non-decoded header
2003-01-28 23:46 jmason
* lib/Mail/SpamAssassin/: Conf.pm, ConfSourceSQL.pm: added patch
from Michael 'Moose' Dinn, dinn at twistedpair.ca, to make SQL
preferences work as part of Horde's SQL preferences.
2003-01-28 23:37 jmason
* lib/Mail/SpamAssassin/Conf.pm: removed obsolete unnamed-rule
support, which was never used
2003-01-27 21:32 jmason
* wordfreqs/README: note for wordfreqs README
2003-01-27 21:31 jmason
* MANIFEST: Malte, you forgot the spamc test's data file ;)
2003-01-27 21:12 jmason
* MANIFEST, lib/Mail/SpamAssassin.pm,
lib/Mail/SpamAssassin/BayesStore.pm,
lib/Mail/SpamAssassin/DBBasedAddrList.pm,
lib/Mail/SpamAssassin/Locker.pm,
lib/Mail/SpamAssassin/UnixLocker.pm, lib/Mail/SpamAssassin/Util.pm,
lib/Mail/SpamAssassin/Win32Locker.pm: abstracted UNIX-specific
locking code into new UnixLocker class; implemented Dan's win32
code as Win32Locker, loaded on win32 only: bug 1246
2003-01-27 20:19 jmason
* spamd/libspamc.h: bug 1330: spamc license fixed
2003-01-27 18:46 jmason
* lib/Mail/SpamAssassin/Conf.pm: finally did away with all that
[-_] crap
2003-01-27 18:30 jmason
* lib/Mail/SpamAssassin/: Conf.pm, Dns.pm: pyzor_options conf item
added, thanks to sabat
2003-01-27 17:56 jmason
* lib/Mail/SpamAssassin/Bayes.pm: oops, spotted typo in
IGNORED_HDRS
2003-01-27 10:21 msquadrat
* t/: SATest.pm, spamc.t: Fix some warnings in spamc test
2003-01-26 17:02 jmason
* lib/Mail/SpamAssassin/PerMsgStatus.pm: bug 1398: cannot use $-[0]
on perl 5.005
2003-01-26 16:54 jmason
* spamd/: libspamc.c, spamc.c, utils.c: bug 1330: spamc license
fixup, spamc and libspamc can now legally be used outside
SpamAssassin
2003-01-26 16:49 msquadrat
* t/: spamc.t, data/etc/hello.txt: Added spamc test to
differentiate between spamc and spamd errors
2003-01-26 16:45 jmason
* spamd/: libspamc.c, spamc.c, spamc.pod: added debug
instrumentation to spamc to track down spamc coredumps
2003-01-25 23:28 msquadrat
* lib/Mail/SpamAssassin/Dns.pm: Remove a spurious dot from the DCC
command line
2003-01-25 22:55 msquadrat
* lib/Mail/SpamAssassin.pm: Fix some _really_ weird and buggy code
2003-01-25 22:16 msquadrat
* lib/Mail/SpamAssassin/Dns.pm: Some more pipe fixing (bug 1410)
2003-01-25 04:15 felicity
* rules/50_scores.cf: reset scores to first set0 run
2003-01-25 02:43 felicity
* rules/70_broken_rules.cf: these rules have both low hits and more
ham hits than spam hits ...
2003-01-24 16:38 felicity
* spamassassin.raw: warn about -S in spamassassin too
2003-01-24 16:37 felicity
* spamd/spamd.raw: we should only warn people that -S was removed,
don't die yet
2003-01-24 15:57 msquadrat
* INSTALL: Added lock file to procmail "tutorial"
2003-01-24 15:01 felicity
* lib/Mail/SpamAssassin/Conf.pm: corrected output statement
2003-01-24 12:16 msquadrat
* Makefile.PL: Just empty (don't remove and recreate)
PKG_LOCAL_RULES_DIR on every install
2003-01-24 12:01 msquadrat
* lib/Mail/SpamAssassin/Dns.pm: Do what I mean not what I think.
