Kevin A. McGrail writes: > Daryl: > > I mentioned this a few days ago but I'm concerned about the promotion of > KAM_STOCKTIP14. That's a partial rule from my toying in the sandbox and > learning about the rules-qa process. > > Promoting it to a full, non-meta rule is not good though. I'll update the > STOCKTIPs with a better version of the rule that relays on partial rules.
Hey Kevin -- As Daryl noted, if you have some rules that you want to block promotion of, the way to do that is to add a "tflags NAMEOFRULE nopublish" line, or to name them T_SOMETHING with a "T_" prefix. (see http://wiki.apache.org/spamassassin/RuleLifeCycle ) > BTW, what's the best way to lint rules on minotaur? There's no need to run a lint there -- if you use an SVN checkout on your own machine, and lint that (or run "make test"), that will do the trick. e.g.: svn co https://svn.apache.org/repos/asf/spamassassin/trunk cd trunk perl Makefile.PL make make test t/basic_lint................[6752] warn: config: warning: score set for non-existent rule KAM_STOCKTIP21 [6752] warn: config: warning: score set for non-existent rule KAM_STOCKTIP15 [6752] warn: config: warning: score set for non-existent rule KAM_STOCKTIP6 [6752] warn: config: warning: score set for non-existent rule KAM_STOCKTIP4 [6752] warn: config: warning: score set for non-existent rule KAM_STOCKTIP20 [6752] warn: config: warning: score set for non-existent rule KAM_STOCKOTC [6752] warn: lint: 6 issues detected, please rerun with debug enabled for more information Not found: anything = # Failed test 1 in t/SATest.pm at line 633 Output can be examined in: t/basic_lint................FAILED test 1 Failed 1/1 tests, 0.00% okay Note -- it has to be a checkout of SVN. by the way, I've created a file called 'rulesrc/sandbox/kmcgrail/80_deadrules.cf', which contains "fake" rules for those now-deleted rules listed above. (Instead of just deleting them it replaces them with meta NAMEOFRULE (0) which is effectively a no-op. This is necessary because the scores in rules/50_scores.cf in the 3.2 maintainance branch refer to those now-obsolete rules, because they were active when the GA was run. Changes to rules/50_scores.cf require a bug and votes, because it's under Review-Then-Commit. It's easier to just keep the rule *names* alive, replacing them with rules that hit no mail, instead of deleting them outright. The correct thing to do, in hindsight, would have been to mark those rules as T_ rules, or tflags nopublish, so that the GA wouldn't have given them a score... remember for next time ;) --j. > Regards, > KAM > > > The following are the new rules that were promoted (some of them hit quite > > a bit of mail) and their scores (as in the update): > > > > score AXB_MIME_IMG830 1.000 > > score AXB_XTIDX_CHAIN 1.000 > > score CTYPE_1SPACE_GIF 1.000 > > score DOS_PROVISION4 1.5 > > score DOS_REPORT_FIN_INC 0.5 > > score DOS_STOCK_CDYV_GENERIC 2.5 > > score DOS_STOCK_INCOME_STATEMENT 1.5 > > score FB_MED1CAT 1.000 > > score FB_MEDS_PERCENT 1.000 > > score FB_WORD1_END_DOLLAR 1.000 > > score FH_XMAIL_RND_833 1.000 > > score FM_MORTGAGE4PLUS 1.000 > > score FRT_OPPORTUN1 1.000 > > score FS_LOWER_YOUR 1.000 > > score FS_WEIGHT_LOSS 1.000 > > score KAM_LOTTO1 1.000 > > score KAM_LOTTO2 1.000 > > score KAM_LOTTO3 1.000 > > score KAM_STOCKTIP14 1.000 > > score STOX_RCVD_N_NN_N 1.000 > > score TEMPLATE_203_RCVD 1.000 > > score TVD_APP_LOAN 1.000 > > score TVD_FW_GRAPHIC_ID1 1.000 > > score TVD_FW_GRAPHIC_ID2 1.000 > > score TVD_PH_SUBJ_META 1.000
