Radical I know, but since email can be sent from any machine, and on a number of ports if proxies are used, it's the only way to go. This will also stop DDOS attacks and port scanning because of the cost involved.
Okay, let's imagine this idiotic scheme was tried. (Tough, I know.) How do they plan it implement it? Any machine can be a mail sender. Even machines with port 25 outgoing blocked can still send to an open proxy via a number of other ports. So how do they plan to meter and collect the tax?
I guess we'll have to wait a while until someone develops a psychic postage meter that can be bolted to every computer. (The improved super-psychic meter can detect the difference between email to the network at large, and internal company email.)
It will improve the situation, for your mailbox. This is not a bad thing. Few people want to join the Don Quixote Liberation Army. It won't stop the loads of crud jamming ISPs -- end-user filters don't stop that.
There are no answers, just more tools in the toolbox.
Yes, but a more industrial minimalist booth. The kind you would put in subway stations (with a hope of survival) rather than shopping mauls. Also, your PDA would control the vertical -- bring your environment with you. (Ooh, just had an SF "The Long Run" moment: "Goddamn slow hardware boss!")
In some apartment buildings, the lobby security camera is available via cable tv so that you can see who you're letting in. (Or see who everyone else is letting in.;^)
Re:SMTP keeps originating IP in the mail header
on
Spam Meeting Wrap-up
·
· Score: 1
Actually, it doesn't. You can only be (almost) certain that the last Received line (topmost) in the header is correct. Every Received before that has to be evaluated for correctness. How much do you trust each mail server each hop further away? Spammers frequently add fake Received lines to headers.
And spammers have been switching to open proxies rather than relays. You might track the spam back to the open proxy, but there's no way to trace it past the proxy without internal logs. And since many open proxies are people that don't even know that they're running one, the chance of finding out anything more is slight.
Frequently the best attack on a spammer is from the other end: They almost always want you to contact them. The spam includes a payload of an email drop-box, a web page, a phone number or a physical address.
Re:They needed three days to figure this out?
on
Spam Meeting Wrap-up
·
· Score: 1
The future is Bayesian.
Umm, no. While Bayesian filtering will do an excellent job of keeping your email box clear, it won't stop the spam. You were never a possible customer of theirs. All their spam to you was a waste in the first place, filtering won't change that. The small fraction of a percent of the spammed who actually bought stuff from spam are [total vocabulary failure] unlikely to use Bayesian filtering or to train it properly. So the spammers will keep spamming to reach those people -- and crank it up for good measure!
And filtering doesn't make that spam to you disappear. It still has to reach your personal filters (since Bayesian filtering doesn't really work globally). Up to that point, it still has to pass through the Internet usually through abused relays and proxies, and hit your ISP. (Hit is an understatement when the spammers flood every box they know of on an ISP at the same time.)
Filters will keep your box clear, but not stop the spammers. The only way to do that is to hit them either financially or loss of connectivity. Educating the morons who buy from spammers in unlikely to happen in this universe. Loss of connectivity? Make it uneconomic for ISPs to offer services to spammers. Blocklists are a controversial approach, but spam-fighters are always willing to listen to any new suggestions. (Note: eCash sender pays and rewriting SMTP are not new.)
Don't worry about DDOS attacks to space probes, worry about DDOS attacks coming from space probes...
How long before we see a Slashdot post: "Warning parent post's venus link is really goatse.uranus"?
.. will be with those killer roaming charges on Grand Tour missions.
Oh, I dunno. Just think of the fun that they have have by being able to yank or otherwise diddle with someone's email certification.
Sure it'll work .. on Uranus!
I guess we'll have to wait a while until someone develops a psychic postage meter that can be bolted to every computer. (The improved super-psychic meter can detect the difference between email to the network at large, and internal company email.)
A good place to start are the ads in Curcuit Cellar I *think* they still do printed copies. I haven't been round to a good magazine store in ages.
There are no answers, just more tools in the toolbox.
Yes, but a more industrial minimalist booth. The kind you would put in subway stations (with a hope of survival) rather than shopping mauls. Also, your PDA would control the vertical -- bring your environment with you. (Ooh, just had an SF "The Long Run" moment: "Goddamn slow hardware boss!")
Get with the neo-retro dude!
I see a niche for "box booths" that provide a keyboard, screen and connectivity for people on-the-go.
As a security feature they should make a sound when being accessed on-the-go. The *SLOORP* as you walk past would be a give-away.
In some apartment buildings, the lobby security camera is available via cable tv so that you can see who you're letting in. (Or see who everyone else is letting in. ;^)
Oh, not because of any ethical issues about spying. (War-tomming?) But mainly because I refuse to buy anything from X10 after all those pop-up ads.
And best of all, you could get servers with funky coloured cases!
Anyone who aquires a momento must now register a duplicate with the FBI as part of their Collector system.
What are they going to do if you're running Linux of BSD?
When I lose an important memento, I don't have to worry because I kept all the serial number and insurance info in a file which... DAMN!
It's called The Boulder Pledge, by Roger Ebert.
(Riing! "Hello?" "This is the genetic reject anti-defamation league. We want an apology for that insult!")
You mean, sort of like this incident?
And spammers have been switching to open proxies rather than relays. You might track the spam back to the open proxy, but there's no way to trace it past the proxy without internal logs. And since many open proxies are people that don't even know that they're running one, the chance of finding out anything more is slight.
Frequently the best attack on a spammer is from the other end: They almost always want you to contact them. The spam includes a payload of an email drop-box, a web page, a phone number or a physical address.
Umm, no. While Bayesian filtering will do an excellent job of keeping your email box clear, it won't stop the spam. You were never a possible customer of theirs. All their spam to you was a waste in the first place, filtering won't change that. The small fraction of a percent of the spammed who actually bought stuff from spam are [total vocabulary failure] unlikely to use Bayesian filtering or to train it properly. So the spammers will keep spamming to reach those people -- and crank it up for good measure!
And filtering doesn't make that spam to you disappear. It still has to reach your personal filters (since Bayesian filtering doesn't really work globally). Up to that point, it still has to pass through the Internet usually through abused relays and proxies, and hit your ISP. (Hit is an understatement when the spammers flood every box they know of on an ISP at the same time.)
Filters will keep your box clear, but not stop the spammers. The only way to do that is to hit them either financially or loss of connectivity. Educating the morons who buy from spammers in unlikely to happen in this universe. Loss of connectivity? Make it uneconomic for ISPs to offer services to spammers. Blocklists are a controversial approach, but spam-fighters are always willing to listen to any new suggestions. (Note: eCash sender pays and rewriting SMTP are not new.)
I wave this t-shirt in your general direction!
It'd better be reliable. You wouldn't a system that would drop packets.