Slashdot Mirror
Unplugging Email To Combat Spam
monkeyserver.com writes "from Reuters (via CNN) we hear that 'Consumers who allow their infected computers to send out millions of 'spam' messages could be unplugged from the Internet under a proposal released Tuesday by six large e-mail providers.' They are looking at 100 per hour or 500 per day; this doesn't really sound like a bad idea, though it could cause problems for a few people trying to run companies from their basement..." On the other side of the coin, rastakid writes "It appears that Microsoft is taking its actions against spamming a little bit too far: Hotmail accounts which are suspected of sending spam are closed without a single investigation. This article states that Maariv International registered a new Hotmail account and sent an abuse message about spamming activities from that account, while not a single message was sent from it. Microsoft closed the account immediately, without investigating."
perhaps something like SPF could be used? That would ensure that spam was sent from the correct mail servers, and hence not spoofed. This would reduce a large amount of spam appearing to come from other users.
Once this has been done we can then clamp down on the dodgey email servers.
Doesn't this pose a risk for effectively DOS'ing all hotmail users? Just create a script to aggregate Hotmail accounts through google and send complaints? Thats mildly annoying.
--Kevin
That doesn't mean that it isn't a stupid or asinine thing to do. It just means that they wouldn't be held culpable.
I gave up on Hotmail a long time ago, not because of spam sent from those accounts, but because any time I opened up a hotmail account, it was immediately deluged by SPAM
The CB App. What's your 20?
I think something like this could work, but not on it's own.
ISPs should send a letter or e-mail to all their customers (i.e.
make sure they get it) stating that they are about to introduce
rate-limiting both from their smtp servers for that IP address/subnet
and from port 25 from the IP(s).
Customers who don't know what this means or who aren't bothered will
ignore it, and will be rate-limited (so they basically won't be
affected since they either a) aren't bothered, or b) aren't heavy
e-mail users).
Customers who know they will be affected or otherwise want to be
rate-unlimited can e-mail the ISP and request the rate be removed.
Perhaps they could be asked to prove they are worthy by describing
what they've done ("I've patched and secured my Windows box, and
my other boxen run BSD and run no mail daemons").
This way, no one has their service unfairly cut back, and unknowledgable
users (those responsible for zombie-Windows systems) will be protected
(or everyone else protected from them..).
The poster makes the claim that 100 per hour or 500 per day would only cause problems for people running companies from their basements. I heartily disagree. Think of people who run mailing lists from their home servers, these can easily send out more than 500 messages a day. Another example, when I recently got engaged, I sent out an email to a LOT of people. Probably over the course of that first hour after I sent out the original notice I sent out well more than 100 emails. I wasn't doing anything wrong.
The real fact of the matter is that this will do nothing to stem the tide of spam when one considers that most spam is now generated by zombies. Also, don't think they won't just find a way around it. This is like the DMCA, it only stops the honest people.
Fortunately, there has been some movement on SPF.
I suppose I can be happy about that.
My Slashdot account is old enough to drink...
So if I email an event notice to my club membership list of 208 addresses, (given freely for this purpose) I'll be labelled a spammer unless I split the mailing up over 3 hours? There are other ways to find spammers besides shear output.
~~~~~
If you throw it, it will come.
A lot of people's hijacked systems could have been kept clean were they fully patched:
I've been saying for a while now, if an ISPs sign-up disk had all current Windows service packs and critical patches loaded into it and installed them as part of the setup procedure -"You consent to Windows update patches being applied to your system during install"- then I'm sure a lot of network and support load could be lifted off the ISP and the net as a whole. If they could broker a deal to install Zonealarm or Sygate Personal firewall at the same time even better.
It isn't an unreasonable expectation that a machine connecting to a public network shouldn't have gaping security gaps. In fact, IMO, it is a public duty that it should not.
I have been a user for about 10 years. This ends Feb 2014. The site's been ruined. I'm off. Dice, FU
You overlook some of the more obvious effects of this policy. Basicly if I disliked you, and you had a fairly important hotmail account... It would become trivial to mess with you out of spite.
Many people use their email accounts for very critical information or personal correspondence. Getting them shutdown because somebody said it was used for spam is wrong.
Another thing. What if spammers took to spamming the support mail with huge amounts of this account used for spamming messages while using some accounts for spamming. Backlogging the folks while raising heck on the side. The more you think about any solution to spam the more you think of ways around it, if you were a spammer.
It is no longer uncommon to be uncommon.
While $35-$50 isn't much of a reconnect fee, disconnecting probably shouldn't be the first step. Ideally the process would go something like this:
ISP notices a lot of email generated from your node.
Emails registered address inquiring if the volume of email (send statistics) is known to the user.
User responds to confirm they are legitimately sending the volume of email or they respond that they are unaware of the volume.
NOTE: If user does not respond, follow with registered letter or a phone call to the registered user.
If user resolves the problem (patch/removal) system remains intact. If user is unable to resolve the problem, provide options for resolving it. This may include free support, charged support, or recommendation to other support services (The DC metro area has a company called "Geeks on Call").
If user doesn't resolve the problem within an alloted time period, disconnect them.
Charge a reconnect fee.