Why did Avast fail Virus Bulletin June 2006

[JEfromCanada]

Can someone from Alwil tell me why Virus Bulletin did not give a VB100% award to avast in its June 2006 test of avast for Windows XP.

I have confidence in this product, but my reputation is on the line for recommending it to several of my friends.  Thanks for your response.




Jack

[igor]

Can someone from Alwil tell me why Virus Bulletin did not give a VB100% award to avast in its June 2006 test of avast for Windows XP.

Actually, that's not [going to be] correct

[DavidR]

I look at real life experiences and not just tests which as far as I'm aware only test the on-demand scan. As a happy avast users for over two years, I have confidence in avast, however no single program is going to provide full protection, so the use of multi-level compatible security programs will beef up your overall protection.

If you haven't already got this software (freeware), download, install, update and run it.
1. Ad-Aware
2. Spybot Search and Destroy
3. Spywareblaster Don't install this until you are clean.
4. Ewido Security Suite If using winXP. or a-Squared free if using win98/ME.

[Vlk]

Avast will get the VB100% mark for the June 06 XP test, together with an apology, in the July 2006 issue (due in a couple of days).

During the test, avast was blamed for having three false positives; however, it turned out that these were in fact not false...

[RejZoR]

Rofl

[JEfromCanada]

Avast will get the VB100% mark for the June 06 XP test, together with an apology, in the July 2006 issue (due in a couple of days).

During the test, avast was blamed for having three false positives; however, it turned out that these were in fact not false...

Thanks, that makes me feel better.

Does that mean that all other products that did NOT report those "false positives" as actual positives will be downgraded to a failure?   

[JEfromCanada]

[snip] however no single program is going to provide full protection, so the use of multi-level compatible security programs will beef up your overall protection [snip]

I agree, and I urge a multi-level approach to security - I just wanted to ensure that my current recommendation for virus protection was still valid.

[Vlk]

Does that mean that all other products that did NOT report those "false positives" as actual positives will be downgraded to a failure?

No, because they're not "ItW" samples (VB100% pass or failure is only given by the ability to detect ItW viruses)

[YLAP]

During the test, avast was blamed for having three false positives; however, it turned out that these were in fact not false...

LOL!  That's great! 

[JerryM]

Vlk,
That is good to know. I admit that I was surprised to see that Avast had not passed.

I wonder why they thought the three were FP? It seems to me that does not speak well for the tests.

Jerry

[igor]

Well, let's put it this way: the files were very questionable. They were not really malicious, but it was not clear whether they should be detected or not. Such files should not be in a cleanset for an official test.

[JerryM]

Well, let's put it this way: the files were very questionable. They were not really malicious, but it was not clear whether they should be detected or not. Such files should not be in a cleanset for an official test.

I agree, and that again, causes me to wonder as to the expertise of the testers. It is a sad fact that bad news travels fast, and an erroneous report is sometimes remembered while the correct version either not noticed or forgotten.

Best,
Jerry

[Vlk]

It has happened a couple of times in the past, too.
Last time, it was Kasperky (if I remember correctly), about a year ago.

Once, it even happened to Avast (I think that was in 2001).


The problem is that the VB cleanset is not maintained too well - and sometimes, some questionable files slip in. This time, the files that avast was detecting were tools for adding the current user to the admin group, with the aid of a security hole in a Windows component. These were not false positives from avast - avast was detecting them intentionally (they were part of our database of samples). Strange that no other AV actually reported them as malicious...

[JerryM]

Thanks for the additional information. Avast is too good an AV to suffer from such errors. However, I do not think a lot of folks are even aware of the test, and those of us who are can understand the problem, and that it is not an Avast shortcoming.

Best,
Jerry

[chocholo]

http://www.avast.com/eng/100-virus-bulletin-award.html
http://www.virusbtn.com/vb100/archive/2006/06 (free registration required)
http://www.virusbtn.com/news/vb_news/2006/07_01b.xml