AV-Comparatives June (May 2007) Results (Retrospective / Proactive Tests)

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Oops , sorry . I read it again and noticed it .

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I'm deeply offended by such a mature post from you!........................Chris you need to grow up and stop being such a fanboy, i had to egt that worked up by idiot posts about Kaspersky i would get a heart attack, i was pulling your leg and as always your insecurity got the best of you, i should have been aware of that and never tried to be funny with you...............in case you didn't get it, it was a friendly joke which you completely missed.

P.s. It was you who posted that you were drunk while posting a while ack, so don't get too worked up over that.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I think Kaspersky did well. 35% is higher than msot and it also got very few false positives.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I wonder if the new heuristic engine will once again make Kaspersky undisputed master of detection.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I don't think so.
Kaspersky = smoking fast release of signatures. In the meantime, use the PDM.
The new heuristic engine should steal a bit of work from the PDM.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Nice job to IBK and team for the results.

I looked through the results of the tests from 2005 onwards, and realised that not all of the products in the market are tested. I understand that they don't test every product because some of them score below 80% in the real world detection. However, from what I understand, those tests were done at least a year or two ago, meaning they could have improved tremendously. Also, how about new AVs such as Rising that seems to be gaining ground? I would suggest that AV Comparative run one full length test on each available AV out there every year or once in 2 years.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

IBk already tested rising. and it only got 71% He also tested ArcaVir, Clamwin, Comodo, CounterSpy, Ikarus, UNA, and VBA32 all of them failed

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Rising (alongside Arcavir, VBA32, Ikarus, ClamAV, Comodo, Sunbelt) is tested in "Others AVs" test
There's also a special test of CA eTrust and Trend Micro.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Heuristics heuristics -- NOD's heuristics surpasses them all. KAV, DrWeb, etc, are relegated to arguing about which of them is further from last place. To wit -- like the old NY Giants -- the other AVs can only say "Wait till next year."

The license for my present AV (don't ask) expires 7/7/2007. IF I decide to continue using any blacklist-based security program, I guess that NOD must be at the top of my "buy list." (Sigh) It has most definitly earned that position, I think.

Concerning heuristics, Weird Willy sez: "Ah well, heur today, Guam to Maui."

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

You're right, . My bad, I didn't notice there were such a test, and kept on wondering why it is the same product that kept on be tested. I also failed to realise that there were so many other tests there.... Thanks!

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I want to see an immediate end to posts containing personal remarks/attacks or whatever you might choose to call them. AV tests can - and will - be discussed without such garbage. I'm not removing anything now but if admin feels otherwise I'll be back to do as they wish. In the meantime I really think we all have a better idea than this in regards to what is and is not acceptable. Let us be adults and have discussion/debate etc deserving of the intellect possessed by Wilders' members.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

As true as that is, it doesn't save them in the detection/removal category.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

If heuristics are so important and useful, then why do not those AVs that have the best heuristics score higher on the on-demand tests?
I suppose the answer might be that the on-demand tests use malware that is known instead of zero day malware, and frequent updates offset any advantage heuristics would provide.

But if heuristics are so good, it would seem that even if the AV did not have a particular sample in its data base then heuristics would tend to pick it up. Isn't that what it is supposed to do? So if heuristics were turned off on the AV would it score lower on the on-demand tests? Since NOD has the best, overall, heuristics would it score even lower, 92% instead of 96%, on the on-demand tests than it did on the Feb AVC test?

Thanks,
Jerry

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

If you have a signature, you can do "variant detection" (heuristics) around it.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I don't know what that means.
Best,
Jerry

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

It means that if the behaviour of an infected file has some similarity with another malware for which the AV has a signature, and a specific family of malware at that, then it can report a suspicious detection as "variant of <blah blah>"

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Hi Firecat,
Thanks. Would that increase the on-demand detection percentages in AVC tests?

Best,
Jerry

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Yes, it would. To what extent is, however, unknown to me.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

In this latest test, NOD32 scored a 68% rate on the sample set used, and gained an Advanced+ category stamping. AVIRA scored 3% more, but got classified as Standard due to false positives.

This is all based on a sample set of over 20,000; are we to assume these and other AVs will score roughly the same percentages in the real world?

People are saying they expected KL to do better even though its new emulator is still in its infancy. In fact, if truth be told I think we should be saying ALL the vendors could do better if we really want heuristics to shine as a form of protection against zero-day malware.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

In my day to day computing, I never get any FP's with Avira. That isn't saying what the test found is wrong, just that if you don't go out trying every piece of software and open every email, you are more than likely not going to see them. I do have the heuristic set to HIGH. I ran NOD for a couple years prior, but decided on Avira once I went to Vista. It just ran better for me.

EDIT: Avira also scanned over 20GB of programs on my USB storage, not a single FP. It actually found a trojan that it took Norton 3 weeks to get signatures for.( I have the data mirrored on a 2nd pc with Norton 2007)

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I have Avira AntiVir's heuristic set to medium. In my old XP home puter it never found anything by scan or by real time detection. It was clean and no fp's either.
With my new pc it is the same.
It is highly unlikely that my computer would have had anyways those 18 false positives, 18 omg so many, lol.

Even my good old Avast found Processguard once as a trojan. So I know it can be a bother if some system files or security programs are found as FP's.

But as IBK gave Symantec an advanced and also in that report telling about quality assurance while he sure very well knew what happened to those people in China, last week I think? Symantec FP made their puters unbootable.
http://www.dslreports.com/forum/remark,18395883
One cannot say quality assurance if some critical system file is found as a fp.

One has to appreciate Avira's configuration abilities, great detection and a no nonsense clear help file. The update problems for free version users from time to time not.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

AntiVir is still my AV of choice, too, but I have had at least one false positive and 2-3 other iffy detections. It found a supposed Zlob trojan in my Gateway Games Console folder. I read from searching Google that someone else had the same detection, and I'm not sure yet if it is a FP.

Still think I'll stick with AntiVir because of its test results at AV Comparatives and elsewhere. I will try Comodo for firewall when it's ready for Vista.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

The congratulations and my satisfactions for the av-comparative team.

I believe I will "defend" the results of KAV7. From my point of view, the output of 35% is surprisingly good. Till these result, many praise him in laudations, as when he shall appear KAV you'll be best. And they were such of secure. Now as the results appeared, they don't praise him... Furthermore, they are disappointed ...And they criticize him.

Honestly, i expect a less 20%, at the very most. 35% is charm well. Why?
-because it's just "born"
-"it's in the house", it was not buy from another vendor and it has no past in the field
-"at first call" (first test) it came out on fifth place, it's half of Avira's detection

Others av tried for long time to exceed 20-30% and they fail to.

Allow me forwards to be pessimist ... i don't think that KAV7 heuristic will exceed 50% in tests very soon.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

This test measures proactive DETECTION, i.e. if Kaspersky had signature detection for many of these threats at the time of testing, this is also regarded as proactive detection. It would seem that a large part of Kaspersky's 35% score consisted of signature and not heuristic detections, which impressively shows how fast and reactive Kaspersky is with respect to adding signatures!

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Why u say that? It means that detection for some of the malware from this test were already on the data base of KAV?
If that's so, this isn't look like a heuristic detection

I think an heuristic test should be like this:
-every av on test up-to-date and "freeze"
-till the test will be done, colect all kind of NEW malware, that "appear" in this time
-test av only with this malware