On 08/08/02 20:24, Manuel Lemos Replied As Follows: --- Original Message ---
> Hello, > > On 08/08/2002 04:45 PM, Jay Garcia wrote: >>>>though I am living in Hong Kong, Asia, i want to filter those >>>>big5/GB-code spam as well via Mozilla's rules. The biggest spam wave >>>>came from, sadly, China - the new owner of Hong Kong. >>>> >>>> >>>> >>>>>I noticed that the character set is BIG5. I noticed that in the subject >>>>>and in the body parts. The problem is that I can't see a way in Mozilla >>>>>mail filters to specify such criteria. Anybody has any ideas? >>>> >>>Yes, many spammers are also moving to China where they are away from the >>>European and North American laws/ >>> >>>However, I am afraid that Mozilla filtering power is very week. >>> >>>I learned that Netscape 4 had Javascript based filtering support. I >>>wonder if Mozilla has or will have that so we can filter Big5 messages. >> >> If you can, at the server level, block out the entire Class "C" 211 >> domain and you'll eliminate most of the Asia Pacific Network that is now >> responsible for over 50% of all spam ... I did just that and all of my >> 100+ email customers are rejoicing. > > That does not sound like filtering but rather like descrimination. > Personally I am not interesting in droping messages before I have the > chance to examining them. Yes, it definitely is descrimination, you betcha it is. And I have the right and reponsiblity to do so as long as it doesn't adversely affect my clients (it doesn't). 100% of all the mail that we've received from the 211 domain in the past year has been spam. And approximately 80% has been written in Asian Characters (I forget the correct designation) that we have no idea what they mean. You have to understand server-side blocking/filtering to know that we can also ALLOW as well as REJECT. If an entire class C domain, for instance, is blocked, we can ALLOW certain specific addresses and/or portions of the domain to pass through the filter. > Chances are that you will be dropping solicited messages that may be > eventually business or other opportunities by eventually excluding > innocent people. There are no "innocent" people that will be sending me or any of my Corporate clients from that domain or from any "known" spammer's domain. > I prefer client side filtering to be safe, so I can verify that messages > are really SPAM before I discard them. In that case, the spammer has won the game. You MUST read the spam message or at least a portion of it no matter how small. I'll take server side blocking any day as long as I am confident and responsible enough to block "known" spam addresses. Every blocked address (over 5,000 to date) is a "known" spammer address or domain, etc. > Another point, is that my public address is handled by a e-mail > forwarding service run by ACM that I obviously do not control. Recently > they have added anti-spam filters that are making me loose a lot of > personal e-mail. Then notify the server admin/isp/whatever. Or switch mail providers. > I had to request to drop all filters for my address because that does > not suite me. I also seen many people in the same situation complaining > of the same. Sounds like an irresponsible provider to me. That's not my game. > Conclusion, I do not trust many spam filtering systems because it is not > feasible to make them guess right what is spam. I prefer to receive all > spam in my mailbox and sort them with client side filters. That's your choice and you have to live with the caveats. > Unfortunately Mozilla is not yet very capable for this purpose. I wonder > if are there any real efforts in this direction. Mozilla is quite capable if the filters are configured correctly. There are some shortcomings to client side filtering but it works quite well for the most part. There is more work to be done yet in this area for Moz. > Manuel Lemos > -- Jay Garcia - Netscape Champion Novell MCNE-5/CNI-Networking Technologies-OSI UFAQ - http://www.UFAQ.org ** Post To Group ONLY, do NOT email **