2003-01-24 11:01 msquadrat
* lib/Mail/SpamAssassin/Dns.pm: Fix some unchecked broken pipes for
Pyzor & DCC
2003-01-23 13:17 jmason
* Makefile.PL, masses/logs-to-c: spamd now not built on Windows,
since it doesn't. ;) suggestion from Roland Bauer, roland.bauer AT
fff.at
2003-01-23 10:56 jmason
* lib/Mail/SpamAssassin/BayesStore.pm: avoid 'argument '' isn't
numeric in addition' warning
2003-01-23 04:12 felicity
* rules/50_scores.cf: new set1 scores
2003-01-23 00:58 jmason
* spamd/: libspamc.c, libspamc.h, spamc.c, utils.c, utils.h: bug
1212: spamc -t timeout code from eugene AT invision.net
2003-01-22 15:06 jmason
* spamd/libspamc.c: removed debugs, left 1 in at LOG_DEBUG level
2003-01-22 14:50 felicity
* rules/50_scores.cf: New set1 GA run completed
2003-01-22 14:35 jmason
* spamd/libspamc.c: reworked the failover-fix code a little to
avoid coredump reported by Malte
2003-01-21 22:15 felicity
* lib/Mail/SpamAssassin/: Conf.pm, PerMsgStatus.pm: Fix for bug
1400. If a rule was defined with no score, only a single scoreset
would get a score set for it. We now set a score for all scoresets
unless one is currently defined. Changed the autolearn code back
to not checking for existence.
2003-01-21 21:48 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: For some stupid reason,
scoresets may have different rules from each other, so we need to
check for existence. :(
2003-01-21 21:02 felicity
* lib/Mail/SpamAssassin/Bayes.pm: give debug info if we can't use a
bayes db due to insufficient messages
2003-01-21 19:38 felicity
* lib/Mail/SpamAssassin/Conf.pm: always_add_report should be 0, not
1 by default
2003-01-21 19:36 felicity
* rules/23_bayes.cf: a bunch of BAYES rules didn't have tflags
learn ...
2003-01-21 18:44 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: let autolearn ignore
userconf too
2003-01-21 18:08 felicity
* lib/Mail/SpamAssassin/PerMsgStatus.pm: Code skips 0 score sets in
autolearning We now ignore tflag "learn" rules in autolearning
instead of /BAYES/
2003-01-21 17:57 jmason
* lib/Mail/SpamAssassin/ArchiveIterator.pm: leftover crud. sorry
about that
2003-01-21 17:33 jmason
* sa-learn-nonspam.raw, lib/Mail/SpamAssassin/ArchiveIterator.pm,
lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm,
masses/corpora/mass-find-nonspam: added always_add_report param,
patch from sabat (sabat /at/ eloan /dot/ com)
2003-01-21 16:09 felicity
* lib/Mail/SpamAssassin/: Conf.pm, PerMsgStatus.pm: autolearning
will now only take place with scoresets 0 and 1. if we're using 2
or 3, the autolearn code will revert to 0 or 1 and recompute the
score, then check about autolearning. this gives us consistency in
autolearning on the same scoresets all the time.
2003-01-21 15:13 felicity
* rules/50_scores.cf: Updated scores after set1 run
2003-01-21 01:55 felicity
* rules/20_head_tests.cf: whitelists are supposed to be nice!
2003-01-21 01:37 felicity
* rules/50_scores.cf: setup for set1 GA run
2003-01-21 01:36 felicity
* masses/corpora/remove-tests-from-logs, rules/20_head_tests.cf:
fixing some stuff to remove tests, X_OSIRU_DUL_FH is supposed to be
nice
2003-01-20 19:54 felicity
* masses/: hit-frequencies, lint-rules-from-freqs: More work to
make net tests mutable if we're doing a net test (scoreset%2 == 1)
2003-01-20 19:40 felicity
* rules/50_scores.cf: GA set0 scores > 4.8 brought down to 4.0
(except those that are very unlikely to FP)
2003-01-20 19:28 felicity
* rules/50_scores.cf: GA scored set 0
2003-01-20 19:05 jmason
* spamd/libspamc.c: struct hostent was overwritten in syslog call
so failover was not working; now fixed, I think
2003-01-20 18:52 felicity
* masses/Makefile: Makefile now knows about scoresets. Do a "make
SCORESET=#" or it'll default to 0.
2003-01-20 18:46 felicity
* masses/logs-to-c: put back includes, lets evolve compile on
macosx. :(
2003-01-20 16:50 felicity
* masses/score-ranges-from-freqs: make net tests mutable if
scoreset indicates a net run
2003-01-20 15:49 felicity
* lib/Mail/SpamAssassin/EvalTests.pm, masses/hit-frequencies,
masses/logs-to-c, masses/parse-rules-for-masses,
masses/score-ranges-from-freqs: Comment out CARRIAGE_RETURN eval
code. masses code now understands scoresets, at least enough. :)
2003-01-20 01:47 felicity
* rules/: 20_body_tests.cf, 30_text_fr.cf, 50_scores.cf,
70_broken_rules.cf: got rid of CARRIAGE_RETURNS
2003-01-19 20:23 felicity
* masses/corpora/remove-tests-from-logs: fixed a few more bugs
2003-01-19 19:50 felicity
* masses/corpora/remove-tests-from-logs: fixed the code some more
to deal with -t option
2003-01-19 17:28 felicity
* rules/50_scores.cf: boo boo with HABEAS_SWE
2003-01-19 05:59 felicity
* masses/corpora/remove-tests-from-logs: take a -t option to
specify a timestamp. rule removal will only happen up to that
timestamp. this is useful for net tests where you only want to
keep X months.
