VB100 RAP Test Feb-Aug 2010

I guess it depends on the user and the conditions. There are times when I have to plug my device into a network that does not have an active internet connection to grab files, etc.

 

For quite a while there have been malware apps that are designed to "kill" security applications. Security applications all have implemented measures to protect themselves from such kill methods.

But it is no longer essential to kill the security app itself. Just kill the user's internet connection & security apps that solely or primarily are cloud-based will lose a major portion of their effectiveness.

And the beat goes on . . .

 

Well said

You NEED the signatures on your PC, you just can't trust in all these cloud tech

 

That is absolutely wrong and most importantly is not valid for all products! I won't comment on others but Norton is designed to well protect user's machine , it won't even allow malware install on the computer or let it kill the internet connection. So you won't even come to a stage with no connection. Nowadays there are not many anvirus program that rely mainly on updates and not other technologies.

Additionally , Norton includes SONAR which is not in-the-cloud protection if you come to a moment with no internet connection. However , due to the complete lack of internet in the VB tests , this one cannot even update itself even once. Even allowing initial internet update of the products would be much better than it is now.

 

Damn, you got me shut
I think i should stop posting my non sense opinions

 

Good job Eset!! Keep up the great work !!

 

I don't think you should.

Obviously Symantec still believes that sigs are important. Sure they have adopted and integrated other great technologies into their software but, they haven't abandoned the whole signature concept and I can't see them doing so anytime soon.

 

now that I look at this graph more closely, Eset really did come out well.

Hope they used more then 16 samples.

 

You can't be serious if you think that the methodology used by VT is really the ONE (a.k.a. absolutely correct)

 

no I dont think any of them are 100 percent nor are their methods. I own most of them and soon Norton. I take what is shown here more as proof then test sites.

 

Please explain "really the ONE".
Is AV-Comparatives "really the ONE"?
Thanks.

 

maybe its the bulls eye target in my avatar that some just dont like.

 

No , I am not saying that . Just like trjam posted above , I think they all should be changed . But do I stand behind Dynamic type of testing not just on-demand testing or testing without Internet connection . In my opinion AV-Test's methods currently are the most accurate while VB's ones are least . I have given few examples above in this thread - you can't really test products for just outdated samples of in-the-wild malware and do this with no Internet and no updates when there are so many products which rely on Internet to get updates and when we all have active internet connection , malware is Internet-borne type.
I am not a Panda Securiy supported but I do agree with everything written here:
http://research.pandasecurity.com/wildlist-virus-bulletin-100-and-other-battles/
http://research.pandasecurity.com/b...TheDisconnectBetweentheWildListandReality.pdf

 

AV-Comparatives ran a whole product "dynamic" test December 2009. Eset and Avira did considerably better with the AV-Comparatives test than with AV-test.
I am not sure AV-test uses the term "dynamic" the same way AV-Comparatives does. It doesn't seem so.

 

Hmmm Actually the results of AVC's dynamic tests and AV-Test's dynamic tests are pretty much the same . This one discussed here is NOT a dynamic test - it is a complete test (all kind of testing) for being certified .

AV-Test's latest dynamic test:
http://blogs.pcmag.com/securitywatch/2009/12/av-testorg_releases_real-world.php

AV-Comparative's latest dynamic test:
http://av-comparatives.org/comparativesreviews/dynamic-tests

 

For example, AVIRA scored 97/100 with the AV-Comparatives dynamic test
and only 38% with the AV-Test dynamic test. I would say there is quite a discrepancy there.

 

I gave you links above . When the web page opens on my screen in the AV-Test.org I see
Avira Premium Security Suite 9.0 87.7%
Avira Premium Security Suite 9.0 87.2%

Please , be acurate and compare the results of both organisations of the previous year (end of 2009) . The current report discussed in another topic is all type (complete) testing and although it includes some dynamic test , it is not like the one of the previous year . AVC still hasn't released a dynamic test this year .

Moreover , we see this:

I think that both line 1 and line 2 can be called dynamic tests (real world testing = dynamic testing)

Additionally if you compare the dynamic test with other lines , it all proves another truth - AVIRA uses outdated technologies and is very good but not that good in zero-day malware (a.k.a. almost 100% of all malware in the past months and less in the current):




And to comment further about AVIRA's outdated technologies - it doesn't include cloud scanning , HIPS ... that kind of stuff

 

term is used differently. we would call that probably "behavioral test" (as it focuses on "blocking malware on or post execution").
what av-c means with whole-product dynamic testing would be that "real-world-testing"+"dynamic detection testing".

 

O.K. thanks for being honest. You're changing the definition of "dynamic" from AV-test. Even though the AV-Comparatives was run 8 months ago, that should explain the difference between 97% and 38%? Really?
Let's stay on topic, VB100 RAP test Feb-Aug 2010.

 

Agreed. It doesnt explain a thing. I think AV-C was right on spot with their test.

 

But it will include the proactive modul in a few months or years ...

 

Hmm disagree, what is 100 samples? When 50000 new samples are added to repository every day ? Neither dynamic tests are statistically relevant, and i agree with 3GUSER, AV-Test methodology is better since more complete. Regarding RAP here (note: some people are confusing it with VB100 certification) it's simply not realistic but i'm not saying it's useless.

Or never since it was anounced for Q1, postponed to Q2 and we're almost at Q3, no sign of it and they don't want to talk about it

 

myself has a problem to see eset such high rated and still failing on new rogues.
so the heuristic/proactive cant be such good!? (raising nr. of threads here)

 

Agreed.
Any attempt to draw conclusions from statistical analysis based on a selection of 16 when the population is >100.000, would cause homeric laughter at every university.
How come, it's reason enough for serious debate here?

Do some of you guys have the full report, does it perhaps offer a valid explanation for this specific test?

 

Yes , I have . I am a VB subscriber. What kind of explanation , would you be more specific ?