[ingo] Whitelist/Blacklist behavior

[Jan Schneider]

Zitat von Jason Felice <jfelice at cronosys.com>:

> I'm adding a test suite to Ingo, and have encountered a difference in
> behavior between the 'imap' driver and the 'sieve' driver.
>
> If you whitelist @example.com and blacklist spammer at example.com with the
> 'imap' driver, messages from spammer at example.com are kept.
>
> With the sieve driver, the messages are discarded.
>
> (Sieve whitelist is testing that the address is identical; imap whitelist is
> testing that the address contains the value.)
>
> How should this work?  How do people use it?  Just from IMP's
> blacklist/whitelist?  (If so, how does that have any effect?)

I personally used it almost exclusively from IMP which means that the  
full address goes to the list and strict checking is required.

I have no problem being convinced otherwise if experiences from large  
user bases are different and I don't use this feature a lot anyway,  
but the Sieve implementation makes more sense to me. Especially if you  
don't want to check for domains but full addresses and still have  
substring checking. E.g. if I want to block jan at horde.org, but not  
karajan at horde.org.

Jan.

-- 
Do you need professional PHP or Horde consulting?
http://horde.org/consulting/