Tiger Mail Filter - getting subject to work on *base text* (including character encoding strings?)

[Jim Witte]

Hi,

   Is is possible in Tiger Mail 2.1.2 (or Leopard - another reason to  
upgrade), to get the spam Subject Contains rule to work on the BASE  
TEXT of the subject header, so I can scan for a character encoding  
string?  I want to set up a filter to scan for '?koi8-r?' so I can  
get spam with a subject like this: 'B?8MXU0iDkxc7J08/Xyd4=?=  
<chester at timberwise.co.uk>'

   But on the other hand, if I tell it to scan for the text *in the  
russian character set* - it works find.  But of course I don't know  
Russian, and the subject lines probably change slightly with every  
spam, so the more reliable marker is the embedded char-set command.

   I'm really tempted to just run all my mail through gMail - I've  
heard it's spam filtering is pretty good, as they simply have  
thousands of users all marking spam, or just either run an  
applescript on every message (but that's probably slow as heck unless  
they did the OSA-bridge implementation is reasonable)..  I have NO  
idea how my IU account name got out, but now it's in the hands of the  
zombie-computer-army being sent to non-existent email addresses (and  
*I* get the bounces! - why the $#%@# can't the RFC process people  
find a better way to do bounces?  I thought DomainKeys or whatever  
MS's competing scheme is was supposed to keep this from happening -  
of course, it didn't - or the spammers found a way around it)

Jim