Why? Simple. They start to spam from their fixed location, and they get blackholed just as quick. Then they have to move to another IP. Which is a pain.
Better to fire your spam off through open relays (preferably anonymizing the true source), or through open http/socks proxies (inherently anonymizing), and just move to the next when you think the open relay has been burned.
There are spammer-providers who sell lists of open relays/socks/http proxies to spammers.
Or it could be their intended goal is something else, like providing a service, building widgets, surfing web sites.
Said within the context of a paper on email, that carries about as much meaning as the "intended goal" being to shovel snow off the driveway. It should be clarified or removed.
While I like seeing reasonable balanced presentations of the pros-and-cons (having to operate spam filters for a very large corporation), his paper is not only riddled with factual errors, it's clear he doesn't understand the subject matter at all.
The most egregious mistake:
1) He talks about RBLs (in his terms, "open relays"). The minority of DNSBLs are open relay lists. By his terms, several of his sample RBLs (ie: Spamhaus), are _not_ RBLs. His paper should have been about DNSBLs in general, not RBLs specifically, and indicated DNSBLs have different listing criteria. Ie: spam sources (ie: SPEWS, Spamhaus), open relays (his "RBL": ie: RSL, OSIRUS inputs, ORDB, ORBL, etc), open http/socks proxies (BOPM, MONKEYS, OSIRUS socks and proxy), DHCP pools (eg: PDL).
Given the above _extreme_ defect, the paper is essentially useless.
Here are defects in his coverage of RBLs:
2) He talks as if RBL listings for open relays should be "appealable". An open relay is either open and abuseable or it isn't. Most RBLs mechanically test servers for open relay - there is no subjective judgement here.
3) Claims that getting delisted by RBLs is difficult and rare - a little research will show that most RBLs retest (either on demand or by time schedule). ORDB and OSIRUS inputs are _particularly_ good at delisting relays that now test closed within a very short period of time.
4) He implies that open relays are desirable. There is no legitimate reason of _any_ kind for an unrestricted open relay. By policy, we will simply not accept email from an open relay/http/socks proxy, because virtually all of it is spam.
5) Unaccountable? If they were, we wouldn't use them. Undocumented? Ditto. DNSBLs have to have predictable behaviour before they're safe to use. Several of the ones he lists are very professionally run and quite trustworthy. Some of them are the opposite. Whether they're good or not is a defect in the implementation, not the concept of DNSBLs.
6) His research on alternate techniques is quite deficient - no mention of DCC, CloudMark, Postini etc.
My favourite remark in the paper:
RBL mechanisms frequently cause a lot of trouble for legitimate Internet users who are trying to send non-spam email in addition to their intended goal.
This implies that the intended goal is to send spam. Oops.
Not at all. If it were true, then most of the model rockets sold would be in violation. Ie: most have at least one metal part, many are over.6 inches in diameter.
There are no laws explicitly setting out rocket limits, except for the maximum limit of model rocketry (impulus and weight limit) to be classified as a "model rocket". All other rules are implemented by the rocketry association (CAR, NAR or TRA). Which aren't enforced thru the penal code. Such as building materials, launch angle limits etc.
The concrete rocket doesn't sound like it would qualify under association model rocket rules, because it probably needed a motor bigger than a G.
Amateur rocketry rules are considerably different (my largest so far is almost 20 pounds, 6" in diameter, and flies to 1400 feet on a J motor). While those rules may not cover a concrete body tube (they only went so far as to ban "non malleable metals"), if it was done responsibly, most Range Safety Officers would probably approve it at a high power launch. If not, it would certainly still be legal at an "official experimental" launch.
level3rockethead, Certified to level 3 high power rocketry Almost a RSO level 1;-)
Slashdot Mirror
Better to fire your spam off through open relays (preferably anonymizing the true source), or through open http/socks proxies (inherently anonymizing), and just move to the next when you think the open relay has been burned.
There are spammer-providers who sell lists of open relays/socks/http proxies to spammers.
Said within the context of a paper on email, that carries about as much meaning as the "intended goal" being to shovel snow off the driveway. It should be clarified or removed.
While I like seeing reasonable balanced presentations of the pros-and-cons (having to operate spam filters for a very large corporation), his paper is not only riddled with factual errors, it's clear he doesn't understand the subject matter at all. The most egregious mistake: 1) He talks about RBLs (in his terms, "open relays"). The minority of DNSBLs are open relay lists. By his terms, several of his sample RBLs (ie: Spamhaus), are _not_ RBLs. His paper should have been about DNSBLs in general, not RBLs specifically, and indicated DNSBLs have different listing criteria. Ie: spam sources (ie: SPEWS, Spamhaus), open relays (his "RBL": ie: RSL, OSIRUS inputs, ORDB, ORBL, etc), open http/socks proxies (BOPM, MONKEYS, OSIRUS socks and proxy), DHCP pools (eg: PDL). Given the above _extreme_ defect, the paper is essentially useless. Here are defects in his coverage of RBLs: 2) He talks as if RBL listings for open relays should be "appealable". An open relay is either open and abuseable or it isn't. Most RBLs mechanically test servers for open relay - there is no subjective judgement here. 3) Claims that getting delisted by RBLs is difficult and rare - a little research will show that most RBLs retest (either on demand or by time schedule). ORDB and OSIRUS inputs are _particularly_ good at delisting relays that now test closed within a very short period of time. 4) He implies that open relays are desirable. There is no legitimate reason of _any_ kind for an unrestricted open relay. By policy, we will simply not accept email from an open relay/http/socks proxy, because virtually all of it is spam. 5) Unaccountable? If they were, we wouldn't use them. Undocumented? Ditto. DNSBLs have to have predictable behaviour before they're safe to use. Several of the ones he lists are very professionally run and quite trustworthy. Some of them are the opposite. Whether they're good or not is a defect in the implementation, not the concept of DNSBLs. 6) His research on alternate techniques is quite deficient - no mention of DCC, CloudMark, Postini etc. My favourite remark in the paper: RBL mechanisms frequently cause a lot of trouble for legitimate Internet users who are trying to send non-spam email in addition to their intended goal. This implies that the intended goal is to send spam. Oops.
Not at all. If it were true, then most of the model rockets sold would be in violation. Ie: most have at least one metal part, many are over .6 inches in diameter.
;-)
There are no laws explicitly setting out rocket limits, except for the maximum limit of model rocketry (impulus and weight limit) to be classified as a "model rocket". All other rules are implemented by the rocketry association (CAR, NAR or TRA). Which aren't enforced thru the penal code. Such as building materials, launch angle limits etc.
The concrete rocket doesn't sound like it would qualify under association model rocket rules, because it probably needed a motor bigger than a G.
Amateur rocketry rules are considerably different (my largest so far is almost 20 pounds, 6" in diameter, and flies to 1400 feet on a J motor). While those rules may not cover a concrete body tube (they only went so far as to ban "non malleable metals"), if it was done responsibly, most Range Safety Officers would probably approve it at a high power launch. If not, it would certainly still be legal at an "official experimental" launch.
level3rockethead,
Certified to level 3 high power rocketry
Almost a RSO level 1