Slashdot Mirror

On the other hand these stats are interesting:
http://www.ciphertrust.com/resources/statistics/

They tell me a few things.

1. Don't use citibank.
2. We're not doing as well as it seems to me

Actually, I'll concede I was wrong about most of the worlds spam coming from the US... but there are numerous studies indicating that the USA is responsible for more of it than any other single country.

i dont have a problem with us being cut off from the rest of the world. It woud be a good start in the right direction for the US.

You think the USA would benefit from being more isolationist?! I'm not even going to ask - you're probably a fundamental religionist or something. By the way I was *joking* before - I think the best thing about the internet is its global nature. The sooner we start to see other people from all over the world as our peers and equals, the sooner retarded things like wars will stop happening.

We won't have to ask if it runs Linux. Since mankind will be enslaved, we'll be running Linux for it, and generating a whole bunch of Soviet Russia jokes on its behalf.

Kinda like Slashdot, now that I think of it...

(I'm wondering who that girl on that CipherTrust book is...)

except IronPort actually copied the name

Proof -

Ciphertrust earliest press release Sept 2001 - mentions IronMail - http://www.ciphertrust.com/company/press_and_event s/article.php?id=0000137

Ironport earliest press release in June 2002 that mentions IronPort

http://www.ironport.com/pdf/ironport_2002-06-25b.p df

CipherTrust http://www.ciphertrust.com/resources/statistics/zo mbie.php has a live ZombieMeter by country.

You shouldn't base your assumption on population, for example China has a huge population but isn't even in the top 30 for the number of broadband subscribers per 100 inhabitants. It would be better to base the assumption on the number of internet/broadband subscribers per country. Lets compare the US & South Korea as suggested:

From zombieMeter, we see that the US has 19.08% of the 'zombie market' with 964,020 new zombie machines found in May. By comparison South Korea has 9.61% of the market with 485,492 new zombies.

In 2004 OECD reported that South Korea had the highest number of broadband subscribers per 100 inhabitants with 24.9% or 11,921,439 subscribers. By comparison America came in 12th with 12.8% or 37,258,608 subscribers.

So, combining the 2 sets of figures we can calculate the ratio/percentage of broadband subscribers that were new zombies in each country during May:

Korea: 485,492 / 11,921,439 = 4.0%
US: 964,020 / 37,258,608 = 2.6%

Conclusion: Your assumption is correct based on broadband figures only. However, as South Korea has a higher percentage of broadband users the figures may change if we included dial-up internet users - although dialup is not 'always-on' so this may not effect the results too greatly.

You can determine whether the above is lies, damn lies or statistics though...

Populations:
EU : 457million
USA: 296million

Zombies in May:
EU : 26.16% (1320985)
USA: 19.08% (964020)

So, zombies per capita:
EU : 0.00289 (1 zombie per 346 people)
USA: 0.00326 (1 zombie per 307 people)

Sources:
http://www.cia.gov/cia/publications/factbook/ranko rder/2119rank.html
http://www.ciphertrust.com/resources/statistics/zo mbie.php

A continent has more zombie PCs than a country ...
Shocking that ..
http://www.ciphertrust.com/resources/statistics/in dex.php
Event hough the statistics infact disagree with the report..
I think we have some odd reporting here , IT should be that may saw the largest rise in Zombie PCs in Europe .

Two points about spam:
1) The most important feature of any spam filtering system is that it should have a few as possible false positives (genuine email but incorrectly regarded as spam, which are then lost.)

2) In the end, we need to stop spam being _sent_. At the moment most effort is going into filtering the spam out after it's been _received_. Clearly we can spend forever writing spam filtering rules, with the spammers a few steps ahead all the time.

Having said that, we work with what we can get. At the moment where I am we're using Ironmail v4.5.2.
http://www.ciphertrust.com/products/index.php

This is a big improvement over the last version of Ironmail, which wasn't half bad.

