AV software put to the test!

[Mercutio]

Clicky for results.

Note the surprisingly good showing by Symantec/Norton AV and Kapersky.

 

[time]

Complete bullsh*t.

 

[Mercutio]

And what makes you say that? Looking at the list of programs and their rankings through that site's semi-annual testing, I can see real changes over time and a consistent testing methodology.

 

[time]

Where to start? How about:

This test was made by the only Greek virus collector, known as VirusP, webmaster of www.virus.gr, whose collection consists of approximately 210,000 virus samples (crc32 different files) and is one of the biggest virus collections worldwide. On the contrary, all other greek tests' database never exceeds the 6,000 virus samples, while internationally most tests' database doesn't exceed the 20,000 virus samples.

:roll:

or:

The 91202 virus samples were chosen using VS2000 according to Kaspersky, F-Prot, RAV, Nod32, Dr.Web, Sweep, BitDefender and McAfee antivirus programs. Each virus sample was unique by virus name, meaning that AT LEAST 1 antivirus program detected it as a new virus.

Wow, what a scientific selection method. Which ones are actually in the wild, I wonder: 5%? So if say, out of the collection hand-picked by the author, Kaspersky finds 10 'new' variants of a zoo virus, other products fail because they don't detect it? Assuming it really is a virus in the first place?

And then there's the guides, tips and tricks on how to collect viruses, including how to trade viruses.

Sorry, I can't even begin to take this crap seriously. It's retarnishing the reputation of antivirus software testing all over again.

And how much sense does it make to test a firewall's resistance to DOS viruses?

 

[Mercutio]

He's testing a very large sample (about half is collection, apparently - more statistically necessary, but perhaps he needs that many to distinguish between the top performers), and he's requiring that at least one of the programs he listed be able to detect it. Given the size of sample I imagine it would be tough to skew the results overmuch. Do you really think he hand-picked 91,000 viruses?

The other important point is that I don't recall seeing an independent test of this kind at any time before. McAffee and Symantec butt heads a couple times a year in industry press. That's about it. If there is other worthwhile testing, please, share it with me.

It's interesting to note that most of the virus scanners he's listing perform substantially better for virus detection than what we consider acceptable for an anti-spyware program.

 

[time]

You're missing the point. Only a tiny fraction of zoo viruses are in the wild, but certain A/V vendors (Kaspersky) add them to their definitions so they look good in tests like this (at the expense of speed and bloat). Detection by "at least one" A/V product doesn't mean a damn thing.

Try www.westcoastlabs.org, www.virusbtn.com or www.icsalabs.com.