AV-Comparatives Results Online

Ahoi Captain Spuck!

 

Yes you do. And i have to agree with you on that. Heuristic can basically only score "well" if you can inspect the "RAW" files. (Except for file infector viruses and heuristic "methods" which do not need unpacked files, maybe detecting of strange double packed files (UPX+YodaProt etc))

That said if you include "new" runtime unpackers that the heuristical score might look completely different. Example: MEW / UPack / NSPack and the like.

Lots of "brandnew" malware is packed with this (Mytob worms for instance, ITW). Now if you added recently a unpacker for this - you can score much more better than just by using "normal" heuristic methods (except emulation)

 

Happy Bytes

why nod32 has 0% in script virus ?

 

Hi Mele20,

I have Firefox 1.5 with the PDF Download 0.5.1.2 and can confirm that the on-line link views as nicely as the alternative of downloading the PDF version in the link next to it and viewing it locally with the Adobe32 Reader.

I only have two extensions the say they are disabled and are grayed out, but appear to be working quite nicely - SpoofStick and Netcraft Toolbar.

-- Tom

 

Isn't it the August versions?

 

I think he is talking about retrospective/proactive Test. But to answer your question, No it isn't (http://www.avast.com/eng/av4_revision_history.html). The same applies to DrWeb. Version 4.33 has been much improved.

Anyway good work IBK!



tECHNODROME

 

If the online results looks weird, it usually helps if you refresh the page (F5 or the refresh button).

 

Hi tECHNODROME,

I was just noting that vlk had posted:

"The point is, the test doesn't really tell how the scanners perform NOW but how they performed 6 months ago. So it would be even more precise to call it a "May 2005" test instead of "November 2005" - as this is how the scanners worked in May."

Looks to me like IBK used the scanner versions available as of around August 4 or 5th. The same as the August "On Demand". So it seemed in error to say it was 6 months ago, or how they worked for May for all the scanners in the test?

But instead how they worked around the first part of August with the dates and versions used as noted by IBK on the test results.

 

Stan, you're right, sorry for this. I was assuming Andreas used 6-month old versions as usual - whereas in this particular test he used versions just 3 months old.

Anyway, the idea stays the same.


But OTOH I sort of have to agree with Happy Bytes, his point is also valid. New unpacker versions, for packers that didn't even EXIST 6 (or 3) months ago, can hardly count...

 

As usual a great test. I do hope you will include VBA next year

 

Hi vlk,

In what test did he use "6-month old versions"?

 

Think he meant 3 months, 08/05/2005. If you reverse it it will be 6 months however
Well .. Almost 6

 

Hi Brian N,

I was just replying to this:

Originally Posted by vlk
"Stan, you're right, sorry for this. I was assuming Andreas used 6-month old versions as usual"

 

3 months is usual. never did anything with 6 months.

 

A bit scary to see how many failed to pro-actively detect any of the "in-the-wild" samples.

Also, thanks for your tests, IBK!

 

Refreshing didn't help the look.

 

Like stated in the report (and what can be also seen by the grey numbers in the tables) this time the ITW category says nothing, as too few samples were used and only the wildlist of septmeber was available.

 

Is it possible to re-run a refresh test using only the latest engines from each vendor when there might be a case of radical improvements to some of the products within the testing timeframe as this case may be?

I think if there are radically changed or improved scan engine products that were out and available at the same time this testing was occuring than it really seems the results of all your hard work don't give us an accurate picture of what the products would have really scored itw during that same time.

Don't get me wrong on this as you have all the respect in the world from me on your skills and wonderful reports but to have current apps would only add more value to your testing efforts.

 

Nice work IBK

Surprising to see NOD32 surpass Kaspersky.

 

For the second consecutive time in the retrospective test.

 

And seeing BitDefender and NOD32 maintain their Advanced+ status, with Kaspersky losing it and going to Advanced.

 

I wonder if other "not too good with computers" type folks are confused, as I am.

KAV did so well in the on-demand test but falls short on the retro test. I wonder if the retro test is as useful as the on-demand test. I am not a KAV user, but instead have Bit Defender. I thought that detection results in the Aug test made KAV a clear winner, but in this retro test NOD is the clear winner by a wide margin. BD does not do badly in either test.

I had thought of going to KAV when my BD license expired, but now I am not convinced that KAV is better all around.

Are the two tests of equal value in the real world?

Jerry

 

Kav didn't do bad to have it's database out of date three months.

 

Eugene commented here:
http://forum.kaspersky.com/index.php?showtopic=6798

 

I'm not sure I understand, aren't all the tested AV databases either from 08/04/05 or 08/05/05?