Check out these numbers:
MAIL: wk to 07 Apr -
505365 in;
16148 viruses; 3%
399415 spam; 79%
89802 clear 18%

March: 2475384 in;
74300 viruses; 3%
1995630 spam; 81%
405454 clear; 16%

Those are hard and fast numbers from the box. The important number (false positives) is of course not so easy to generate. And there's my point.

The University where I work has introduced
1) Censorship of the Web, using Websense http://ww2.websense.com/global/en/.
2) Throttling bandwidth on network ports using Storm Control http://www.cisco.com/en/US/products/hw/switches/ps 708/products_configuration_guide_chapter09186a0080 160a9f.html
3) Filtering out spam using Ironmail http://www.ciphertrust.com/products/index.php
Each these measures have had a negative impact on genuine study and research.

Our Computer Centre Director, who is in the invidious position of having to balance academic freedom against meeting JANET http://www.ja.net/ regulations, released this message which I reproduce here to show what Universities are dealing with.

-END OF QUOTE-
The introduction of restrictions is not something
that we have come to lightly. We certainly have
no desire to apply censorship to our users;
however, unlike Internet Service Providers,
we have somewhat more legal responsibility for
the material that is carried over or stored
within our network. In particular, the University
can be held 'vicariously liable' for a number
of offences relating to, for example, the
display or storage of pornography. Similarly,
material relating to religion or race that is
capable of offending is a potential threat, in
a legal sense, to the University. There are others.

On the matter of websites that just plainly offer
no business value to the University, we need to
strike the right balance between the various
interests. We have real concerns about the
capacity of our network and to compromise academic
and business activity on the network because we
are hosting a flood of dubious traffic does not
make good sense. However, under this specific
concern, clearly there may be scope for relaxing
restrictions outside the 'working day'.
-END OF QUOTE-

Over the Hols, our Ironmail box was dropping 17 out of every 18 emails sent to our domain.

I'm sick and tired of people holding our Barracuda as some kind of measuring stick for anti-spam methods.

1.) Barracuda is just a SpamAssassin bundle with a nice interface and some REGEX updates. There isn't any contrast between SpamAssassing and Barracuda, the one is simply a super-set containing the other.

2.) Barracuda has less than 2 percent of the market share, while other companies have significantly more! (link is to IDC data, I couldn't readily find it other than on CipherTrust's site). Further, most of Barracuda's customers are educational insituations and other tiny entities that don't have the budget to spend on a real solution. They have virtually none of the Fortune 500 (maybe two or three customers in that bracket, total). If they were so great, a lot more large enterprises would have adopted their solution.

For some reason OSS zealots have a love affair with Barracuda, because their product is nearly 100% an OSS bundle and because they advertise on /., I guess. Curiously, there are other products out there primarily built on OSS as well, such as McAfee's e250/500/1000 product line (who is also interestingly #2 in the secure content management appliance marketshare with nearly 20%, or 10 times Barracuda's share)

You should all should quit being blind zealots and do some objective research, the Barracuda product isn't even close to being the best in it's space. On the other hand, that would be so un-Slashdot to actually take an objective look at an issue and not comment on things you don't understand.

The FUD may have already started.

Anyone know about CipherTrust? Are they in bed with MS?

CipherTrust has a breakdown of by country stats by message volume and IPs on their webpage

Brunel University (London UK) use Ironmail. Current stats: Week to 28 Apr: 719504 messages; 65456 viruses; 562642 spam; 91406 clear

Most of the proposals are probably patented (as ridiculous as that may sound). No doubt the recent spam proposals are being pushed by folks with an agenda totally unrelated to spam. There is no way they would get this much media attention without a commercial PR department. Which begs the question, who is behind ASRG? The guy in charge has six pending patents on this very subject.

To stop spam, we should use less invasive approaches such as bayesian filtering and common sense legislation (mandatory headers and spam-hunting boundies aren't a bad idea). We do not need privacy-invading, censorship centers which outlaw open-source solutions.

-----

[1] I fully expect that if we adopt authentication systems, the certificate authorities will permit paying marketeers to spam anyone they choose.