October 09, 2003
Long Live Email!
Brother Shirky joins the email doomsayers.
I also have the same pit in my stomach about email in 2003 that I did in 1997 about usenet.
I never liked usenet, not even in 1994. It was one of those things, like Gopher, that I was expecting to go away once the Web got rolling.
I survived the spam avalanche from the worms with almost no loss of productivity. I survived the response to an "instalanche" when Glenn linked to my "open letter" essay. In that case, my filters had to cope with a bunch of good emails from people who had never written to me before.
My filters worked fine.
I think that the issue with email is that we are doing more with it. That means that we have to adapt our habits. We have to adopt better sorting tools.
I continue to believe that putting limitations on the ability to cc emails with attachments may be worthwhile. But I don't see it going down the path of usenet.
Posted by Arnold at 6:06 PM |
Email this entry
| Category:
spam wars
Effective filters... I don't use Bayesian filters, nor do I rely on the "Junk Mail" feature in Apple Mail for more than an indication of what it thinks is junk. I have found that after I sort my lists and divert friends and family and people I work with into appropriate mailboxes, a rule that simply diverts e-mail with content-type header "text/html" to a "likely SPAM" box solves the problem with manageable type 1 and type 2 errors.
Nightly, I go through the SPAM messages and make sure I didn't miss anything important. Since I run our servers, I also scan at the Received headers to see if there is another domain to add to my private blacklist. I used to have a few users automatically forward me candidate SPAM so that I could harvest more blacklist domains. SPAM probably costs me 15 minutes a day, but if I were to let it go, we'd have overloaded mail servers and network connections.
Lately, the two biggest SPAM problems I've noticed are people who set up SPAM servers on DSL and cable accounts and bounced messages from people who have misappropriated my domains. I'm guessing the former aren't "open relays" so much as they are a lot of people who want to get rich quick sending SPAM. When analyzing a month's worth of SPAM of this type, I don't see a lot of repetition of addresses in the Received header.
It would be most helpful if the high speed internet providers would make reporting of SPAM from their customers easy and/or standard, have clear policies about setting up mail relays, and deal with customers who pollute the net. As for domain hijacking... A California state law kicks in in January which would give me the ability to sue anyone in CA who decides to do this with my domains. A small but substantial portion of the instances I've seen (through returned e-mails) in the past year are likely from people operating in CA. I imagine that a mail-savvy CA attorney will make a fortune on this law. I'd be happy to help.
-Brad
Posted by Brad Hutchings on October 9, 2003 08:01 PM | Permalink to Comment
The spam filters on my Hotmail account work supremely well. I set the filters to exclude every single e-mail address that I had not previously approved. The result is that I merely have to briefly peruse all of the new incoming e-mail addresses---and this only takes about thirty seconds! I then delete the garbage in one fell swoop. Effectively dealing with spam should not take more than a few minutes of your day. It’s a very minor problem if you employ common sense.
Posted by David Thomson on October 10, 2003 09:59 AM | Permalink to Comment
Effective filters... I don't use Bayesian filters, nor do I rely on the "Junk Mail" feature in Apple Mail for more than an indication of what it thinks is junk. I have found that after I sort my lists and divert friends and family and people I work with into appropriate mailboxes, a rule that simply diverts e-mail with content-type header "text/html" to a "likely SPAM" box solves the problem with manageable type 1 and type 2 errors.
Nightly, I go through the SPAM messages and make sure I didn't miss anything important. Since I run our servers, I also scan at the Received headers to see if there is another domain to add to my private blacklist. I used to have a few users automatically forward me candidate SPAM so that I could harvest more blacklist domains. SPAM probably costs me 15 minutes a day, but if I were to let it go, we'd have overloaded mail servers and network connections.
Lately, the two biggest SPAM problems I've noticed are people who set up SPAM servers on DSL and cable accounts and bounced messages from people who have misappropriated my domains. I'm guessing the former aren't "open relays" so much as they are a lot of people who want to get rich quick sending SPAM. When analyzing a month's worth of SPAM of this type, I don't see a lot of repetition of addresses in the Received header.
It would be most helpful if the high speed internet providers would make reporting of SPAM from their customers easy and/or standard, have clear policies about setting up mail relays, and deal with customers who pollute the net. As for domain hijacking... A California state law kicks in in January which would give me the ability to sue anyone in CA who decides to do this with my domains. A small but substantial portion of the instances I've seen (through returned e-mails) in the past year are likely from people operating in CA. I imagine that a mail-savvy CA attorney will make a fortune on this law. I'd be happy to help.
-Brad
Posted by Brad Hutchings on October 9, 2003 08:01 PM | Permalink to Comment