http://issues.apache.org/SpamAssassin/show_bug.cgi?id=4347
------- Additional Comments From [EMAIL PROTECTED] 2006-08-02 21:23 ------- (In reply to comment #41) > I think the best fix (as I said above) is to provide a new setting that > indicates "DCC_CHECK is a valid rule name, even though it may be undefined for > various reasons"; then if DCC_CHECK appears in a meta declaration, it won't > generate this warning. something like "tflags DCC_CHECK exists" maybe? Let's please stop overcomplicating things and overloading tflags with stuff it doesn't need. The problem is that people are complaining about the code doing what it's supposed to be doing ... DCC_CHECK, for example, *DOES NOT EXIST* if the plugin isn't loaded -- I really don't see the necessity to put a "tflag RULE exists" for every rule inside of an ifplugin statement, that's ridiculous imo. It seems as if the easiest course of action is: if --lint, do info possibly a more complex version: if --lint, do info else if debug channel rules is specified, do dbg then most people will never notice it, unless they run "--lint -D" or "-D rules", wherein they really should be told about this stuff because they're looking for things like this. FWIW, something that came up in IRC while chatting w/ wtogami was that the "rule is undefined" message could be made nicer if it were "rule is undefined because plugin X is not loaded". Which is doable, except we'd have to completely change how ifplugin and such is handled (like, don't completely skip everything in the block, and parse it enough to get rule names out but don't actually include the rule in the conf object, and ...) ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
