On Monday, July 12, 2004, 2:48:10 AM, Sean Doherty wrote: > On Fri, 2004-07-09 at 19:50, Fred wrote: >> For a test to be considered a net test, it needs to have a tflag set to net. >> For your surbl rules, they should all have: >> tflags SURBL_WHATEVER net >> This will force them to be skipped like you want.
> Thanks for the answer Fred. > However these tests still get run when skip_rbl_checks = 1; > even if the tflags 'net' flag is set. > The tests *do* get skipped if I specify the spamassassin -L > flag (local tests only). > After a bit more digging I found that the spamCop-uri code > doesn't check the skip_rbl_checks settings as is done for > other rbl tests. If this is by design or not is a question > for the spamcop-uri developers. That's probably a question for the spamcopuri author Eric Kolve, though a partial hedge is that we tend not to think of SURBLs in quite the same way as other RBLs since the usage is fairly different. (Two other uses of RBLs are more common: match sender IP addresses or domains in message headers or match resolved URI IP addresses in message bodies.) Chris' answer is reasonable that the 3.0 code hopefully may treat SURBLs like other RBLs in that skip_rbl_checks may work as expected. Reading the source code would probably reveal this if so. Jeff C. -- Jeff Chan mailto:[EMAIL PROTECTED] http://www.surbl.org/
