http://bugzilla.spamassassin.org/show_bug.cgi?id=3483
------- Additional Comments From [EMAIL PROTECTED] 2004-07-14 14:02 ------- Created an attachment (id=2126) --> (http://bugzilla.spamassassin.org/attachment.cgi?id=2126&action=view) suggested patch for is_scan_available() issue ok, this one overrides the global caller_will_untie flag during a scan(). it also makes scan check the global flag before it tries to untie. the theory here is that if the global setting is that the caller will untie, then the module functions shouldn't untie ever. if the global setting is 0 or otherwise undefined, however, then the modules should do the right thing(tm) appropriately. so in this case, scan() will have all the modules that it calls know that at a minimum, it will untie the db so the modules don't need to worry about it. then, at the end, scan() checks to see if its caller will untie and do the right thing(tm). :) in our code, btw, the only thing that sets the global caller_will_untie is sa-learn, and learn_trapped (which does the same kluge that I'm proposing for scan() so that forget() doesn't untie during a learn()...) ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