2003-01-19 05:20 felicity
* masses/mass-check: total_check wasn't being incremented before
progress, so at 100% done, we were one off. :(
2003-01-19 04:25 felicity
* lib/Mail/SpamAssassin/: BayesStore.pm, Conf.pm,
DBBasedAddrList.pm, NoMailAudit.pm, PerMsgStatus.pm, Util.pm: Put
umask around any open or tie commands. This will 1) let the *_mode
options work as expected, and 2) keep some of our temp files more
secure.
2003-01-19 00:56 felicity
* masses/score-ranges-from-freqs: it didn't know how to deal with
multiple scoresets, so I let it default to set0
2003-01-19 00:41 felicity
* rules/: 20_compensate.cf, 20_head_tests.cf, 20_phrases.cf,
20_ratware.cf, 20_uri_tests.cf, 50_scores.cf, 70_broken_rules.cf:
pushed rules to 70_broken from the first 2.50 mass-check with bad
results.
2003-01-18 22:40 felicity
* rules/20_compensate.cf: [no log message]
2003-01-18 20:58 felicity
* rules/: 20_meta_tests.cf, 20_phrases.cf: NMS_CGI_NOT_BUGGY wasn't
hitting anything, so I made it part of a meta rule with BUGGY_CGI
per suggestion from eallen.
2003-01-18 16:51 felicity
* masses/: craig-evolve.c, logs-to-c: Applied Allen's GA patches
from bug 1088.
2003-01-18 11:56 msquadrat
* spamd/spamd.raw: Removed small logging glitch (more code cleanup
coming for 2.60)
2003-01-18 06:19 felicity
* masses/hit-frequencies: hit-frequencies doesn't know what to do
with multiple scoresets, so I let it default to scoreset 0...
2003-01-18 06:08 felicity
* rules/50_scores.cf: setting pre-GA scores for bayes and net tests
2003-01-17 01:45 felicity
* masses/mass-check: $date is passed undef to progress, so I added
in a "set to 0 if undefined" to at least stop the warning.
$total_count is set to 0 in the parent, then each child increases
their copy, then the parent calls progress which still sees
$total_count==0. So I moved the increment into the parent.
2003-01-16 20:28 felicity
* lib/Mail/SpamAssassin/Dns.pm: added a debug statement in case the
object format changes incompatibly again
2003-01-16 20:22 felicity
* lib/Mail/SpamAssassin/: Dns.pm, EvalTests.pm: Fixed up Razor2
check code some more. We now support both razor object layouts
(<2.14 and >=2.14). Got rid of extra variable in Razor2 code.
Fixed comment for SEMIFORGED_HOTMAIL_RCVD.
2003-01-16 16:10 felicity
* masses/mass-check: mass-check now shows status every $updates
messages and at 100%
2003-01-16 14:42 felicity
* masses/mass-check: fixed status displays to not give "illegal
modulus 0" errors on small checks.
2003-01-16 12:06 msquadrat
* masses/mass-check: Some tweaks to the --mid code (again): Make
the mid meta RE safe and remove the time part (we don't need to
create unique MIDs, this is just for logging)
2003-01-16 08:07 quinlan
* lib/Mail/SpamAssassin/ArchiveIterator.pm: remove comment
2003-01-16 07:55 quinlan
* lib/Mail/SpamAssassin/ArchiveIterator.pm, masses/mass-check: fix
mass-check progress indicator fix mid so it will be consistent
every run and use "." for whitespace so regexp search for the
Message-ID: will work
2003-01-16 07:04 climent
* rules/25_body_tests_es.cf: New entry
2003-01-15 20:43 jmason
* masses/mass-check: mass-check eye candy ;) -- now reports
progress and timestamps to STDERR if --showdots is off.
2003-01-15 19:59 jmason
* lib/Mail/SpamAssassin/Conf.pm: [no log message]
2003-01-15 19:56 jmason
* lib/Mail/SpamAssassin/ArchiveIterator.pm,
masses/corpora/mass-find-nonspam: minor tweak: differentiate
between MH and cyrus maildir formats
2003-01-15 19:47 msquadrat
* spamd/spamd.raw: Fixing some possible undef errors related to
logging
2003-01-15 19:31 quinlan
* masses/hit-frequencies: add -t option for looking at certain
tflags
2003-01-15 19:09 felicity
* lib/Mail/SpamAssassin/Dns.pm: Fixed 1386, if people use engines
other than the default (1,3,4), the response array is different
than expected.
2003-01-15 16:56 felicity
* masses/rewrite-cf-with-new-scores: rewrite-cf-with-new-scores now
supports score lines with multiple scoresets. bug 1384.
2003-01-15 15:48 msquadrat
* spamd/spamd.raw: Save a variable
2003-01-15 15:36 felicity
* rules/50_scores.cf: fixed small typo
2003-01-15 15:30 msquadrat
* spamd/spamd.raw: Fixed logging issues
2003-01-15 14:42 msquadrat
* masses/mass-check: Fix a cute typo while I'm here
2003-01-15 14:34 msquadrat
* masses/mass-check: Remove path from generated MIDs in --mid mode
2003-01-15 07:20 quinlan
* masses/freqdiff: update for new mass-check format
2003-01-14 23:03 felicity
* lib/Mail/SpamAssassin.pm: Added documentation about options to
Reporter.
2003-01-14 22:55 jmason
* lib/Mail/: SpamAssassin.pm, SpamAssassin/PerMsgStatus.pm: fixed
bug 1356: spamd tests all failed if run as root
2003-01-14 22:45 felicity
* lib/Mail/SpamAssassin/Dns.pm: use_* was looking at the wrong
place ($self->{main}, not $self->{conf})
2003-01-14 20:48 jmason
* t/SATest.pm: dump spamd log in case it fails to start
2003-01-14 20:43 jmason
* lib/Mail/SpamAssassin/Conf.pm: more config item rearranging to
allow more settings to be set in user_prefs
2003-01-14 20:38 jmason
* sa-learn-nonspam.raw, lib/Mail/SpamAssassin/PerMsgStatus.pm:
fixed 2 pod bugs
2003-01-14 20:36 jmason
* masses/mass-check: re-added remove_spamassassin_markup() code,
working this time
2003-01-14 19:43 felicity
* lib/Mail/SpamAssassin/Reporter.pm: don't need temp copy of
fulltext in report, it's already a copy
2003-01-14 19:36 jmason
* lib/Mail/SpamAssassin/AutoWhitelist.pm,
lib/Mail/SpamAssassin/Bayes.pm, lib/Mail/SpamAssassin/Conf.pm,
lib/Mail/SpamAssassin/DBBasedAddrList.pm,
lib/Mail/SpamAssassin/Dns.pm,
lib/Mail/SpamAssassin/PerMsgStatus.pm, t/razor2.t: bug 1354:
add-addr-to-foolist broke after we removed support for IPless
entries in the AWL, now fixed again.
2003-01-14 17:49 felicity
* lib/Mail/SpamAssassin/: Dns.pm, Reporter.pm: standardize our
razor text, use a copy of fulltext to avoid munging
2003-01-14 16:49 felicity
* lib/Mail/SpamAssassin/Dns.pm: Fix what we pass to Razor, and
don't fsck up the $$fulltext string
2003-01-14 15:56 felicity
* lib/Mail/SpamAssassin/EvalTests.pm: fixes a problem where parsing
an IP out with '[' (or any other odd RE-ism) causes the check to
die.
2003-01-14 14:19 felicity
* lib/Mail/SpamAssassin/Dns.pm: fixed typo in comment
2003-01-14 14:17 jmason
* masses/hit-frequencies: oops
2003-01-14 14:16 jmason
* masses/: Makefile, hit-frequencies: allow hit-freqs to display
net test results. surely someone did this before? can't find
it...
2003-01-13 21:26 quinlan
* masses/mass-check: chomp $host while we're at it
2003-01-13 21:13 quinlan
* masses/mass-check: privacy concern: add --mid option so
Message-ID is not reported unless requested
2003-01-13 21:06 quinlan
* masses/mass-check: add message-id to last field, make people
happy again
2003-01-13 16:48 quinlan
* lib/Mail/SpamAssassin/BayesStore.pm: changing "breaking lock"
debug message to "unlocking lock" as it should be
2003-01-13 16:09 quinlan
* masses/: logs-to-c, post-ga-analysis.pl: allow for time=